go/src/cmd/compile/internal/gc/walk.go

4070 lines
102 KiB
Go
Raw Normal View History

// Copyright 2009 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package gc
import (
"cmd/compile/internal/types"
"cmd/internal/obj"
"cmd/internal/objabi"
"cmd/internal/sys"
"encoding/binary"
"fmt"
"strings"
)
// The constant is known to runtime.
const tmpstringbufsize = 32
const zeroValSize = 1024 // must match value of runtime/map.go:maxZero
func walk(fn *Node) {
Curfn = fn
if Debug['W'] != 0 {
s := fmt.Sprintf("\nbefore walk %v", Curfn.Func.Nname.Sym)
dumplist(s, Curfn.Nbody)
}
lno := lineno
// Final typecheck for any unused variables.
for i, ln := range fn.Func.Dcl {
cmd/compile: move Node.Class to flags Put it at position zero, since it is fairly hot. This shrinks gc.Node into a smaller size class on 64 bit systems. name old time/op new time/op delta Template 193ms ± 5% 192ms ± 3% ~ (p=0.353 n=94+93) Unicode 86.1ms ± 5% 85.0ms ± 4% -1.23% (p=0.000 n=95+98) GoTypes 546ms ± 3% 544ms ± 4% -0.40% (p=0.007 n=94+97) Compiler 2.56s ± 3% 2.54s ± 3% -0.67% (p=0.000 n=99+97) SSA 5.13s ± 2% 5.10s ± 3% -0.55% (p=0.000 n=94+98) Flate 122ms ± 6% 121ms ± 4% -0.75% (p=0.002 n=97+95) GoParser 144ms ± 5% 144ms ± 4% ~ (p=0.298 n=98+97) Reflect 348ms ± 4% 349ms ± 4% ~ (p=0.350 n=98+97) Tar 105ms ± 5% 104ms ± 5% ~ (p=0.154 n=96+98) XML 200ms ± 5% 198ms ± 4% -0.71% (p=0.015 n=97+98) [Geo mean] 330ms 328ms -0.52% name old user-time/op new user-time/op delta Template 229ms ±11% 224ms ± 7% -2.16% (p=0.001 n=100+87) Unicode 109ms ± 5% 109ms ± 6% ~ (p=0.897 n=96+91) GoTypes 712ms ± 4% 709ms ± 4% ~ (p=0.085 n=96+98) Compiler 3.41s ± 3% 3.36s ± 3% -1.43% (p=0.000 n=98+98) SSA 7.46s ± 3% 7.31s ± 3% -2.02% (p=0.000 n=100+99) Flate 145ms ± 6% 143ms ± 6% -1.11% (p=0.001 n=99+97) GoParser 177ms ± 5% 176ms ± 5% -0.78% (p=0.018 n=95+95) Reflect 432ms ± 7% 435ms ± 9% ~ (p=0.296 n=100+100) Tar 121ms ± 7% 121ms ± 5% ~ (p=0.072 n=100+95) XML 241ms ± 4% 239ms ± 5% ~ (p=0.085 n=97+99) [Geo mean] 413ms 410ms -0.73% name old alloc/op new alloc/op delta Template 38.4MB ± 0% 37.7MB ± 0% -1.85% (p=0.008 n=5+5) Unicode 30.1MB ± 0% 28.8MB ± 0% -4.09% (p=0.008 n=5+5) GoTypes 112MB ± 0% 110MB ± 0% -1.69% (p=0.008 n=5+5) Compiler 470MB ± 0% 461MB ± 0% -1.91% (p=0.008 n=5+5) SSA 1.13GB ± 0% 1.11GB ± 0% -1.70% (p=0.008 n=5+5) Flate 25.0MB ± 0% 24.6MB ± 0% -1.67% (p=0.008 n=5+5) GoParser 31.6MB ± 0% 31.1MB ± 0% -1.66% (p=0.008 n=5+5) Reflect 77.1MB ± 0% 75.8MB ± 0% -1.69% (p=0.008 n=5+5) Tar 26.3MB ± 0% 25.7MB ± 0% -2.06% (p=0.008 n=5+5) XML 41.9MB ± 0% 41.1MB ± 0% -1.93% (p=0.008 n=5+5) [Geo mean] 73.5MB 72.0MB -2.03% name old allocs/op new allocs/op delta Template 383k ± 0% 383k ± 0% ~ (p=0.690 n=5+5) Unicode 343k ± 0% 343k ± 0% ~ (p=0.841 n=5+5) GoTypes 1.16M ± 0% 1.16M ± 0% ~ (p=0.310 n=5+5) Compiler 4.43M ± 0% 4.42M ± 0% -0.17% (p=0.008 n=5+5) SSA 9.85M ± 0% 9.85M ± 0% ~ (p=0.310 n=5+5) Flate 236k ± 0% 236k ± 1% ~ (p=0.841 n=5+5) GoParser 320k ± 0% 320k ± 0% ~ (p=0.421 n=5+5) Reflect 988k ± 0% 987k ± 0% ~ (p=0.690 n=5+5) Tar 252k ± 0% 251k ± 0% ~ (p=0.095 n=5+5) XML 399k ± 0% 399k ± 0% ~ (p=1.000 n=5+5) [Geo mean] 741k 740k -0.07% Change-Id: I9e952b58a98e30a12494304db9ce50d0a85e459c Reviewed-on: https://go-review.googlesource.com/41797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com>
2017-04-25 18:14:12 -07:00
if ln.Op == ONAME && (ln.Class() == PAUTO || ln.Class() == PAUTOHEAP) {
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
ln = typecheck(ln, ctxExpr|ctxAssign)
fn.Func.Dcl[i] = ln
}
}
// Propagate the used flag for typeswitch variables up to the NONAME in its definition.
for _, ln := range fn.Func.Dcl {
if ln.Op == ONAME && (ln.Class() == PAUTO || ln.Class() == PAUTOHEAP) && ln.Name.Defn != nil && ln.Name.Defn.Op == OTYPESW && ln.Name.Used() {
ln.Name.Defn.Left.Name.SetUsed(true)
}
}
for _, ln := range fn.Func.Dcl {
if ln.Op != ONAME || (ln.Class() != PAUTO && ln.Class() != PAUTOHEAP) || ln.Sym.Name[0] == '&' || ln.Name.Used() {
continue
}
if defn := ln.Name.Defn; defn != nil && defn.Op == OTYPESW {
if defn.Left.Name.Used() {
continue
}
yyerrorl(defn.Left.Pos, "%v declared but not used", ln.Sym)
defn.Left.Name.SetUsed(true) // suppress repeats
} else {
yyerrorl(ln.Pos, "%v declared but not used", ln.Sym)
}
}
lineno = lno
if nerrors != 0 {
return
}
walkstmtlist(Curfn.Nbody.Slice())
if Debug['W'] != 0 {
s := fmt.Sprintf("after walk %v", Curfn.Func.Nname.Sym)
dumplist(s, Curfn.Nbody)
}
zeroResults()
heapmoves()
if Debug['W'] != 0 && Curfn.Func.Enter.Len() > 0 {
s := fmt.Sprintf("enter %v", Curfn.Func.Nname.Sym)
dumplist(s, Curfn.Func.Enter)
}
}
func walkstmtlist(s []*Node) {
for i := range s {
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
s[i] = walkstmt(s[i])
}
}
func paramoutheap(fn *Node) bool {
for _, ln := range fn.Func.Dcl {
cmd/compile: move Node.Class to flags Put it at position zero, since it is fairly hot. This shrinks gc.Node into a smaller size class on 64 bit systems. name old time/op new time/op delta Template 193ms ± 5% 192ms ± 3% ~ (p=0.353 n=94+93) Unicode 86.1ms ± 5% 85.0ms ± 4% -1.23% (p=0.000 n=95+98) GoTypes 546ms ± 3% 544ms ± 4% -0.40% (p=0.007 n=94+97) Compiler 2.56s ± 3% 2.54s ± 3% -0.67% (p=0.000 n=99+97) SSA 5.13s ± 2% 5.10s ± 3% -0.55% (p=0.000 n=94+98) Flate 122ms ± 6% 121ms ± 4% -0.75% (p=0.002 n=97+95) GoParser 144ms ± 5% 144ms ± 4% ~ (p=0.298 n=98+97) Reflect 348ms ± 4% 349ms ± 4% ~ (p=0.350 n=98+97) Tar 105ms ± 5% 104ms ± 5% ~ (p=0.154 n=96+98) XML 200ms ± 5% 198ms ± 4% -0.71% (p=0.015 n=97+98) [Geo mean] 330ms 328ms -0.52% name old user-time/op new user-time/op delta Template 229ms ±11% 224ms ± 7% -2.16% (p=0.001 n=100+87) Unicode 109ms ± 5% 109ms ± 6% ~ (p=0.897 n=96+91) GoTypes 712ms ± 4% 709ms ± 4% ~ (p=0.085 n=96+98) Compiler 3.41s ± 3% 3.36s ± 3% -1.43% (p=0.000 n=98+98) SSA 7.46s ± 3% 7.31s ± 3% -2.02% (p=0.000 n=100+99) Flate 145ms ± 6% 143ms ± 6% -1.11% (p=0.001 n=99+97) GoParser 177ms ± 5% 176ms ± 5% -0.78% (p=0.018 n=95+95) Reflect 432ms ± 7% 435ms ± 9% ~ (p=0.296 n=100+100) Tar 121ms ± 7% 121ms ± 5% ~ (p=0.072 n=100+95) XML 241ms ± 4% 239ms ± 5% ~ (p=0.085 n=97+99) [Geo mean] 413ms 410ms -0.73% name old alloc/op new alloc/op delta Template 38.4MB ± 0% 37.7MB ± 0% -1.85% (p=0.008 n=5+5) Unicode 30.1MB ± 0% 28.8MB ± 0% -4.09% (p=0.008 n=5+5) GoTypes 112MB ± 0% 110MB ± 0% -1.69% (p=0.008 n=5+5) Compiler 470MB ± 0% 461MB ± 0% -1.91% (p=0.008 n=5+5) SSA 1.13GB ± 0% 1.11GB ± 0% -1.70% (p=0.008 n=5+5) Flate 25.0MB ± 0% 24.6MB ± 0% -1.67% (p=0.008 n=5+5) GoParser 31.6MB ± 0% 31.1MB ± 0% -1.66% (p=0.008 n=5+5) Reflect 77.1MB ± 0% 75.8MB ± 0% -1.69% (p=0.008 n=5+5) Tar 26.3MB ± 0% 25.7MB ± 0% -2.06% (p=0.008 n=5+5) XML 41.9MB ± 0% 41.1MB ± 0% -1.93% (p=0.008 n=5+5) [Geo mean] 73.5MB 72.0MB -2.03% name old allocs/op new allocs/op delta Template 383k ± 0% 383k ± 0% ~ (p=0.690 n=5+5) Unicode 343k ± 0% 343k ± 0% ~ (p=0.841 n=5+5) GoTypes 1.16M ± 0% 1.16M ± 0% ~ (p=0.310 n=5+5) Compiler 4.43M ± 0% 4.42M ± 0% -0.17% (p=0.008 n=5+5) SSA 9.85M ± 0% 9.85M ± 0% ~ (p=0.310 n=5+5) Flate 236k ± 0% 236k ± 1% ~ (p=0.841 n=5+5) GoParser 320k ± 0% 320k ± 0% ~ (p=0.421 n=5+5) Reflect 988k ± 0% 987k ± 0% ~ (p=0.690 n=5+5) Tar 252k ± 0% 251k ± 0% ~ (p=0.095 n=5+5) XML 399k ± 0% 399k ± 0% ~ (p=1.000 n=5+5) [Geo mean] 741k 740k -0.07% Change-Id: I9e952b58a98e30a12494304db9ce50d0a85e459c Reviewed-on: https://go-review.googlesource.com/41797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com>
2017-04-25 18:14:12 -07:00
switch ln.Class() {
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
case PPARAMOUT:
if ln.isParamStackCopy() || ln.Name.Addrtaken() {
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
return true
}
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
case PAUTO:
// stop early - parameters are over
return false
}
}
return false
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
// The result of walkstmt MUST be assigned back to n, e.g.
// n.Left = walkstmt(n.Left)
func walkstmt(n *Node) *Node {
if n == nil {
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
return n
}
setlineno(n)
walkstmtlist(n.Ninit.Slice())
switch n.Op {
default:
if n.Op == ONAME {
yyerror("%v is not a top level statement", n.Sym)
} else {
yyerror("%v is not a top level statement", n.Op)
}
Dump("nottop", n)
case OAS,
OASOP,
OAS2,
OAS2DOTTYPE,
OAS2RECV,
OAS2FUNC,
OAS2MAPR,
OCLOSE,
OCOPY,
OCALLMETH,
OCALLINTER,
OCALL,
OCALLFUNC,
ODELETE,
OSEND,
OPRINT,
OPRINTN,
OPANIC,
OEMPTY,
cmd/internal/gc: inline runtime.getg This more closely restores what the old C runtime did. (In C, g was an 'extern register' with the same effective implementation as in this CL.) On a late 2012 MacBookPro10,2, best of 5 old vs best of 5 new: benchmark old ns/op new ns/op delta BenchmarkBinaryTree17 4981312777 4463426605 -10.40% BenchmarkFannkuch11 3046495712 3006819428 -1.30% BenchmarkFmtFprintfEmpty 89.3 79.8 -10.64% BenchmarkFmtFprintfString 284 262 -7.75% BenchmarkFmtFprintfInt 282 262 -7.09% BenchmarkFmtFprintfIntInt 480 448 -6.67% BenchmarkFmtFprintfPrefixedInt 382 358 -6.28% BenchmarkFmtFprintfFloat 529 486 -8.13% BenchmarkFmtManyArgs 1849 1773 -4.11% BenchmarkGobDecode 12835963 11794385 -8.11% BenchmarkGobEncode 10527170 10288422 -2.27% BenchmarkGzip 436109569 438422516 +0.53% BenchmarkGunzip 110121663 109843648 -0.25% BenchmarkHTTPClientServer 81930 85446 +4.29% BenchmarkJSONEncode 24638574 24280603 -1.45% BenchmarkJSONDecode 93022423 85753546 -7.81% BenchmarkMandelbrot200 4703899 4735407 +0.67% BenchmarkGoParse 5319853 5086843 -4.38% BenchmarkRegexpMatchEasy0_32 151 151 +0.00% BenchmarkRegexpMatchEasy0_1K 452 453 +0.22% BenchmarkRegexpMatchEasy1_32 131 132 +0.76% BenchmarkRegexpMatchEasy1_1K 761 722 -5.12% BenchmarkRegexpMatchMedium_32 228 224 -1.75% BenchmarkRegexpMatchMedium_1K 63751 64296 +0.85% BenchmarkRegexpMatchHard_32 3188 3238 +1.57% BenchmarkRegexpMatchHard_1K 95396 96756 +1.43% BenchmarkRevcomp 661587262 687107364 +3.86% BenchmarkTemplate 108312598 104008540 -3.97% BenchmarkTimeParse 453 459 +1.32% BenchmarkTimeFormat 475 441 -7.16% The garbage benchmark from the benchmarks subrepo gets 2.6% faster as well. Change-Id: I320aeda332db81012688b26ffab23f6581c59cfa Reviewed-on: https://go-review.googlesource.com/8460 Reviewed-by: Rick Hudson <rlh@golang.org> Run-TryBot: Rick Hudson <rlh@golang.org> Reviewed-by: Austin Clements <austin@google.com>
2015-04-03 12:23:28 -04:00
ORECOVER,
OGETG:
if n.Typecheck() == 0 {
Fatalf("missing typecheck: %+v", n)
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
wascopy := n.Op == OCOPY
init := n.Ninit
n.Ninit.Set(nil)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n = walkexpr(n, &init)
n = addinit(n, init.Slice())
if wascopy && n.Op == OCONVNOP {
n.Op = OEMPTY // don't leave plain values as statements.
}
// special case for a receive where we throw away
// the value received.
case ORECV:
if n.Typecheck() == 0 {
Fatalf("missing typecheck: %+v", n)
}
init := n.Ninit
n.Ninit.Set(nil)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, &init)
n = mkcall1(chanfn("chanrecv1", 2, n.Left.Type), nil, &init, n.Left, nodnil())
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n = walkexpr(n, &init)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n = addinit(n, init.Slice())
case OBREAK,
OCONTINUE,
OFALL,
OGOTO,
OLABEL,
ODCLCONST,
ODCLTYPE,
OCHECKNIL,
OVARDEF,
cmd/compile: recognize Syscall-like functions for liveness analysis Consider this code: func f(*int) func g() { p := new(int) f(p) } where f is an assembly function. In general liveness analysis assumes that during the call to f, p is dead in this frame. If f has retained p, p will be found alive in f's frame and keep the new(int) from being garbage collected. This is all correct and works. We use the Go func declaration for f to give the assembly function liveness information (the arguments are assumed live for the entire call). Now consider this code: func h1() { p := new(int) syscall.Syscall(1, 2, 3, uintptr(unsafe.Pointer(p))) } Here syscall.Syscall is taking the place of f, but because its arguments are uintptr, the liveness analysis and the garbage collector ignore them. Since p is no longer live in h once the call starts, if the garbage collector scans the stack while the system call is blocked, it will find no reference to the new(int) and reclaim it. If the kernel is going to write to *p once the call finishes, reclaiming the memory is a mistake. We can't change the arguments or the liveness information for syscall.Syscall itself, both for compatibility and because sometimes the arguments really are integers, and the garbage collector will get quite upset if it finds an integer where it expects a pointer. The problem is that these arguments are fundamentally untyped. The solution we have taken in the syscall package's wrappers in past releases is to insert a call to a dummy function named "use", to make it look like the argument is live during the call to syscall.Syscall: func h2() { p := new(int) syscall.Syscall(1, 2, 3, uintptr(unsafe.Pointer(p))) use(unsafe.Pointer(p)) } Keeping p alive during the call means that if the garbage collector scans the stack during the system call now, it will find the reference to p. Unfortunately, this approach is not available to users outside syscall, because 'use' is unexported, and people also have to realize they need to use it and do so. There is much existing code using syscall.Syscall without a 'use'-like function. That code will fail very occasionally in mysterious ways (see #13372). This CL fixes all that existing code by making the compiler do the right thing automatically, without any code modifications. That is, it takes h1 above, which is incorrect code today, and makes it correct code. Specifically, if the compiler sees a foreign func definition (one without a body) that has uintptr arguments, it marks those arguments as "unsafe uintptrs". If it later sees the function being called with uintptr(unsafe.Pointer(x)) as an argument, it arranges to mark x as having escaped, and it makes sure to hold x in a live temporary variable until the call returns, so that the garbage collector cannot reclaim whatever heap memory x points to. For now I am leaving the explicit calls to use in package syscall, but they can be removed early in a future cycle (likely Go 1.7). The rule has no effect on escape analysis, only on liveness analysis. Fixes #13372. Change-Id: I2addb83f70d08db08c64d394f9d06ff0a063c500 Reviewed-on: https://go-review.googlesource.com/18584 Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-01-13 00:46:28 -05:00
OVARKILL,
OVARLIVE:
break
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
case ODCL:
v := n.Left
cmd/compile: move Node.Class to flags Put it at position zero, since it is fairly hot. This shrinks gc.Node into a smaller size class on 64 bit systems. name old time/op new time/op delta Template 193ms ± 5% 192ms ± 3% ~ (p=0.353 n=94+93) Unicode 86.1ms ± 5% 85.0ms ± 4% -1.23% (p=0.000 n=95+98) GoTypes 546ms ± 3% 544ms ± 4% -0.40% (p=0.007 n=94+97) Compiler 2.56s ± 3% 2.54s ± 3% -0.67% (p=0.000 n=99+97) SSA 5.13s ± 2% 5.10s ± 3% -0.55% (p=0.000 n=94+98) Flate 122ms ± 6% 121ms ± 4% -0.75% (p=0.002 n=97+95) GoParser 144ms ± 5% 144ms ± 4% ~ (p=0.298 n=98+97) Reflect 348ms ± 4% 349ms ± 4% ~ (p=0.350 n=98+97) Tar 105ms ± 5% 104ms ± 5% ~ (p=0.154 n=96+98) XML 200ms ± 5% 198ms ± 4% -0.71% (p=0.015 n=97+98) [Geo mean] 330ms 328ms -0.52% name old user-time/op new user-time/op delta Template 229ms ±11% 224ms ± 7% -2.16% (p=0.001 n=100+87) Unicode 109ms ± 5% 109ms ± 6% ~ (p=0.897 n=96+91) GoTypes 712ms ± 4% 709ms ± 4% ~ (p=0.085 n=96+98) Compiler 3.41s ± 3% 3.36s ± 3% -1.43% (p=0.000 n=98+98) SSA 7.46s ± 3% 7.31s ± 3% -2.02% (p=0.000 n=100+99) Flate 145ms ± 6% 143ms ± 6% -1.11% (p=0.001 n=99+97) GoParser 177ms ± 5% 176ms ± 5% -0.78% (p=0.018 n=95+95) Reflect 432ms ± 7% 435ms ± 9% ~ (p=0.296 n=100+100) Tar 121ms ± 7% 121ms ± 5% ~ (p=0.072 n=100+95) XML 241ms ± 4% 239ms ± 5% ~ (p=0.085 n=97+99) [Geo mean] 413ms 410ms -0.73% name old alloc/op new alloc/op delta Template 38.4MB ± 0% 37.7MB ± 0% -1.85% (p=0.008 n=5+5) Unicode 30.1MB ± 0% 28.8MB ± 0% -4.09% (p=0.008 n=5+5) GoTypes 112MB ± 0% 110MB ± 0% -1.69% (p=0.008 n=5+5) Compiler 470MB ± 0% 461MB ± 0% -1.91% (p=0.008 n=5+5) SSA 1.13GB ± 0% 1.11GB ± 0% -1.70% (p=0.008 n=5+5) Flate 25.0MB ± 0% 24.6MB ± 0% -1.67% (p=0.008 n=5+5) GoParser 31.6MB ± 0% 31.1MB ± 0% -1.66% (p=0.008 n=5+5) Reflect 77.1MB ± 0% 75.8MB ± 0% -1.69% (p=0.008 n=5+5) Tar 26.3MB ± 0% 25.7MB ± 0% -2.06% (p=0.008 n=5+5) XML 41.9MB ± 0% 41.1MB ± 0% -1.93% (p=0.008 n=5+5) [Geo mean] 73.5MB 72.0MB -2.03% name old allocs/op new allocs/op delta Template 383k ± 0% 383k ± 0% ~ (p=0.690 n=5+5) Unicode 343k ± 0% 343k ± 0% ~ (p=0.841 n=5+5) GoTypes 1.16M ± 0% 1.16M ± 0% ~ (p=0.310 n=5+5) Compiler 4.43M ± 0% 4.42M ± 0% -0.17% (p=0.008 n=5+5) SSA 9.85M ± 0% 9.85M ± 0% ~ (p=0.310 n=5+5) Flate 236k ± 0% 236k ± 1% ~ (p=0.841 n=5+5) GoParser 320k ± 0% 320k ± 0% ~ (p=0.421 n=5+5) Reflect 988k ± 0% 987k ± 0% ~ (p=0.690 n=5+5) Tar 252k ± 0% 251k ± 0% ~ (p=0.095 n=5+5) XML 399k ± 0% 399k ± 0% ~ (p=1.000 n=5+5) [Geo mean] 741k 740k -0.07% Change-Id: I9e952b58a98e30a12494304db9ce50d0a85e459c Reviewed-on: https://go-review.googlesource.com/41797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com>
2017-04-25 18:14:12 -07:00
if v.Class() == PAUTOHEAP {
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
if compiling_runtime {
yyerror("%v escapes to heap, not allowed in runtime", v)
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
}
if prealloc[v] == nil {
prealloc[v] = callnew(v.Type)
}
nn := nod(OAS, v.Name.Param.Heapaddr, prealloc[v])
cmd/compile: pack bool fields in Node, Name, Func and Type structs to bitsets This reduces compiler memory usage by up to 4% - see compilebench results below. name old time/op new time/op delta Template 245ms ± 4% 241ms ± 2% -1.88% (p=0.029 n=10+10) Unicode 126ms ± 3% 124ms ± 3% ~ (p=0.105 n=10+10) GoTypes 805ms ± 2% 813ms ± 3% ~ (p=0.515 n=8+10) Compiler 3.95s ± 2% 3.83s ± 1% -2.96% (p=0.000 n=9+10) MakeBash 47.4s ± 4% 46.6s ± 1% -1.59% (p=0.028 n=9+10) name old user-ns/op new user-ns/op delta Template 324M ± 5% 326M ± 3% ~ (p=0.935 n=10+10) Unicode 186M ± 5% 178M ±10% ~ (p=0.067 n=9+10) GoTypes 1.08G ± 7% 1.09G ± 4% ~ (p=0.956 n=10+10) Compiler 5.34G ± 4% 5.31G ± 1% ~ (p=0.501 n=10+8) name old alloc/op new alloc/op delta Template 41.0MB ± 0% 39.8MB ± 0% -3.03% (p=0.000 n=10+10) Unicode 32.3MB ± 0% 31.0MB ± 0% -4.13% (p=0.000 n=10+10) GoTypes 119MB ± 0% 116MB ± 0% -2.39% (p=0.000 n=10+10) Compiler 499MB ± 0% 487MB ± 0% -2.48% (p=0.000 n=10+10) name old allocs/op new allocs/op delta Template 380k ± 1% 379k ± 1% ~ (p=0.436 n=10+10) Unicode 324k ± 1% 324k ± 0% ~ (p=0.853 n=10+10) GoTypes 1.15M ± 0% 1.15M ± 0% ~ (p=0.481 n=10+10) Compiler 4.41M ± 0% 4.41M ± 0% -0.12% (p=0.007 n=10+10) name old text-bytes new text-bytes delta HelloSize 623k ± 0% 623k ± 0% ~ (all equal) CmdGoSize 6.64M ± 0% 6.64M ± 0% ~ (all equal) name old data-bytes new data-bytes delta HelloSize 5.81k ± 0% 5.81k ± 0% ~ (all equal) CmdGoSize 238k ± 0% 238k ± 0% ~ (all equal) name old bss-bytes new bss-bytes delta HelloSize 134k ± 0% 134k ± 0% ~ (all equal) CmdGoSize 152k ± 0% 152k ± 0% ~ (all equal) name old exe-bytes new exe-bytes delta HelloSize 967k ± 0% 967k ± 0% ~ (all equal) CmdGoSize 10.2M ± 0% 10.2M ± 0% ~ (all equal) Change-Id: I1f40af738254892bd6c8ba2eb43390b175753d52 Reviewed-on: https://go-review.googlesource.com/37445 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-02-27 19:56:38 +02:00
nn.SetColas(true)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
nn = typecheck(nn, ctxStmt)
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
return walkstmt(nn)
}
case OBLOCK:
walkstmtlist(n.List.Slice())
case OCASE:
yyerror("case statement out of place")
case ODEFER:
Curfn.Func.SetHasDefer(true)
cmd/compile, cmd/link, runtime: make defers low-cost through inline code and extra funcdata Generate inline code at defer time to save the args of defer calls to unique (autotmp) stack slots, and generate inline code at exit time to check which defer calls were made and make the associated function/method/interface calls. We remember that a particular defer statement was reached by storing in the deferBits variable (always stored on the stack). At exit time, we check the bits of the deferBits variable to determine which defer function calls to make (in reverse order). These low-cost defers are only used for functions where no defers appear in loops. In addition, we don't do these low-cost defers if there are too many defer statements or too many exits in a function (to limit code increase). When a function uses open-coded defers, we produce extra FUNCDATA_OpenCodedDeferInfo information that specifies the number of defers, and for each defer, the stack slots where the closure and associated args have been stored. The funcdata also includes the location of the deferBits variable. Therefore, for panics, we can use this funcdata to determine exactly which defers are active, and call the appropriate functions/methods/closures with the correct arguments for each active defer. In order to unwind the stack correctly after a recover(), we need to add an extra code segment to functions with open-coded defers that simply calls deferreturn() and returns. This segment is not reachable by the normal function, but is returned to by the runtime during recovery. We set the liveness information of this deferreturn() to be the same as the liveness at the first function call during the last defer exit code (so all return values and all stack slots needed by the defer calls will be live). I needed to increase the stackguard constant from 880 to 896, because of a small amount of new code in deferreturn(). The -N flag disables open-coded defers. '-d defer' prints out the kind of defer being used at each defer statement (heap-allocated, stack-allocated, or open-coded). Cost of defer statement [ go test -run NONE -bench BenchmarkDefer$ runtime ] With normal (stack-allocated) defers only: 35.4 ns/op With open-coded defers: 5.6 ns/op Cost of function call alone (remove defer keyword): 4.4 ns/op Text size increase (including funcdata) for go binary without/with open-coded defers: 0.09% The average size increase (including funcdata) for only the functions that use open-coded defers is 1.1%. The cost of a panic followed by a recover got noticeably slower, since panic processing now requires a scan of the stack for open-coded defer frames. This scan is required, even if no frames are using open-coded defers: Cost of panic and recover [ go test -run NONE -bench BenchmarkPanicRecover runtime ] Without open-coded defers: 62.0 ns/op With open-coded defers: 255 ns/op A CGO Go-to-C-to-Go benchmark got noticeably faster because of open-coded defers: CGO Go-to-C-to-Go benchmark [cd misc/cgo/test; go test -run NONE -bench BenchmarkCGoCallback ] Without open-coded defers: 443 ns/op With open-coded defers: 347 ns/op Updates #14939 (defer performance) Updates #34481 (design doc) Change-Id: I63b1a60d1ebf28126f55ee9fd7ecffe9cb23d1ff Reviewed-on: https://go-review.googlesource.com/c/go/+/202340 Reviewed-by: Austin Clements <austin@google.com>
2019-06-24 12:59:22 -07:00
Curfn.Func.numDefers++
if Curfn.Func.numDefers > maxOpenDefers {
// Don't allow open-coded defers if there are more than
// 8 defers in the function, since we use a single
// byte to record active defers.
Curfn.Func.SetOpenCodedDeferDisallowed(true)
}
if n.Esc != EscNever {
// If n.Esc is not EscNever, then this defer occurs in a loop,
// so open-coded defers cannot be used in this function.
Curfn.Func.SetOpenCodedDeferDisallowed(true)
}
fallthrough
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
case OGO:
switch n.Left.Op {
case OPRINT, OPRINTN:
n.Left = wrapCall(n.Left, &n.Ninit)
case ODELETE:
if mapfast(n.Left.List.First().Type) == mapslow {
n.Left = wrapCall(n.Left, &n.Ninit)
} else {
n.Left = walkexpr(n.Left, &n.Ninit)
}
case OCOPY:
n.Left = copyany(n.Left, &n.Ninit, true)
default:
if n.Left.NeedsWrapper() {
n.Left = wrapCall(n.Left, &n.Ninit)
} else {
n.Left = walkexpr(n.Left, &n.Ninit)
}
}
case OFOR, OFORUNTIL:
if n.Left != nil {
walkstmtlist(n.Left.Ninit.Slice())
init := n.Left.Ninit
n.Left.Ninit.Set(nil)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, &init)
n.Left = addinit(n.Left, init.Slice())
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Right = walkstmt(n.Right)
cmd/compile: don't produce a past-the-end pointer in range loops Currently, range loops over slices and arrays are compiled roughly like: for i, x := range s { b } ⇓ for i, _n, _p := 0, len(s), &s[0]; i < _n; i, _p = i+1, _p + unsafe.Sizeof(s[0]) { b } ⇓ i, _n, _p := 0, len(s), &s[0] goto cond body: { b } i, _p = i+1, _p + unsafe.Sizeof(s[0]) cond: if i < _n { goto body } else { goto end } end: The problem with this lowering is that _p may temporarily point past the end of the allocation the moment before the loop terminates. Right now this isn't a problem because there's never a safe-point during this brief moment. We're about to introduce safe-points everywhere, so this bad pointer is going to be a problem. We could mark the increment as an unsafe block, but this inhibits reordering opportunities and could result in infrequent safe-points if the body is short. Instead, this CL fixes this by changing how we compile range loops to never produce this past-the-end pointer. It changes the lowering to roughly: i, _n, _p := 0, len(s), &s[0] if i < _n { goto body } else { goto end } top: _p += unsafe.Sizeof(s[0]) body: { b } i++ if i < _n { goto top } else { goto end } end: Notably, the increment is split into two parts: we increment the index before checking the condition, but increment the pointer only *after* the condition check has succeeded. The implementation builds on the OFORUNTIL construct that was introduced during the loop preemption experiments, since OFORUNTIL places the increment and condition after the loop body. To support the extra "late increment" step, we further define OFORUNTIL's "List" field to contain the late increment statements. This makes all of this a relatively small change. This depends on the improvements to the prove pass in CL 102603. With the current lowering, bounds-check elimination knows that i < _n in the body because the body block is dominated by the cond block. In the new lowering, deriving this fact requires detecting that i < _n on *both* paths into body and hence is true in body. CL 102603 made prove able to detect this. The code size effect of this is minimal. The cmd/go binary on linux/amd64 increases by 0.17%. Performance-wise, this actually appears to be a net win, though it's mostly noise: name old time/op new time/op delta BinaryTree17-12 2.80s ± 0% 2.61s ± 1% -6.88% (p=0.000 n=20+18) Fannkuch11-12 2.41s ± 0% 2.42s ± 0% +0.05% (p=0.005 n=20+20) FmtFprintfEmpty-12 41.6ns ± 5% 41.4ns ± 6% ~ (p=0.765 n=20+19) FmtFprintfString-12 69.4ns ± 3% 69.3ns ± 1% ~ (p=0.084 n=19+17) FmtFprintfInt-12 76.1ns ± 1% 77.3ns ± 1% +1.57% (p=0.000 n=19+19) FmtFprintfIntInt-12 122ns ± 2% 123ns ± 3% +0.95% (p=0.015 n=20+20) FmtFprintfPrefixedInt-12 153ns ± 2% 151ns ± 3% -1.27% (p=0.013 n=20+20) FmtFprintfFloat-12 215ns ± 0% 216ns ± 0% +0.47% (p=0.000 n=20+16) FmtManyArgs-12 486ns ± 1% 498ns ± 0% +2.40% (p=0.000 n=20+17) GobDecode-12 6.43ms ± 0% 6.50ms ± 0% +1.08% (p=0.000 n=18+19) GobEncode-12 5.43ms ± 1% 5.47ms ± 0% +0.76% (p=0.000 n=20+20) Gzip-12 218ms ± 1% 218ms ± 1% ~ (p=0.883 n=20+20) Gunzip-12 38.8ms ± 0% 38.9ms ± 0% ~ (p=0.644 n=19+19) HTTPClientServer-12 76.2µs ± 1% 76.4µs ± 2% ~ (p=0.218 n=20+20) JSONEncode-12 12.2ms ± 0% 12.3ms ± 1% +0.45% (p=0.000 n=19+19) JSONDecode-12 54.2ms ± 1% 53.3ms ± 0% -1.67% (p=0.000 n=20+20) Mandelbrot200-12 3.71ms ± 0% 3.71ms ± 0% ~ (p=0.143 n=19+20) GoParse-12 3.22ms ± 0% 3.19ms ± 1% -0.72% (p=0.000 n=20+20) RegexpMatchEasy0_32-12 76.7ns ± 1% 75.8ns ± 1% -1.19% (p=0.000 n=20+17) RegexpMatchEasy0_1K-12 245ns ± 1% 243ns ± 0% -0.72% (p=0.000 n=18+17) RegexpMatchEasy1_32-12 71.9ns ± 0% 71.7ns ± 1% -0.39% (p=0.006 n=12+18) RegexpMatchEasy1_1K-12 358ns ± 1% 354ns ± 1% -1.13% (p=0.000 n=20+19) RegexpMatchMedium_32-12 105ns ± 2% 105ns ± 1% -0.63% (p=0.007 n=19+20) RegexpMatchMedium_1K-12 31.9µs ± 1% 31.9µs ± 1% ~ (p=1.000 n=17+17) RegexpMatchHard_32-12 1.51µs ± 1% 1.52µs ± 2% +0.46% (p=0.042 n=18+18) RegexpMatchHard_1K-12 45.3µs ± 1% 45.5µs ± 2% +0.44% (p=0.029 n=18+19) Revcomp-12 388ms ± 1% 385ms ± 0% -0.57% (p=0.000 n=19+18) Template-12 63.0ms ± 1% 63.3ms ± 0% +0.50% (p=0.000 n=19+20) TimeParse-12 309ns ± 1% 307ns ± 0% -0.62% (p=0.000 n=20+20) TimeFormat-12 328ns ± 0% 333ns ± 0% +1.35% (p=0.000 n=19+19) [Geo mean] 47.0µs 46.9µs -0.20% (https://perf.golang.org/search?q=upload:20180326.1) For #10958. For #24543. Change-Id: Icbd52e711fdbe7938a1fea3e6baca1104b53ac3a Reviewed-on: https://go-review.googlesource.com/102604 Run-TryBot: Austin Clements <austin@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: David Chase <drchase@google.com>
2018-03-22 12:04:51 -04:00
if n.Op == OFORUNTIL {
walkstmtlist(n.List.Slice())
}
walkstmtlist(n.Nbody.Slice())
case OIF:
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, &n.Ninit)
walkstmtlist(n.Nbody.Slice())
walkstmtlist(n.Rlist.Slice())
case ORETURN:
cmd/compile, cmd/link, runtime: make defers low-cost through inline code and extra funcdata Generate inline code at defer time to save the args of defer calls to unique (autotmp) stack slots, and generate inline code at exit time to check which defer calls were made and make the associated function/method/interface calls. We remember that a particular defer statement was reached by storing in the deferBits variable (always stored on the stack). At exit time, we check the bits of the deferBits variable to determine which defer function calls to make (in reverse order). These low-cost defers are only used for functions where no defers appear in loops. In addition, we don't do these low-cost defers if there are too many defer statements or too many exits in a function (to limit code increase). When a function uses open-coded defers, we produce extra FUNCDATA_OpenCodedDeferInfo information that specifies the number of defers, and for each defer, the stack slots where the closure and associated args have been stored. The funcdata also includes the location of the deferBits variable. Therefore, for panics, we can use this funcdata to determine exactly which defers are active, and call the appropriate functions/methods/closures with the correct arguments for each active defer. In order to unwind the stack correctly after a recover(), we need to add an extra code segment to functions with open-coded defers that simply calls deferreturn() and returns. This segment is not reachable by the normal function, but is returned to by the runtime during recovery. We set the liveness information of this deferreturn() to be the same as the liveness at the first function call during the last defer exit code (so all return values and all stack slots needed by the defer calls will be live). I needed to increase the stackguard constant from 880 to 896, because of a small amount of new code in deferreturn(). The -N flag disables open-coded defers. '-d defer' prints out the kind of defer being used at each defer statement (heap-allocated, stack-allocated, or open-coded). Cost of defer statement [ go test -run NONE -bench BenchmarkDefer$ runtime ] With normal (stack-allocated) defers only: 35.4 ns/op With open-coded defers: 5.6 ns/op Cost of function call alone (remove defer keyword): 4.4 ns/op Text size increase (including funcdata) for go binary without/with open-coded defers: 0.09% The average size increase (including funcdata) for only the functions that use open-coded defers is 1.1%. The cost of a panic followed by a recover got noticeably slower, since panic processing now requires a scan of the stack for open-coded defer frames. This scan is required, even if no frames are using open-coded defers: Cost of panic and recover [ go test -run NONE -bench BenchmarkPanicRecover runtime ] Without open-coded defers: 62.0 ns/op With open-coded defers: 255 ns/op A CGO Go-to-C-to-Go benchmark got noticeably faster because of open-coded defers: CGO Go-to-C-to-Go benchmark [cd misc/cgo/test; go test -run NONE -bench BenchmarkCGoCallback ] Without open-coded defers: 443 ns/op With open-coded defers: 347 ns/op Updates #14939 (defer performance) Updates #34481 (design doc) Change-Id: I63b1a60d1ebf28126f55ee9fd7ecffe9cb23d1ff Reviewed-on: https://go-review.googlesource.com/c/go/+/202340 Reviewed-by: Austin Clements <austin@google.com>
2019-06-24 12:59:22 -07:00
Curfn.Func.numReturns++
if n.List.Len() == 0 {
break
}
cmd/compile: shrink gc.Type in half Many of Type's fields are etype-specific. This CL organizes them into their own auxiliary types, duplicating a few fields as necessary, and adds an Extra field to hold them. It also sorts the remaining fields for better struct packing. It also improves documentation for most fields. This reduces the size of Type at the cost of some extra allocations. There's no CPU impact; memory impact below. It also makes the natural structure of Type clearer. Passes toolstash -cmp on all architectures. Ideas for future work in this vein: (1) Width and Align probably only need to be stored for Struct and Array types. The refactoring to accomplish this would hopefully also eliminate TFUNCARGS and TCHANARGS entirely. (2) Maplineno is sparsely used and could probably better be stored in a separate map[*Type]int32, with mapqueue updated to store both a Node and a line number. (3) The Printed field may be removable once the old (non-binary) importer/exported has been removed. (4) StructType's fields field could be changed from *[]*Field to []*Field, which would remove a common allocation. (5) I believe that Type.Nod can be moved to ForwardType. Separate CL. name old alloc/op new alloc/op delta Template 57.9MB ± 0% 55.9MB ± 0% -3.43% (p=0.000 n=50+50) Unicode 38.3MB ± 0% 37.8MB ± 0% -1.39% (p=0.000 n=50+50) GoTypes 185MB ± 0% 180MB ± 0% -2.56% (p=0.000 n=50+50) Compiler 824MB ± 0% 806MB ± 0% -2.19% (p=0.000 n=50+50) name old allocs/op new allocs/op delta Template 486k ± 0% 497k ± 0% +2.25% (p=0.000 n=50+50) Unicode 377k ± 0% 379k ± 0% +0.55% (p=0.000 n=50+50) GoTypes 1.39M ± 0% 1.42M ± 0% +1.63% (p=0.000 n=50+50) Compiler 5.52M ± 0% 5.57M ± 0% +0.84% (p=0.000 n=47+50) Change-Id: I828488eeb74902b013d5ae4cf844de0b6c0dfc87 Reviewed-on: https://go-review.googlesource.com/21611 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
2016-04-01 20:11:30 -07:00
if (Curfn.Type.FuncType().Outnamed && n.List.Len() > 1) || paramoutheap(Curfn) {
// assign to the function out parameters,
// so that reorder3 can fix up conflicts
var rl []*Node
for _, ln := range Curfn.Func.Dcl {
cl := ln.Class()
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
if cl == PAUTO || cl == PAUTOHEAP {
break
}
if cl == PPARAMOUT {
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
if ln.isParamStackCopy() {
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
ln = walkexpr(typecheck(nod(ODEREF, ln.Name.Param.Heapaddr, nil), ctxExpr), nil)
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
}
rl = append(rl, ln)
}
}
if got, want := n.List.Len(), len(rl); got != want {
// order should have rewritten multi-value function calls
// with explicit OAS2FUNC nodes.
Fatalf("expected %v return arguments, have %v", want, got)
}
// move function calls out, to make reorder3's job easier.
walkexprlistsafe(n.List.Slice(), &n.Ninit)
ll := ascompatee(n.Op, rl, n.List.Slice(), &n.Ninit)
n.List.Set(reorder3(ll))
break
}
walkexprlist(n.List.Slice(), &n.Ninit)
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
// For each return parameter (lhs), assign the corresponding result (rhs).
lhs := Curfn.Type.Results()
rhs := n.List.Slice()
res := make([]*Node, lhs.NumFields())
for i, nl := range lhs.FieldSlice() {
nname := asNode(nl.Nname)
if nname.isParamHeapCopy() {
nname = nname.Name.Param.Stackcopy
}
a := nod(OAS, nname, rhs[i])
res[i] = convas(a, &n.Ninit)
}
n.List.Set(res)
case ORETJMP:
break
case OINLMARK:
break
case OSELECT:
walkselect(n)
case OSWITCH:
walkswitch(n)
case ORANGE:
n = walkrange(n)
}
if n.Op == ONAME {
Fatalf("walkstmt ended up with name: %+v", n)
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
return n
}
func isSmallMakeSlice(n *Node) bool {
if n.Op != OMAKESLICE {
return false
}
r := n.Right
if r == nil {
r = n.Left
}
t := n.Type
return smallintconst(r) && (t.Elem().Width == 0 || r.Int64() < maxImplicitStackVarSize/t.Elem().Width)
}
// walk the whole tree of the body of an
// expression or simple statement.
// the types expressions are calculated.
// compile-time constants are evaluated.
// complex side effects like statements are appended to init
func walkexprlist(s []*Node, init *Nodes) {
for i := range s {
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
s[i] = walkexpr(s[i], init)
}
}
func walkexprlistsafe(s []*Node, init *Nodes) {
for i, n := range s {
s[i] = safeexpr(n, init)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
s[i] = walkexpr(s[i], init)
}
}
func walkexprlistcheap(s []*Node, init *Nodes) {
for i, n := range s {
s[i] = cheapexpr(n, init)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
s[i] = walkexpr(s[i], init)
}
}
// convFuncName builds the runtime function name for interface conversion.
// It also reports whether the function expects the data by address.
// Not all names are possible. For example, we never generate convE2E or convE2I.
func convFuncName(from, to *types.Type) (fnname string, needsaddr bool) {
tkind := to.Tie()
switch from.Tie() {
case 'I':
cmd/compile: shrink specialized convT2x call sites convT2E16 and other specialized type-to-interface routines accept a type/itab argument and return a complete interface value. However, we know enough in the routine to do without the type. And the caller can construct the interface value using the type. Doing so shrinks the call sites of ten of the specialized convT2x routines. It also lets us unify the empty and non-empty interface routines. Cuts 12k off cmd/go. name old time/op new time/op delta ConvT2ESmall-8 2.96ns ± 2% 2.34ns ± 4% -21.01% (p=0.000 n=175+189) ConvT2EUintptr-8 3.00ns ± 3% 2.34ns ± 4% -22.02% (p=0.000 n=189+187) ConvT2ELarge-8 21.3ns ± 7% 21.5ns ± 5% +1.02% (p=0.000 n=200+197) ConvT2ISmall-8 2.99ns ± 4% 2.33ns ± 3% -21.95% (p=0.000 n=193+184) ConvT2IUintptr-8 3.02ns ± 3% 2.33ns ± 3% -22.82% (p=0.000 n=198+190) ConvT2ILarge-8 21.7ns ± 5% 22.2ns ± 4% +2.31% (p=0.000 n=199+198) ConvT2Ezero/zero/16-8 2.96ns ± 2% 2.33ns ± 3% -21.11% (p=0.000 n=174+187) ConvT2Ezero/zero/32-8 2.96ns ± 1% 2.35ns ± 4% -20.62% (p=0.000 n=163+193) ConvT2Ezero/zero/64-8 2.99ns ± 2% 2.34ns ± 4% -21.78% (p=0.000 n=183+188) ConvT2Ezero/zero/str-8 3.27ns ± 3% 2.54ns ± 3% -22.32% (p=0.000 n=195+192) ConvT2Ezero/zero/slice-8 3.46ns ± 4% 2.81ns ± 3% -18.96% (p=0.000 n=197+164) ConvT2Ezero/zero/big-8 88.4ns ±20% 90.0ns ±20% +1.84% (p=0.000 n=196+198) ConvT2Ezero/nonzero/16-8 12.6ns ± 3% 12.3ns ± 3% -2.34% (p=0.000 n=167+196) ConvT2Ezero/nonzero/32-8 12.3ns ± 4% 11.9ns ± 3% -2.95% (p=0.000 n=187+193) ConvT2Ezero/nonzero/64-8 14.2ns ± 6% 13.8ns ± 5% -2.94% (p=0.000 n=198+199) ConvT2Ezero/nonzero/str-8 27.2ns ± 5% 26.8ns ± 5% -1.33% (p=0.000 n=200+198) ConvT2Ezero/nonzero/slice-8 33.3ns ± 8% 33.1ns ± 6% -0.82% (p=0.000 n=199+200) ConvT2Ezero/nonzero/big-8 88.8ns ±22% 90.2ns ±18% +1.58% (p=0.000 n=200+199) Neligible toolspeed impact. name old alloc/op new alloc/op delta Template 35.4MB ± 0% 35.3MB ± 0% -0.06% (p=0.008 n=5+5) Unicode 29.1MB ± 0% 29.1MB ± 0% ~ (p=0.310 n=5+5) GoTypes 122MB ± 0% 122MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 514MB ± 0% 513MB ± 0% -0.02% (p=0.008 n=5+5) SSA 1.94GB ± 0% 1.94GB ± 0% -0.01% (p=0.008 n=5+5) Flate 24.2MB ± 0% 24.2MB ± 0% ~ (p=0.548 n=5+5) GoParser 28.5MB ± 0% 28.5MB ± 0% -0.05% (p=0.016 n=5+5) Reflect 86.3MB ± 0% 86.2MB ± 0% -0.02% (p=0.008 n=5+5) Tar 34.9MB ± 0% 34.9MB ± 0% ~ (p=0.095 n=5+5) XML 47.1MB ± 0% 47.1MB ± 0% -0.05% (p=0.008 n=5+5) [Geo mean] 81.0MB 81.0MB -0.03% name old allocs/op new allocs/op delta Template 349k ± 0% 349k ± 0% -0.08% (p=0.008 n=5+5) Unicode 340k ± 0% 340k ± 0% ~ (p=0.111 n=5+5) GoTypes 1.28M ± 0% 1.28M ± 0% -0.09% (p=0.008 n=5+5) Compiler 4.92M ± 0% 4.92M ± 0% -0.08% (p=0.008 n=5+5) SSA 15.3M ± 0% 15.3M ± 0% -0.03% (p=0.008 n=5+5) Flate 233k ± 0% 233k ± 0% ~ (p=0.500 n=5+5) GoParser 292k ± 0% 292k ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.05M ± 0% 1.05M ± 0% -0.02% (p=0.008 n=5+5) Tar 344k ± 0% 343k ± 0% -0.06% (p=0.008 n=5+5) XML 430k ± 0% 429k ± 0% -0.08% (p=0.008 n=5+5) [Geo mean] 809k 809k -0.05% name old object-bytes new object-bytes delta Template 507kB ± 0% 507kB ± 0% -0.04% (p=0.008 n=5+5) Unicode 225kB ± 0% 225kB ± 0% ~ (all equal) GoTypes 1.85MB ± 0% 1.85MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 6.75MB ± 0% 6.75MB ± 0% +0.01% (p=0.008 n=5+5) SSA 21.4MB ± 0% 21.4MB ± 0% -0.02% (p=0.008 n=5+5) Flate 328kB ± 0% 328kB ± 0% -0.03% (p=0.008 n=5+5) GoParser 403kB ± 0% 402kB ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.41MB ± 0% 1.41MB ± 0% -0.03% (p=0.008 n=5+5) Tar 457kB ± 0% 457kB ± 0% -0.05% (p=0.008 n=5+5) XML 601kB ± 0% 600kB ± 0% -0.16% (p=0.008 n=5+5) [Geo mean] 1.05MB 1.04MB -0.05% Change-Id: I677a4108c0ecd32617549294036aa84f9214c4fe Reviewed-on: https://go-review.googlesource.com/c/147360 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Martin Möhrmann <moehrmann@google.com>
2018-10-29 13:54:24 -07:00
if tkind == 'I' {
return "convI2I", false
}
case 'T':
cmd/compile: shrink specialized convT2x call sites convT2E16 and other specialized type-to-interface routines accept a type/itab argument and return a complete interface value. However, we know enough in the routine to do without the type. And the caller can construct the interface value using the type. Doing so shrinks the call sites of ten of the specialized convT2x routines. It also lets us unify the empty and non-empty interface routines. Cuts 12k off cmd/go. name old time/op new time/op delta ConvT2ESmall-8 2.96ns ± 2% 2.34ns ± 4% -21.01% (p=0.000 n=175+189) ConvT2EUintptr-8 3.00ns ± 3% 2.34ns ± 4% -22.02% (p=0.000 n=189+187) ConvT2ELarge-8 21.3ns ± 7% 21.5ns ± 5% +1.02% (p=0.000 n=200+197) ConvT2ISmall-8 2.99ns ± 4% 2.33ns ± 3% -21.95% (p=0.000 n=193+184) ConvT2IUintptr-8 3.02ns ± 3% 2.33ns ± 3% -22.82% (p=0.000 n=198+190) ConvT2ILarge-8 21.7ns ± 5% 22.2ns ± 4% +2.31% (p=0.000 n=199+198) ConvT2Ezero/zero/16-8 2.96ns ± 2% 2.33ns ± 3% -21.11% (p=0.000 n=174+187) ConvT2Ezero/zero/32-8 2.96ns ± 1% 2.35ns ± 4% -20.62% (p=0.000 n=163+193) ConvT2Ezero/zero/64-8 2.99ns ± 2% 2.34ns ± 4% -21.78% (p=0.000 n=183+188) ConvT2Ezero/zero/str-8 3.27ns ± 3% 2.54ns ± 3% -22.32% (p=0.000 n=195+192) ConvT2Ezero/zero/slice-8 3.46ns ± 4% 2.81ns ± 3% -18.96% (p=0.000 n=197+164) ConvT2Ezero/zero/big-8 88.4ns ±20% 90.0ns ±20% +1.84% (p=0.000 n=196+198) ConvT2Ezero/nonzero/16-8 12.6ns ± 3% 12.3ns ± 3% -2.34% (p=0.000 n=167+196) ConvT2Ezero/nonzero/32-8 12.3ns ± 4% 11.9ns ± 3% -2.95% (p=0.000 n=187+193) ConvT2Ezero/nonzero/64-8 14.2ns ± 6% 13.8ns ± 5% -2.94% (p=0.000 n=198+199) ConvT2Ezero/nonzero/str-8 27.2ns ± 5% 26.8ns ± 5% -1.33% (p=0.000 n=200+198) ConvT2Ezero/nonzero/slice-8 33.3ns ± 8% 33.1ns ± 6% -0.82% (p=0.000 n=199+200) ConvT2Ezero/nonzero/big-8 88.8ns ±22% 90.2ns ±18% +1.58% (p=0.000 n=200+199) Neligible toolspeed impact. name old alloc/op new alloc/op delta Template 35.4MB ± 0% 35.3MB ± 0% -0.06% (p=0.008 n=5+5) Unicode 29.1MB ± 0% 29.1MB ± 0% ~ (p=0.310 n=5+5) GoTypes 122MB ± 0% 122MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 514MB ± 0% 513MB ± 0% -0.02% (p=0.008 n=5+5) SSA 1.94GB ± 0% 1.94GB ± 0% -0.01% (p=0.008 n=5+5) Flate 24.2MB ± 0% 24.2MB ± 0% ~ (p=0.548 n=5+5) GoParser 28.5MB ± 0% 28.5MB ± 0% -0.05% (p=0.016 n=5+5) Reflect 86.3MB ± 0% 86.2MB ± 0% -0.02% (p=0.008 n=5+5) Tar 34.9MB ± 0% 34.9MB ± 0% ~ (p=0.095 n=5+5) XML 47.1MB ± 0% 47.1MB ± 0% -0.05% (p=0.008 n=5+5) [Geo mean] 81.0MB 81.0MB -0.03% name old allocs/op new allocs/op delta Template 349k ± 0% 349k ± 0% -0.08% (p=0.008 n=5+5) Unicode 340k ± 0% 340k ± 0% ~ (p=0.111 n=5+5) GoTypes 1.28M ± 0% 1.28M ± 0% -0.09% (p=0.008 n=5+5) Compiler 4.92M ± 0% 4.92M ± 0% -0.08% (p=0.008 n=5+5) SSA 15.3M ± 0% 15.3M ± 0% -0.03% (p=0.008 n=5+5) Flate 233k ± 0% 233k ± 0% ~ (p=0.500 n=5+5) GoParser 292k ± 0% 292k ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.05M ± 0% 1.05M ± 0% -0.02% (p=0.008 n=5+5) Tar 344k ± 0% 343k ± 0% -0.06% (p=0.008 n=5+5) XML 430k ± 0% 429k ± 0% -0.08% (p=0.008 n=5+5) [Geo mean] 809k 809k -0.05% name old object-bytes new object-bytes delta Template 507kB ± 0% 507kB ± 0% -0.04% (p=0.008 n=5+5) Unicode 225kB ± 0% 225kB ± 0% ~ (all equal) GoTypes 1.85MB ± 0% 1.85MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 6.75MB ± 0% 6.75MB ± 0% +0.01% (p=0.008 n=5+5) SSA 21.4MB ± 0% 21.4MB ± 0% -0.02% (p=0.008 n=5+5) Flate 328kB ± 0% 328kB ± 0% -0.03% (p=0.008 n=5+5) GoParser 403kB ± 0% 402kB ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.41MB ± 0% 1.41MB ± 0% -0.03% (p=0.008 n=5+5) Tar 457kB ± 0% 457kB ± 0% -0.05% (p=0.008 n=5+5) XML 601kB ± 0% 600kB ± 0% -0.16% (p=0.008 n=5+5) [Geo mean] 1.05MB 1.04MB -0.05% Change-Id: I677a4108c0ecd32617549294036aa84f9214c4fe Reviewed-on: https://go-review.googlesource.com/c/147360 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Martin Möhrmann <moehrmann@google.com>
2018-10-29 13:54:24 -07:00
switch {
case from.Size() == 2 && from.Align == 2:
return "convT16", false
case from.Size() == 4 && from.Align == 4 && !from.HasPointers():
cmd/compile: shrink specialized convT2x call sites convT2E16 and other specialized type-to-interface routines accept a type/itab argument and return a complete interface value. However, we know enough in the routine to do without the type. And the caller can construct the interface value using the type. Doing so shrinks the call sites of ten of the specialized convT2x routines. It also lets us unify the empty and non-empty interface routines. Cuts 12k off cmd/go. name old time/op new time/op delta ConvT2ESmall-8 2.96ns ± 2% 2.34ns ± 4% -21.01% (p=0.000 n=175+189) ConvT2EUintptr-8 3.00ns ± 3% 2.34ns ± 4% -22.02% (p=0.000 n=189+187) ConvT2ELarge-8 21.3ns ± 7% 21.5ns ± 5% +1.02% (p=0.000 n=200+197) ConvT2ISmall-8 2.99ns ± 4% 2.33ns ± 3% -21.95% (p=0.000 n=193+184) ConvT2IUintptr-8 3.02ns ± 3% 2.33ns ± 3% -22.82% (p=0.000 n=198+190) ConvT2ILarge-8 21.7ns ± 5% 22.2ns ± 4% +2.31% (p=0.000 n=199+198) ConvT2Ezero/zero/16-8 2.96ns ± 2% 2.33ns ± 3% -21.11% (p=0.000 n=174+187) ConvT2Ezero/zero/32-8 2.96ns ± 1% 2.35ns ± 4% -20.62% (p=0.000 n=163+193) ConvT2Ezero/zero/64-8 2.99ns ± 2% 2.34ns ± 4% -21.78% (p=0.000 n=183+188) ConvT2Ezero/zero/str-8 3.27ns ± 3% 2.54ns ± 3% -22.32% (p=0.000 n=195+192) ConvT2Ezero/zero/slice-8 3.46ns ± 4% 2.81ns ± 3% -18.96% (p=0.000 n=197+164) ConvT2Ezero/zero/big-8 88.4ns ±20% 90.0ns ±20% +1.84% (p=0.000 n=196+198) ConvT2Ezero/nonzero/16-8 12.6ns ± 3% 12.3ns ± 3% -2.34% (p=0.000 n=167+196) ConvT2Ezero/nonzero/32-8 12.3ns ± 4% 11.9ns ± 3% -2.95% (p=0.000 n=187+193) ConvT2Ezero/nonzero/64-8 14.2ns ± 6% 13.8ns ± 5% -2.94% (p=0.000 n=198+199) ConvT2Ezero/nonzero/str-8 27.2ns ± 5% 26.8ns ± 5% -1.33% (p=0.000 n=200+198) ConvT2Ezero/nonzero/slice-8 33.3ns ± 8% 33.1ns ± 6% -0.82% (p=0.000 n=199+200) ConvT2Ezero/nonzero/big-8 88.8ns ±22% 90.2ns ±18% +1.58% (p=0.000 n=200+199) Neligible toolspeed impact. name old alloc/op new alloc/op delta Template 35.4MB ± 0% 35.3MB ± 0% -0.06% (p=0.008 n=5+5) Unicode 29.1MB ± 0% 29.1MB ± 0% ~ (p=0.310 n=5+5) GoTypes 122MB ± 0% 122MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 514MB ± 0% 513MB ± 0% -0.02% (p=0.008 n=5+5) SSA 1.94GB ± 0% 1.94GB ± 0% -0.01% (p=0.008 n=5+5) Flate 24.2MB ± 0% 24.2MB ± 0% ~ (p=0.548 n=5+5) GoParser 28.5MB ± 0% 28.5MB ± 0% -0.05% (p=0.016 n=5+5) Reflect 86.3MB ± 0% 86.2MB ± 0% -0.02% (p=0.008 n=5+5) Tar 34.9MB ± 0% 34.9MB ± 0% ~ (p=0.095 n=5+5) XML 47.1MB ± 0% 47.1MB ± 0% -0.05% (p=0.008 n=5+5) [Geo mean] 81.0MB 81.0MB -0.03% name old allocs/op new allocs/op delta Template 349k ± 0% 349k ± 0% -0.08% (p=0.008 n=5+5) Unicode 340k ± 0% 340k ± 0% ~ (p=0.111 n=5+5) GoTypes 1.28M ± 0% 1.28M ± 0% -0.09% (p=0.008 n=5+5) Compiler 4.92M ± 0% 4.92M ± 0% -0.08% (p=0.008 n=5+5) SSA 15.3M ± 0% 15.3M ± 0% -0.03% (p=0.008 n=5+5) Flate 233k ± 0% 233k ± 0% ~ (p=0.500 n=5+5) GoParser 292k ± 0% 292k ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.05M ± 0% 1.05M ± 0% -0.02% (p=0.008 n=5+5) Tar 344k ± 0% 343k ± 0% -0.06% (p=0.008 n=5+5) XML 430k ± 0% 429k ± 0% -0.08% (p=0.008 n=5+5) [Geo mean] 809k 809k -0.05% name old object-bytes new object-bytes delta Template 507kB ± 0% 507kB ± 0% -0.04% (p=0.008 n=5+5) Unicode 225kB ± 0% 225kB ± 0% ~ (all equal) GoTypes 1.85MB ± 0% 1.85MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 6.75MB ± 0% 6.75MB ± 0% +0.01% (p=0.008 n=5+5) SSA 21.4MB ± 0% 21.4MB ± 0% -0.02% (p=0.008 n=5+5) Flate 328kB ± 0% 328kB ± 0% -0.03% (p=0.008 n=5+5) GoParser 403kB ± 0% 402kB ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.41MB ± 0% 1.41MB ± 0% -0.03% (p=0.008 n=5+5) Tar 457kB ± 0% 457kB ± 0% -0.05% (p=0.008 n=5+5) XML 601kB ± 0% 600kB ± 0% -0.16% (p=0.008 n=5+5) [Geo mean] 1.05MB 1.04MB -0.05% Change-Id: I677a4108c0ecd32617549294036aa84f9214c4fe Reviewed-on: https://go-review.googlesource.com/c/147360 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Martin Möhrmann <moehrmann@google.com>
2018-10-29 13:54:24 -07:00
return "convT32", false
case from.Size() == 8 && from.Align == types.Types[TUINT64].Align && !from.HasPointers():
cmd/compile: shrink specialized convT2x call sites convT2E16 and other specialized type-to-interface routines accept a type/itab argument and return a complete interface value. However, we know enough in the routine to do without the type. And the caller can construct the interface value using the type. Doing so shrinks the call sites of ten of the specialized convT2x routines. It also lets us unify the empty and non-empty interface routines. Cuts 12k off cmd/go. name old time/op new time/op delta ConvT2ESmall-8 2.96ns ± 2% 2.34ns ± 4% -21.01% (p=0.000 n=175+189) ConvT2EUintptr-8 3.00ns ± 3% 2.34ns ± 4% -22.02% (p=0.000 n=189+187) ConvT2ELarge-8 21.3ns ± 7% 21.5ns ± 5% +1.02% (p=0.000 n=200+197) ConvT2ISmall-8 2.99ns ± 4% 2.33ns ± 3% -21.95% (p=0.000 n=193+184) ConvT2IUintptr-8 3.02ns ± 3% 2.33ns ± 3% -22.82% (p=0.000 n=198+190) ConvT2ILarge-8 21.7ns ± 5% 22.2ns ± 4% +2.31% (p=0.000 n=199+198) ConvT2Ezero/zero/16-8 2.96ns ± 2% 2.33ns ± 3% -21.11% (p=0.000 n=174+187) ConvT2Ezero/zero/32-8 2.96ns ± 1% 2.35ns ± 4% -20.62% (p=0.000 n=163+193) ConvT2Ezero/zero/64-8 2.99ns ± 2% 2.34ns ± 4% -21.78% (p=0.000 n=183+188) ConvT2Ezero/zero/str-8 3.27ns ± 3% 2.54ns ± 3% -22.32% (p=0.000 n=195+192) ConvT2Ezero/zero/slice-8 3.46ns ± 4% 2.81ns ± 3% -18.96% (p=0.000 n=197+164) ConvT2Ezero/zero/big-8 88.4ns ±20% 90.0ns ±20% +1.84% (p=0.000 n=196+198) ConvT2Ezero/nonzero/16-8 12.6ns ± 3% 12.3ns ± 3% -2.34% (p=0.000 n=167+196) ConvT2Ezero/nonzero/32-8 12.3ns ± 4% 11.9ns ± 3% -2.95% (p=0.000 n=187+193) ConvT2Ezero/nonzero/64-8 14.2ns ± 6% 13.8ns ± 5% -2.94% (p=0.000 n=198+199) ConvT2Ezero/nonzero/str-8 27.2ns ± 5% 26.8ns ± 5% -1.33% (p=0.000 n=200+198) ConvT2Ezero/nonzero/slice-8 33.3ns ± 8% 33.1ns ± 6% -0.82% (p=0.000 n=199+200) ConvT2Ezero/nonzero/big-8 88.8ns ±22% 90.2ns ±18% +1.58% (p=0.000 n=200+199) Neligible toolspeed impact. name old alloc/op new alloc/op delta Template 35.4MB ± 0% 35.3MB ± 0% -0.06% (p=0.008 n=5+5) Unicode 29.1MB ± 0% 29.1MB ± 0% ~ (p=0.310 n=5+5) GoTypes 122MB ± 0% 122MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 514MB ± 0% 513MB ± 0% -0.02% (p=0.008 n=5+5) SSA 1.94GB ± 0% 1.94GB ± 0% -0.01% (p=0.008 n=5+5) Flate 24.2MB ± 0% 24.2MB ± 0% ~ (p=0.548 n=5+5) GoParser 28.5MB ± 0% 28.5MB ± 0% -0.05% (p=0.016 n=5+5) Reflect 86.3MB ± 0% 86.2MB ± 0% -0.02% (p=0.008 n=5+5) Tar 34.9MB ± 0% 34.9MB ± 0% ~ (p=0.095 n=5+5) XML 47.1MB ± 0% 47.1MB ± 0% -0.05% (p=0.008 n=5+5) [Geo mean] 81.0MB 81.0MB -0.03% name old allocs/op new allocs/op delta Template 349k ± 0% 349k ± 0% -0.08% (p=0.008 n=5+5) Unicode 340k ± 0% 340k ± 0% ~ (p=0.111 n=5+5) GoTypes 1.28M ± 0% 1.28M ± 0% -0.09% (p=0.008 n=5+5) Compiler 4.92M ± 0% 4.92M ± 0% -0.08% (p=0.008 n=5+5) SSA 15.3M ± 0% 15.3M ± 0% -0.03% (p=0.008 n=5+5) Flate 233k ± 0% 233k ± 0% ~ (p=0.500 n=5+5) GoParser 292k ± 0% 292k ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.05M ± 0% 1.05M ± 0% -0.02% (p=0.008 n=5+5) Tar 344k ± 0% 343k ± 0% -0.06% (p=0.008 n=5+5) XML 430k ± 0% 429k ± 0% -0.08% (p=0.008 n=5+5) [Geo mean] 809k 809k -0.05% name old object-bytes new object-bytes delta Template 507kB ± 0% 507kB ± 0% -0.04% (p=0.008 n=5+5) Unicode 225kB ± 0% 225kB ± 0% ~ (all equal) GoTypes 1.85MB ± 0% 1.85MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 6.75MB ± 0% 6.75MB ± 0% +0.01% (p=0.008 n=5+5) SSA 21.4MB ± 0% 21.4MB ± 0% -0.02% (p=0.008 n=5+5) Flate 328kB ± 0% 328kB ± 0% -0.03% (p=0.008 n=5+5) GoParser 403kB ± 0% 402kB ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.41MB ± 0% 1.41MB ± 0% -0.03% (p=0.008 n=5+5) Tar 457kB ± 0% 457kB ± 0% -0.05% (p=0.008 n=5+5) XML 601kB ± 0% 600kB ± 0% -0.16% (p=0.008 n=5+5) [Geo mean] 1.05MB 1.04MB -0.05% Change-Id: I677a4108c0ecd32617549294036aa84f9214c4fe Reviewed-on: https://go-review.googlesource.com/c/147360 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Martin Möhrmann <moehrmann@google.com>
2018-10-29 13:54:24 -07:00
return "convT64", false
}
if sc := from.SoleComponent(); sc != nil {
switch {
case sc.IsString():
return "convTstring", false
case sc.IsSlice():
return "convTslice", false
}
cmd/compile: shrink specialized convT2x call sites convT2E16 and other specialized type-to-interface routines accept a type/itab argument and return a complete interface value. However, we know enough in the routine to do without the type. And the caller can construct the interface value using the type. Doing so shrinks the call sites of ten of the specialized convT2x routines. It also lets us unify the empty and non-empty interface routines. Cuts 12k off cmd/go. name old time/op new time/op delta ConvT2ESmall-8 2.96ns ± 2% 2.34ns ± 4% -21.01% (p=0.000 n=175+189) ConvT2EUintptr-8 3.00ns ± 3% 2.34ns ± 4% -22.02% (p=0.000 n=189+187) ConvT2ELarge-8 21.3ns ± 7% 21.5ns ± 5% +1.02% (p=0.000 n=200+197) ConvT2ISmall-8 2.99ns ± 4% 2.33ns ± 3% -21.95% (p=0.000 n=193+184) ConvT2IUintptr-8 3.02ns ± 3% 2.33ns ± 3% -22.82% (p=0.000 n=198+190) ConvT2ILarge-8 21.7ns ± 5% 22.2ns ± 4% +2.31% (p=0.000 n=199+198) ConvT2Ezero/zero/16-8 2.96ns ± 2% 2.33ns ± 3% -21.11% (p=0.000 n=174+187) ConvT2Ezero/zero/32-8 2.96ns ± 1% 2.35ns ± 4% -20.62% (p=0.000 n=163+193) ConvT2Ezero/zero/64-8 2.99ns ± 2% 2.34ns ± 4% -21.78% (p=0.000 n=183+188) ConvT2Ezero/zero/str-8 3.27ns ± 3% 2.54ns ± 3% -22.32% (p=0.000 n=195+192) ConvT2Ezero/zero/slice-8 3.46ns ± 4% 2.81ns ± 3% -18.96% (p=0.000 n=197+164) ConvT2Ezero/zero/big-8 88.4ns ±20% 90.0ns ±20% +1.84% (p=0.000 n=196+198) ConvT2Ezero/nonzero/16-8 12.6ns ± 3% 12.3ns ± 3% -2.34% (p=0.000 n=167+196) ConvT2Ezero/nonzero/32-8 12.3ns ± 4% 11.9ns ± 3% -2.95% (p=0.000 n=187+193) ConvT2Ezero/nonzero/64-8 14.2ns ± 6% 13.8ns ± 5% -2.94% (p=0.000 n=198+199) ConvT2Ezero/nonzero/str-8 27.2ns ± 5% 26.8ns ± 5% -1.33% (p=0.000 n=200+198) ConvT2Ezero/nonzero/slice-8 33.3ns ± 8% 33.1ns ± 6% -0.82% (p=0.000 n=199+200) ConvT2Ezero/nonzero/big-8 88.8ns ±22% 90.2ns ±18% +1.58% (p=0.000 n=200+199) Neligible toolspeed impact. name old alloc/op new alloc/op delta Template 35.4MB ± 0% 35.3MB ± 0% -0.06% (p=0.008 n=5+5) Unicode 29.1MB ± 0% 29.1MB ± 0% ~ (p=0.310 n=5+5) GoTypes 122MB ± 0% 122MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 514MB ± 0% 513MB ± 0% -0.02% (p=0.008 n=5+5) SSA 1.94GB ± 0% 1.94GB ± 0% -0.01% (p=0.008 n=5+5) Flate 24.2MB ± 0% 24.2MB ± 0% ~ (p=0.548 n=5+5) GoParser 28.5MB ± 0% 28.5MB ± 0% -0.05% (p=0.016 n=5+5) Reflect 86.3MB ± 0% 86.2MB ± 0% -0.02% (p=0.008 n=5+5) Tar 34.9MB ± 0% 34.9MB ± 0% ~ (p=0.095 n=5+5) XML 47.1MB ± 0% 47.1MB ± 0% -0.05% (p=0.008 n=5+5) [Geo mean] 81.0MB 81.0MB -0.03% name old allocs/op new allocs/op delta Template 349k ± 0% 349k ± 0% -0.08% (p=0.008 n=5+5) Unicode 340k ± 0% 340k ± 0% ~ (p=0.111 n=5+5) GoTypes 1.28M ± 0% 1.28M ± 0% -0.09% (p=0.008 n=5+5) Compiler 4.92M ± 0% 4.92M ± 0% -0.08% (p=0.008 n=5+5) SSA 15.3M ± 0% 15.3M ± 0% -0.03% (p=0.008 n=5+5) Flate 233k ± 0% 233k ± 0% ~ (p=0.500 n=5+5) GoParser 292k ± 0% 292k ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.05M ± 0% 1.05M ± 0% -0.02% (p=0.008 n=5+5) Tar 344k ± 0% 343k ± 0% -0.06% (p=0.008 n=5+5) XML 430k ± 0% 429k ± 0% -0.08% (p=0.008 n=5+5) [Geo mean] 809k 809k -0.05% name old object-bytes new object-bytes delta Template 507kB ± 0% 507kB ± 0% -0.04% (p=0.008 n=5+5) Unicode 225kB ± 0% 225kB ± 0% ~ (all equal) GoTypes 1.85MB ± 0% 1.85MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 6.75MB ± 0% 6.75MB ± 0% +0.01% (p=0.008 n=5+5) SSA 21.4MB ± 0% 21.4MB ± 0% -0.02% (p=0.008 n=5+5) Flate 328kB ± 0% 328kB ± 0% -0.03% (p=0.008 n=5+5) GoParser 403kB ± 0% 402kB ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.41MB ± 0% 1.41MB ± 0% -0.03% (p=0.008 n=5+5) Tar 457kB ± 0% 457kB ± 0% -0.05% (p=0.008 n=5+5) XML 601kB ± 0% 600kB ± 0% -0.16% (p=0.008 n=5+5) [Geo mean] 1.05MB 1.04MB -0.05% Change-Id: I677a4108c0ecd32617549294036aa84f9214c4fe Reviewed-on: https://go-review.googlesource.com/c/147360 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Martin Möhrmann <moehrmann@google.com>
2018-10-29 13:54:24 -07:00
}
switch tkind {
case 'E':
if !from.HasPointers() {
return "convT2Enoptr", true
cmd/compile, runtime: specialize convT2x, don't alloc for zero vals Prior to this CL, all runtime conversions from a concrete value to an interface went through one of two runtime calls: convT2E or convT2I. However, in practice, basic types are very common. Specializing convT2x for those basic types allows for a more efficient implementation for those types. For basic scalars and strings, allocation and copying can use the same methods as normal code. For pointer-free types, allocation can occur without zeroing, and copying can take place without GC calls. For slices, copying is cheaper and simpler. This CL adds twelve runtime routines: convT2E16, convT2I16 convT2E32, convT2I32 convT2E64, convT2I64 convT2Estring, convT2Istring convT2Eslice, convT2Islice convT2Enoptr, convT2Inoptr While compiling make.bash, 93% of all convT2x calls are now to one of these specialized convT2x call. Within specialized convT2x routines, it is cheap to check for a zero value, in a way that it is not in general. When we detect a zero value there, we return a pointer to zeroVal, rather than allocating. name old time/op new time/op delta ConvT2Ezero/zero/16-8 17.9ns ± 2% 3.0ns ± 3% -83.20% (p=0.000 n=56+56) ConvT2Ezero/zero/32-8 17.8ns ± 2% 3.0ns ± 3% -83.15% (p=0.000 n=59+60) ConvT2Ezero/zero/64-8 20.1ns ± 1% 3.0ns ± 2% -84.98% (p=0.000 n=57+57) ConvT2Ezero/zero/str-8 32.6ns ± 1% 3.0ns ± 4% -90.70% (p=0.000 n=59+60) ConvT2Ezero/zero/slice-8 36.7ns ± 2% 3.0ns ± 2% -91.78% (p=0.000 n=59+59) ConvT2Ezero/zero/big-8 91.9ns ± 2% 85.9ns ± 2% -6.52% (p=0.000 n=57+57) ConvT2Ezero/nonzero/16-8 17.7ns ± 2% 12.7ns ± 3% -28.38% (p=0.000 n=55+60) ConvT2Ezero/nonzero/32-8 17.8ns ± 1% 12.7ns ± 1% -28.44% (p=0.000 n=54+57) ConvT2Ezero/nonzero/64-8 20.0ns ± 1% 15.0ns ± 1% -24.90% (p=0.000 n=56+58) ConvT2Ezero/nonzero/str-8 32.6ns ± 1% 25.7ns ± 1% -21.17% (p=0.000 n=58+55) ConvT2Ezero/nonzero/slice-8 36.8ns ± 2% 30.4ns ± 1% -17.32% (p=0.000 n=60+52) ConvT2Ezero/nonzero/big-8 92.1ns ± 2% 85.9ns ± 2% -6.70% (p=0.000 n=57+59) Benchmarks on a real program (the compiler): name old time/op new time/op delta Template 227ms ± 5% 221ms ± 2% -2.48% (p=0.000 n=30+26) Unicode 102ms ± 5% 100ms ± 3% -1.30% (p=0.009 n=30+26) GoTypes 656ms ± 5% 659ms ± 4% ~ (p=0.208 n=30+30) Compiler 2.82s ± 2% 2.82s ± 1% ~ (p=0.614 n=29+27) Flate 128ms ± 2% 128ms ± 5% ~ (p=0.783 n=27+28) GoParser 158ms ± 3% 158ms ± 3% ~ (p=0.261 n=28+30) Reflect 408ms ± 7% 401ms ± 3% ~ (p=0.075 n=30+30) Tar 123ms ± 6% 121ms ± 8% ~ (p=0.287 n=29+30) XML 220ms ± 2% 220ms ± 4% ~ (p=0.805 n=29+29) name old user-ns/op new user-ns/op delta Template 281user-ms ± 4% 279user-ms ± 3% -0.87% (p=0.044 n=28+28) Unicode 142user-ms ± 4% 141user-ms ± 3% -1.04% (p=0.015 n=30+27) GoTypes 884user-ms ± 3% 886user-ms ± 2% ~ (p=0.532 n=30+30) Compiler 3.94user-s ± 3% 3.92user-s ± 1% ~ (p=0.185 n=30+28) Flate 165user-ms ± 2% 165user-ms ± 4% ~ (p=0.780 n=27+29) GoParser 209user-ms ± 2% 208user-ms ± 3% ~ (p=0.453 n=28+30) Reflect 533user-ms ± 6% 526user-ms ± 3% ~ (p=0.057 n=30+30) Tar 156user-ms ± 6% 154user-ms ± 6% ~ (p=0.133 n=29+30) XML 288user-ms ± 4% 288user-ms ± 4% ~ (p=0.633 n=30+30) name old alloc/op new alloc/op delta Template 41.0MB ± 0% 40.9MB ± 0% -0.11% (p=0.000 n=29+29) Unicode 32.6MB ± 0% 32.6MB ± 0% ~ (p=0.572 n=29+30) GoTypes 122MB ± 0% 122MB ± 0% -0.10% (p=0.000 n=30+30) Compiler 482MB ± 0% 481MB ± 0% -0.07% (p=0.000 n=30+29) Flate 26.6MB ± 0% 26.6MB ± 0% ~ (p=0.096 n=30+30) GoParser 32.7MB ± 0% 32.6MB ± 0% -0.06% (p=0.011 n=28+28) Reflect 84.2MB ± 0% 84.1MB ± 0% -0.17% (p=0.000 n=29+30) Tar 27.7MB ± 0% 27.7MB ± 0% -0.05% (p=0.032 n=27+28) XML 44.7MB ± 0% 44.7MB ± 0% ~ (p=0.131 n=28+30) name old allocs/op new allocs/op delta Template 373k ± 1% 370k ± 1% -0.76% (p=0.000 n=30+30) Unicode 325k ± 1% 325k ± 1% ~ (p=0.383 n=29+30) GoTypes 1.16M ± 0% 1.15M ± 0% -0.75% (p=0.000 n=29+30) Compiler 4.15M ± 0% 4.13M ± 0% -0.59% (p=0.000 n=30+29) Flate 238k ± 1% 237k ± 1% -0.62% (p=0.000 n=30+30) GoParser 304k ± 1% 302k ± 1% -0.64% (p=0.000 n=30+28) Reflect 1.00M ± 0% 0.99M ± 0% -1.10% (p=0.000 n=29+30) Tar 245k ± 1% 244k ± 1% -0.59% (p=0.000 n=27+29) XML 391k ± 1% 389k ± 1% -0.59% (p=0.000 n=29+30) Change-Id: Id7f456d690567c2b0a96b0d6d64de8784b6e305f Reviewed-on: https://go-review.googlesource.com/36476 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2017-02-03 22:40:56 -08:00
}
return "convT2E", true
case 'I':
if !from.HasPointers() {
return "convT2Inoptr", true
cmd/compile, runtime: specialize convT2x, don't alloc for zero vals Prior to this CL, all runtime conversions from a concrete value to an interface went through one of two runtime calls: convT2E or convT2I. However, in practice, basic types are very common. Specializing convT2x for those basic types allows for a more efficient implementation for those types. For basic scalars and strings, allocation and copying can use the same methods as normal code. For pointer-free types, allocation can occur without zeroing, and copying can take place without GC calls. For slices, copying is cheaper and simpler. This CL adds twelve runtime routines: convT2E16, convT2I16 convT2E32, convT2I32 convT2E64, convT2I64 convT2Estring, convT2Istring convT2Eslice, convT2Islice convT2Enoptr, convT2Inoptr While compiling make.bash, 93% of all convT2x calls are now to one of these specialized convT2x call. Within specialized convT2x routines, it is cheap to check for a zero value, in a way that it is not in general. When we detect a zero value there, we return a pointer to zeroVal, rather than allocating. name old time/op new time/op delta ConvT2Ezero/zero/16-8 17.9ns ± 2% 3.0ns ± 3% -83.20% (p=0.000 n=56+56) ConvT2Ezero/zero/32-8 17.8ns ± 2% 3.0ns ± 3% -83.15% (p=0.000 n=59+60) ConvT2Ezero/zero/64-8 20.1ns ± 1% 3.0ns ± 2% -84.98% (p=0.000 n=57+57) ConvT2Ezero/zero/str-8 32.6ns ± 1% 3.0ns ± 4% -90.70% (p=0.000 n=59+60) ConvT2Ezero/zero/slice-8 36.7ns ± 2% 3.0ns ± 2% -91.78% (p=0.000 n=59+59) ConvT2Ezero/zero/big-8 91.9ns ± 2% 85.9ns ± 2% -6.52% (p=0.000 n=57+57) ConvT2Ezero/nonzero/16-8 17.7ns ± 2% 12.7ns ± 3% -28.38% (p=0.000 n=55+60) ConvT2Ezero/nonzero/32-8 17.8ns ± 1% 12.7ns ± 1% -28.44% (p=0.000 n=54+57) ConvT2Ezero/nonzero/64-8 20.0ns ± 1% 15.0ns ± 1% -24.90% (p=0.000 n=56+58) ConvT2Ezero/nonzero/str-8 32.6ns ± 1% 25.7ns ± 1% -21.17% (p=0.000 n=58+55) ConvT2Ezero/nonzero/slice-8 36.8ns ± 2% 30.4ns ± 1% -17.32% (p=0.000 n=60+52) ConvT2Ezero/nonzero/big-8 92.1ns ± 2% 85.9ns ± 2% -6.70% (p=0.000 n=57+59) Benchmarks on a real program (the compiler): name old time/op new time/op delta Template 227ms ± 5% 221ms ± 2% -2.48% (p=0.000 n=30+26) Unicode 102ms ± 5% 100ms ± 3% -1.30% (p=0.009 n=30+26) GoTypes 656ms ± 5% 659ms ± 4% ~ (p=0.208 n=30+30) Compiler 2.82s ± 2% 2.82s ± 1% ~ (p=0.614 n=29+27) Flate 128ms ± 2% 128ms ± 5% ~ (p=0.783 n=27+28) GoParser 158ms ± 3% 158ms ± 3% ~ (p=0.261 n=28+30) Reflect 408ms ± 7% 401ms ± 3% ~ (p=0.075 n=30+30) Tar 123ms ± 6% 121ms ± 8% ~ (p=0.287 n=29+30) XML 220ms ± 2% 220ms ± 4% ~ (p=0.805 n=29+29) name old user-ns/op new user-ns/op delta Template 281user-ms ± 4% 279user-ms ± 3% -0.87% (p=0.044 n=28+28) Unicode 142user-ms ± 4% 141user-ms ± 3% -1.04% (p=0.015 n=30+27) GoTypes 884user-ms ± 3% 886user-ms ± 2% ~ (p=0.532 n=30+30) Compiler 3.94user-s ± 3% 3.92user-s ± 1% ~ (p=0.185 n=30+28) Flate 165user-ms ± 2% 165user-ms ± 4% ~ (p=0.780 n=27+29) GoParser 209user-ms ± 2% 208user-ms ± 3% ~ (p=0.453 n=28+30) Reflect 533user-ms ± 6% 526user-ms ± 3% ~ (p=0.057 n=30+30) Tar 156user-ms ± 6% 154user-ms ± 6% ~ (p=0.133 n=29+30) XML 288user-ms ± 4% 288user-ms ± 4% ~ (p=0.633 n=30+30) name old alloc/op new alloc/op delta Template 41.0MB ± 0% 40.9MB ± 0% -0.11% (p=0.000 n=29+29) Unicode 32.6MB ± 0% 32.6MB ± 0% ~ (p=0.572 n=29+30) GoTypes 122MB ± 0% 122MB ± 0% -0.10% (p=0.000 n=30+30) Compiler 482MB ± 0% 481MB ± 0% -0.07% (p=0.000 n=30+29) Flate 26.6MB ± 0% 26.6MB ± 0% ~ (p=0.096 n=30+30) GoParser 32.7MB ± 0% 32.6MB ± 0% -0.06% (p=0.011 n=28+28) Reflect 84.2MB ± 0% 84.1MB ± 0% -0.17% (p=0.000 n=29+30) Tar 27.7MB ± 0% 27.7MB ± 0% -0.05% (p=0.032 n=27+28) XML 44.7MB ± 0% 44.7MB ± 0% ~ (p=0.131 n=28+30) name old allocs/op new allocs/op delta Template 373k ± 1% 370k ± 1% -0.76% (p=0.000 n=30+30) Unicode 325k ± 1% 325k ± 1% ~ (p=0.383 n=29+30) GoTypes 1.16M ± 0% 1.15M ± 0% -0.75% (p=0.000 n=29+30) Compiler 4.15M ± 0% 4.13M ± 0% -0.59% (p=0.000 n=30+29) Flate 238k ± 1% 237k ± 1% -0.62% (p=0.000 n=30+30) GoParser 304k ± 1% 302k ± 1% -0.64% (p=0.000 n=30+28) Reflect 1.00M ± 0% 0.99M ± 0% -1.10% (p=0.000 n=29+30) Tar 245k ± 1% 244k ± 1% -0.59% (p=0.000 n=27+29) XML 391k ± 1% 389k ± 1% -0.59% (p=0.000 n=29+30) Change-Id: Id7f456d690567c2b0a96b0d6d64de8784b6e305f Reviewed-on: https://go-review.googlesource.com/36476 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2017-02-03 22:40:56 -08:00
}
return "convT2I", true
}
}
Fatalf("unknown conv func %c2%c", from.Tie(), to.Tie())
panic("unreachable")
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
// The result of walkexpr MUST be assigned back to n, e.g.
// n.Left = walkexpr(n.Left, init)
func walkexpr(n *Node, init *Nodes) *Node {
if n == nil {
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
return n
}
// Eagerly checkwidth all expressions for the back end.
if n.Type != nil && !n.Type.WidthCalculated() {
switch n.Type.Etype {
case TBLANK, TNIL, TIDEAL:
default:
checkwidth(n.Type)
}
}
if init == &n.Ninit {
// not okay to use n->ninit when walking n,
// because we might replace n with some other node
// and would lose the init list.
Fatalf("walkexpr init == &n->ninit")
}
if n.Ninit.Len() != 0 {
walkstmtlist(n.Ninit.Slice())
init.AppendNodes(&n.Ninit)
}
lno := setlineno(n)
if Debug['w'] > 1 {
Dump("before walk expr", n)
}
if n.Typecheck() != 1 {
Fatalf("missed typecheck: %+v", n)
}
if n.Type.IsUntyped() {
Fatalf("expression has untyped type: %+v", n)
}
cmd/compile: move Node.Class to flags Put it at position zero, since it is fairly hot. This shrinks gc.Node into a smaller size class on 64 bit systems. name old time/op new time/op delta Template 193ms ± 5% 192ms ± 3% ~ (p=0.353 n=94+93) Unicode 86.1ms ± 5% 85.0ms ± 4% -1.23% (p=0.000 n=95+98) GoTypes 546ms ± 3% 544ms ± 4% -0.40% (p=0.007 n=94+97) Compiler 2.56s ± 3% 2.54s ± 3% -0.67% (p=0.000 n=99+97) SSA 5.13s ± 2% 5.10s ± 3% -0.55% (p=0.000 n=94+98) Flate 122ms ± 6% 121ms ± 4% -0.75% (p=0.002 n=97+95) GoParser 144ms ± 5% 144ms ± 4% ~ (p=0.298 n=98+97) Reflect 348ms ± 4% 349ms ± 4% ~ (p=0.350 n=98+97) Tar 105ms ± 5% 104ms ± 5% ~ (p=0.154 n=96+98) XML 200ms ± 5% 198ms ± 4% -0.71% (p=0.015 n=97+98) [Geo mean] 330ms 328ms -0.52% name old user-time/op new user-time/op delta Template 229ms ±11% 224ms ± 7% -2.16% (p=0.001 n=100+87) Unicode 109ms ± 5% 109ms ± 6% ~ (p=0.897 n=96+91) GoTypes 712ms ± 4% 709ms ± 4% ~ (p=0.085 n=96+98) Compiler 3.41s ± 3% 3.36s ± 3% -1.43% (p=0.000 n=98+98) SSA 7.46s ± 3% 7.31s ± 3% -2.02% (p=0.000 n=100+99) Flate 145ms ± 6% 143ms ± 6% -1.11% (p=0.001 n=99+97) GoParser 177ms ± 5% 176ms ± 5% -0.78% (p=0.018 n=95+95) Reflect 432ms ± 7% 435ms ± 9% ~ (p=0.296 n=100+100) Tar 121ms ± 7% 121ms ± 5% ~ (p=0.072 n=100+95) XML 241ms ± 4% 239ms ± 5% ~ (p=0.085 n=97+99) [Geo mean] 413ms 410ms -0.73% name old alloc/op new alloc/op delta Template 38.4MB ± 0% 37.7MB ± 0% -1.85% (p=0.008 n=5+5) Unicode 30.1MB ± 0% 28.8MB ± 0% -4.09% (p=0.008 n=5+5) GoTypes 112MB ± 0% 110MB ± 0% -1.69% (p=0.008 n=5+5) Compiler 470MB ± 0% 461MB ± 0% -1.91% (p=0.008 n=5+5) SSA 1.13GB ± 0% 1.11GB ± 0% -1.70% (p=0.008 n=5+5) Flate 25.0MB ± 0% 24.6MB ± 0% -1.67% (p=0.008 n=5+5) GoParser 31.6MB ± 0% 31.1MB ± 0% -1.66% (p=0.008 n=5+5) Reflect 77.1MB ± 0% 75.8MB ± 0% -1.69% (p=0.008 n=5+5) Tar 26.3MB ± 0% 25.7MB ± 0% -2.06% (p=0.008 n=5+5) XML 41.9MB ± 0% 41.1MB ± 0% -1.93% (p=0.008 n=5+5) [Geo mean] 73.5MB 72.0MB -2.03% name old allocs/op new allocs/op delta Template 383k ± 0% 383k ± 0% ~ (p=0.690 n=5+5) Unicode 343k ± 0% 343k ± 0% ~ (p=0.841 n=5+5) GoTypes 1.16M ± 0% 1.16M ± 0% ~ (p=0.310 n=5+5) Compiler 4.43M ± 0% 4.42M ± 0% -0.17% (p=0.008 n=5+5) SSA 9.85M ± 0% 9.85M ± 0% ~ (p=0.310 n=5+5) Flate 236k ± 0% 236k ± 1% ~ (p=0.841 n=5+5) GoParser 320k ± 0% 320k ± 0% ~ (p=0.421 n=5+5) Reflect 988k ± 0% 987k ± 0% ~ (p=0.690 n=5+5) Tar 252k ± 0% 251k ± 0% ~ (p=0.095 n=5+5) XML 399k ± 0% 399k ± 0% ~ (p=1.000 n=5+5) [Geo mean] 741k 740k -0.07% Change-Id: I9e952b58a98e30a12494304db9ce50d0a85e459c Reviewed-on: https://go-review.googlesource.com/41797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com>
2017-04-25 18:14:12 -07:00
if n.Op == ONAME && n.Class() == PAUTOHEAP {
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
nn := nod(ODEREF, n.Name.Param.Heapaddr, nil)
nn = typecheck(nn, ctxExpr)
nn = walkexpr(nn, init)
nn.Left.MarkNonNil()
return nn
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
}
opswitch:
switch n.Op {
default:
Dump("walk", n)
Fatalf("walkexpr: switch 1 unknown op %+S", n)
case ONONAME, OEMPTY, OGETG, ONEWOBJ:
case OTYPE, ONAME, OLITERAL:
// TODO(mdempsky): Just return n; see discussion on CL 38655.
// Perhaps refactor to use Node.mayBeShared for these instead.
// If these return early, make sure to still call
// stringsym for constant strings.
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
case ONOT, ONEG, OPLUS, OBITNOT, OREAL, OIMAG, ODOTMETH, ODOTINTER,
ODEREF, OSPTR, OITAB, OIDATA, OADDR:
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
case OEFACE, OAND, OSUB, OMUL, OADD, OOR, OXOR, OLSH, ORSH:
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
n.Right = walkexpr(n.Right, init)
case ODOT, ODOTPTR:
usefield(n)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
case ODOTTYPE, ODOTTYPE2:
n.Left = walkexpr(n.Left, init)
// Set up interface type addresses for back end.
n.Right = typename(n.Type)
if n.Op == ODOTTYPE {
n.Right.Right = typename(n.Left.Type)
}
cmd/compile: evaluate itabname during walk instead of SSA For backend concurrency safety. Follow-up to CL 38721. This does introduce a Nodes where there wasn't one before, but these are so rare that the performance impact is negligible. Does not pass toolstash-check, but the only change is line numbers, and the new line numbers appear preferable. Updates #15756 name old alloc/op new alloc/op delta Template 39.9MB ± 0% 39.9MB ± 0% ~ (p=0.841 n=5+5) Unicode 29.8MB ± 0% 29.8MB ± 0% ~ (p=0.690 n=5+5) GoTypes 113MB ± 0% 113MB ± 0% +0.09% (p=0.008 n=5+5) SSA 854MB ± 0% 855MB ± 0% ~ (p=0.222 n=5+5) Flate 25.3MB ± 0% 25.3MB ± 0% ~ (p=0.690 n=5+5) GoParser 31.8MB ± 0% 31.9MB ± 0% ~ (p=0.421 n=5+5) Reflect 78.2MB ± 0% 78.3MB ± 0% ~ (p=0.548 n=5+5) Tar 26.7MB ± 0% 26.7MB ± 0% ~ (p=0.690 n=5+5) XML 42.3MB ± 0% 42.3MB ± 0% ~ (p=0.222 n=5+5) name old allocs/op new allocs/op delta Template 391k ± 1% 391k ± 0% ~ (p=0.841 n=5+5) Unicode 320k ± 0% 320k ± 0% ~ (p=0.841 n=5+5) GoTypes 1.14M ± 0% 1.14M ± 0% +0.26% (p=0.008 n=5+5) SSA 7.60M ± 0% 7.60M ± 0% ~ (p=0.548 n=5+5) Flate 234k ± 0% 234k ± 1% ~ (p=1.000 n=5+5) GoParser 316k ± 1% 317k ± 0% ~ (p=0.841 n=5+5) Reflect 979k ± 0% 980k ± 0% ~ (p=0.690 n=5+5) Tar 251k ± 1% 251k ± 0% ~ (p=0.595 n=5+5) XML 394k ± 0% 393k ± 0% ~ (p=0.222 n=5+5) Change-Id: I237ae5502db4560f78ce021dc62f6d289797afd6 Reviewed-on: https://go-review.googlesource.com/39197 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Robert Griesemer <gri@golang.org>
2017-03-28 13:52:33 -07:00
if !n.Type.IsInterface() && !n.Left.Type.IsEmptyInterface() {
n.List.Set1(itabname(n.Type, n.Left.Type))
}
case OLEN, OCAP:
if isRuneCount(n) {
// Replace len([]rune(string)) with runtime.countrunes(string).
n = mkcall("countrunes", n.Type, init, conv(n.Left.Left, types.Types[TSTRING]))
break
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
// replace len(*[10]int) with 10.
// delayed until now to preserve side effects.
t := n.Left.Type
if t.IsPtr() {
t = t.Elem()
}
if t.IsArray() {
safeexpr(n.Left, init)
setintconst(n, t.NumElem())
n.SetTypecheck(1)
}
case OCOMPLEX:
// Use results from call expression as arguments for complex.
if n.Left == nil && n.Right == nil {
n.Left = n.List.First()
n.Right = n.List.Second()
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
n.Right = walkexpr(n.Right, init)
case OEQ, ONE, OLT, OLE, OGT, OGE:
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n = walkcompare(n, init)
case OANDAND, OOROR:
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
// cannot put side effects from n.Right on init,
// because they cannot run before n.Left is checked.
// save elsewhere and store on the eventual n.Right.
var ll Nodes
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Right = walkexpr(n.Right, &ll)
n.Right = addinit(n.Right, ll.Slice())
case OPRINT, OPRINTN:
n = walkprint(n, init)
case OPANIC:
n = mkcall("gopanic", nil, init, n.Left)
case ORECOVER:
n = mkcall("gorecover", n.Type, init, nod(OADDR, nodfp, nil))
case OCLOSUREVAR, OCFUNC:
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
case OCALLINTER, OCALLFUNC, OCALLMETH:
if n.Op == OCALLINTER {
usemethod(n)
}
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
if n.Op == OCALLFUNC && n.Left.Op == OCLOSURE {
// Transform direct call of a closure to call of a normal function.
// transformclosure already did all preparation work.
// Prepend captured variables to argument list.
n.List.Prepend(n.Left.Func.Enter.Slice()...)
n.Left.Func.Enter.Set(nil)
// Replace OCLOSURE with ONAME/PFUNC.
n.Left = n.Left.Func.Closure.Func.Nname
// Update type of OCALLFUNC node.
// Output arguments had not changed, but their offsets could.
if n.Left.Type.NumResults() == 1 {
n.Type = n.Left.Type.Results().Field(0).Type
} else {
n.Type = n.Left.Type.Results()
}
}
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
walkCall(n, init)
case OAS, OASOP:
init.AppendNodes(&n.Ninit)
cmd/compile: avoid mapaccess at m[k]=append(m[k].. Currently rvalue m[k] is transformed during walk into: tmp1 := *mapaccess(m, k) tmp2 := append(tmp1, ...) *mapassign(m, k) = tmp2 However, this is suboptimal, as we could instead produce just: tmp := mapassign(m, k) *tmp := append(*tmp, ...) Optimization is possible only if during Order it may tell that m[k] is exactly the same at left and right part of assignment. It doesn't work: 1) m[f(k)] = append(m[f(k)], ...) 2) sink, m[k] = sink, append(m[k]...) 3) m[k] = append(..., m[k],...) Benchmark: name old time/op new time/op delta MapAppendAssign/Int32/256-8 33.5ns ± 3% 22.4ns ±10% -33.24% (p=0.000 n=16+18) MapAppendAssign/Int32/65536-8 68.2ns ± 6% 48.5ns ±29% -28.90% (p=0.000 n=20+20) MapAppendAssign/Int64/256-8 34.3ns ± 4% 23.3ns ± 5% -32.23% (p=0.000 n=17+18) MapAppendAssign/Int64/65536-8 65.9ns ± 7% 61.2ns ±19% -7.06% (p=0.002 n=18+20) MapAppendAssign/Str/256-8 116ns ±12% 79ns ±16% -31.70% (p=0.000 n=20+19) MapAppendAssign/Str/65536-8 134ns ±15% 111ns ±45% -16.95% (p=0.000 n=19+20) name old alloc/op new alloc/op delta MapAppendAssign/Int32/256-8 47.0B ± 0% 46.0B ± 0% -2.13% (p=0.000 n=19+18) MapAppendAssign/Int32/65536-8 27.0B ± 0% 20.7B ±30% -23.33% (p=0.000 n=20+20) MapAppendAssign/Int64/256-8 47.0B ± 0% 46.0B ± 0% -2.13% (p=0.000 n=20+17) MapAppendAssign/Int64/65536-8 27.0B ± 0% 27.0B ± 0% ~ (all equal) MapAppendAssign/Str/256-8 94.0B ± 0% 78.0B ± 0% -17.02% (p=0.000 n=20+16) MapAppendAssign/Str/65536-8 54.0B ± 0% 54.0B ± 0% ~ (all equal) Fixes #24364 Updates #5147 Change-Id: Id257d052b75b9a445b4885dc571bf06ce6f6b409 Reviewed-on: https://go-review.googlesource.com/100838 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2018-03-14 01:16:43 -07:00
// Recognize m[k] = append(m[k], ...) so we can reuse
// the mapassign call.
mapAppend := n.Left.Op == OINDEXMAP && n.Right.Op == OAPPEND
if mapAppend && !samesafeexpr(n.Left, n.Right.List.First()) {
Fatalf("not same expressions: %v != %v", n.Left, n.Right.List.First())
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
n.Left = safeexpr(n.Left, init)
cmd/compile: avoid mapaccess at m[k]=append(m[k].. Currently rvalue m[k] is transformed during walk into: tmp1 := *mapaccess(m, k) tmp2 := append(tmp1, ...) *mapassign(m, k) = tmp2 However, this is suboptimal, as we could instead produce just: tmp := mapassign(m, k) *tmp := append(*tmp, ...) Optimization is possible only if during Order it may tell that m[k] is exactly the same at left and right part of assignment. It doesn't work: 1) m[f(k)] = append(m[f(k)], ...) 2) sink, m[k] = sink, append(m[k]...) 3) m[k] = append(..., m[k],...) Benchmark: name old time/op new time/op delta MapAppendAssign/Int32/256-8 33.5ns ± 3% 22.4ns ±10% -33.24% (p=0.000 n=16+18) MapAppendAssign/Int32/65536-8 68.2ns ± 6% 48.5ns ±29% -28.90% (p=0.000 n=20+20) MapAppendAssign/Int64/256-8 34.3ns ± 4% 23.3ns ± 5% -32.23% (p=0.000 n=17+18) MapAppendAssign/Int64/65536-8 65.9ns ± 7% 61.2ns ±19% -7.06% (p=0.002 n=18+20) MapAppendAssign/Str/256-8 116ns ±12% 79ns ±16% -31.70% (p=0.000 n=20+19) MapAppendAssign/Str/65536-8 134ns ±15% 111ns ±45% -16.95% (p=0.000 n=19+20) name old alloc/op new alloc/op delta MapAppendAssign/Int32/256-8 47.0B ± 0% 46.0B ± 0% -2.13% (p=0.000 n=19+18) MapAppendAssign/Int32/65536-8 27.0B ± 0% 20.7B ±30% -23.33% (p=0.000 n=20+20) MapAppendAssign/Int64/256-8 47.0B ± 0% 46.0B ± 0% -2.13% (p=0.000 n=20+17) MapAppendAssign/Int64/65536-8 27.0B ± 0% 27.0B ± 0% ~ (all equal) MapAppendAssign/Str/256-8 94.0B ± 0% 78.0B ± 0% -17.02% (p=0.000 n=20+16) MapAppendAssign/Str/65536-8 54.0B ± 0% 54.0B ± 0% ~ (all equal) Fixes #24364 Updates #5147 Change-Id: Id257d052b75b9a445b4885dc571bf06ce6f6b409 Reviewed-on: https://go-review.googlesource.com/100838 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2018-03-14 01:16:43 -07:00
if mapAppend {
n.Right.List.SetFirst(n.Left)
}
if n.Op == OASOP {
// Rewrite x op= y into x = x op y.
n.Right = nod(n.SubOp(), n.Left, n.Right)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
n.Right = typecheck(n.Right, ctxExpr)
n.Op = OAS
n.ResetAux()
}
if oaslit(n, init) {
break
}
cmd/compile: disable various write barrier optimizations Several of our current write barrier elision optimizations are invalid with the hybrid barrier. Eliding the hybrid barrier requires that *both* the current and new pointer be already shaded and, since we don't have the flow analysis to figure out anything about the slot's current value, for now we have to just disable several of these optimizations. This has a slight impact on binary size. On linux/amd64, the go tool binary increases by 0.7% and the compile binary increases by 1.5%. It also has a slight impact on performance, as one would expect. We'll win some of this back in subsequent commits. name old time/op new time/op delta BinaryTree17-12 2.38s ± 1% 2.40s ± 1% +0.82% (p=0.000 n=18+20) Fannkuch11-12 2.84s ± 1% 2.70s ± 0% -4.97% (p=0.000 n=18+18) FmtFprintfEmpty-12 44.2ns ± 1% 46.4ns ± 2% +4.89% (p=0.000 n=16+18) FmtFprintfString-12 131ns ± 0% 134ns ± 1% +2.05% (p=0.000 n=12+19) FmtFprintfInt-12 114ns ± 1% 117ns ± 1% +3.26% (p=0.000 n=19+20) FmtFprintfIntInt-12 176ns ± 1% 181ns ± 1% +3.25% (p=0.000 n=20+20) FmtFprintfPrefixedInt-12 185ns ± 1% 190ns ± 1% +2.77% (p=0.000 n=19+18) FmtFprintfFloat-12 249ns ± 1% 254ns ± 1% +1.71% (p=0.000 n=18+20) FmtManyArgs-12 747ns ± 1% 743ns ± 1% -0.58% (p=0.000 n=19+18) GobDecode-12 6.57ms ± 1% 6.61ms ± 0% +0.73% (p=0.000 n=19+20) GobEncode-12 5.58ms ± 1% 5.60ms ± 0% +0.27% (p=0.001 n=18+18) Gzip-12 223ms ± 1% 223ms ± 1% ~ (p=0.351 n=19+20) Gunzip-12 37.9ms ± 0% 37.9ms ± 1% ~ (p=0.095 n=16+20) HTTPClientServer-12 77.8µs ± 1% 78.5µs ± 1% +0.97% (p=0.000 n=19+20) JSONEncode-12 14.8ms ± 1% 14.8ms ± 1% ~ (p=0.079 n=20+19) JSONDecode-12 53.7ms ± 1% 54.2ms ± 1% +0.92% (p=0.000 n=20+19) Mandelbrot200-12 3.81ms ± 1% 3.81ms ± 0% ~ (p=0.916 n=19+18) GoParse-12 3.19ms ± 1% 3.19ms ± 1% ~ (p=0.175 n=20+19) RegexpMatchEasy0_32-12 71.9ns ± 1% 70.6ns ± 1% -1.87% (p=0.000 n=19+20) RegexpMatchEasy0_1K-12 946ns ± 0% 944ns ± 0% -0.22% (p=0.000 n=19+16) RegexpMatchEasy1_32-12 67.3ns ± 2% 66.8ns ± 1% -0.72% (p=0.008 n=20+20) RegexpMatchEasy1_1K-12 374ns ± 1% 384ns ± 1% +2.69% (p=0.000 n=18+20) RegexpMatchMedium_32-12 107ns ± 1% 107ns ± 1% ~ (p=1.000 n=20+20) RegexpMatchMedium_1K-12 34.3µs ± 1% 34.6µs ± 1% +0.90% (p=0.000 n=20+20) RegexpMatchHard_32-12 1.78µs ± 1% 1.80µs ± 1% +1.45% (p=0.000 n=20+19) RegexpMatchHard_1K-12 53.6µs ± 0% 54.5µs ± 1% +1.52% (p=0.000 n=19+18) Revcomp-12 417ms ± 5% 391ms ± 1% -6.42% (p=0.000 n=16+19) Template-12 61.1ms ± 1% 64.2ms ± 0% +5.07% (p=0.000 n=19+20) TimeParse-12 302ns ± 1% 305ns ± 1% +0.90% (p=0.000 n=18+18) TimeFormat-12 319ns ± 1% 315ns ± 1% -1.25% (p=0.000 n=18+18) [Geo mean] 54.0µs 54.3µs +0.58% name old time/op new time/op delta XGarbage-12 2.24ms ± 2% 2.28ms ± 1% +1.68% (p=0.000 n=18+17) XHTTP-12 11.4µs ± 1% 11.6µs ± 2% +1.63% (p=0.000 n=18+18) XJSON-12 11.6ms ± 0% 12.5ms ± 0% +7.84% (p=0.000 n=18+17) Updates #17503. Change-Id: I1899f8e35662971e24bf692b517dfbe2b533c00c Reviewed-on: https://go-review.googlesource.com/31572 Reviewed-by: Keith Randall <khr@golang.org>
2016-10-18 10:26:28 -04:00
if n.Right == nil {
// TODO(austin): Check all "implicit zeroing"
break
}
if !instrumenting && isZero(n.Right) {
break
}
switch n.Right.Op {
default:
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Right = walkexpr(n.Right, init)
case ORECV:
// x = <-c; n.Left is x, n.Right.Left is c.
// order.stmt made sure x is addressable.
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Right.Left = walkexpr(n.Right.Left, init)
n1 := nod(OADDR, n.Left, nil)
r := n.Right.Left // the channel
n = mkcall1(chanfn("chanrecv1", 2, r.Type), nil, init, r, n1)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n = walkexpr(n, init)
break opswitch
cmd/internal/gc: optimize append + write barrier The code generated for x = append(x, v) is roughly: t := x if len(t)+1 > cap(t) { t = grow(t) } t[len(t)] = v len(t)++ x = t We used to generate this code as Go pseudocode during walk. Generate it instead as actual instructions during gen. Doing so lets us apply a few optimizations. The most important is that when, as in the above example, the source slice and the destination slice are the same, the code can instead do: t := x if len(t)+1 > cap(t) { t = grow(t) x = {base(t), len(t)+1, cap(t)} } else { len(x)++ } t[len(t)] = v That is, in the fast path that does not reallocate the array, only the updated length needs to be written back to x, not the array pointer and not the capacity. This is more like what you'd write by hand in C. It's faster in general, since the fast path elides two of the three stores, but it's especially faster when the form of x is such that the base pointer write would turn into a write barrier. No write, no barrier. name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.81s × (0.98,1.03) +2.35% (p=0.023) Fannkuch11 4.41s × (0.98,1.03) 4.35s × (1.00,1.00) ~ (p=0.090) FmtFprintfEmpty 92.7ns × (0.91,1.16) 86.0ns × (0.94,1.11) -7.31% (p=0.038) FmtFprintfString 281ns × (0.96,1.08) 276ns × (0.98,1.04) ~ (p=0.219) FmtFprintfInt 288ns × (0.97,1.06) 274ns × (0.98,1.06) -4.94% (p=0.002) FmtFprintfIntInt 493ns × (0.97,1.04) 506ns × (0.99,1.01) +2.65% (p=0.009) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 391ns × (0.99,1.01) -7.52% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 566ns × (0.99,1.01) -5.27% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.91µs × (0.99,1.01) ~ (p=0.231) GobDecode 14.8ms × (0.98,1.03) 15.3ms × (0.99,1.02) +3.01% (p=0.000) GobEncode 12.3ms × (0.98,1.01) 11.5ms × (0.97,1.03) -5.93% (p=0.000) Gzip 656ms × (0.99,1.05) 645ms × (0.99,1.01) ~ (p=0.055) Gunzip 142ms × (1.00,1.00) 142ms × (1.00,1.00) -0.32% (p=0.034) HTTPClientServer 91.2µs × (0.97,1.04) 90.5µs × (0.97,1.04) ~ (p=0.468) JSONEncode 32.6ms × (0.97,1.08) 32.0ms × (0.98,1.03) ~ (p=0.190) JSONDecode 114ms × (0.97,1.05) 114ms × (0.99,1.01) ~ (p=0.887) Mandelbrot200 6.11ms × (0.98,1.04) 6.04ms × (1.00,1.01) ~ (p=0.167) GoParse 6.66ms × (0.97,1.04) 6.47ms × (0.97,1.05) -2.81% (p=0.014) RegexpMatchEasy0_32 159ns × (0.99,1.00) 171ns × (0.93,1.07) +7.19% (p=0.002) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 550ns × (0.98,1.01) +2.30% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 135ns × (0.99,1.02) -1.60% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 879ns × (1.00,1.01) +1.08% (p=0.000) RegexpMatchMedium_32 252ns × (0.99,1.01) 243ns × (1.00,1.00) -3.71% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 70.3µs × (1.00,1.00) -3.34% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.82µs × (1.00,1.01) -0.81% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 117µs × (1.00,1.00) -0.56% (p=0.000) Revcomp 920ms × (0.97,1.07) 917ms × (0.97,1.04) ~ (p=0.808) Template 129ms × (0.98,1.03) 114ms × (0.99,1.01) -12.06% (p=0.000) TimeParse 619ns × (0.99,1.01) 622ns × (0.99,1.01) ~ (p=0.062) TimeFormat 661ns × (0.98,1.04) 665ns × (0.99,1.01) ~ (p=0.524) See next CL for combination with a similar optimization for slice. The benchmarks that are slower in this CL are still faster overall with the combination of the two. Change-Id: I2a7421658091b2488c64741b4db15ab6c3b4cb7e Reviewed-on: https://go-review.googlesource.com/9812 Reviewed-by: David Chase <drchase@google.com>
2015-05-06 12:34:30 -04:00
case OAPPEND:
// x = append(...)
r := n.Right
cmd/compile: pack bool fields in Node, Name, Func and Type structs to bitsets This reduces compiler memory usage by up to 4% - see compilebench results below. name old time/op new time/op delta Template 245ms ± 4% 241ms ± 2% -1.88% (p=0.029 n=10+10) Unicode 126ms ± 3% 124ms ± 3% ~ (p=0.105 n=10+10) GoTypes 805ms ± 2% 813ms ± 3% ~ (p=0.515 n=8+10) Compiler 3.95s ± 2% 3.83s ± 1% -2.96% (p=0.000 n=9+10) MakeBash 47.4s ± 4% 46.6s ± 1% -1.59% (p=0.028 n=9+10) name old user-ns/op new user-ns/op delta Template 324M ± 5% 326M ± 3% ~ (p=0.935 n=10+10) Unicode 186M ± 5% 178M ±10% ~ (p=0.067 n=9+10) GoTypes 1.08G ± 7% 1.09G ± 4% ~ (p=0.956 n=10+10) Compiler 5.34G ± 4% 5.31G ± 1% ~ (p=0.501 n=10+8) name old alloc/op new alloc/op delta Template 41.0MB ± 0% 39.8MB ± 0% -3.03% (p=0.000 n=10+10) Unicode 32.3MB ± 0% 31.0MB ± 0% -4.13% (p=0.000 n=10+10) GoTypes 119MB ± 0% 116MB ± 0% -2.39% (p=0.000 n=10+10) Compiler 499MB ± 0% 487MB ± 0% -2.48% (p=0.000 n=10+10) name old allocs/op new allocs/op delta Template 380k ± 1% 379k ± 1% ~ (p=0.436 n=10+10) Unicode 324k ± 1% 324k ± 0% ~ (p=0.853 n=10+10) GoTypes 1.15M ± 0% 1.15M ± 0% ~ (p=0.481 n=10+10) Compiler 4.41M ± 0% 4.41M ± 0% -0.12% (p=0.007 n=10+10) name old text-bytes new text-bytes delta HelloSize 623k ± 0% 623k ± 0% ~ (all equal) CmdGoSize 6.64M ± 0% 6.64M ± 0% ~ (all equal) name old data-bytes new data-bytes delta HelloSize 5.81k ± 0% 5.81k ± 0% ~ (all equal) CmdGoSize 238k ± 0% 238k ± 0% ~ (all equal) name old bss-bytes new bss-bytes delta HelloSize 134k ± 0% 134k ± 0% ~ (all equal) CmdGoSize 152k ± 0% 152k ± 0% ~ (all equal) name old exe-bytes new exe-bytes delta HelloSize 967k ± 0% 967k ± 0% ~ (all equal) CmdGoSize 10.2M ± 0% 10.2M ± 0% ~ (all equal) Change-Id: I1f40af738254892bd6c8ba2eb43390b175753d52 Reviewed-on: https://go-review.googlesource.com/37445 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-02-27 19:56:38 +02:00
if r.Type.Elem().NotInHeap() {
cmd/compile: add go:notinheap type pragma This adds a //go:notinheap pragma for declarations of types that must not be heap allocated. We ensure these rules by disallowing new(T), make([]T), append([]T), or implicit allocation of T, by disallowing conversions to notinheap types, and by propagating notinheap to any struct or array that contains notinheap elements. The utility of this pragma is that we can eliminate write barriers for writes to pointers to go:notinheap types, since the write barrier is guaranteed to be a no-op. This will let us mark several scheduler and memory allocator structures as go:notinheap, which will let us disallow write barriers in the scheduler and memory allocator much more thoroughly and also eliminate some problematic hybrid write barriers. This also makes go:nowritebarrierrec and go:yeswritebarrierrec much more powerful. Currently we use go:nowritebarrier all over the place, but it's almost never what you actually want: when write barriers are illegal, they're typically illegal for a whole dynamic scope. Partly this is because go:nowritebarrier has been around longer, but it's also because go:nowritebarrierrec couldn't be used in situations that had no-op write barriers or where some nested scope did allow write barriers. go:notinheap eliminates many no-op write barriers and go:yeswritebarrierrec makes it possible to opt back in to write barriers, so these two changes will let us use go:nowritebarrierrec far more liberally. This updates #13386, which is about controlling pointers from non-GC'd memory to GC'd memory. That would require some additional pragma (or pragmas), but could build on this pragma. Change-Id: I6314f8f4181535dd166887c9ec239977b54940bd Reviewed-on: https://go-review.googlesource.com/30939 Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-10-11 22:53:27 -04:00
yyerror("%v is go:notinheap; heap allocation disallowed", r.Type.Elem())
}
switch {
case isAppendOfMake(r):
// x = append(y, make([]T, y)...)
r = extendslice(r, init)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
case r.IsDDD():
cmd/internal/gc: optimize append + write barrier The code generated for x = append(x, v) is roughly: t := x if len(t)+1 > cap(t) { t = grow(t) } t[len(t)] = v len(t)++ x = t We used to generate this code as Go pseudocode during walk. Generate it instead as actual instructions during gen. Doing so lets us apply a few optimizations. The most important is that when, as in the above example, the source slice and the destination slice are the same, the code can instead do: t := x if len(t)+1 > cap(t) { t = grow(t) x = {base(t), len(t)+1, cap(t)} } else { len(x)++ } t[len(t)] = v That is, in the fast path that does not reallocate the array, only the updated length needs to be written back to x, not the array pointer and not the capacity. This is more like what you'd write by hand in C. It's faster in general, since the fast path elides two of the three stores, but it's especially faster when the form of x is such that the base pointer write would turn into a write barrier. No write, no barrier. name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.81s × (0.98,1.03) +2.35% (p=0.023) Fannkuch11 4.41s × (0.98,1.03) 4.35s × (1.00,1.00) ~ (p=0.090) FmtFprintfEmpty 92.7ns × (0.91,1.16) 86.0ns × (0.94,1.11) -7.31% (p=0.038) FmtFprintfString 281ns × (0.96,1.08) 276ns × (0.98,1.04) ~ (p=0.219) FmtFprintfInt 288ns × (0.97,1.06) 274ns × (0.98,1.06) -4.94% (p=0.002) FmtFprintfIntInt 493ns × (0.97,1.04) 506ns × (0.99,1.01) +2.65% (p=0.009) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 391ns × (0.99,1.01) -7.52% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 566ns × (0.99,1.01) -5.27% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.91µs × (0.99,1.01) ~ (p=0.231) GobDecode 14.8ms × (0.98,1.03) 15.3ms × (0.99,1.02) +3.01% (p=0.000) GobEncode 12.3ms × (0.98,1.01) 11.5ms × (0.97,1.03) -5.93% (p=0.000) Gzip 656ms × (0.99,1.05) 645ms × (0.99,1.01) ~ (p=0.055) Gunzip 142ms × (1.00,1.00) 142ms × (1.00,1.00) -0.32% (p=0.034) HTTPClientServer 91.2µs × (0.97,1.04) 90.5µs × (0.97,1.04) ~ (p=0.468) JSONEncode 32.6ms × (0.97,1.08) 32.0ms × (0.98,1.03) ~ (p=0.190) JSONDecode 114ms × (0.97,1.05) 114ms × (0.99,1.01) ~ (p=0.887) Mandelbrot200 6.11ms × (0.98,1.04) 6.04ms × (1.00,1.01) ~ (p=0.167) GoParse 6.66ms × (0.97,1.04) 6.47ms × (0.97,1.05) -2.81% (p=0.014) RegexpMatchEasy0_32 159ns × (0.99,1.00) 171ns × (0.93,1.07) +7.19% (p=0.002) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 550ns × (0.98,1.01) +2.30% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 135ns × (0.99,1.02) -1.60% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 879ns × (1.00,1.01) +1.08% (p=0.000) RegexpMatchMedium_32 252ns × (0.99,1.01) 243ns × (1.00,1.00) -3.71% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 70.3µs × (1.00,1.00) -3.34% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.82µs × (1.00,1.01) -0.81% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 117µs × (1.00,1.00) -0.56% (p=0.000) Revcomp 920ms × (0.97,1.07) 917ms × (0.97,1.04) ~ (p=0.808) Template 129ms × (0.98,1.03) 114ms × (0.99,1.01) -12.06% (p=0.000) TimeParse 619ns × (0.99,1.01) 622ns × (0.99,1.01) ~ (p=0.062) TimeFormat 661ns × (0.98,1.04) 665ns × (0.99,1.01) ~ (p=0.524) See next CL for combination with a similar optimization for slice. The benchmarks that are slower in this CL are still faster overall with the combination of the two. Change-Id: I2a7421658091b2488c64741b4db15ab6c3b4cb7e Reviewed-on: https://go-review.googlesource.com/9812 Reviewed-by: David Chase <drchase@google.com>
2015-05-06 12:34:30 -04:00
r = appendslice(r, init) // also works for append(slice, string).
default:
cmd/internal/gc: optimize append + write barrier The code generated for x = append(x, v) is roughly: t := x if len(t)+1 > cap(t) { t = grow(t) } t[len(t)] = v len(t)++ x = t We used to generate this code as Go pseudocode during walk. Generate it instead as actual instructions during gen. Doing so lets us apply a few optimizations. The most important is that when, as in the above example, the source slice and the destination slice are the same, the code can instead do: t := x if len(t)+1 > cap(t) { t = grow(t) x = {base(t), len(t)+1, cap(t)} } else { len(x)++ } t[len(t)] = v That is, in the fast path that does not reallocate the array, only the updated length needs to be written back to x, not the array pointer and not the capacity. This is more like what you'd write by hand in C. It's faster in general, since the fast path elides two of the three stores, but it's especially faster when the form of x is such that the base pointer write would turn into a write barrier. No write, no barrier. name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.81s × (0.98,1.03) +2.35% (p=0.023) Fannkuch11 4.41s × (0.98,1.03) 4.35s × (1.00,1.00) ~ (p=0.090) FmtFprintfEmpty 92.7ns × (0.91,1.16) 86.0ns × (0.94,1.11) -7.31% (p=0.038) FmtFprintfString 281ns × (0.96,1.08) 276ns × (0.98,1.04) ~ (p=0.219) FmtFprintfInt 288ns × (0.97,1.06) 274ns × (0.98,1.06) -4.94% (p=0.002) FmtFprintfIntInt 493ns × (0.97,1.04) 506ns × (0.99,1.01) +2.65% (p=0.009) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 391ns × (0.99,1.01) -7.52% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 566ns × (0.99,1.01) -5.27% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.91µs × (0.99,1.01) ~ (p=0.231) GobDecode 14.8ms × (0.98,1.03) 15.3ms × (0.99,1.02) +3.01% (p=0.000) GobEncode 12.3ms × (0.98,1.01) 11.5ms × (0.97,1.03) -5.93% (p=0.000) Gzip 656ms × (0.99,1.05) 645ms × (0.99,1.01) ~ (p=0.055) Gunzip 142ms × (1.00,1.00) 142ms × (1.00,1.00) -0.32% (p=0.034) HTTPClientServer 91.2µs × (0.97,1.04) 90.5µs × (0.97,1.04) ~ (p=0.468) JSONEncode 32.6ms × (0.97,1.08) 32.0ms × (0.98,1.03) ~ (p=0.190) JSONDecode 114ms × (0.97,1.05) 114ms × (0.99,1.01) ~ (p=0.887) Mandelbrot200 6.11ms × (0.98,1.04) 6.04ms × (1.00,1.01) ~ (p=0.167) GoParse 6.66ms × (0.97,1.04) 6.47ms × (0.97,1.05) -2.81% (p=0.014) RegexpMatchEasy0_32 159ns × (0.99,1.00) 171ns × (0.93,1.07) +7.19% (p=0.002) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 550ns × (0.98,1.01) +2.30% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 135ns × (0.99,1.02) -1.60% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 879ns × (1.00,1.01) +1.08% (p=0.000) RegexpMatchMedium_32 252ns × (0.99,1.01) 243ns × (1.00,1.00) -3.71% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 70.3µs × (1.00,1.00) -3.34% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.82µs × (1.00,1.01) -0.81% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 117µs × (1.00,1.00) -0.56% (p=0.000) Revcomp 920ms × (0.97,1.07) 917ms × (0.97,1.04) ~ (p=0.808) Template 129ms × (0.98,1.03) 114ms × (0.99,1.01) -12.06% (p=0.000) TimeParse 619ns × (0.99,1.01) 622ns × (0.99,1.01) ~ (p=0.062) TimeFormat 661ns × (0.98,1.04) 665ns × (0.99,1.01) ~ (p=0.524) See next CL for combination with a similar optimization for slice. The benchmarks that are slower in this CL are still faster overall with the combination of the two. Change-Id: I2a7421658091b2488c64741b4db15ab6c3b4cb7e Reviewed-on: https://go-review.googlesource.com/9812 Reviewed-by: David Chase <drchase@google.com>
2015-05-06 12:34:30 -04:00
r = walkappend(r, init, n)
}
n.Right = r
if r.Op == OAPPEND {
// Left in place for back end.
// Do not add a new write barrier.
// Set up address of type for back end.
r.Left = typename(r.Type.Elem())
break opswitch
cmd/internal/gc: optimize append + write barrier The code generated for x = append(x, v) is roughly: t := x if len(t)+1 > cap(t) { t = grow(t) } t[len(t)] = v len(t)++ x = t We used to generate this code as Go pseudocode during walk. Generate it instead as actual instructions during gen. Doing so lets us apply a few optimizations. The most important is that when, as in the above example, the source slice and the destination slice are the same, the code can instead do: t := x if len(t)+1 > cap(t) { t = grow(t) x = {base(t), len(t)+1, cap(t)} } else { len(x)++ } t[len(t)] = v That is, in the fast path that does not reallocate the array, only the updated length needs to be written back to x, not the array pointer and not the capacity. This is more like what you'd write by hand in C. It's faster in general, since the fast path elides two of the three stores, but it's especially faster when the form of x is such that the base pointer write would turn into a write barrier. No write, no barrier. name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.81s × (0.98,1.03) +2.35% (p=0.023) Fannkuch11 4.41s × (0.98,1.03) 4.35s × (1.00,1.00) ~ (p=0.090) FmtFprintfEmpty 92.7ns × (0.91,1.16) 86.0ns × (0.94,1.11) -7.31% (p=0.038) FmtFprintfString 281ns × (0.96,1.08) 276ns × (0.98,1.04) ~ (p=0.219) FmtFprintfInt 288ns × (0.97,1.06) 274ns × (0.98,1.06) -4.94% (p=0.002) FmtFprintfIntInt 493ns × (0.97,1.04) 506ns × (0.99,1.01) +2.65% (p=0.009) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 391ns × (0.99,1.01) -7.52% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 566ns × (0.99,1.01) -5.27% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.91µs × (0.99,1.01) ~ (p=0.231) GobDecode 14.8ms × (0.98,1.03) 15.3ms × (0.99,1.02) +3.01% (p=0.000) GobEncode 12.3ms × (0.98,1.01) 11.5ms × (0.97,1.03) -5.93% (p=0.000) Gzip 656ms × (0.99,1.05) 645ms × (0.99,1.01) ~ (p=0.055) Gunzip 142ms × (1.00,1.00) 142ms × (1.00,1.00) -0.32% (p=0.034) HTTPClientServer 91.2µs × (0.97,1.04) 90.5µs × (0.97,1.04) ~ (p=0.468) JSONEncode 32.6ms × (0.97,1.08) 32.0ms × (0.98,1.03) ~ (p=0.190) JSONDecode 114ms × (0.97,1.05) 114ms × (0.99,1.01) ~ (p=0.887) Mandelbrot200 6.11ms × (0.98,1.04) 6.04ms × (1.00,1.01) ~ (p=0.167) GoParse 6.66ms × (0.97,1.04) 6.47ms × (0.97,1.05) -2.81% (p=0.014) RegexpMatchEasy0_32 159ns × (0.99,1.00) 171ns × (0.93,1.07) +7.19% (p=0.002) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 550ns × (0.98,1.01) +2.30% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 135ns × (0.99,1.02) -1.60% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 879ns × (1.00,1.01) +1.08% (p=0.000) RegexpMatchMedium_32 252ns × (0.99,1.01) 243ns × (1.00,1.00) -3.71% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 70.3µs × (1.00,1.00) -3.34% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.82µs × (1.00,1.01) -0.81% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 117µs × (1.00,1.00) -0.56% (p=0.000) Revcomp 920ms × (0.97,1.07) 917ms × (0.97,1.04) ~ (p=0.808) Template 129ms × (0.98,1.03) 114ms × (0.99,1.01) -12.06% (p=0.000) TimeParse 619ns × (0.99,1.01) 622ns × (0.99,1.01) ~ (p=0.062) TimeFormat 661ns × (0.98,1.04) 665ns × (0.99,1.01) ~ (p=0.524) See next CL for combination with a similar optimization for slice. The benchmarks that are slower in this CL are still faster overall with the combination of the two. Change-Id: I2a7421658091b2488c64741b4db15ab6c3b4cb7e Reviewed-on: https://go-review.googlesource.com/9812 Reviewed-by: David Chase <drchase@google.com>
2015-05-06 12:34:30 -04:00
}
// Otherwise, lowered for race detector.
// Treat as ordinary assignment.
}
if n.Left != nil && n.Right != nil {
n = convas(n, init)
}
case OAS2:
init.AppendNodes(&n.Ninit)
walkexprlistsafe(n.List.Slice(), init)
walkexprlistsafe(n.Rlist.Slice(), init)
ll := ascompatee(OAS, n.List.Slice(), n.Rlist.Slice(), init)
ll = reorder3(ll)
n = liststmt(ll)
// a,b,... = fn()
case OAS2FUNC:
init.AppendNodes(&n.Ninit)
r := n.Right
walkexprlistsafe(n.List.Slice(), init)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
r = walkexpr(r, init)
if isIntrinsicCall(r) {
n.Right = r
break
}
init.Append(r)
ll := ascompatet(n.List, r.Type)
n = liststmt(ll)
// x, y = <-c
// order.stmt made sure x is addressable or blank.
case OAS2RECV:
init.AppendNodes(&n.Ninit)
r := n.Right
walkexprlistsafe(n.List.Slice(), init)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
r.Left = walkexpr(r.Left, init)
var n1 *Node
if n.List.First().isBlank() {
n1 = nodnil()
} else {
n1 = nod(OADDR, n.List.First(), nil)
}
fn := chanfn("chanrecv2", 2, r.Left.Type)
ok := n.List.Second()
call := mkcall1(fn, types.Types[TBOOL], init, r.Left, n1)
n = nod(OAS, ok, call)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
n = typecheck(n, ctxStmt)
// a,b = m[i]
case OAS2MAPR:
init.AppendNodes(&n.Ninit)
r := n.Right
walkexprlistsafe(n.List.Slice(), init)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
r.Left = walkexpr(r.Left, init)
r.Right = walkexpr(r.Right, init)
t := r.Left.Type
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
fast := mapfast(t)
var key *Node
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
if fast != mapslow {
// fast versions take key by value
key = r.Right
} else {
// standard version takes key by reference
// order.expr made sure key is addressable.
key = nod(OADDR, r.Right, nil)
}
// from:
// a,b = m[i]
// to:
// var,b = mapaccess2*(t, m, i)
// a = *var
a := n.List.First()
if w := t.Elem().Width; w <= zeroValSize {
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
fn := mapfn(mapaccess2[fast], t)
r = mkcall1(fn, fn.Type.Results(), init, typename(t), r.Left, key)
} else {
fn := mapfn("mapaccess2_fat", t)
z := zeroaddr(w)
r = mkcall1(fn, fn.Type.Results(), init, typename(t), r.Left, key, z)
}
// mapaccess2* returns a typed bool, but due to spec changes,
// the boolean result of i.(T) is now untyped so we make it the
// same type as the variable on the lhs.
if ok := n.List.Second(); !ok.isBlank() && ok.Type.IsBoolean() {
r.Type.Field(1).Type = ok.Type
}
n.Right = r
n.Op = OAS2FUNC
// don't generate a = *var if a is _
if !a.isBlank() {
var_ := temp(types.NewPtr(t.Elem()))
var_.SetTypecheck(1)
var_.MarkNonNil() // mapaccess always returns a non-nil pointer
n.List.SetFirst(var_)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n = walkexpr(n, init)
init.Append(n)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
n = nod(OAS, a, nod(ODEREF, var_, nil))
}
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
n = typecheck(n, ctxStmt)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n = walkexpr(n, init)
case ODELETE:
init.AppendNodes(&n.Ninit)
map_ := n.List.First()
key := n.List.Second()
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
map_ = walkexpr(map_, init)
key = walkexpr(key, init)
t := map_.Type
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
fast := mapfast(t)
if fast == mapslow {
// order.stmt made sure key is addressable.
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
key = nod(OADDR, key, nil)
}
n = mkcall1(mapfndel(mapdelete[fast], t), nil, init, typename(t), map_, key)
case OAS2DOTTYPE:
walkexprlistsafe(n.List.Slice(), init)
n.Right = walkexpr(n.Right, init)
case OCONVIFACE:
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
fromType := n.Left.Type
toType := n.Type
if !fromType.IsInterface() {
markTypeUsedInInterface(fromType)
}
// typeword generates the type word of the interface value.
typeword := func() *Node {
if toType.IsEmptyInterface() {
return typename(fromType)
}
return itabname(fromType, toType)
}
// Optimize convT2E or convT2I as a two-word copy when T is pointer-shaped.
if isdirectiface(fromType) {
l := nod(OEFACE, typeword(), n.Left)
l.Type = toType
l.SetTypecheck(n.Typecheck())
n = l
break
}
cmd/compile: convert constants to interfaces without allocating The order pass is responsible for ensuring that values passed to runtime functions, including convT2E/convT2I, are addressable. Prior to this CL, this was always accomplished by creating a temp, which frequently escaped to the heap, causing allocations, perhaps most notably in code like: fmt.Println(1, 2, 3) // allocates three times None of the runtime routines modify the contents of the pointers they receive, so in the case of constants, instead of creating a temp value, we can create a static value. (Marking the static value as read-only provides protection against accidental attempts by the runtime to modify the constant data.) This improves code generation for code like: panic("abc") c <- 2 // c is a chan int which can now simply refer to "abc" and 2, rather than going by way of a temporary. It also allows us to optimize convT2E/convT2I, by recognizing static readonly values and directly constructing the interface. This CL adds ~0.5% to binary size, despite decreasing the size of many functions, because it also adds many static symbols. This binary size regression could be recovered in future (but currently unplanned) work. There is a lot of content-duplication in these symbols; this statement generates six new symbols, three containing an int 1 and three containing a pointer to the string "a": fmt.Println(1, 1, 1, "a", "a", "a") These symbols could be made content-addressable. Furthermore, these symbols are small, so the alignment and naming overhead is large. As with the go.strings section, these symbols could be hidden and have their alignment reduced. The changes to test/live.go make it impossible (at least with current optimization techniques) to place the values being passed to the runtime in static symbols, preserving autotmp creation. Fixes #18704 Benchmarks from fmt and go-kit's logging package: github.com/go-kit/kit/log name old time/op new time/op delta JSONLoggerSimple-8 1.91µs ± 2% 2.11µs ±22% ~ (p=1.000 n=9+10) JSONLoggerContextual-8 2.60µs ± 6% 2.43µs ± 2% -6.29% (p=0.000 n=9+10) Discard-8 101ns ± 2% 34ns ±14% -66.33% (p=0.000 n=10+9) OneWith-8 161ns ± 1% 102ns ±16% -36.78% (p=0.000 n=10+10) TwoWith-8 175ns ± 3% 106ns ± 7% -39.36% (p=0.000 n=10+9) TenWith-8 293ns ± 3% 227ns ±15% -22.44% (p=0.000 n=9+10) LogfmtLoggerSimple-8 704ns ± 2% 608ns ± 2% -13.65% (p=0.000 n=10+9) LogfmtLoggerContextual-8 962ns ± 1% 860ns ±17% -10.57% (p=0.003 n=9+10) NopLoggerSimple-8 188ns ± 1% 120ns ± 1% -36.39% (p=0.000 n=9+10) NopLoggerContextual-8 379ns ± 1% 243ns ± 0% -35.77% (p=0.000 n=9+10) ValueBindingTimestamp-8 577ns ± 1% 499ns ± 1% -13.51% (p=0.000 n=10+10) ValueBindingCaller-8 898ns ± 2% 844ns ± 2% -6.00% (p=0.000 n=10+10) name old alloc/op new alloc/op delta JSONLoggerSimple-8 904B ± 0% 872B ± 0% -3.54% (p=0.000 n=10+10) JSONLoggerContextual-8 1.20kB ± 0% 1.14kB ± 0% -5.33% (p=0.000 n=10+10) Discard-8 64.0B ± 0% 32.0B ± 0% -50.00% (p=0.000 n=10+10) OneWith-8 96.0B ± 0% 64.0B ± 0% -33.33% (p=0.000 n=10+10) TwoWith-8 160B ± 0% 128B ± 0% -20.00% (p=0.000 n=10+10) TenWith-8 672B ± 0% 640B ± 0% -4.76% (p=0.000 n=10+10) LogfmtLoggerSimple-8 128B ± 0% 96B ± 0% -25.00% (p=0.000 n=10+10) LogfmtLoggerContextual-8 304B ± 0% 240B ± 0% -21.05% (p=0.000 n=10+10) NopLoggerSimple-8 128B ± 0% 96B ± 0% -25.00% (p=0.000 n=10+10) NopLoggerContextual-8 304B ± 0% 240B ± 0% -21.05% (p=0.000 n=10+10) ValueBindingTimestamp-8 159B ± 0% 127B ± 0% -20.13% (p=0.000 n=10+10) ValueBindingCaller-8 112B ± 0% 80B ± 0% -28.57% (p=0.000 n=10+10) name old allocs/op new allocs/op delta JSONLoggerSimple-8 19.0 ± 0% 17.0 ± 0% -10.53% (p=0.000 n=10+10) JSONLoggerContextual-8 25.0 ± 0% 21.0 ± 0% -16.00% (p=0.000 n=10+10) Discard-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) OneWith-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) TwoWith-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) TenWith-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) LogfmtLoggerSimple-8 4.00 ± 0% 2.00 ± 0% -50.00% (p=0.000 n=10+10) LogfmtLoggerContextual-8 7.00 ± 0% 3.00 ± 0% -57.14% (p=0.000 n=10+10) NopLoggerSimple-8 4.00 ± 0% 2.00 ± 0% -50.00% (p=0.000 n=10+10) NopLoggerContextual-8 7.00 ± 0% 3.00 ± 0% -57.14% (p=0.000 n=10+10) ValueBindingTimestamp-8 5.00 ± 0% 3.00 ± 0% -40.00% (p=0.000 n=10+10) ValueBindingCaller-8 4.00 ± 0% 2.00 ± 0% -50.00% (p=0.000 n=10+10) fmt name old time/op new time/op delta SprintfPadding-8 88.9ns ± 3% 79.1ns ± 1% -11.09% (p=0.000 n=10+7) SprintfEmpty-8 12.6ns ± 3% 12.8ns ± 3% ~ (p=0.136 n=10+10) SprintfString-8 38.7ns ± 5% 26.9ns ± 6% -30.65% (p=0.000 n=10+10) SprintfTruncateString-8 56.7ns ± 2% 47.0ns ± 3% -17.05% (p=0.000 n=10+10) SprintfQuoteString-8 164ns ± 2% 153ns ± 2% -7.01% (p=0.000 n=10+10) SprintfInt-8 38.9ns ±15% 26.5ns ± 2% -31.93% (p=0.000 n=10+9) SprintfIntInt-8 60.3ns ± 9% 38.2ns ± 1% -36.67% (p=0.000 n=10+8) SprintfPrefixedInt-8 58.6ns ±13% 51.2ns ±11% -12.66% (p=0.001 n=10+10) SprintfFloat-8 71.4ns ± 3% 64.2ns ± 3% -10.08% (p=0.000 n=8+10) SprintfComplex-8 175ns ± 3% 159ns ± 2% -9.03% (p=0.000 n=10+10) SprintfBoolean-8 33.5ns ± 4% 25.7ns ± 5% -23.28% (p=0.000 n=10+10) SprintfHexString-8 65.3ns ± 3% 51.7ns ± 5% -20.86% (p=0.000 n=10+9) SprintfHexBytes-8 67.2ns ± 5% 67.9ns ± 4% ~ (p=0.383 n=10+10) SprintfBytes-8 129ns ± 7% 124ns ± 7% ~ (p=0.074 n=9+10) SprintfStringer-8 127ns ± 4% 126ns ± 8% ~ (p=0.506 n=9+10) SprintfStructure-8 357ns ± 3% 359ns ± 3% ~ (p=0.469 n=10+10) ManyArgs-8 203ns ± 6% 126ns ± 3% -37.94% (p=0.000 n=10+10) FprintInt-8 119ns ±10% 74ns ± 3% -37.54% (p=0.000 n=10+10) FprintfBytes-8 122ns ± 4% 120ns ± 3% ~ (p=0.124 n=10+10) FprintIntNoAlloc-8 78.2ns ± 5% 74.1ns ± 3% -5.28% (p=0.000 n=10+10) ScanInts-8 349µs ± 1% 349µs ± 0% ~ (p=0.606 n=9+8) ScanRecursiveInt-8 43.8ms ± 7% 40.1ms ± 2% -8.42% (p=0.000 n=10+10) ScanRecursiveIntReaderWrapper-8 43.5ms ± 4% 40.4ms ± 2% -7.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta SprintfPadding-8 24.0B ± 0% 16.0B ± 0% -33.33% (p=0.000 n=10+10) SprintfEmpty-8 0.00B 0.00B ~ (all equal) SprintfString-8 21.0B ± 0% 5.0B ± 0% -76.19% (p=0.000 n=10+10) SprintfTruncateString-8 32.0B ± 0% 16.0B ± 0% -50.00% (p=0.000 n=10+10) SprintfQuoteString-8 48.0B ± 0% 32.0B ± 0% -33.33% (p=0.000 n=10+10) SprintfInt-8 16.0B ± 0% 1.0B ± 0% -93.75% (p=0.000 n=10+10) SprintfIntInt-8 24.0B ± 0% 3.0B ± 0% -87.50% (p=0.000 n=10+10) SprintfPrefixedInt-8 72.0B ± 0% 64.0B ± 0% -11.11% (p=0.000 n=10+10) SprintfFloat-8 16.0B ± 0% 8.0B ± 0% -50.00% (p=0.000 n=10+10) SprintfComplex-8 48.0B ± 0% 32.0B ± 0% -33.33% (p=0.000 n=10+10) SprintfBoolean-8 8.00B ± 0% 4.00B ± 0% -50.00% (p=0.000 n=10+10) SprintfHexString-8 96.0B ± 0% 80.0B ± 0% -16.67% (p=0.000 n=10+10) SprintfHexBytes-8 112B ± 0% 112B ± 0% ~ (all equal) SprintfBytes-8 96.0B ± 0% 96.0B ± 0% ~ (all equal) SprintfStringer-8 32.0B ± 0% 32.0B ± 0% ~ (all equal) SprintfStructure-8 256B ± 0% 256B ± 0% ~ (all equal) ManyArgs-8 80.0B ± 0% 0.0B -100.00% (p=0.000 n=10+10) FprintInt-8 8.00B ± 0% 0.00B -100.00% (p=0.000 n=10+10) FprintfBytes-8 32.0B ± 0% 32.0B ± 0% ~ (all equal) FprintIntNoAlloc-8 0.00B 0.00B ~ (all equal) ScanInts-8 15.2kB ± 0% 15.2kB ± 0% ~ (p=0.248 n=9+10) ScanRecursiveInt-8 21.6kB ± 0% 21.6kB ± 0% ~ (all equal) ScanRecursiveIntReaderWrapper-8 21.7kB ± 0% 21.7kB ± 0% ~ (all equal) name old allocs/op new allocs/op delta SprintfPadding-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfEmpty-8 0.00 0.00 ~ (all equal) SprintfString-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfTruncateString-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfQuoteString-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfInt-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfIntInt-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) SprintfPrefixedInt-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfFloat-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfComplex-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfBoolean-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfHexString-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfHexBytes-8 2.00 ± 0% 2.00 ± 0% ~ (all equal) SprintfBytes-8 2.00 ± 0% 2.00 ± 0% ~ (all equal) SprintfStringer-8 4.00 ± 0% 4.00 ± 0% ~ (all equal) SprintfStructure-8 7.00 ± 0% 7.00 ± 0% ~ (all equal) ManyArgs-8 8.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10) FprintInt-8 1.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10) FprintfBytes-8 1.00 ± 0% 1.00 ± 0% ~ (all equal) FprintIntNoAlloc-8 0.00 0.00 ~ (all equal) ScanInts-8 1.60k ± 0% 1.60k ± 0% ~ (all equal) ScanRecursiveInt-8 1.71k ± 0% 1.71k ± 0% ~ (all equal) ScanRecursiveIntReaderWrapper-8 1.71k ± 0% 1.71k ± 0% ~ (all equal) Change-Id: I7ba72a25fea4140a0ba40a9f443103ed87cc69b5 Reviewed-on: https://go-review.googlesource.com/35554 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2017-01-21 13:41:06 -08:00
if staticuint64s == nil {
staticuint64s = newname(Runtimepkg.Lookup("staticuint64s"))
staticuint64s.SetClass(PEXTERN)
// The actual type is [256]uint64, but we use [256*8]uint8 so we can address
// individual bytes.
staticuint64s.Type = types.NewArray(types.Types[TUINT8], 256*8)
zerobase = newname(Runtimepkg.Lookup("zerobase"))
cmd/compile: move Node.Class to flags Put it at position zero, since it is fairly hot. This shrinks gc.Node into a smaller size class on 64 bit systems. name old time/op new time/op delta Template 193ms ± 5% 192ms ± 3% ~ (p=0.353 n=94+93) Unicode 86.1ms ± 5% 85.0ms ± 4% -1.23% (p=0.000 n=95+98) GoTypes 546ms ± 3% 544ms ± 4% -0.40% (p=0.007 n=94+97) Compiler 2.56s ± 3% 2.54s ± 3% -0.67% (p=0.000 n=99+97) SSA 5.13s ± 2% 5.10s ± 3% -0.55% (p=0.000 n=94+98) Flate 122ms ± 6% 121ms ± 4% -0.75% (p=0.002 n=97+95) GoParser 144ms ± 5% 144ms ± 4% ~ (p=0.298 n=98+97) Reflect 348ms ± 4% 349ms ± 4% ~ (p=0.350 n=98+97) Tar 105ms ± 5% 104ms ± 5% ~ (p=0.154 n=96+98) XML 200ms ± 5% 198ms ± 4% -0.71% (p=0.015 n=97+98) [Geo mean] 330ms 328ms -0.52% name old user-time/op new user-time/op delta Template 229ms ±11% 224ms ± 7% -2.16% (p=0.001 n=100+87) Unicode 109ms ± 5% 109ms ± 6% ~ (p=0.897 n=96+91) GoTypes 712ms ± 4% 709ms ± 4% ~ (p=0.085 n=96+98) Compiler 3.41s ± 3% 3.36s ± 3% -1.43% (p=0.000 n=98+98) SSA 7.46s ± 3% 7.31s ± 3% -2.02% (p=0.000 n=100+99) Flate 145ms ± 6% 143ms ± 6% -1.11% (p=0.001 n=99+97) GoParser 177ms ± 5% 176ms ± 5% -0.78% (p=0.018 n=95+95) Reflect 432ms ± 7% 435ms ± 9% ~ (p=0.296 n=100+100) Tar 121ms ± 7% 121ms ± 5% ~ (p=0.072 n=100+95) XML 241ms ± 4% 239ms ± 5% ~ (p=0.085 n=97+99) [Geo mean] 413ms 410ms -0.73% name old alloc/op new alloc/op delta Template 38.4MB ± 0% 37.7MB ± 0% -1.85% (p=0.008 n=5+5) Unicode 30.1MB ± 0% 28.8MB ± 0% -4.09% (p=0.008 n=5+5) GoTypes 112MB ± 0% 110MB ± 0% -1.69% (p=0.008 n=5+5) Compiler 470MB ± 0% 461MB ± 0% -1.91% (p=0.008 n=5+5) SSA 1.13GB ± 0% 1.11GB ± 0% -1.70% (p=0.008 n=5+5) Flate 25.0MB ± 0% 24.6MB ± 0% -1.67% (p=0.008 n=5+5) GoParser 31.6MB ± 0% 31.1MB ± 0% -1.66% (p=0.008 n=5+5) Reflect 77.1MB ± 0% 75.8MB ± 0% -1.69% (p=0.008 n=5+5) Tar 26.3MB ± 0% 25.7MB ± 0% -2.06% (p=0.008 n=5+5) XML 41.9MB ± 0% 41.1MB ± 0% -1.93% (p=0.008 n=5+5) [Geo mean] 73.5MB 72.0MB -2.03% name old allocs/op new allocs/op delta Template 383k ± 0% 383k ± 0% ~ (p=0.690 n=5+5) Unicode 343k ± 0% 343k ± 0% ~ (p=0.841 n=5+5) GoTypes 1.16M ± 0% 1.16M ± 0% ~ (p=0.310 n=5+5) Compiler 4.43M ± 0% 4.42M ± 0% -0.17% (p=0.008 n=5+5) SSA 9.85M ± 0% 9.85M ± 0% ~ (p=0.310 n=5+5) Flate 236k ± 0% 236k ± 1% ~ (p=0.841 n=5+5) GoParser 320k ± 0% 320k ± 0% ~ (p=0.421 n=5+5) Reflect 988k ± 0% 987k ± 0% ~ (p=0.690 n=5+5) Tar 252k ± 0% 251k ± 0% ~ (p=0.095 n=5+5) XML 399k ± 0% 399k ± 0% ~ (p=1.000 n=5+5) [Geo mean] 741k 740k -0.07% Change-Id: I9e952b58a98e30a12494304db9ce50d0a85e459c Reviewed-on: https://go-review.googlesource.com/41797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com>
2017-04-25 18:14:12 -07:00
zerobase.SetClass(PEXTERN)
zerobase.Type = types.Types[TUINTPTR]
}
// Optimize convT2{E,I} for many cases in which T is not pointer-shaped,
// by using an existing addressable value identical to n.Left
// or creating one on the stack.
cmd/compile: convert constants to interfaces without allocating The order pass is responsible for ensuring that values passed to runtime functions, including convT2E/convT2I, are addressable. Prior to this CL, this was always accomplished by creating a temp, which frequently escaped to the heap, causing allocations, perhaps most notably in code like: fmt.Println(1, 2, 3) // allocates three times None of the runtime routines modify the contents of the pointers they receive, so in the case of constants, instead of creating a temp value, we can create a static value. (Marking the static value as read-only provides protection against accidental attempts by the runtime to modify the constant data.) This improves code generation for code like: panic("abc") c <- 2 // c is a chan int which can now simply refer to "abc" and 2, rather than going by way of a temporary. It also allows us to optimize convT2E/convT2I, by recognizing static readonly values and directly constructing the interface. This CL adds ~0.5% to binary size, despite decreasing the size of many functions, because it also adds many static symbols. This binary size regression could be recovered in future (but currently unplanned) work. There is a lot of content-duplication in these symbols; this statement generates six new symbols, three containing an int 1 and three containing a pointer to the string "a": fmt.Println(1, 1, 1, "a", "a", "a") These symbols could be made content-addressable. Furthermore, these symbols are small, so the alignment and naming overhead is large. As with the go.strings section, these symbols could be hidden and have their alignment reduced. The changes to test/live.go make it impossible (at least with current optimization techniques) to place the values being passed to the runtime in static symbols, preserving autotmp creation. Fixes #18704 Benchmarks from fmt and go-kit's logging package: github.com/go-kit/kit/log name old time/op new time/op delta JSONLoggerSimple-8 1.91µs ± 2% 2.11µs ±22% ~ (p=1.000 n=9+10) JSONLoggerContextual-8 2.60µs ± 6% 2.43µs ± 2% -6.29% (p=0.000 n=9+10) Discard-8 101ns ± 2% 34ns ±14% -66.33% (p=0.000 n=10+9) OneWith-8 161ns ± 1% 102ns ±16% -36.78% (p=0.000 n=10+10) TwoWith-8 175ns ± 3% 106ns ± 7% -39.36% (p=0.000 n=10+9) TenWith-8 293ns ± 3% 227ns ±15% -22.44% (p=0.000 n=9+10) LogfmtLoggerSimple-8 704ns ± 2% 608ns ± 2% -13.65% (p=0.000 n=10+9) LogfmtLoggerContextual-8 962ns ± 1% 860ns ±17% -10.57% (p=0.003 n=9+10) NopLoggerSimple-8 188ns ± 1% 120ns ± 1% -36.39% (p=0.000 n=9+10) NopLoggerContextual-8 379ns ± 1% 243ns ± 0% -35.77% (p=0.000 n=9+10) ValueBindingTimestamp-8 577ns ± 1% 499ns ± 1% -13.51% (p=0.000 n=10+10) ValueBindingCaller-8 898ns ± 2% 844ns ± 2% -6.00% (p=0.000 n=10+10) name old alloc/op new alloc/op delta JSONLoggerSimple-8 904B ± 0% 872B ± 0% -3.54% (p=0.000 n=10+10) JSONLoggerContextual-8 1.20kB ± 0% 1.14kB ± 0% -5.33% (p=0.000 n=10+10) Discard-8 64.0B ± 0% 32.0B ± 0% -50.00% (p=0.000 n=10+10) OneWith-8 96.0B ± 0% 64.0B ± 0% -33.33% (p=0.000 n=10+10) TwoWith-8 160B ± 0% 128B ± 0% -20.00% (p=0.000 n=10+10) TenWith-8 672B ± 0% 640B ± 0% -4.76% (p=0.000 n=10+10) LogfmtLoggerSimple-8 128B ± 0% 96B ± 0% -25.00% (p=0.000 n=10+10) LogfmtLoggerContextual-8 304B ± 0% 240B ± 0% -21.05% (p=0.000 n=10+10) NopLoggerSimple-8 128B ± 0% 96B ± 0% -25.00% (p=0.000 n=10+10) NopLoggerContextual-8 304B ± 0% 240B ± 0% -21.05% (p=0.000 n=10+10) ValueBindingTimestamp-8 159B ± 0% 127B ± 0% -20.13% (p=0.000 n=10+10) ValueBindingCaller-8 112B ± 0% 80B ± 0% -28.57% (p=0.000 n=10+10) name old allocs/op new allocs/op delta JSONLoggerSimple-8 19.0 ± 0% 17.0 ± 0% -10.53% (p=0.000 n=10+10) JSONLoggerContextual-8 25.0 ± 0% 21.0 ± 0% -16.00% (p=0.000 n=10+10) Discard-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) OneWith-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) TwoWith-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) TenWith-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) LogfmtLoggerSimple-8 4.00 ± 0% 2.00 ± 0% -50.00% (p=0.000 n=10+10) LogfmtLoggerContextual-8 7.00 ± 0% 3.00 ± 0% -57.14% (p=0.000 n=10+10) NopLoggerSimple-8 4.00 ± 0% 2.00 ± 0% -50.00% (p=0.000 n=10+10) NopLoggerContextual-8 7.00 ± 0% 3.00 ± 0% -57.14% (p=0.000 n=10+10) ValueBindingTimestamp-8 5.00 ± 0% 3.00 ± 0% -40.00% (p=0.000 n=10+10) ValueBindingCaller-8 4.00 ± 0% 2.00 ± 0% -50.00% (p=0.000 n=10+10) fmt name old time/op new time/op delta SprintfPadding-8 88.9ns ± 3% 79.1ns ± 1% -11.09% (p=0.000 n=10+7) SprintfEmpty-8 12.6ns ± 3% 12.8ns ± 3% ~ (p=0.136 n=10+10) SprintfString-8 38.7ns ± 5% 26.9ns ± 6% -30.65% (p=0.000 n=10+10) SprintfTruncateString-8 56.7ns ± 2% 47.0ns ± 3% -17.05% (p=0.000 n=10+10) SprintfQuoteString-8 164ns ± 2% 153ns ± 2% -7.01% (p=0.000 n=10+10) SprintfInt-8 38.9ns ±15% 26.5ns ± 2% -31.93% (p=0.000 n=10+9) SprintfIntInt-8 60.3ns ± 9% 38.2ns ± 1% -36.67% (p=0.000 n=10+8) SprintfPrefixedInt-8 58.6ns ±13% 51.2ns ±11% -12.66% (p=0.001 n=10+10) SprintfFloat-8 71.4ns ± 3% 64.2ns ± 3% -10.08% (p=0.000 n=8+10) SprintfComplex-8 175ns ± 3% 159ns ± 2% -9.03% (p=0.000 n=10+10) SprintfBoolean-8 33.5ns ± 4% 25.7ns ± 5% -23.28% (p=0.000 n=10+10) SprintfHexString-8 65.3ns ± 3% 51.7ns ± 5% -20.86% (p=0.000 n=10+9) SprintfHexBytes-8 67.2ns ± 5% 67.9ns ± 4% ~ (p=0.383 n=10+10) SprintfBytes-8 129ns ± 7% 124ns ± 7% ~ (p=0.074 n=9+10) SprintfStringer-8 127ns ± 4% 126ns ± 8% ~ (p=0.506 n=9+10) SprintfStructure-8 357ns ± 3% 359ns ± 3% ~ (p=0.469 n=10+10) ManyArgs-8 203ns ± 6% 126ns ± 3% -37.94% (p=0.000 n=10+10) FprintInt-8 119ns ±10% 74ns ± 3% -37.54% (p=0.000 n=10+10) FprintfBytes-8 122ns ± 4% 120ns ± 3% ~ (p=0.124 n=10+10) FprintIntNoAlloc-8 78.2ns ± 5% 74.1ns ± 3% -5.28% (p=0.000 n=10+10) ScanInts-8 349µs ± 1% 349µs ± 0% ~ (p=0.606 n=9+8) ScanRecursiveInt-8 43.8ms ± 7% 40.1ms ± 2% -8.42% (p=0.000 n=10+10) ScanRecursiveIntReaderWrapper-8 43.5ms ± 4% 40.4ms ± 2% -7.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta SprintfPadding-8 24.0B ± 0% 16.0B ± 0% -33.33% (p=0.000 n=10+10) SprintfEmpty-8 0.00B 0.00B ~ (all equal) SprintfString-8 21.0B ± 0% 5.0B ± 0% -76.19% (p=0.000 n=10+10) SprintfTruncateString-8 32.0B ± 0% 16.0B ± 0% -50.00% (p=0.000 n=10+10) SprintfQuoteString-8 48.0B ± 0% 32.0B ± 0% -33.33% (p=0.000 n=10+10) SprintfInt-8 16.0B ± 0% 1.0B ± 0% -93.75% (p=0.000 n=10+10) SprintfIntInt-8 24.0B ± 0% 3.0B ± 0% -87.50% (p=0.000 n=10+10) SprintfPrefixedInt-8 72.0B ± 0% 64.0B ± 0% -11.11% (p=0.000 n=10+10) SprintfFloat-8 16.0B ± 0% 8.0B ± 0% -50.00% (p=0.000 n=10+10) SprintfComplex-8 48.0B ± 0% 32.0B ± 0% -33.33% (p=0.000 n=10+10) SprintfBoolean-8 8.00B ± 0% 4.00B ± 0% -50.00% (p=0.000 n=10+10) SprintfHexString-8 96.0B ± 0% 80.0B ± 0% -16.67% (p=0.000 n=10+10) SprintfHexBytes-8 112B ± 0% 112B ± 0% ~ (all equal) SprintfBytes-8 96.0B ± 0% 96.0B ± 0% ~ (all equal) SprintfStringer-8 32.0B ± 0% 32.0B ± 0% ~ (all equal) SprintfStructure-8 256B ± 0% 256B ± 0% ~ (all equal) ManyArgs-8 80.0B ± 0% 0.0B -100.00% (p=0.000 n=10+10) FprintInt-8 8.00B ± 0% 0.00B -100.00% (p=0.000 n=10+10) FprintfBytes-8 32.0B ± 0% 32.0B ± 0% ~ (all equal) FprintIntNoAlloc-8 0.00B 0.00B ~ (all equal) ScanInts-8 15.2kB ± 0% 15.2kB ± 0% ~ (p=0.248 n=9+10) ScanRecursiveInt-8 21.6kB ± 0% 21.6kB ± 0% ~ (all equal) ScanRecursiveIntReaderWrapper-8 21.7kB ± 0% 21.7kB ± 0% ~ (all equal) name old allocs/op new allocs/op delta SprintfPadding-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfEmpty-8 0.00 0.00 ~ (all equal) SprintfString-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfTruncateString-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfQuoteString-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfInt-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfIntInt-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) SprintfPrefixedInt-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfFloat-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfComplex-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfBoolean-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfHexString-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfHexBytes-8 2.00 ± 0% 2.00 ± 0% ~ (all equal) SprintfBytes-8 2.00 ± 0% 2.00 ± 0% ~ (all equal) SprintfStringer-8 4.00 ± 0% 4.00 ± 0% ~ (all equal) SprintfStructure-8 7.00 ± 0% 7.00 ± 0% ~ (all equal) ManyArgs-8 8.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10) FprintInt-8 1.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10) FprintfBytes-8 1.00 ± 0% 1.00 ± 0% ~ (all equal) FprintIntNoAlloc-8 0.00 0.00 ~ (all equal) ScanInts-8 1.60k ± 0% 1.60k ± 0% ~ (all equal) ScanRecursiveInt-8 1.71k ± 0% 1.71k ± 0% ~ (all equal) ScanRecursiveIntReaderWrapper-8 1.71k ± 0% 1.71k ± 0% ~ (all equal) Change-Id: I7ba72a25fea4140a0ba40a9f443103ed87cc69b5 Reviewed-on: https://go-review.googlesource.com/35554 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2017-01-21 13:41:06 -08:00
var value *Node
switch {
case fromType.Size() == 0:
// n.Left is zero-sized. Use zerobase.
cheapexpr(n.Left, init) // Evaluate n.Left for side-effects. See issue 19246.
value = zerobase
case fromType.IsBoolean() || (fromType.Size() == 1 && fromType.IsInteger()):
// n.Left is a bool/byte. Use staticuint64s[n.Left * 8] on little-endian
// and staticuint64s[n.Left * 8 + 7] on big-endian.
n.Left = cheapexpr(n.Left, init)
// byteindex widens n.Left so that the multiplication doesn't overflow.
index := nod(OLSH, byteindex(n.Left), nodintconst(3))
if thearch.LinkArch.ByteOrder == binary.BigEndian {
index = nod(OADD, index, nodintconst(7))
}
value = nod(OINDEX, staticuint64s, index)
cmd/compile: pack bool fields in Node, Name, Func and Type structs to bitsets This reduces compiler memory usage by up to 4% - see compilebench results below. name old time/op new time/op delta Template 245ms ± 4% 241ms ± 2% -1.88% (p=0.029 n=10+10) Unicode 126ms ± 3% 124ms ± 3% ~ (p=0.105 n=10+10) GoTypes 805ms ± 2% 813ms ± 3% ~ (p=0.515 n=8+10) Compiler 3.95s ± 2% 3.83s ± 1% -2.96% (p=0.000 n=9+10) MakeBash 47.4s ± 4% 46.6s ± 1% -1.59% (p=0.028 n=9+10) name old user-ns/op new user-ns/op delta Template 324M ± 5% 326M ± 3% ~ (p=0.935 n=10+10) Unicode 186M ± 5% 178M ±10% ~ (p=0.067 n=9+10) GoTypes 1.08G ± 7% 1.09G ± 4% ~ (p=0.956 n=10+10) Compiler 5.34G ± 4% 5.31G ± 1% ~ (p=0.501 n=10+8) name old alloc/op new alloc/op delta Template 41.0MB ± 0% 39.8MB ± 0% -3.03% (p=0.000 n=10+10) Unicode 32.3MB ± 0% 31.0MB ± 0% -4.13% (p=0.000 n=10+10) GoTypes 119MB ± 0% 116MB ± 0% -2.39% (p=0.000 n=10+10) Compiler 499MB ± 0% 487MB ± 0% -2.48% (p=0.000 n=10+10) name old allocs/op new allocs/op delta Template 380k ± 1% 379k ± 1% ~ (p=0.436 n=10+10) Unicode 324k ± 1% 324k ± 0% ~ (p=0.853 n=10+10) GoTypes 1.15M ± 0% 1.15M ± 0% ~ (p=0.481 n=10+10) Compiler 4.41M ± 0% 4.41M ± 0% -0.12% (p=0.007 n=10+10) name old text-bytes new text-bytes delta HelloSize 623k ± 0% 623k ± 0% ~ (all equal) CmdGoSize 6.64M ± 0% 6.64M ± 0% ~ (all equal) name old data-bytes new data-bytes delta HelloSize 5.81k ± 0% 5.81k ± 0% ~ (all equal) CmdGoSize 238k ± 0% 238k ± 0% ~ (all equal) name old bss-bytes new bss-bytes delta HelloSize 134k ± 0% 134k ± 0% ~ (all equal) CmdGoSize 152k ± 0% 152k ± 0% ~ (all equal) name old exe-bytes new exe-bytes delta HelloSize 967k ± 0% 967k ± 0% ~ (all equal) CmdGoSize 10.2M ± 0% 10.2M ± 0% ~ (all equal) Change-Id: I1f40af738254892bd6c8ba2eb43390b175753d52 Reviewed-on: https://go-review.googlesource.com/37445 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-02-27 19:56:38 +02:00
value.SetBounded(true)
cmd/compile: move Node.Class to flags Put it at position zero, since it is fairly hot. This shrinks gc.Node into a smaller size class on 64 bit systems. name old time/op new time/op delta Template 193ms ± 5% 192ms ± 3% ~ (p=0.353 n=94+93) Unicode 86.1ms ± 5% 85.0ms ± 4% -1.23% (p=0.000 n=95+98) GoTypes 546ms ± 3% 544ms ± 4% -0.40% (p=0.007 n=94+97) Compiler 2.56s ± 3% 2.54s ± 3% -0.67% (p=0.000 n=99+97) SSA 5.13s ± 2% 5.10s ± 3% -0.55% (p=0.000 n=94+98) Flate 122ms ± 6% 121ms ± 4% -0.75% (p=0.002 n=97+95) GoParser 144ms ± 5% 144ms ± 4% ~ (p=0.298 n=98+97) Reflect 348ms ± 4% 349ms ± 4% ~ (p=0.350 n=98+97) Tar 105ms ± 5% 104ms ± 5% ~ (p=0.154 n=96+98) XML 200ms ± 5% 198ms ± 4% -0.71% (p=0.015 n=97+98) [Geo mean] 330ms 328ms -0.52% name old user-time/op new user-time/op delta Template 229ms ±11% 224ms ± 7% -2.16% (p=0.001 n=100+87) Unicode 109ms ± 5% 109ms ± 6% ~ (p=0.897 n=96+91) GoTypes 712ms ± 4% 709ms ± 4% ~ (p=0.085 n=96+98) Compiler 3.41s ± 3% 3.36s ± 3% -1.43% (p=0.000 n=98+98) SSA 7.46s ± 3% 7.31s ± 3% -2.02% (p=0.000 n=100+99) Flate 145ms ± 6% 143ms ± 6% -1.11% (p=0.001 n=99+97) GoParser 177ms ± 5% 176ms ± 5% -0.78% (p=0.018 n=95+95) Reflect 432ms ± 7% 435ms ± 9% ~ (p=0.296 n=100+100) Tar 121ms ± 7% 121ms ± 5% ~ (p=0.072 n=100+95) XML 241ms ± 4% 239ms ± 5% ~ (p=0.085 n=97+99) [Geo mean] 413ms 410ms -0.73% name old alloc/op new alloc/op delta Template 38.4MB ± 0% 37.7MB ± 0% -1.85% (p=0.008 n=5+5) Unicode 30.1MB ± 0% 28.8MB ± 0% -4.09% (p=0.008 n=5+5) GoTypes 112MB ± 0% 110MB ± 0% -1.69% (p=0.008 n=5+5) Compiler 470MB ± 0% 461MB ± 0% -1.91% (p=0.008 n=5+5) SSA 1.13GB ± 0% 1.11GB ± 0% -1.70% (p=0.008 n=5+5) Flate 25.0MB ± 0% 24.6MB ± 0% -1.67% (p=0.008 n=5+5) GoParser 31.6MB ± 0% 31.1MB ± 0% -1.66% (p=0.008 n=5+5) Reflect 77.1MB ± 0% 75.8MB ± 0% -1.69% (p=0.008 n=5+5) Tar 26.3MB ± 0% 25.7MB ± 0% -2.06% (p=0.008 n=5+5) XML 41.9MB ± 0% 41.1MB ± 0% -1.93% (p=0.008 n=5+5) [Geo mean] 73.5MB 72.0MB -2.03% name old allocs/op new allocs/op delta Template 383k ± 0% 383k ± 0% ~ (p=0.690 n=5+5) Unicode 343k ± 0% 343k ± 0% ~ (p=0.841 n=5+5) GoTypes 1.16M ± 0% 1.16M ± 0% ~ (p=0.310 n=5+5) Compiler 4.43M ± 0% 4.42M ± 0% -0.17% (p=0.008 n=5+5) SSA 9.85M ± 0% 9.85M ± 0% ~ (p=0.310 n=5+5) Flate 236k ± 0% 236k ± 1% ~ (p=0.841 n=5+5) GoParser 320k ± 0% 320k ± 0% ~ (p=0.421 n=5+5) Reflect 988k ± 0% 987k ± 0% ~ (p=0.690 n=5+5) Tar 252k ± 0% 251k ± 0% ~ (p=0.095 n=5+5) XML 399k ± 0% 399k ± 0% ~ (p=1.000 n=5+5) [Geo mean] 741k 740k -0.07% Change-Id: I9e952b58a98e30a12494304db9ce50d0a85e459c Reviewed-on: https://go-review.googlesource.com/41797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com>
2017-04-25 18:14:12 -07:00
case n.Left.Class() == PEXTERN && n.Left.Name != nil && n.Left.Name.Readonly():
// n.Left is a readonly global; use it directly.
value = n.Left
case !fromType.IsInterface() && n.Esc == EscNone && fromType.Width <= 1024:
// n.Left does not escape. Use a stack temporary initialized to n.Left.
value = temp(fromType)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
init.Append(typecheck(nod(OAS, value, n.Left), ctxStmt))
cmd/compile: convert constants to interfaces without allocating The order pass is responsible for ensuring that values passed to runtime functions, including convT2E/convT2I, are addressable. Prior to this CL, this was always accomplished by creating a temp, which frequently escaped to the heap, causing allocations, perhaps most notably in code like: fmt.Println(1, 2, 3) // allocates three times None of the runtime routines modify the contents of the pointers they receive, so in the case of constants, instead of creating a temp value, we can create a static value. (Marking the static value as read-only provides protection against accidental attempts by the runtime to modify the constant data.) This improves code generation for code like: panic("abc") c <- 2 // c is a chan int which can now simply refer to "abc" and 2, rather than going by way of a temporary. It also allows us to optimize convT2E/convT2I, by recognizing static readonly values and directly constructing the interface. This CL adds ~0.5% to binary size, despite decreasing the size of many functions, because it also adds many static symbols. This binary size regression could be recovered in future (but currently unplanned) work. There is a lot of content-duplication in these symbols; this statement generates six new symbols, three containing an int 1 and three containing a pointer to the string "a": fmt.Println(1, 1, 1, "a", "a", "a") These symbols could be made content-addressable. Furthermore, these symbols are small, so the alignment and naming overhead is large. As with the go.strings section, these symbols could be hidden and have their alignment reduced. The changes to test/live.go make it impossible (at least with current optimization techniques) to place the values being passed to the runtime in static symbols, preserving autotmp creation. Fixes #18704 Benchmarks from fmt and go-kit's logging package: github.com/go-kit/kit/log name old time/op new time/op delta JSONLoggerSimple-8 1.91µs ± 2% 2.11µs ±22% ~ (p=1.000 n=9+10) JSONLoggerContextual-8 2.60µs ± 6% 2.43µs ± 2% -6.29% (p=0.000 n=9+10) Discard-8 101ns ± 2% 34ns ±14% -66.33% (p=0.000 n=10+9) OneWith-8 161ns ± 1% 102ns ±16% -36.78% (p=0.000 n=10+10) TwoWith-8 175ns ± 3% 106ns ± 7% -39.36% (p=0.000 n=10+9) TenWith-8 293ns ± 3% 227ns ±15% -22.44% (p=0.000 n=9+10) LogfmtLoggerSimple-8 704ns ± 2% 608ns ± 2% -13.65% (p=0.000 n=10+9) LogfmtLoggerContextual-8 962ns ± 1% 860ns ±17% -10.57% (p=0.003 n=9+10) NopLoggerSimple-8 188ns ± 1% 120ns ± 1% -36.39% (p=0.000 n=9+10) NopLoggerContextual-8 379ns ± 1% 243ns ± 0% -35.77% (p=0.000 n=9+10) ValueBindingTimestamp-8 577ns ± 1% 499ns ± 1% -13.51% (p=0.000 n=10+10) ValueBindingCaller-8 898ns ± 2% 844ns ± 2% -6.00% (p=0.000 n=10+10) name old alloc/op new alloc/op delta JSONLoggerSimple-8 904B ± 0% 872B ± 0% -3.54% (p=0.000 n=10+10) JSONLoggerContextual-8 1.20kB ± 0% 1.14kB ± 0% -5.33% (p=0.000 n=10+10) Discard-8 64.0B ± 0% 32.0B ± 0% -50.00% (p=0.000 n=10+10) OneWith-8 96.0B ± 0% 64.0B ± 0% -33.33% (p=0.000 n=10+10) TwoWith-8 160B ± 0% 128B ± 0% -20.00% (p=0.000 n=10+10) TenWith-8 672B ± 0% 640B ± 0% -4.76% (p=0.000 n=10+10) LogfmtLoggerSimple-8 128B ± 0% 96B ± 0% -25.00% (p=0.000 n=10+10) LogfmtLoggerContextual-8 304B ± 0% 240B ± 0% -21.05% (p=0.000 n=10+10) NopLoggerSimple-8 128B ± 0% 96B ± 0% -25.00% (p=0.000 n=10+10) NopLoggerContextual-8 304B ± 0% 240B ± 0% -21.05% (p=0.000 n=10+10) ValueBindingTimestamp-8 159B ± 0% 127B ± 0% -20.13% (p=0.000 n=10+10) ValueBindingCaller-8 112B ± 0% 80B ± 0% -28.57% (p=0.000 n=10+10) name old allocs/op new allocs/op delta JSONLoggerSimple-8 19.0 ± 0% 17.0 ± 0% -10.53% (p=0.000 n=10+10) JSONLoggerContextual-8 25.0 ± 0% 21.0 ± 0% -16.00% (p=0.000 n=10+10) Discard-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) OneWith-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) TwoWith-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) TenWith-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) LogfmtLoggerSimple-8 4.00 ± 0% 2.00 ± 0% -50.00% (p=0.000 n=10+10) LogfmtLoggerContextual-8 7.00 ± 0% 3.00 ± 0% -57.14% (p=0.000 n=10+10) NopLoggerSimple-8 4.00 ± 0% 2.00 ± 0% -50.00% (p=0.000 n=10+10) NopLoggerContextual-8 7.00 ± 0% 3.00 ± 0% -57.14% (p=0.000 n=10+10) ValueBindingTimestamp-8 5.00 ± 0% 3.00 ± 0% -40.00% (p=0.000 n=10+10) ValueBindingCaller-8 4.00 ± 0% 2.00 ± 0% -50.00% (p=0.000 n=10+10) fmt name old time/op new time/op delta SprintfPadding-8 88.9ns ± 3% 79.1ns ± 1% -11.09% (p=0.000 n=10+7) SprintfEmpty-8 12.6ns ± 3% 12.8ns ± 3% ~ (p=0.136 n=10+10) SprintfString-8 38.7ns ± 5% 26.9ns ± 6% -30.65% (p=0.000 n=10+10) SprintfTruncateString-8 56.7ns ± 2% 47.0ns ± 3% -17.05% (p=0.000 n=10+10) SprintfQuoteString-8 164ns ± 2% 153ns ± 2% -7.01% (p=0.000 n=10+10) SprintfInt-8 38.9ns ±15% 26.5ns ± 2% -31.93% (p=0.000 n=10+9) SprintfIntInt-8 60.3ns ± 9% 38.2ns ± 1% -36.67% (p=0.000 n=10+8) SprintfPrefixedInt-8 58.6ns ±13% 51.2ns ±11% -12.66% (p=0.001 n=10+10) SprintfFloat-8 71.4ns ± 3% 64.2ns ± 3% -10.08% (p=0.000 n=8+10) SprintfComplex-8 175ns ± 3% 159ns ± 2% -9.03% (p=0.000 n=10+10) SprintfBoolean-8 33.5ns ± 4% 25.7ns ± 5% -23.28% (p=0.000 n=10+10) SprintfHexString-8 65.3ns ± 3% 51.7ns ± 5% -20.86% (p=0.000 n=10+9) SprintfHexBytes-8 67.2ns ± 5% 67.9ns ± 4% ~ (p=0.383 n=10+10) SprintfBytes-8 129ns ± 7% 124ns ± 7% ~ (p=0.074 n=9+10) SprintfStringer-8 127ns ± 4% 126ns ± 8% ~ (p=0.506 n=9+10) SprintfStructure-8 357ns ± 3% 359ns ± 3% ~ (p=0.469 n=10+10) ManyArgs-8 203ns ± 6% 126ns ± 3% -37.94% (p=0.000 n=10+10) FprintInt-8 119ns ±10% 74ns ± 3% -37.54% (p=0.000 n=10+10) FprintfBytes-8 122ns ± 4% 120ns ± 3% ~ (p=0.124 n=10+10) FprintIntNoAlloc-8 78.2ns ± 5% 74.1ns ± 3% -5.28% (p=0.000 n=10+10) ScanInts-8 349µs ± 1% 349µs ± 0% ~ (p=0.606 n=9+8) ScanRecursiveInt-8 43.8ms ± 7% 40.1ms ± 2% -8.42% (p=0.000 n=10+10) ScanRecursiveIntReaderWrapper-8 43.5ms ± 4% 40.4ms ± 2% -7.16% (p=0.000 n=10+9) name old alloc/op new alloc/op delta SprintfPadding-8 24.0B ± 0% 16.0B ± 0% -33.33% (p=0.000 n=10+10) SprintfEmpty-8 0.00B 0.00B ~ (all equal) SprintfString-8 21.0B ± 0% 5.0B ± 0% -76.19% (p=0.000 n=10+10) SprintfTruncateString-8 32.0B ± 0% 16.0B ± 0% -50.00% (p=0.000 n=10+10) SprintfQuoteString-8 48.0B ± 0% 32.0B ± 0% -33.33% (p=0.000 n=10+10) SprintfInt-8 16.0B ± 0% 1.0B ± 0% -93.75% (p=0.000 n=10+10) SprintfIntInt-8 24.0B ± 0% 3.0B ± 0% -87.50% (p=0.000 n=10+10) SprintfPrefixedInt-8 72.0B ± 0% 64.0B ± 0% -11.11% (p=0.000 n=10+10) SprintfFloat-8 16.0B ± 0% 8.0B ± 0% -50.00% (p=0.000 n=10+10) SprintfComplex-8 48.0B ± 0% 32.0B ± 0% -33.33% (p=0.000 n=10+10) SprintfBoolean-8 8.00B ± 0% 4.00B ± 0% -50.00% (p=0.000 n=10+10) SprintfHexString-8 96.0B ± 0% 80.0B ± 0% -16.67% (p=0.000 n=10+10) SprintfHexBytes-8 112B ± 0% 112B ± 0% ~ (all equal) SprintfBytes-8 96.0B ± 0% 96.0B ± 0% ~ (all equal) SprintfStringer-8 32.0B ± 0% 32.0B ± 0% ~ (all equal) SprintfStructure-8 256B ± 0% 256B ± 0% ~ (all equal) ManyArgs-8 80.0B ± 0% 0.0B -100.00% (p=0.000 n=10+10) FprintInt-8 8.00B ± 0% 0.00B -100.00% (p=0.000 n=10+10) FprintfBytes-8 32.0B ± 0% 32.0B ± 0% ~ (all equal) FprintIntNoAlloc-8 0.00B 0.00B ~ (all equal) ScanInts-8 15.2kB ± 0% 15.2kB ± 0% ~ (p=0.248 n=9+10) ScanRecursiveInt-8 21.6kB ± 0% 21.6kB ± 0% ~ (all equal) ScanRecursiveIntReaderWrapper-8 21.7kB ± 0% 21.7kB ± 0% ~ (all equal) name old allocs/op new allocs/op delta SprintfPadding-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfEmpty-8 0.00 0.00 ~ (all equal) SprintfString-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfTruncateString-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfQuoteString-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfInt-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfIntInt-8 3.00 ± 0% 1.00 ± 0% -66.67% (p=0.000 n=10+10) SprintfPrefixedInt-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfFloat-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfComplex-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfBoolean-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfHexString-8 2.00 ± 0% 1.00 ± 0% -50.00% (p=0.000 n=10+10) SprintfHexBytes-8 2.00 ± 0% 2.00 ± 0% ~ (all equal) SprintfBytes-8 2.00 ± 0% 2.00 ± 0% ~ (all equal) SprintfStringer-8 4.00 ± 0% 4.00 ± 0% ~ (all equal) SprintfStructure-8 7.00 ± 0% 7.00 ± 0% ~ (all equal) ManyArgs-8 8.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10) FprintInt-8 1.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10) FprintfBytes-8 1.00 ± 0% 1.00 ± 0% ~ (all equal) FprintIntNoAlloc-8 0.00 0.00 ~ (all equal) ScanInts-8 1.60k ± 0% 1.60k ± 0% ~ (all equal) ScanRecursiveInt-8 1.71k ± 0% 1.71k ± 0% ~ (all equal) ScanRecursiveIntReaderWrapper-8 1.71k ± 0% 1.71k ± 0% ~ (all equal) Change-Id: I7ba72a25fea4140a0ba40a9f443103ed87cc69b5 Reviewed-on: https://go-review.googlesource.com/35554 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2017-01-21 13:41:06 -08:00
}
if value != nil {
// Value is identical to n.Left.
// Construct the interface directly: {type/itab, &value}.
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
l := nod(OEFACE, typeword(), typecheck(nod(OADDR, value, nil), ctxExpr))
l.Type = toType
l.SetTypecheck(n.Typecheck())
n = l
break
}
// Implement interface to empty interface conversion.
// tmp = i.itab
// if tmp != nil {
// tmp = tmp.type
// }
// e = iface{tmp, i.data}
if toType.IsEmptyInterface() && fromType.IsInterface() && !fromType.IsEmptyInterface() {
// Evaluate the input interface.
c := temp(fromType)
init.Append(nod(OAS, c, n.Left))
// Get the itab out of the interface.
tmp := temp(types.NewPtr(types.Types[TUINT8]))
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
init.Append(nod(OAS, tmp, typecheck(nod(OITAB, c, nil), ctxExpr)))
// Get the type out of the itab.
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
nif := nod(OIF, typecheck(nod(ONE, tmp, nodnil()), ctxExpr), nil)
nif.Nbody.Set1(nod(OAS, tmp, itabType(tmp)))
init.Append(nif)
// Build the result.
e := nod(OEFACE, tmp, ifaceData(n.Pos, c, types.NewPtr(types.Types[TUINT8])))
e.Type = toType // assign type manually, typecheck doesn't understand OEFACE.
e.SetTypecheck(1)
n = e
break
}
fnname, needsaddr := convFuncName(fromType, toType)
cmd/compile: shrink specialized convT2x call sites convT2E16 and other specialized type-to-interface routines accept a type/itab argument and return a complete interface value. However, we know enough in the routine to do without the type. And the caller can construct the interface value using the type. Doing so shrinks the call sites of ten of the specialized convT2x routines. It also lets us unify the empty and non-empty interface routines. Cuts 12k off cmd/go. name old time/op new time/op delta ConvT2ESmall-8 2.96ns ± 2% 2.34ns ± 4% -21.01% (p=0.000 n=175+189) ConvT2EUintptr-8 3.00ns ± 3% 2.34ns ± 4% -22.02% (p=0.000 n=189+187) ConvT2ELarge-8 21.3ns ± 7% 21.5ns ± 5% +1.02% (p=0.000 n=200+197) ConvT2ISmall-8 2.99ns ± 4% 2.33ns ± 3% -21.95% (p=0.000 n=193+184) ConvT2IUintptr-8 3.02ns ± 3% 2.33ns ± 3% -22.82% (p=0.000 n=198+190) ConvT2ILarge-8 21.7ns ± 5% 22.2ns ± 4% +2.31% (p=0.000 n=199+198) ConvT2Ezero/zero/16-8 2.96ns ± 2% 2.33ns ± 3% -21.11% (p=0.000 n=174+187) ConvT2Ezero/zero/32-8 2.96ns ± 1% 2.35ns ± 4% -20.62% (p=0.000 n=163+193) ConvT2Ezero/zero/64-8 2.99ns ± 2% 2.34ns ± 4% -21.78% (p=0.000 n=183+188) ConvT2Ezero/zero/str-8 3.27ns ± 3% 2.54ns ± 3% -22.32% (p=0.000 n=195+192) ConvT2Ezero/zero/slice-8 3.46ns ± 4% 2.81ns ± 3% -18.96% (p=0.000 n=197+164) ConvT2Ezero/zero/big-8 88.4ns ±20% 90.0ns ±20% +1.84% (p=0.000 n=196+198) ConvT2Ezero/nonzero/16-8 12.6ns ± 3% 12.3ns ± 3% -2.34% (p=0.000 n=167+196) ConvT2Ezero/nonzero/32-8 12.3ns ± 4% 11.9ns ± 3% -2.95% (p=0.000 n=187+193) ConvT2Ezero/nonzero/64-8 14.2ns ± 6% 13.8ns ± 5% -2.94% (p=0.000 n=198+199) ConvT2Ezero/nonzero/str-8 27.2ns ± 5% 26.8ns ± 5% -1.33% (p=0.000 n=200+198) ConvT2Ezero/nonzero/slice-8 33.3ns ± 8% 33.1ns ± 6% -0.82% (p=0.000 n=199+200) ConvT2Ezero/nonzero/big-8 88.8ns ±22% 90.2ns ±18% +1.58% (p=0.000 n=200+199) Neligible toolspeed impact. name old alloc/op new alloc/op delta Template 35.4MB ± 0% 35.3MB ± 0% -0.06% (p=0.008 n=5+5) Unicode 29.1MB ± 0% 29.1MB ± 0% ~ (p=0.310 n=5+5) GoTypes 122MB ± 0% 122MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 514MB ± 0% 513MB ± 0% -0.02% (p=0.008 n=5+5) SSA 1.94GB ± 0% 1.94GB ± 0% -0.01% (p=0.008 n=5+5) Flate 24.2MB ± 0% 24.2MB ± 0% ~ (p=0.548 n=5+5) GoParser 28.5MB ± 0% 28.5MB ± 0% -0.05% (p=0.016 n=5+5) Reflect 86.3MB ± 0% 86.2MB ± 0% -0.02% (p=0.008 n=5+5) Tar 34.9MB ± 0% 34.9MB ± 0% ~ (p=0.095 n=5+5) XML 47.1MB ± 0% 47.1MB ± 0% -0.05% (p=0.008 n=5+5) [Geo mean] 81.0MB 81.0MB -0.03% name old allocs/op new allocs/op delta Template 349k ± 0% 349k ± 0% -0.08% (p=0.008 n=5+5) Unicode 340k ± 0% 340k ± 0% ~ (p=0.111 n=5+5) GoTypes 1.28M ± 0% 1.28M ± 0% -0.09% (p=0.008 n=5+5) Compiler 4.92M ± 0% 4.92M ± 0% -0.08% (p=0.008 n=5+5) SSA 15.3M ± 0% 15.3M ± 0% -0.03% (p=0.008 n=5+5) Flate 233k ± 0% 233k ± 0% ~ (p=0.500 n=5+5) GoParser 292k ± 0% 292k ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.05M ± 0% 1.05M ± 0% -0.02% (p=0.008 n=5+5) Tar 344k ± 0% 343k ± 0% -0.06% (p=0.008 n=5+5) XML 430k ± 0% 429k ± 0% -0.08% (p=0.008 n=5+5) [Geo mean] 809k 809k -0.05% name old object-bytes new object-bytes delta Template 507kB ± 0% 507kB ± 0% -0.04% (p=0.008 n=5+5) Unicode 225kB ± 0% 225kB ± 0% ~ (all equal) GoTypes 1.85MB ± 0% 1.85MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 6.75MB ± 0% 6.75MB ± 0% +0.01% (p=0.008 n=5+5) SSA 21.4MB ± 0% 21.4MB ± 0% -0.02% (p=0.008 n=5+5) Flate 328kB ± 0% 328kB ± 0% -0.03% (p=0.008 n=5+5) GoParser 403kB ± 0% 402kB ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.41MB ± 0% 1.41MB ± 0% -0.03% (p=0.008 n=5+5) Tar 457kB ± 0% 457kB ± 0% -0.05% (p=0.008 n=5+5) XML 601kB ± 0% 600kB ± 0% -0.16% (p=0.008 n=5+5) [Geo mean] 1.05MB 1.04MB -0.05% Change-Id: I677a4108c0ecd32617549294036aa84f9214c4fe Reviewed-on: https://go-review.googlesource.com/c/147360 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Martin Möhrmann <moehrmann@google.com>
2018-10-29 13:54:24 -07:00
if !needsaddr && !fromType.IsInterface() {
cmd/compile: shrink specialized convT2x call sites convT2E16 and other specialized type-to-interface routines accept a type/itab argument and return a complete interface value. However, we know enough in the routine to do without the type. And the caller can construct the interface value using the type. Doing so shrinks the call sites of ten of the specialized convT2x routines. It also lets us unify the empty and non-empty interface routines. Cuts 12k off cmd/go. name old time/op new time/op delta ConvT2ESmall-8 2.96ns ± 2% 2.34ns ± 4% -21.01% (p=0.000 n=175+189) ConvT2EUintptr-8 3.00ns ± 3% 2.34ns ± 4% -22.02% (p=0.000 n=189+187) ConvT2ELarge-8 21.3ns ± 7% 21.5ns ± 5% +1.02% (p=0.000 n=200+197) ConvT2ISmall-8 2.99ns ± 4% 2.33ns ± 3% -21.95% (p=0.000 n=193+184) ConvT2IUintptr-8 3.02ns ± 3% 2.33ns ± 3% -22.82% (p=0.000 n=198+190) ConvT2ILarge-8 21.7ns ± 5% 22.2ns ± 4% +2.31% (p=0.000 n=199+198) ConvT2Ezero/zero/16-8 2.96ns ± 2% 2.33ns ± 3% -21.11% (p=0.000 n=174+187) ConvT2Ezero/zero/32-8 2.96ns ± 1% 2.35ns ± 4% -20.62% (p=0.000 n=163+193) ConvT2Ezero/zero/64-8 2.99ns ± 2% 2.34ns ± 4% -21.78% (p=0.000 n=183+188) ConvT2Ezero/zero/str-8 3.27ns ± 3% 2.54ns ± 3% -22.32% (p=0.000 n=195+192) ConvT2Ezero/zero/slice-8 3.46ns ± 4% 2.81ns ± 3% -18.96% (p=0.000 n=197+164) ConvT2Ezero/zero/big-8 88.4ns ±20% 90.0ns ±20% +1.84% (p=0.000 n=196+198) ConvT2Ezero/nonzero/16-8 12.6ns ± 3% 12.3ns ± 3% -2.34% (p=0.000 n=167+196) ConvT2Ezero/nonzero/32-8 12.3ns ± 4% 11.9ns ± 3% -2.95% (p=0.000 n=187+193) ConvT2Ezero/nonzero/64-8 14.2ns ± 6% 13.8ns ± 5% -2.94% (p=0.000 n=198+199) ConvT2Ezero/nonzero/str-8 27.2ns ± 5% 26.8ns ± 5% -1.33% (p=0.000 n=200+198) ConvT2Ezero/nonzero/slice-8 33.3ns ± 8% 33.1ns ± 6% -0.82% (p=0.000 n=199+200) ConvT2Ezero/nonzero/big-8 88.8ns ±22% 90.2ns ±18% +1.58% (p=0.000 n=200+199) Neligible toolspeed impact. name old alloc/op new alloc/op delta Template 35.4MB ± 0% 35.3MB ± 0% -0.06% (p=0.008 n=5+5) Unicode 29.1MB ± 0% 29.1MB ± 0% ~ (p=0.310 n=5+5) GoTypes 122MB ± 0% 122MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 514MB ± 0% 513MB ± 0% -0.02% (p=0.008 n=5+5) SSA 1.94GB ± 0% 1.94GB ± 0% -0.01% (p=0.008 n=5+5) Flate 24.2MB ± 0% 24.2MB ± 0% ~ (p=0.548 n=5+5) GoParser 28.5MB ± 0% 28.5MB ± 0% -0.05% (p=0.016 n=5+5) Reflect 86.3MB ± 0% 86.2MB ± 0% -0.02% (p=0.008 n=5+5) Tar 34.9MB ± 0% 34.9MB ± 0% ~ (p=0.095 n=5+5) XML 47.1MB ± 0% 47.1MB ± 0% -0.05% (p=0.008 n=5+5) [Geo mean] 81.0MB 81.0MB -0.03% name old allocs/op new allocs/op delta Template 349k ± 0% 349k ± 0% -0.08% (p=0.008 n=5+5) Unicode 340k ± 0% 340k ± 0% ~ (p=0.111 n=5+5) GoTypes 1.28M ± 0% 1.28M ± 0% -0.09% (p=0.008 n=5+5) Compiler 4.92M ± 0% 4.92M ± 0% -0.08% (p=0.008 n=5+5) SSA 15.3M ± 0% 15.3M ± 0% -0.03% (p=0.008 n=5+5) Flate 233k ± 0% 233k ± 0% ~ (p=0.500 n=5+5) GoParser 292k ± 0% 292k ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.05M ± 0% 1.05M ± 0% -0.02% (p=0.008 n=5+5) Tar 344k ± 0% 343k ± 0% -0.06% (p=0.008 n=5+5) XML 430k ± 0% 429k ± 0% -0.08% (p=0.008 n=5+5) [Geo mean] 809k 809k -0.05% name old object-bytes new object-bytes delta Template 507kB ± 0% 507kB ± 0% -0.04% (p=0.008 n=5+5) Unicode 225kB ± 0% 225kB ± 0% ~ (all equal) GoTypes 1.85MB ± 0% 1.85MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 6.75MB ± 0% 6.75MB ± 0% +0.01% (p=0.008 n=5+5) SSA 21.4MB ± 0% 21.4MB ± 0% -0.02% (p=0.008 n=5+5) Flate 328kB ± 0% 328kB ± 0% -0.03% (p=0.008 n=5+5) GoParser 403kB ± 0% 402kB ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.41MB ± 0% 1.41MB ± 0% -0.03% (p=0.008 n=5+5) Tar 457kB ± 0% 457kB ± 0% -0.05% (p=0.008 n=5+5) XML 601kB ± 0% 600kB ± 0% -0.16% (p=0.008 n=5+5) [Geo mean] 1.05MB 1.04MB -0.05% Change-Id: I677a4108c0ecd32617549294036aa84f9214c4fe Reviewed-on: https://go-review.googlesource.com/c/147360 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Martin Möhrmann <moehrmann@google.com>
2018-10-29 13:54:24 -07:00
// Use a specialized conversion routine that only returns a data pointer.
// ptr = convT2X(val)
// e = iface{typ/tab, ptr}
fn := syslook(fnname)
dowidth(fromType)
fn = substArgTypes(fn, fromType)
cmd/compile: shrink specialized convT2x call sites convT2E16 and other specialized type-to-interface routines accept a type/itab argument and return a complete interface value. However, we know enough in the routine to do without the type. And the caller can construct the interface value using the type. Doing so shrinks the call sites of ten of the specialized convT2x routines. It also lets us unify the empty and non-empty interface routines. Cuts 12k off cmd/go. name old time/op new time/op delta ConvT2ESmall-8 2.96ns ± 2% 2.34ns ± 4% -21.01% (p=0.000 n=175+189) ConvT2EUintptr-8 3.00ns ± 3% 2.34ns ± 4% -22.02% (p=0.000 n=189+187) ConvT2ELarge-8 21.3ns ± 7% 21.5ns ± 5% +1.02% (p=0.000 n=200+197) ConvT2ISmall-8 2.99ns ± 4% 2.33ns ± 3% -21.95% (p=0.000 n=193+184) ConvT2IUintptr-8 3.02ns ± 3% 2.33ns ± 3% -22.82% (p=0.000 n=198+190) ConvT2ILarge-8 21.7ns ± 5% 22.2ns ± 4% +2.31% (p=0.000 n=199+198) ConvT2Ezero/zero/16-8 2.96ns ± 2% 2.33ns ± 3% -21.11% (p=0.000 n=174+187) ConvT2Ezero/zero/32-8 2.96ns ± 1% 2.35ns ± 4% -20.62% (p=0.000 n=163+193) ConvT2Ezero/zero/64-8 2.99ns ± 2% 2.34ns ± 4% -21.78% (p=0.000 n=183+188) ConvT2Ezero/zero/str-8 3.27ns ± 3% 2.54ns ± 3% -22.32% (p=0.000 n=195+192) ConvT2Ezero/zero/slice-8 3.46ns ± 4% 2.81ns ± 3% -18.96% (p=0.000 n=197+164) ConvT2Ezero/zero/big-8 88.4ns ±20% 90.0ns ±20% +1.84% (p=0.000 n=196+198) ConvT2Ezero/nonzero/16-8 12.6ns ± 3% 12.3ns ± 3% -2.34% (p=0.000 n=167+196) ConvT2Ezero/nonzero/32-8 12.3ns ± 4% 11.9ns ± 3% -2.95% (p=0.000 n=187+193) ConvT2Ezero/nonzero/64-8 14.2ns ± 6% 13.8ns ± 5% -2.94% (p=0.000 n=198+199) ConvT2Ezero/nonzero/str-8 27.2ns ± 5% 26.8ns ± 5% -1.33% (p=0.000 n=200+198) ConvT2Ezero/nonzero/slice-8 33.3ns ± 8% 33.1ns ± 6% -0.82% (p=0.000 n=199+200) ConvT2Ezero/nonzero/big-8 88.8ns ±22% 90.2ns ±18% +1.58% (p=0.000 n=200+199) Neligible toolspeed impact. name old alloc/op new alloc/op delta Template 35.4MB ± 0% 35.3MB ± 0% -0.06% (p=0.008 n=5+5) Unicode 29.1MB ± 0% 29.1MB ± 0% ~ (p=0.310 n=5+5) GoTypes 122MB ± 0% 122MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 514MB ± 0% 513MB ± 0% -0.02% (p=0.008 n=5+5) SSA 1.94GB ± 0% 1.94GB ± 0% -0.01% (p=0.008 n=5+5) Flate 24.2MB ± 0% 24.2MB ± 0% ~ (p=0.548 n=5+5) GoParser 28.5MB ± 0% 28.5MB ± 0% -0.05% (p=0.016 n=5+5) Reflect 86.3MB ± 0% 86.2MB ± 0% -0.02% (p=0.008 n=5+5) Tar 34.9MB ± 0% 34.9MB ± 0% ~ (p=0.095 n=5+5) XML 47.1MB ± 0% 47.1MB ± 0% -0.05% (p=0.008 n=5+5) [Geo mean] 81.0MB 81.0MB -0.03% name old allocs/op new allocs/op delta Template 349k ± 0% 349k ± 0% -0.08% (p=0.008 n=5+5) Unicode 340k ± 0% 340k ± 0% ~ (p=0.111 n=5+5) GoTypes 1.28M ± 0% 1.28M ± 0% -0.09% (p=0.008 n=5+5) Compiler 4.92M ± 0% 4.92M ± 0% -0.08% (p=0.008 n=5+5) SSA 15.3M ± 0% 15.3M ± 0% -0.03% (p=0.008 n=5+5) Flate 233k ± 0% 233k ± 0% ~ (p=0.500 n=5+5) GoParser 292k ± 0% 292k ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.05M ± 0% 1.05M ± 0% -0.02% (p=0.008 n=5+5) Tar 344k ± 0% 343k ± 0% -0.06% (p=0.008 n=5+5) XML 430k ± 0% 429k ± 0% -0.08% (p=0.008 n=5+5) [Geo mean] 809k 809k -0.05% name old object-bytes new object-bytes delta Template 507kB ± 0% 507kB ± 0% -0.04% (p=0.008 n=5+5) Unicode 225kB ± 0% 225kB ± 0% ~ (all equal) GoTypes 1.85MB ± 0% 1.85MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 6.75MB ± 0% 6.75MB ± 0% +0.01% (p=0.008 n=5+5) SSA 21.4MB ± 0% 21.4MB ± 0% -0.02% (p=0.008 n=5+5) Flate 328kB ± 0% 328kB ± 0% -0.03% (p=0.008 n=5+5) GoParser 403kB ± 0% 402kB ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.41MB ± 0% 1.41MB ± 0% -0.03% (p=0.008 n=5+5) Tar 457kB ± 0% 457kB ± 0% -0.05% (p=0.008 n=5+5) XML 601kB ± 0% 600kB ± 0% -0.16% (p=0.008 n=5+5) [Geo mean] 1.05MB 1.04MB -0.05% Change-Id: I677a4108c0ecd32617549294036aa84f9214c4fe Reviewed-on: https://go-review.googlesource.com/c/147360 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Martin Möhrmann <moehrmann@google.com>
2018-10-29 13:54:24 -07:00
dowidth(fn.Type)
call := nod(OCALL, fn, nil)
call.List.Set1(n.Left)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
call = typecheck(call, ctxExpr)
cmd/compile: shrink specialized convT2x call sites convT2E16 and other specialized type-to-interface routines accept a type/itab argument and return a complete interface value. However, we know enough in the routine to do without the type. And the caller can construct the interface value using the type. Doing so shrinks the call sites of ten of the specialized convT2x routines. It also lets us unify the empty and non-empty interface routines. Cuts 12k off cmd/go. name old time/op new time/op delta ConvT2ESmall-8 2.96ns ± 2% 2.34ns ± 4% -21.01% (p=0.000 n=175+189) ConvT2EUintptr-8 3.00ns ± 3% 2.34ns ± 4% -22.02% (p=0.000 n=189+187) ConvT2ELarge-8 21.3ns ± 7% 21.5ns ± 5% +1.02% (p=0.000 n=200+197) ConvT2ISmall-8 2.99ns ± 4% 2.33ns ± 3% -21.95% (p=0.000 n=193+184) ConvT2IUintptr-8 3.02ns ± 3% 2.33ns ± 3% -22.82% (p=0.000 n=198+190) ConvT2ILarge-8 21.7ns ± 5% 22.2ns ± 4% +2.31% (p=0.000 n=199+198) ConvT2Ezero/zero/16-8 2.96ns ± 2% 2.33ns ± 3% -21.11% (p=0.000 n=174+187) ConvT2Ezero/zero/32-8 2.96ns ± 1% 2.35ns ± 4% -20.62% (p=0.000 n=163+193) ConvT2Ezero/zero/64-8 2.99ns ± 2% 2.34ns ± 4% -21.78% (p=0.000 n=183+188) ConvT2Ezero/zero/str-8 3.27ns ± 3% 2.54ns ± 3% -22.32% (p=0.000 n=195+192) ConvT2Ezero/zero/slice-8 3.46ns ± 4% 2.81ns ± 3% -18.96% (p=0.000 n=197+164) ConvT2Ezero/zero/big-8 88.4ns ±20% 90.0ns ±20% +1.84% (p=0.000 n=196+198) ConvT2Ezero/nonzero/16-8 12.6ns ± 3% 12.3ns ± 3% -2.34% (p=0.000 n=167+196) ConvT2Ezero/nonzero/32-8 12.3ns ± 4% 11.9ns ± 3% -2.95% (p=0.000 n=187+193) ConvT2Ezero/nonzero/64-8 14.2ns ± 6% 13.8ns ± 5% -2.94% (p=0.000 n=198+199) ConvT2Ezero/nonzero/str-8 27.2ns ± 5% 26.8ns ± 5% -1.33% (p=0.000 n=200+198) ConvT2Ezero/nonzero/slice-8 33.3ns ± 8% 33.1ns ± 6% -0.82% (p=0.000 n=199+200) ConvT2Ezero/nonzero/big-8 88.8ns ±22% 90.2ns ±18% +1.58% (p=0.000 n=200+199) Neligible toolspeed impact. name old alloc/op new alloc/op delta Template 35.4MB ± 0% 35.3MB ± 0% -0.06% (p=0.008 n=5+5) Unicode 29.1MB ± 0% 29.1MB ± 0% ~ (p=0.310 n=5+5) GoTypes 122MB ± 0% 122MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 514MB ± 0% 513MB ± 0% -0.02% (p=0.008 n=5+5) SSA 1.94GB ± 0% 1.94GB ± 0% -0.01% (p=0.008 n=5+5) Flate 24.2MB ± 0% 24.2MB ± 0% ~ (p=0.548 n=5+5) GoParser 28.5MB ± 0% 28.5MB ± 0% -0.05% (p=0.016 n=5+5) Reflect 86.3MB ± 0% 86.2MB ± 0% -0.02% (p=0.008 n=5+5) Tar 34.9MB ± 0% 34.9MB ± 0% ~ (p=0.095 n=5+5) XML 47.1MB ± 0% 47.1MB ± 0% -0.05% (p=0.008 n=5+5) [Geo mean] 81.0MB 81.0MB -0.03% name old allocs/op new allocs/op delta Template 349k ± 0% 349k ± 0% -0.08% (p=0.008 n=5+5) Unicode 340k ± 0% 340k ± 0% ~ (p=0.111 n=5+5) GoTypes 1.28M ± 0% 1.28M ± 0% -0.09% (p=0.008 n=5+5) Compiler 4.92M ± 0% 4.92M ± 0% -0.08% (p=0.008 n=5+5) SSA 15.3M ± 0% 15.3M ± 0% -0.03% (p=0.008 n=5+5) Flate 233k ± 0% 233k ± 0% ~ (p=0.500 n=5+5) GoParser 292k ± 0% 292k ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.05M ± 0% 1.05M ± 0% -0.02% (p=0.008 n=5+5) Tar 344k ± 0% 343k ± 0% -0.06% (p=0.008 n=5+5) XML 430k ± 0% 429k ± 0% -0.08% (p=0.008 n=5+5) [Geo mean] 809k 809k -0.05% name old object-bytes new object-bytes delta Template 507kB ± 0% 507kB ± 0% -0.04% (p=0.008 n=5+5) Unicode 225kB ± 0% 225kB ± 0% ~ (all equal) GoTypes 1.85MB ± 0% 1.85MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 6.75MB ± 0% 6.75MB ± 0% +0.01% (p=0.008 n=5+5) SSA 21.4MB ± 0% 21.4MB ± 0% -0.02% (p=0.008 n=5+5) Flate 328kB ± 0% 328kB ± 0% -0.03% (p=0.008 n=5+5) GoParser 403kB ± 0% 402kB ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.41MB ± 0% 1.41MB ± 0% -0.03% (p=0.008 n=5+5) Tar 457kB ± 0% 457kB ± 0% -0.05% (p=0.008 n=5+5) XML 601kB ± 0% 600kB ± 0% -0.16% (p=0.008 n=5+5) [Geo mean] 1.05MB 1.04MB -0.05% Change-Id: I677a4108c0ecd32617549294036aa84f9214c4fe Reviewed-on: https://go-review.googlesource.com/c/147360 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Martin Möhrmann <moehrmann@google.com>
2018-10-29 13:54:24 -07:00
call = walkexpr(call, init)
call = safeexpr(call, init)
e := nod(OEFACE, typeword(), call)
e.Type = toType
cmd/compile: shrink specialized convT2x call sites convT2E16 and other specialized type-to-interface routines accept a type/itab argument and return a complete interface value. However, we know enough in the routine to do without the type. And the caller can construct the interface value using the type. Doing so shrinks the call sites of ten of the specialized convT2x routines. It also lets us unify the empty and non-empty interface routines. Cuts 12k off cmd/go. name old time/op new time/op delta ConvT2ESmall-8 2.96ns ± 2% 2.34ns ± 4% -21.01% (p=0.000 n=175+189) ConvT2EUintptr-8 3.00ns ± 3% 2.34ns ± 4% -22.02% (p=0.000 n=189+187) ConvT2ELarge-8 21.3ns ± 7% 21.5ns ± 5% +1.02% (p=0.000 n=200+197) ConvT2ISmall-8 2.99ns ± 4% 2.33ns ± 3% -21.95% (p=0.000 n=193+184) ConvT2IUintptr-8 3.02ns ± 3% 2.33ns ± 3% -22.82% (p=0.000 n=198+190) ConvT2ILarge-8 21.7ns ± 5% 22.2ns ± 4% +2.31% (p=0.000 n=199+198) ConvT2Ezero/zero/16-8 2.96ns ± 2% 2.33ns ± 3% -21.11% (p=0.000 n=174+187) ConvT2Ezero/zero/32-8 2.96ns ± 1% 2.35ns ± 4% -20.62% (p=0.000 n=163+193) ConvT2Ezero/zero/64-8 2.99ns ± 2% 2.34ns ± 4% -21.78% (p=0.000 n=183+188) ConvT2Ezero/zero/str-8 3.27ns ± 3% 2.54ns ± 3% -22.32% (p=0.000 n=195+192) ConvT2Ezero/zero/slice-8 3.46ns ± 4% 2.81ns ± 3% -18.96% (p=0.000 n=197+164) ConvT2Ezero/zero/big-8 88.4ns ±20% 90.0ns ±20% +1.84% (p=0.000 n=196+198) ConvT2Ezero/nonzero/16-8 12.6ns ± 3% 12.3ns ± 3% -2.34% (p=0.000 n=167+196) ConvT2Ezero/nonzero/32-8 12.3ns ± 4% 11.9ns ± 3% -2.95% (p=0.000 n=187+193) ConvT2Ezero/nonzero/64-8 14.2ns ± 6% 13.8ns ± 5% -2.94% (p=0.000 n=198+199) ConvT2Ezero/nonzero/str-8 27.2ns ± 5% 26.8ns ± 5% -1.33% (p=0.000 n=200+198) ConvT2Ezero/nonzero/slice-8 33.3ns ± 8% 33.1ns ± 6% -0.82% (p=0.000 n=199+200) ConvT2Ezero/nonzero/big-8 88.8ns ±22% 90.2ns ±18% +1.58% (p=0.000 n=200+199) Neligible toolspeed impact. name old alloc/op new alloc/op delta Template 35.4MB ± 0% 35.3MB ± 0% -0.06% (p=0.008 n=5+5) Unicode 29.1MB ± 0% 29.1MB ± 0% ~ (p=0.310 n=5+5) GoTypes 122MB ± 0% 122MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 514MB ± 0% 513MB ± 0% -0.02% (p=0.008 n=5+5) SSA 1.94GB ± 0% 1.94GB ± 0% -0.01% (p=0.008 n=5+5) Flate 24.2MB ± 0% 24.2MB ± 0% ~ (p=0.548 n=5+5) GoParser 28.5MB ± 0% 28.5MB ± 0% -0.05% (p=0.016 n=5+5) Reflect 86.3MB ± 0% 86.2MB ± 0% -0.02% (p=0.008 n=5+5) Tar 34.9MB ± 0% 34.9MB ± 0% ~ (p=0.095 n=5+5) XML 47.1MB ± 0% 47.1MB ± 0% -0.05% (p=0.008 n=5+5) [Geo mean] 81.0MB 81.0MB -0.03% name old allocs/op new allocs/op delta Template 349k ± 0% 349k ± 0% -0.08% (p=0.008 n=5+5) Unicode 340k ± 0% 340k ± 0% ~ (p=0.111 n=5+5) GoTypes 1.28M ± 0% 1.28M ± 0% -0.09% (p=0.008 n=5+5) Compiler 4.92M ± 0% 4.92M ± 0% -0.08% (p=0.008 n=5+5) SSA 15.3M ± 0% 15.3M ± 0% -0.03% (p=0.008 n=5+5) Flate 233k ± 0% 233k ± 0% ~ (p=0.500 n=5+5) GoParser 292k ± 0% 292k ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.05M ± 0% 1.05M ± 0% -0.02% (p=0.008 n=5+5) Tar 344k ± 0% 343k ± 0% -0.06% (p=0.008 n=5+5) XML 430k ± 0% 429k ± 0% -0.08% (p=0.008 n=5+5) [Geo mean] 809k 809k -0.05% name old object-bytes new object-bytes delta Template 507kB ± 0% 507kB ± 0% -0.04% (p=0.008 n=5+5) Unicode 225kB ± 0% 225kB ± 0% ~ (all equal) GoTypes 1.85MB ± 0% 1.85MB ± 0% -0.08% (p=0.008 n=5+5) Compiler 6.75MB ± 0% 6.75MB ± 0% +0.01% (p=0.008 n=5+5) SSA 21.4MB ± 0% 21.4MB ± 0% -0.02% (p=0.008 n=5+5) Flate 328kB ± 0% 328kB ± 0% -0.03% (p=0.008 n=5+5) GoParser 403kB ± 0% 402kB ± 0% -0.06% (p=0.008 n=5+5) Reflect 1.41MB ± 0% 1.41MB ± 0% -0.03% (p=0.008 n=5+5) Tar 457kB ± 0% 457kB ± 0% -0.05% (p=0.008 n=5+5) XML 601kB ± 0% 600kB ± 0% -0.16% (p=0.008 n=5+5) [Geo mean] 1.05MB 1.04MB -0.05% Change-Id: I677a4108c0ecd32617549294036aa84f9214c4fe Reviewed-on: https://go-review.googlesource.com/c/147360 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Martin Möhrmann <moehrmann@google.com>
2018-10-29 13:54:24 -07:00
e.SetTypecheck(1)
n = e
break
}
var tab *Node
if fromType.IsInterface() {
// convI2I
tab = typename(toType)
cmd/compile: optimize remaining convT2I calls See #14874 Updates #6853 This change adds a compiler optimization for non pointer shaped convT2I. Since itab symbols are now emitted by the compiler, the itab address can be passed directly to convT2I instead of passing the iface type and a cache pointer argument. Compilebench results for the 5-commits series ending here: name old time/op new time/op delta Template 336ms ± 4% 344ms ± 4% +2.61% (p=0.027 n=9+8) Unicode 165ms ± 6% 173ms ± 7% +5.11% (p=0.014 n=9+9) GoTypes 1.09s ± 1% 1.06s ± 2% -3.29% (p=0.000 n=9+9) Compiler 5.09s ±10% 4.75s ±10% -6.64% (p=0.011 n=10+10) MakeBash 31.1s ± 5% 30.3s ± 3% ~ (p=0.089 n=10+10) name old text-bytes new text-bytes delta HelloSize 558k ± 0% 558k ± 0% +0.02% (p=0.000 n=10+10) CmdGoSize 6.24M ± 0% 6.11M ± 0% -2.11% (p=0.000 n=10+10) name old data-bytes new data-bytes delta HelloSize 3.66k ± 0% 3.74k ± 0% +2.41% (p=0.000 n=10+10) CmdGoSize 134k ± 0% 162k ± 0% +20.76% (p=0.000 n=10+10) name old bss-bytes new bss-bytes delta HelloSize 126k ± 0% 126k ± 0% ~ (all samples are equal) CmdGoSize 149k ± 0% 146k ± 0% -2.17% (p=0.000 n=10+10) name old exe-bytes new exe-bytes delta HelloSize 924k ± 0% 924k ± 0% +0.05% (p=0.000 n=10+10) CmdGoSize 9.77M ± 0% 9.62M ± 0% -1.47% (p=0.000 n=10+10) Change-Id: Ib230ddc04988824035c32287ae544a965fedd344 Reviewed-on: https://go-review.googlesource.com/20902 Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: David Crawshaw <crawshaw@golang.org> Run-TryBot: Michel Lespinasse <walken@google.com>
2016-03-18 17:21:33 -07:00
} else {
// convT2x
tab = typeword()
}
v := n.Left
if needsaddr {
// Types of large or unknown size are passed by reference.
// Orderexpr arranged for n.Left to be a temporary for all
// the conversions it could see. Comparison of an interface
// with a non-interface, especially in a switch on interface value
// with non-interface cases, is not visible to order.stmt, so we
// have to fall back on allocating a temp here.
if !islvalue(v) {
v = copyexpr(v, v.Type, init)
}
v = nod(OADDR, v, nil)
}
dowidth(fromType)
fn := syslook(fnname)
fn = substArgTypes(fn, fromType, toType)
dowidth(fn.Type)
n = nod(OCALL, fn, nil)
n.List.Set2(tab, v)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
n = typecheck(n, ctxExpr)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n = walkexpr(n, init)
case OCONV, OCONVNOP:
n.Left = walkexpr(n.Left, init)
if n.Op == OCONVNOP && checkPtr(Curfn, 1) {
if n.Type.IsPtr() && n.Left.Type.IsUnsafePtr() { // unsafe.Pointer to *T
n = walkCheckPtrAlignment(n, init, nil)
break
}
if n.Type.IsUnsafePtr() && n.Left.Type.Etype == TUINTPTR { // uintptr to unsafe.Pointer
n = walkCheckPtrArithmetic(n, init)
break
}
}
param, result := rtconvfn(n.Left.Type, n.Type)
if param == Txxx {
break
}
fn := basicnames[param] + "to" + basicnames[result]
n = conv(mkcall(fn, types.Types[result], init, conv(n.Left, types.Types[param])), n.Type)
case OANDNOT:
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
n.Op = OAND
n.SetImplicit(true) // for walkCheckPtrArithmetic
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
n.Right = nod(OBITNOT, n.Right, nil)
n.Right = typecheck(n.Right, ctxExpr)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Right = walkexpr(n.Right, init)
case ODIV, OMOD:
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
n.Right = walkexpr(n.Right, init)
// rewrite complex div into function call.
et := n.Left.Type.Etype
if isComplex[et] && n.Op == ODIV {
t := n.Type
n = mkcall("complex128div", types.Types[TCOMPLEX128], init, conv(n.Left, types.Types[TCOMPLEX128]), conv(n.Right, types.Types[TCOMPLEX128]))
n = conv(n, t)
break
}
// Nothing to do for float divisions.
if isFloat[et] {
break
}
// rewrite 64-bit div and mod on 32-bit architectures.
// TODO: Remove this code once we can introduce
// runtime calls late in SSA processing.
if Widthreg < 8 && (et == TINT64 || et == TUINT64) {
if n.Right.Op == OLITERAL {
// Leave div/mod by constant powers of 2.
// The SSA backend will handle those.
switch et {
case TINT64:
c := n.Right.Int64()
if c < 0 {
c = -c
}
if c != 0 && c&(c-1) == 0 {
break opswitch
}
case TUINT64:
c := uint64(n.Right.Int64())
if c != 0 && c&(c-1) == 0 {
break opswitch
}
}
}
var fn string
if et == TINT64 {
fn = "int64"
} else {
fn = "uint64"
}
if n.Op == ODIV {
fn += "div"
} else {
fn += "mod"
}
n = mkcall(fn, n.Type, init, conv(n.Left, types.Types[et]), conv(n.Right, types.Types[et]))
}
case OINDEX:
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
// save the original node for bounds checking elision.
// If it was a ODIV/OMOD walk might rewrite it.
r := n.Right
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Right = walkexpr(n.Right, init)
// if range of type cannot exceed static array bound,
// disable bounds check.
cmd/compile: pack bool fields in Node, Name, Func and Type structs to bitsets This reduces compiler memory usage by up to 4% - see compilebench results below. name old time/op new time/op delta Template 245ms ± 4% 241ms ± 2% -1.88% (p=0.029 n=10+10) Unicode 126ms ± 3% 124ms ± 3% ~ (p=0.105 n=10+10) GoTypes 805ms ± 2% 813ms ± 3% ~ (p=0.515 n=8+10) Compiler 3.95s ± 2% 3.83s ± 1% -2.96% (p=0.000 n=9+10) MakeBash 47.4s ± 4% 46.6s ± 1% -1.59% (p=0.028 n=9+10) name old user-ns/op new user-ns/op delta Template 324M ± 5% 326M ± 3% ~ (p=0.935 n=10+10) Unicode 186M ± 5% 178M ±10% ~ (p=0.067 n=9+10) GoTypes 1.08G ± 7% 1.09G ± 4% ~ (p=0.956 n=10+10) Compiler 5.34G ± 4% 5.31G ± 1% ~ (p=0.501 n=10+8) name old alloc/op new alloc/op delta Template 41.0MB ± 0% 39.8MB ± 0% -3.03% (p=0.000 n=10+10) Unicode 32.3MB ± 0% 31.0MB ± 0% -4.13% (p=0.000 n=10+10) GoTypes 119MB ± 0% 116MB ± 0% -2.39% (p=0.000 n=10+10) Compiler 499MB ± 0% 487MB ± 0% -2.48% (p=0.000 n=10+10) name old allocs/op new allocs/op delta Template 380k ± 1% 379k ± 1% ~ (p=0.436 n=10+10) Unicode 324k ± 1% 324k ± 0% ~ (p=0.853 n=10+10) GoTypes 1.15M ± 0% 1.15M ± 0% ~ (p=0.481 n=10+10) Compiler 4.41M ± 0% 4.41M ± 0% -0.12% (p=0.007 n=10+10) name old text-bytes new text-bytes delta HelloSize 623k ± 0% 623k ± 0% ~ (all equal) CmdGoSize 6.64M ± 0% 6.64M ± 0% ~ (all equal) name old data-bytes new data-bytes delta HelloSize 5.81k ± 0% 5.81k ± 0% ~ (all equal) CmdGoSize 238k ± 0% 238k ± 0% ~ (all equal) name old bss-bytes new bss-bytes delta HelloSize 134k ± 0% 134k ± 0% ~ (all equal) CmdGoSize 152k ± 0% 152k ± 0% ~ (all equal) name old exe-bytes new exe-bytes delta HelloSize 967k ± 0% 967k ± 0% ~ (all equal) CmdGoSize 10.2M ± 0% 10.2M ± 0% ~ (all equal) Change-Id: I1f40af738254892bd6c8ba2eb43390b175753d52 Reviewed-on: https://go-review.googlesource.com/37445 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-02-27 19:56:38 +02:00
if n.Bounded() {
break
}
t := n.Left.Type
if t != nil && t.IsPtr() {
t = t.Elem()
}
if t.IsArray() {
cmd/compile: pack bool fields in Node, Name, Func and Type structs to bitsets This reduces compiler memory usage by up to 4% - see compilebench results below. name old time/op new time/op delta Template 245ms ± 4% 241ms ± 2% -1.88% (p=0.029 n=10+10) Unicode 126ms ± 3% 124ms ± 3% ~ (p=0.105 n=10+10) GoTypes 805ms ± 2% 813ms ± 3% ~ (p=0.515 n=8+10) Compiler 3.95s ± 2% 3.83s ± 1% -2.96% (p=0.000 n=9+10) MakeBash 47.4s ± 4% 46.6s ± 1% -1.59% (p=0.028 n=9+10) name old user-ns/op new user-ns/op delta Template 324M ± 5% 326M ± 3% ~ (p=0.935 n=10+10) Unicode 186M ± 5% 178M ±10% ~ (p=0.067 n=9+10) GoTypes 1.08G ± 7% 1.09G ± 4% ~ (p=0.956 n=10+10) Compiler 5.34G ± 4% 5.31G ± 1% ~ (p=0.501 n=10+8) name old alloc/op new alloc/op delta Template 41.0MB ± 0% 39.8MB ± 0% -3.03% (p=0.000 n=10+10) Unicode 32.3MB ± 0% 31.0MB ± 0% -4.13% (p=0.000 n=10+10) GoTypes 119MB ± 0% 116MB ± 0% -2.39% (p=0.000 n=10+10) Compiler 499MB ± 0% 487MB ± 0% -2.48% (p=0.000 n=10+10) name old allocs/op new allocs/op delta Template 380k ± 1% 379k ± 1% ~ (p=0.436 n=10+10) Unicode 324k ± 1% 324k ± 0% ~ (p=0.853 n=10+10) GoTypes 1.15M ± 0% 1.15M ± 0% ~ (p=0.481 n=10+10) Compiler 4.41M ± 0% 4.41M ± 0% -0.12% (p=0.007 n=10+10) name old text-bytes new text-bytes delta HelloSize 623k ± 0% 623k ± 0% ~ (all equal) CmdGoSize 6.64M ± 0% 6.64M ± 0% ~ (all equal) name old data-bytes new data-bytes delta HelloSize 5.81k ± 0% 5.81k ± 0% ~ (all equal) CmdGoSize 238k ± 0% 238k ± 0% ~ (all equal) name old bss-bytes new bss-bytes delta HelloSize 134k ± 0% 134k ± 0% ~ (all equal) CmdGoSize 152k ± 0% 152k ± 0% ~ (all equal) name old exe-bytes new exe-bytes delta HelloSize 967k ± 0% 967k ± 0% ~ (all equal) CmdGoSize 10.2M ± 0% 10.2M ± 0% ~ (all equal) Change-Id: I1f40af738254892bd6c8ba2eb43390b175753d52 Reviewed-on: https://go-review.googlesource.com/37445 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-02-27 19:56:38 +02:00
n.SetBounded(bounded(r, t.NumElem()))
if Debug['m'] != 0 && n.Bounded() && !Isconst(n.Right, CTINT) {
Warn("index bounds check elided")
}
cmd/compile: pack bool fields in Node, Name, Func and Type structs to bitsets This reduces compiler memory usage by up to 4% - see compilebench results below. name old time/op new time/op delta Template 245ms ± 4% 241ms ± 2% -1.88% (p=0.029 n=10+10) Unicode 126ms ± 3% 124ms ± 3% ~ (p=0.105 n=10+10) GoTypes 805ms ± 2% 813ms ± 3% ~ (p=0.515 n=8+10) Compiler 3.95s ± 2% 3.83s ± 1% -2.96% (p=0.000 n=9+10) MakeBash 47.4s ± 4% 46.6s ± 1% -1.59% (p=0.028 n=9+10) name old user-ns/op new user-ns/op delta Template 324M ± 5% 326M ± 3% ~ (p=0.935 n=10+10) Unicode 186M ± 5% 178M ±10% ~ (p=0.067 n=9+10) GoTypes 1.08G ± 7% 1.09G ± 4% ~ (p=0.956 n=10+10) Compiler 5.34G ± 4% 5.31G ± 1% ~ (p=0.501 n=10+8) name old alloc/op new alloc/op delta Template 41.0MB ± 0% 39.8MB ± 0% -3.03% (p=0.000 n=10+10) Unicode 32.3MB ± 0% 31.0MB ± 0% -4.13% (p=0.000 n=10+10) GoTypes 119MB ± 0% 116MB ± 0% -2.39% (p=0.000 n=10+10) Compiler 499MB ± 0% 487MB ± 0% -2.48% (p=0.000 n=10+10) name old allocs/op new allocs/op delta Template 380k ± 1% 379k ± 1% ~ (p=0.436 n=10+10) Unicode 324k ± 1% 324k ± 0% ~ (p=0.853 n=10+10) GoTypes 1.15M ± 0% 1.15M ± 0% ~ (p=0.481 n=10+10) Compiler 4.41M ± 0% 4.41M ± 0% -0.12% (p=0.007 n=10+10) name old text-bytes new text-bytes delta HelloSize 623k ± 0% 623k ± 0% ~ (all equal) CmdGoSize 6.64M ± 0% 6.64M ± 0% ~ (all equal) name old data-bytes new data-bytes delta HelloSize 5.81k ± 0% 5.81k ± 0% ~ (all equal) CmdGoSize 238k ± 0% 238k ± 0% ~ (all equal) name old bss-bytes new bss-bytes delta HelloSize 134k ± 0% 134k ± 0% ~ (all equal) CmdGoSize 152k ± 0% 152k ± 0% ~ (all equal) name old exe-bytes new exe-bytes delta HelloSize 967k ± 0% 967k ± 0% ~ (all equal) CmdGoSize 10.2M ± 0% 10.2M ± 0% ~ (all equal) Change-Id: I1f40af738254892bd6c8ba2eb43390b175753d52 Reviewed-on: https://go-review.googlesource.com/37445 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-02-27 19:56:38 +02:00
if smallintconst(n.Right) && !n.Bounded() {
yyerror("index out of bounds")
}
} else if Isconst(n.Left, CTSTR) {
n.SetBounded(bounded(r, int64(len(strlit(n.Left)))))
cmd/compile: pack bool fields in Node, Name, Func and Type structs to bitsets This reduces compiler memory usage by up to 4% - see compilebench results below. name old time/op new time/op delta Template 245ms ± 4% 241ms ± 2% -1.88% (p=0.029 n=10+10) Unicode 126ms ± 3% 124ms ± 3% ~ (p=0.105 n=10+10) GoTypes 805ms ± 2% 813ms ± 3% ~ (p=0.515 n=8+10) Compiler 3.95s ± 2% 3.83s ± 1% -2.96% (p=0.000 n=9+10) MakeBash 47.4s ± 4% 46.6s ± 1% -1.59% (p=0.028 n=9+10) name old user-ns/op new user-ns/op delta Template 324M ± 5% 326M ± 3% ~ (p=0.935 n=10+10) Unicode 186M ± 5% 178M ±10% ~ (p=0.067 n=9+10) GoTypes 1.08G ± 7% 1.09G ± 4% ~ (p=0.956 n=10+10) Compiler 5.34G ± 4% 5.31G ± 1% ~ (p=0.501 n=10+8) name old alloc/op new alloc/op delta Template 41.0MB ± 0% 39.8MB ± 0% -3.03% (p=0.000 n=10+10) Unicode 32.3MB ± 0% 31.0MB ± 0% -4.13% (p=0.000 n=10+10) GoTypes 119MB ± 0% 116MB ± 0% -2.39% (p=0.000 n=10+10) Compiler 499MB ± 0% 487MB ± 0% -2.48% (p=0.000 n=10+10) name old allocs/op new allocs/op delta Template 380k ± 1% 379k ± 1% ~ (p=0.436 n=10+10) Unicode 324k ± 1% 324k ± 0% ~ (p=0.853 n=10+10) GoTypes 1.15M ± 0% 1.15M ± 0% ~ (p=0.481 n=10+10) Compiler 4.41M ± 0% 4.41M ± 0% -0.12% (p=0.007 n=10+10) name old text-bytes new text-bytes delta HelloSize 623k ± 0% 623k ± 0% ~ (all equal) CmdGoSize 6.64M ± 0% 6.64M ± 0% ~ (all equal) name old data-bytes new data-bytes delta HelloSize 5.81k ± 0% 5.81k ± 0% ~ (all equal) CmdGoSize 238k ± 0% 238k ± 0% ~ (all equal) name old bss-bytes new bss-bytes delta HelloSize 134k ± 0% 134k ± 0% ~ (all equal) CmdGoSize 152k ± 0% 152k ± 0% ~ (all equal) name old exe-bytes new exe-bytes delta HelloSize 967k ± 0% 967k ± 0% ~ (all equal) CmdGoSize 10.2M ± 0% 10.2M ± 0% ~ (all equal) Change-Id: I1f40af738254892bd6c8ba2eb43390b175753d52 Reviewed-on: https://go-review.googlesource.com/37445 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-02-27 19:56:38 +02:00
if Debug['m'] != 0 && n.Bounded() && !Isconst(n.Right, CTINT) {
Warn("index bounds check elided")
}
cmd/compile: pack bool fields in Node, Name, Func and Type structs to bitsets This reduces compiler memory usage by up to 4% - see compilebench results below. name old time/op new time/op delta Template 245ms ± 4% 241ms ± 2% -1.88% (p=0.029 n=10+10) Unicode 126ms ± 3% 124ms ± 3% ~ (p=0.105 n=10+10) GoTypes 805ms ± 2% 813ms ± 3% ~ (p=0.515 n=8+10) Compiler 3.95s ± 2% 3.83s ± 1% -2.96% (p=0.000 n=9+10) MakeBash 47.4s ± 4% 46.6s ± 1% -1.59% (p=0.028 n=9+10) name old user-ns/op new user-ns/op delta Template 324M ± 5% 326M ± 3% ~ (p=0.935 n=10+10) Unicode 186M ± 5% 178M ±10% ~ (p=0.067 n=9+10) GoTypes 1.08G ± 7% 1.09G ± 4% ~ (p=0.956 n=10+10) Compiler 5.34G ± 4% 5.31G ± 1% ~ (p=0.501 n=10+8) name old alloc/op new alloc/op delta Template 41.0MB ± 0% 39.8MB ± 0% -3.03% (p=0.000 n=10+10) Unicode 32.3MB ± 0% 31.0MB ± 0% -4.13% (p=0.000 n=10+10) GoTypes 119MB ± 0% 116MB ± 0% -2.39% (p=0.000 n=10+10) Compiler 499MB ± 0% 487MB ± 0% -2.48% (p=0.000 n=10+10) name old allocs/op new allocs/op delta Template 380k ± 1% 379k ± 1% ~ (p=0.436 n=10+10) Unicode 324k ± 1% 324k ± 0% ~ (p=0.853 n=10+10) GoTypes 1.15M ± 0% 1.15M ± 0% ~ (p=0.481 n=10+10) Compiler 4.41M ± 0% 4.41M ± 0% -0.12% (p=0.007 n=10+10) name old text-bytes new text-bytes delta HelloSize 623k ± 0% 623k ± 0% ~ (all equal) CmdGoSize 6.64M ± 0% 6.64M ± 0% ~ (all equal) name old data-bytes new data-bytes delta HelloSize 5.81k ± 0% 5.81k ± 0% ~ (all equal) CmdGoSize 238k ± 0% 238k ± 0% ~ (all equal) name old bss-bytes new bss-bytes delta HelloSize 134k ± 0% 134k ± 0% ~ (all equal) CmdGoSize 152k ± 0% 152k ± 0% ~ (all equal) name old exe-bytes new exe-bytes delta HelloSize 967k ± 0% 967k ± 0% ~ (all equal) CmdGoSize 10.2M ± 0% 10.2M ± 0% ~ (all equal) Change-Id: I1f40af738254892bd6c8ba2eb43390b175753d52 Reviewed-on: https://go-review.googlesource.com/37445 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-02-27 19:56:38 +02:00
if smallintconst(n.Right) && !n.Bounded() {
yyerror("index out of bounds")
}
}
if Isconst(n.Right, CTINT) {
if n.Right.Val().U.(*Mpint).CmpInt64(0) < 0 || n.Right.Val().U.(*Mpint).Cmp(maxintval[TINT]) > 0 {
yyerror("index out of bounds")
}
}
case OINDEXMAP:
// Replace m[k] with *map{access1,assign}(maptype, m, &k)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
n.Right = walkexpr(n.Right, init)
map_ := n.Left
key := n.Right
t := map_.Type
if n.IndexMapLValue() {
// This m[k] expression is on the left-hand side of an assignment.
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
fast := mapfast(t)
if fast == mapslow {
runtime: add mapassign_fast* Add benchmarks for map assignment with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapAssignInt32_255-8 24.7ns ± 3% 17.4ns ± 2% -29.75% (p=0.000 n=10+10) MapAssignInt32_64k-8 45.5ns ± 4% 37.6ns ± 4% -17.18% (p=0.000 n=10+10) MapAssignInt64_255-8 26.0ns ± 3% 17.9ns ± 4% -31.03% (p=0.000 n=10+10) MapAssignInt64_64k-8 46.9ns ± 5% 38.7ns ± 2% -17.53% (p=0.000 n=9+10) MapAssignStr_255-8 47.8ns ± 3% 24.8ns ± 4% -48.01% (p=0.000 n=10+10) MapAssignStr_64k-8 83.0ns ± 3% 51.9ns ± 3% -37.45% (p=0.000 n=10+9) name old time/op new time/op delta BinaryTree17-8 3.11s ±19% 2.78s ± 3% ~ (p=0.095 n=5+5) Fannkuch11-8 3.26s ± 1% 3.21s ± 2% ~ (p=0.056 n=5+5) FmtFprintfEmpty-8 50.3ns ± 1% 50.8ns ± 2% ~ (p=0.246 n=5+5) FmtFprintfString-8 82.7ns ± 4% 80.1ns ± 5% ~ (p=0.238 n=5+5) FmtFprintfInt-8 82.6ns ± 2% 81.9ns ± 3% ~ (p=0.508 n=5+5) FmtFprintfIntInt-8 124ns ± 4% 121ns ± 3% ~ (p=0.111 n=5+5) FmtFprintfPrefixedInt-8 158ns ± 6% 160ns ± 2% ~ (p=0.341 n=5+5) FmtFprintfFloat-8 249ns ± 2% 245ns ± 2% ~ (p=0.095 n=5+5) FmtManyArgs-8 513ns ± 2% 519ns ± 3% ~ (p=0.151 n=5+5) GobDecode-8 7.48ms ±12% 7.11ms ± 2% ~ (p=0.222 n=5+5) GobEncode-8 6.25ms ± 1% 6.03ms ± 2% -3.56% (p=0.008 n=5+5) Gzip-8 252ms ± 4% 252ms ± 4% ~ (p=1.000 n=5+5) Gunzip-8 38.4ms ± 3% 38.6ms ± 2% ~ (p=0.690 n=5+5) HTTPClientServer-8 76.9µs ±41% 66.4µs ± 6% ~ (p=0.310 n=5+5) JSONEncode-8 16.5ms ± 3% 16.7ms ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 54.6ms ± 1% 54.3ms ± 2% ~ (p=0.548 n=5+5) Mandelbrot200-8 4.45ms ± 3% 4.47ms ± 1% ~ (p=0.841 n=5+5) GoParse-8 3.43ms ± 1% 3.32ms ± 2% -3.28% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 88.2ns ± 3% 89.4ns ± 2% ~ (p=0.333 n=5+5) RegexpMatchEasy0_1K-8 205ns ± 1% 206ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchEasy1_32-8 85.1ns ± 1% 85.5ns ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 365ns ± 1% 371ns ± 9% ~ (p=1.000 n=5+5) RegexpMatchMedium_32-8 129ns ± 2% 128ns ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 39.8µs ± 0% 39.7µs ± 4% ~ (p=0.730 n=4+5) RegexpMatchHard_32-8 1.99µs ± 3% 2.05µs ±16% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 59.3µs ± 1% 60.3µs ± 7% ~ (p=1.000 n=5+5) Revcomp-8 1.36s ±63% 0.52s ± 5% ~ (p=0.095 n=5+5) Template-8 62.6ms ±14% 60.5ms ± 5% ~ (p=0.690 n=5+5) TimeParse-8 330ns ± 2% 324ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 350ns ± 3% 340ns ± 1% -2.86% (p=0.008 n=5+5) name old speed new speed delta GobDecode-8 103MB/s ±11% 108MB/s ± 2% ~ (p=0.222 n=5+5) GobEncode-8 123MB/s ± 1% 127MB/s ± 2% +3.71% (p=0.008 n=5+5) Gzip-8 77.1MB/s ± 4% 76.9MB/s ± 3% ~ (p=1.000 n=5+5) Gunzip-8 505MB/s ± 3% 503MB/s ± 2% ~ (p=0.690 n=5+5) JSONEncode-8 118MB/s ± 3% 116MB/s ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 35.5MB/s ± 1% 35.8MB/s ± 2% ~ (p=0.397 n=5+5) GoParse-8 16.9MB/s ± 1% 17.4MB/s ± 2% +3.45% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 363MB/s ± 3% 358MB/s ± 2% ~ (p=0.421 n=5+5) RegexpMatchEasy0_1K-8 4.98GB/s ± 1% 4.97GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 376MB/s ± 1% 375MB/s ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 2.80GB/s ± 1% 2.76GB/s ± 9% ~ (p=0.841 n=5+5) RegexpMatchMedium_32-8 7.73MB/s ± 1% 7.76MB/s ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 25.8MB/s ± 0% 25.8MB/s ± 4% ~ (p=0.651 n=4+5) RegexpMatchHard_32-8 16.1MB/s ± 3% 15.7MB/s ±14% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 17.3MB/s ± 1% 17.0MB/s ± 7% ~ (p=0.984 n=5+5) Revcomp-8 273MB/s ±83% 488MB/s ± 5% ~ (p=0.095 n=5+5) Template-8 31.1MB/s ±13% 32.1MB/s ± 5% ~ (p=0.690 n=5+5) Updates #19495 Change-Id: I116e9a2a4594769318b22d736464de8a98499909 Reviewed-on: https://go-review.googlesource.com/38091 Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-12 14:47:59 -07:00
// standard version takes key by reference.
// order.expr made sure key is addressable.
runtime: add mapassign_fast* Add benchmarks for map assignment with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapAssignInt32_255-8 24.7ns ± 3% 17.4ns ± 2% -29.75% (p=0.000 n=10+10) MapAssignInt32_64k-8 45.5ns ± 4% 37.6ns ± 4% -17.18% (p=0.000 n=10+10) MapAssignInt64_255-8 26.0ns ± 3% 17.9ns ± 4% -31.03% (p=0.000 n=10+10) MapAssignInt64_64k-8 46.9ns ± 5% 38.7ns ± 2% -17.53% (p=0.000 n=9+10) MapAssignStr_255-8 47.8ns ± 3% 24.8ns ± 4% -48.01% (p=0.000 n=10+10) MapAssignStr_64k-8 83.0ns ± 3% 51.9ns ± 3% -37.45% (p=0.000 n=10+9) name old time/op new time/op delta BinaryTree17-8 3.11s ±19% 2.78s ± 3% ~ (p=0.095 n=5+5) Fannkuch11-8 3.26s ± 1% 3.21s ± 2% ~ (p=0.056 n=5+5) FmtFprintfEmpty-8 50.3ns ± 1% 50.8ns ± 2% ~ (p=0.246 n=5+5) FmtFprintfString-8 82.7ns ± 4% 80.1ns ± 5% ~ (p=0.238 n=5+5) FmtFprintfInt-8 82.6ns ± 2% 81.9ns ± 3% ~ (p=0.508 n=5+5) FmtFprintfIntInt-8 124ns ± 4% 121ns ± 3% ~ (p=0.111 n=5+5) FmtFprintfPrefixedInt-8 158ns ± 6% 160ns ± 2% ~ (p=0.341 n=5+5) FmtFprintfFloat-8 249ns ± 2% 245ns ± 2% ~ (p=0.095 n=5+5) FmtManyArgs-8 513ns ± 2% 519ns ± 3% ~ (p=0.151 n=5+5) GobDecode-8 7.48ms ±12% 7.11ms ± 2% ~ (p=0.222 n=5+5) GobEncode-8 6.25ms ± 1% 6.03ms ± 2% -3.56% (p=0.008 n=5+5) Gzip-8 252ms ± 4% 252ms ± 4% ~ (p=1.000 n=5+5) Gunzip-8 38.4ms ± 3% 38.6ms ± 2% ~ (p=0.690 n=5+5) HTTPClientServer-8 76.9µs ±41% 66.4µs ± 6% ~ (p=0.310 n=5+5) JSONEncode-8 16.5ms ± 3% 16.7ms ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 54.6ms ± 1% 54.3ms ± 2% ~ (p=0.548 n=5+5) Mandelbrot200-8 4.45ms ± 3% 4.47ms ± 1% ~ (p=0.841 n=5+5) GoParse-8 3.43ms ± 1% 3.32ms ± 2% -3.28% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 88.2ns ± 3% 89.4ns ± 2% ~ (p=0.333 n=5+5) RegexpMatchEasy0_1K-8 205ns ± 1% 206ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchEasy1_32-8 85.1ns ± 1% 85.5ns ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 365ns ± 1% 371ns ± 9% ~ (p=1.000 n=5+5) RegexpMatchMedium_32-8 129ns ± 2% 128ns ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 39.8µs ± 0% 39.7µs ± 4% ~ (p=0.730 n=4+5) RegexpMatchHard_32-8 1.99µs ± 3% 2.05µs ±16% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 59.3µs ± 1% 60.3µs ± 7% ~ (p=1.000 n=5+5) Revcomp-8 1.36s ±63% 0.52s ± 5% ~ (p=0.095 n=5+5) Template-8 62.6ms ±14% 60.5ms ± 5% ~ (p=0.690 n=5+5) TimeParse-8 330ns ± 2% 324ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 350ns ± 3% 340ns ± 1% -2.86% (p=0.008 n=5+5) name old speed new speed delta GobDecode-8 103MB/s ±11% 108MB/s ± 2% ~ (p=0.222 n=5+5) GobEncode-8 123MB/s ± 1% 127MB/s ± 2% +3.71% (p=0.008 n=5+5) Gzip-8 77.1MB/s ± 4% 76.9MB/s ± 3% ~ (p=1.000 n=5+5) Gunzip-8 505MB/s ± 3% 503MB/s ± 2% ~ (p=0.690 n=5+5) JSONEncode-8 118MB/s ± 3% 116MB/s ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 35.5MB/s ± 1% 35.8MB/s ± 2% ~ (p=0.397 n=5+5) GoParse-8 16.9MB/s ± 1% 17.4MB/s ± 2% +3.45% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 363MB/s ± 3% 358MB/s ± 2% ~ (p=0.421 n=5+5) RegexpMatchEasy0_1K-8 4.98GB/s ± 1% 4.97GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 376MB/s ± 1% 375MB/s ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 2.80GB/s ± 1% 2.76GB/s ± 9% ~ (p=0.841 n=5+5) RegexpMatchMedium_32-8 7.73MB/s ± 1% 7.76MB/s ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 25.8MB/s ± 0% 25.8MB/s ± 4% ~ (p=0.651 n=4+5) RegexpMatchHard_32-8 16.1MB/s ± 3% 15.7MB/s ±14% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 17.3MB/s ± 1% 17.0MB/s ± 7% ~ (p=0.984 n=5+5) Revcomp-8 273MB/s ±83% 488MB/s ± 5% ~ (p=0.095 n=5+5) Template-8 31.1MB/s ±13% 32.1MB/s ± 5% ~ (p=0.690 n=5+5) Updates #19495 Change-Id: I116e9a2a4594769318b22d736464de8a98499909 Reviewed-on: https://go-review.googlesource.com/38091 Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-12 14:47:59 -07:00
key = nod(OADDR, key, nil)
}
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
n = mkcall1(mapfn(mapassign[fast], t), nil, init, typename(t), map_, key)
} else {
// m[k] is not the target of an assignment.
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
fast := mapfast(t)
if fast == mapslow {
// standard version takes key by reference.
// order.expr made sure key is addressable.
key = nod(OADDR, key, nil)
}
if w := t.Elem().Width; w <= zeroValSize {
n = mkcall1(mapfn(mapaccess1[fast], t), types.NewPtr(t.Elem()), init, typename(t), map_, key)
} else {
z := zeroaddr(w)
n = mkcall1(mapfn("mapaccess1_fat", t), types.NewPtr(t.Elem()), init, typename(t), map_, key, z)
}
}
n.Type = types.NewPtr(t.Elem())
n.MarkNonNil() // mapaccess1* and mapassign always return non-nil pointers.
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
n = nod(ODEREF, n, nil)
n.Type = t.Elem()
n.SetTypecheck(1)
case ORECV:
Fatalf("walkexpr ORECV") // should see inside OAS only
case OSLICEHEADER:
n.Left = walkexpr(n.Left, init)
n.List.SetFirst(walkexpr(n.List.First(), init))
n.List.SetSecond(walkexpr(n.List.Second(), init))
case OSLICE, OSLICEARR, OSLICESTR, OSLICE3, OSLICE3ARR:
checkSlice := checkPtr(Curfn, 1) && n.Op == OSLICE3ARR && n.Left.Op == OCONVNOP && n.Left.Left.Type.IsUnsafePtr()
if checkSlice {
n.Left.Left = walkexpr(n.Left.Left, init)
} else {
n.Left = walkexpr(n.Left, init)
}
low, high, max := n.SliceBounds()
low = walkexpr(low, init)
if low != nil && isZero(low) {
// Reduce x[0:j] to x[:j] and x[0:j:k] to x[:j:k].
low = nil
}
high = walkexpr(high, init)
max = walkexpr(max, init)
n.SetSliceBounds(low, high, max)
if checkSlice {
n.Left = walkCheckPtrAlignment(n.Left, init, max)
}
if n.Op.IsSlice3() {
if max != nil && max.Op == OCAP && samesafeexpr(n.Left, max.Left) {
// Reduce x[i:j:cap(x)] to x[i:j].
if n.Op == OSLICE3 {
n.Op = OSLICE
} else {
n.Op = OSLICEARR
}
n = reduceSlice(n)
cmd/internal/gc: optimize slice + write barrier The code generated for a slice x[i:j] or x[i:j:k] computes the entire new slice (base, len, cap) and then uses it as the evaluation of the slice expression. If the slice is part of an update x = x[i:j] or x = x[i:j:k], there are opportunities to avoid computing some of these fields. For x = x[0:i], we know that only the len is changing; base can be ignored completely, and cap can be left unmodified. For x = x[0:i:j], we know that only len and cap are changing; base can be ignored completely. For x = x[i:i], we know that the resulting cap is zero, and we don't adjust the base during a slice producing a zero-cap result, so again base can be ignored completely. No write to base, no write barrier. The old slice code was trying to work at a Go syntax level, mainly because that was how you wrote code just once instead of once per architecture. Now the compiler is factored a bit better and we can implement slice during code generation but still have one copy of the code. So the new code is working at that lower level. (It must, to update only parts of the result.) This CL by itself: name old mean new mean delta BinaryTree17 5.81s × (0.98,1.03) 5.71s × (0.96,1.05) ~ (p=0.101) Fannkuch11 4.35s × (1.00,1.00) 4.39s × (1.00,1.00) +0.79% (p=0.000) FmtFprintfEmpty 86.0ns × (0.94,1.11) 82.6ns × (0.98,1.04) -3.86% (p=0.048) FmtFprintfString 276ns × (0.98,1.04) 273ns × (0.98,1.02) ~ (p=0.235) FmtFprintfInt 274ns × (0.98,1.06) 270ns × (0.99,1.01) ~ (p=0.119) FmtFprintfIntInt 506ns × (0.99,1.01) 475ns × (0.99,1.01) -6.02% (p=0.000) FmtFprintfPrefixedInt 391ns × (0.99,1.01) 393ns × (1.00,1.01) ~ (p=0.139) FmtFprintfFloat 566ns × (0.99,1.01) 574ns × (1.00,1.01) +1.33% (p=0.001) FmtManyArgs 1.91µs × (0.99,1.01) 1.87µs × (0.99,1.02) -1.83% (p=0.000) GobDecode 15.3ms × (0.99,1.02) 15.0ms × (0.98,1.05) -1.84% (p=0.042) GobEncode 11.5ms × (0.97,1.03) 11.4ms × (0.99,1.03) ~ (p=0.152) Gzip 645ms × (0.99,1.01) 647ms × (0.99,1.01) ~ (p=0.265) Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) +0.90% (p=0.000) HTTPClientServer 90.5µs × (0.97,1.04) 88.5µs × (0.99,1.03) -2.27% (p=0.014) JSONEncode 32.0ms × (0.98,1.03) 29.6ms × (0.98,1.01) -7.51% (p=0.000) JSONDecode 114ms × (0.99,1.01) 104ms × (1.00,1.01) -8.60% (p=0.000) Mandelbrot200 6.04ms × (1.00,1.01) 6.02ms × (1.00,1.00) ~ (p=0.057) GoParse 6.47ms × (0.97,1.05) 6.37ms × (0.97,1.04) ~ (p=0.105) RegexpMatchEasy0_32 171ns × (0.93,1.07) 152ns × (0.99,1.01) -11.09% (p=0.000) RegexpMatchEasy0_1K 550ns × (0.98,1.01) 530ns × (1.00,1.00) -3.78% (p=0.000) RegexpMatchEasy1_32 135ns × (0.99,1.02) 134ns × (0.99,1.01) -1.33% (p=0.002) RegexpMatchEasy1_1K 879ns × (1.00,1.01) 865ns × (1.00,1.00) -1.58% (p=0.000) RegexpMatchMedium_32 243ns × (1.00,1.00) 233ns × (1.00,1.00) -4.30% (p=0.000) RegexpMatchMedium_1K 70.3µs × (1.00,1.00) 69.5µs × (1.00,1.00) -1.13% (p=0.000) RegexpMatchHard_32 3.82µs × (1.00,1.01) 3.74µs × (1.00,1.00) -1.95% (p=0.000) RegexpMatchHard_1K 117µs × (1.00,1.00) 115µs × (1.00,1.00) -1.69% (p=0.000) Revcomp 917ms × (0.97,1.04) 920ms × (0.97,1.04) ~ (p=0.786) Template 114ms × (0.99,1.01) 117ms × (0.99,1.01) +2.58% (p=0.000) TimeParse 622ns × (0.99,1.01) 615ns × (0.99,1.00) -1.06% (p=0.000) TimeFormat 665ns × (0.99,1.01) 654ns × (0.99,1.00) -1.70% (p=0.000) This CL and previous CL (append) combined: name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.71s × (0.96,1.05) ~ (p=0.638) Fannkuch11 4.41s × (0.98,1.03) 4.39s × (1.00,1.00) ~ (p=0.474) FmtFprintfEmpty 92.7ns × (0.91,1.16) 82.6ns × (0.98,1.04) -10.89% (p=0.004) FmtFprintfString 281ns × (0.96,1.08) 273ns × (0.98,1.02) ~ (p=0.078) FmtFprintfInt 288ns × (0.97,1.06) 270ns × (0.99,1.01) -6.37% (p=0.000) FmtFprintfIntInt 493ns × (0.97,1.04) 475ns × (0.99,1.01) -3.53% (p=0.002) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 393ns × (1.00,1.01) -7.07% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 574ns × (1.00,1.01) -4.02% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.87µs × (0.99,1.02) ~ (p=0.305) GobDecode 14.8ms × (0.98,1.03) 15.0ms × (0.98,1.05) ~ (p=0.237) GobEncode 12.3ms × (0.98,1.01) 11.4ms × (0.99,1.03) -6.95% (p=0.000) Gzip 656ms × (0.99,1.05) 647ms × (0.99,1.01) ~ (p=0.101) Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) +0.58% (p=0.001) HTTPClientServer 91.2µs × (0.97,1.04) 88.5µs × (0.99,1.03) -3.02% (p=0.003) JSONEncode 32.6ms × (0.97,1.08) 29.6ms × (0.98,1.01) -9.10% (p=0.000) JSONDecode 114ms × (0.97,1.05) 104ms × (1.00,1.01) -8.74% (p=0.000) Mandelbrot200 6.11ms × (0.98,1.04) 6.02ms × (1.00,1.00) ~ (p=0.090) GoParse 6.66ms × (0.97,1.04) 6.37ms × (0.97,1.04) -4.41% (p=0.000) RegexpMatchEasy0_32 159ns × (0.99,1.00) 152ns × (0.99,1.01) -4.69% (p=0.000) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 530ns × (1.00,1.00) -1.57% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 134ns × (0.99,1.01) -2.91% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 865ns × (1.00,1.00) -0.51% (p=0.012) RegexpMatchMedium_32 252ns × (0.99,1.01) 233ns × (1.00,1.00) -7.85% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 69.5µs × (1.00,1.00) -4.43% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.74µs × (1.00,1.00) -2.74% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 115µs × (1.00,1.00) -2.24% (p=0.000) Revcomp 920ms × (0.97,1.07) 920ms × (0.97,1.04) ~ (p=0.998) Template 129ms × (0.98,1.03) 117ms × (0.99,1.01) -9.79% (p=0.000) TimeParse 619ns × (0.99,1.01) 615ns × (0.99,1.00) -0.57% (p=0.011) TimeFormat 661ns × (0.98,1.04) 654ns × (0.99,1.00) ~ (p=0.223) Change-Id: If054d81ab2c71d8d62cf54b5b1fac2af66b387fc Reviewed-on: https://go-review.googlesource.com/9813 Reviewed-by: David Chase <drchase@google.com> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2015-05-06 12:35:53 -04:00
}
} else {
cmd/internal/gc: optimize slice + write barrier The code generated for a slice x[i:j] or x[i:j:k] computes the entire new slice (base, len, cap) and then uses it as the evaluation of the slice expression. If the slice is part of an update x = x[i:j] or x = x[i:j:k], there are opportunities to avoid computing some of these fields. For x = x[0:i], we know that only the len is changing; base can be ignored completely, and cap can be left unmodified. For x = x[0:i:j], we know that only len and cap are changing; base can be ignored completely. For x = x[i:i], we know that the resulting cap is zero, and we don't adjust the base during a slice producing a zero-cap result, so again base can be ignored completely. No write to base, no write barrier. The old slice code was trying to work at a Go syntax level, mainly because that was how you wrote code just once instead of once per architecture. Now the compiler is factored a bit better and we can implement slice during code generation but still have one copy of the code. So the new code is working at that lower level. (It must, to update only parts of the result.) This CL by itself: name old mean new mean delta BinaryTree17 5.81s × (0.98,1.03) 5.71s × (0.96,1.05) ~ (p=0.101) Fannkuch11 4.35s × (1.00,1.00) 4.39s × (1.00,1.00) +0.79% (p=0.000) FmtFprintfEmpty 86.0ns × (0.94,1.11) 82.6ns × (0.98,1.04) -3.86% (p=0.048) FmtFprintfString 276ns × (0.98,1.04) 273ns × (0.98,1.02) ~ (p=0.235) FmtFprintfInt 274ns × (0.98,1.06) 270ns × (0.99,1.01) ~ (p=0.119) FmtFprintfIntInt 506ns × (0.99,1.01) 475ns × (0.99,1.01) -6.02% (p=0.000) FmtFprintfPrefixedInt 391ns × (0.99,1.01) 393ns × (1.00,1.01) ~ (p=0.139) FmtFprintfFloat 566ns × (0.99,1.01) 574ns × (1.00,1.01) +1.33% (p=0.001) FmtManyArgs 1.91µs × (0.99,1.01) 1.87µs × (0.99,1.02) -1.83% (p=0.000) GobDecode 15.3ms × (0.99,1.02) 15.0ms × (0.98,1.05) -1.84% (p=0.042) GobEncode 11.5ms × (0.97,1.03) 11.4ms × (0.99,1.03) ~ (p=0.152) Gzip 645ms × (0.99,1.01) 647ms × (0.99,1.01) ~ (p=0.265) Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) +0.90% (p=0.000) HTTPClientServer 90.5µs × (0.97,1.04) 88.5µs × (0.99,1.03) -2.27% (p=0.014) JSONEncode 32.0ms × (0.98,1.03) 29.6ms × (0.98,1.01) -7.51% (p=0.000) JSONDecode 114ms × (0.99,1.01) 104ms × (1.00,1.01) -8.60% (p=0.000) Mandelbrot200 6.04ms × (1.00,1.01) 6.02ms × (1.00,1.00) ~ (p=0.057) GoParse 6.47ms × (0.97,1.05) 6.37ms × (0.97,1.04) ~ (p=0.105) RegexpMatchEasy0_32 171ns × (0.93,1.07) 152ns × (0.99,1.01) -11.09% (p=0.000) RegexpMatchEasy0_1K 550ns × (0.98,1.01) 530ns × (1.00,1.00) -3.78% (p=0.000) RegexpMatchEasy1_32 135ns × (0.99,1.02) 134ns × (0.99,1.01) -1.33% (p=0.002) RegexpMatchEasy1_1K 879ns × (1.00,1.01) 865ns × (1.00,1.00) -1.58% (p=0.000) RegexpMatchMedium_32 243ns × (1.00,1.00) 233ns × (1.00,1.00) -4.30% (p=0.000) RegexpMatchMedium_1K 70.3µs × (1.00,1.00) 69.5µs × (1.00,1.00) -1.13% (p=0.000) RegexpMatchHard_32 3.82µs × (1.00,1.01) 3.74µs × (1.00,1.00) -1.95% (p=0.000) RegexpMatchHard_1K 117µs × (1.00,1.00) 115µs × (1.00,1.00) -1.69% (p=0.000) Revcomp 917ms × (0.97,1.04) 920ms × (0.97,1.04) ~ (p=0.786) Template 114ms × (0.99,1.01) 117ms × (0.99,1.01) +2.58% (p=0.000) TimeParse 622ns × (0.99,1.01) 615ns × (0.99,1.00) -1.06% (p=0.000) TimeFormat 665ns × (0.99,1.01) 654ns × (0.99,1.00) -1.70% (p=0.000) This CL and previous CL (append) combined: name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.71s × (0.96,1.05) ~ (p=0.638) Fannkuch11 4.41s × (0.98,1.03) 4.39s × (1.00,1.00) ~ (p=0.474) FmtFprintfEmpty 92.7ns × (0.91,1.16) 82.6ns × (0.98,1.04) -10.89% (p=0.004) FmtFprintfString 281ns × (0.96,1.08) 273ns × (0.98,1.02) ~ (p=0.078) FmtFprintfInt 288ns × (0.97,1.06) 270ns × (0.99,1.01) -6.37% (p=0.000) FmtFprintfIntInt 493ns × (0.97,1.04) 475ns × (0.99,1.01) -3.53% (p=0.002) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 393ns × (1.00,1.01) -7.07% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 574ns × (1.00,1.01) -4.02% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.87µs × (0.99,1.02) ~ (p=0.305) GobDecode 14.8ms × (0.98,1.03) 15.0ms × (0.98,1.05) ~ (p=0.237) GobEncode 12.3ms × (0.98,1.01) 11.4ms × (0.99,1.03) -6.95% (p=0.000) Gzip 656ms × (0.99,1.05) 647ms × (0.99,1.01) ~ (p=0.101) Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) +0.58% (p=0.001) HTTPClientServer 91.2µs × (0.97,1.04) 88.5µs × (0.99,1.03) -3.02% (p=0.003) JSONEncode 32.6ms × (0.97,1.08) 29.6ms × (0.98,1.01) -9.10% (p=0.000) JSONDecode 114ms × (0.97,1.05) 104ms × (1.00,1.01) -8.74% (p=0.000) Mandelbrot200 6.11ms × (0.98,1.04) 6.02ms × (1.00,1.00) ~ (p=0.090) GoParse 6.66ms × (0.97,1.04) 6.37ms × (0.97,1.04) -4.41% (p=0.000) RegexpMatchEasy0_32 159ns × (0.99,1.00) 152ns × (0.99,1.01) -4.69% (p=0.000) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 530ns × (1.00,1.00) -1.57% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 134ns × (0.99,1.01) -2.91% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 865ns × (1.00,1.00) -0.51% (p=0.012) RegexpMatchMedium_32 252ns × (0.99,1.01) 233ns × (1.00,1.00) -7.85% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 69.5µs × (1.00,1.00) -4.43% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.74µs × (1.00,1.00) -2.74% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 115µs × (1.00,1.00) -2.24% (p=0.000) Revcomp 920ms × (0.97,1.07) 920ms × (0.97,1.04) ~ (p=0.998) Template 129ms × (0.98,1.03) 117ms × (0.99,1.01) -9.79% (p=0.000) TimeParse 619ns × (0.99,1.01) 615ns × (0.99,1.00) -0.57% (p=0.011) TimeFormat 661ns × (0.98,1.04) 654ns × (0.99,1.00) ~ (p=0.223) Change-Id: If054d81ab2c71d8d62cf54b5b1fac2af66b387fc Reviewed-on: https://go-review.googlesource.com/9813 Reviewed-by: David Chase <drchase@google.com> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2015-05-06 12:35:53 -04:00
n = reduceSlice(n)
}
case ONEW:
if n.Type.Elem().NotInHeap() {
yyerror("%v is go:notinheap; heap allocation disallowed", n.Type.Elem())
}
if n.Esc == EscNone {
if n.Type.Elem().Width >= maxImplicitStackVarSize {
Fatalf("large ONEW with EscNone: %v", n)
}
r := temp(n.Type.Elem())
r = nod(OAS, r, nil) // zero temp
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
r = typecheck(r, ctxStmt)
init.Append(r)
r = nod(OADDR, r.Left, nil)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
r = typecheck(r, ctxExpr)
n = r
} else {
n = callnew(n.Type.Elem())
}
case OADDSTR:
n = addstr(n, init)
case OAPPEND:
cmd/internal/gc: optimize append + write barrier The code generated for x = append(x, v) is roughly: t := x if len(t)+1 > cap(t) { t = grow(t) } t[len(t)] = v len(t)++ x = t We used to generate this code as Go pseudocode during walk. Generate it instead as actual instructions during gen. Doing so lets us apply a few optimizations. The most important is that when, as in the above example, the source slice and the destination slice are the same, the code can instead do: t := x if len(t)+1 > cap(t) { t = grow(t) x = {base(t), len(t)+1, cap(t)} } else { len(x)++ } t[len(t)] = v That is, in the fast path that does not reallocate the array, only the updated length needs to be written back to x, not the array pointer and not the capacity. This is more like what you'd write by hand in C. It's faster in general, since the fast path elides two of the three stores, but it's especially faster when the form of x is such that the base pointer write would turn into a write barrier. No write, no barrier. name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.81s × (0.98,1.03) +2.35% (p=0.023) Fannkuch11 4.41s × (0.98,1.03) 4.35s × (1.00,1.00) ~ (p=0.090) FmtFprintfEmpty 92.7ns × (0.91,1.16) 86.0ns × (0.94,1.11) -7.31% (p=0.038) FmtFprintfString 281ns × (0.96,1.08) 276ns × (0.98,1.04) ~ (p=0.219) FmtFprintfInt 288ns × (0.97,1.06) 274ns × (0.98,1.06) -4.94% (p=0.002) FmtFprintfIntInt 493ns × (0.97,1.04) 506ns × (0.99,1.01) +2.65% (p=0.009) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 391ns × (0.99,1.01) -7.52% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 566ns × (0.99,1.01) -5.27% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.91µs × (0.99,1.01) ~ (p=0.231) GobDecode 14.8ms × (0.98,1.03) 15.3ms × (0.99,1.02) +3.01% (p=0.000) GobEncode 12.3ms × (0.98,1.01) 11.5ms × (0.97,1.03) -5.93% (p=0.000) Gzip 656ms × (0.99,1.05) 645ms × (0.99,1.01) ~ (p=0.055) Gunzip 142ms × (1.00,1.00) 142ms × (1.00,1.00) -0.32% (p=0.034) HTTPClientServer 91.2µs × (0.97,1.04) 90.5µs × (0.97,1.04) ~ (p=0.468) JSONEncode 32.6ms × (0.97,1.08) 32.0ms × (0.98,1.03) ~ (p=0.190) JSONDecode 114ms × (0.97,1.05) 114ms × (0.99,1.01) ~ (p=0.887) Mandelbrot200 6.11ms × (0.98,1.04) 6.04ms × (1.00,1.01) ~ (p=0.167) GoParse 6.66ms × (0.97,1.04) 6.47ms × (0.97,1.05) -2.81% (p=0.014) RegexpMatchEasy0_32 159ns × (0.99,1.00) 171ns × (0.93,1.07) +7.19% (p=0.002) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 550ns × (0.98,1.01) +2.30% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 135ns × (0.99,1.02) -1.60% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 879ns × (1.00,1.01) +1.08% (p=0.000) RegexpMatchMedium_32 252ns × (0.99,1.01) 243ns × (1.00,1.00) -3.71% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 70.3µs × (1.00,1.00) -3.34% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.82µs × (1.00,1.01) -0.81% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 117µs × (1.00,1.00) -0.56% (p=0.000) Revcomp 920ms × (0.97,1.07) 917ms × (0.97,1.04) ~ (p=0.808) Template 129ms × (0.98,1.03) 114ms × (0.99,1.01) -12.06% (p=0.000) TimeParse 619ns × (0.99,1.01) 622ns × (0.99,1.01) ~ (p=0.062) TimeFormat 661ns × (0.98,1.04) 665ns × (0.99,1.01) ~ (p=0.524) See next CL for combination with a similar optimization for slice. The benchmarks that are slower in this CL are still faster overall with the combination of the two. Change-Id: I2a7421658091b2488c64741b4db15ab6c3b4cb7e Reviewed-on: https://go-review.googlesource.com/9812 Reviewed-by: David Chase <drchase@google.com>
2015-05-06 12:34:30 -04:00
// order should make sure we only see OAS(node, OAPPEND), which we handle above.
Fatalf("append outside assignment")
case OCOPY:
n = copyany(n, init, instrumenting && !compiling_runtime)
// cannot use chanfn - closechan takes any, not chan any
case OCLOSE:
fn := syslook("closechan")
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
fn = substArgTypes(fn, n.Left.Type)
n = mkcall1(fn, nil, init, n.Left)
case OMAKECHAN:
cmd/compile: generate makechan calls with int arguments Where possible generate calls to runtime makechan with int arguments during compile time instead of makechan with int64 arguments. This eliminates converting arguments for calls to makechan with int64 arguments for platforms where int64 values do not fit into arguments of type int. A similar optimization for makeslice was introduced in CL golang.org/cl/27851. 386: name old time/op new time/op delta MakeChan/Byte 52.4ns ± 6% 45.0ns ± 1% -14.14% (p=0.000 n=10+10) MakeChan/Int 54.5ns ± 1% 49.1ns ± 1% -9.87% (p=0.000 n=10+10) MakeChan/Ptr 150ns ± 1% 143ns ± 0% -4.38% (p=0.000 n=9+7) MakeChan/Struct/0 49.2ns ± 2% 43.2ns ± 2% -12.27% (p=0.000 n=10+10) MakeChan/Struct/32 81.7ns ± 2% 76.2ns ± 1% -6.71% (p=0.000 n=10+10) MakeChan/Struct/40 88.4ns ± 2% 82.5ns ± 2% -6.60% (p=0.000 n=10+10) AMD64: name old time/op new time/op delta MakeChan/Byte 83.4ns ± 8% 80.8ns ± 3% ~ (p=0.171 n=10+10) MakeChan/Int 101ns ± 3% 101ns ± 2% ~ (p=0.412 n=10+10) MakeChan/Ptr 128ns ± 1% 128ns ± 1% ~ (p=0.191 n=10+10) MakeChan/Struct/0 67.6ns ± 3% 68.7ns ± 4% ~ (p=0.224 n=10+10) MakeChan/Struct/32 138ns ± 1% 139ns ± 1% ~ (p=0.185 n=10+9) MakeChan/Struct/40 154ns ± 1% 154ns ± 1% -0.55% (p=0.027 n=10+9) Change-Id: Ie854cb066007232c5e9f71ea7d6fe27e81a9c050 Reviewed-on: https://go-review.googlesource.com/55140 Run-TryBot: Martin Möhrmann <moehrmann@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
2017-08-13 20:03:02 +02:00
// When size fits into int, use makechan instead of
// makechan64, which is faster and shorter on 32 bit platforms.
size := n.Left
fnname := "makechan64"
argtype := types.Types[TINT64]
// Type checking guarantees that TIDEAL size is positive and fits in an int.
// The case of size overflow when converting TUINT or TUINTPTR to TINT
// will be handled by the negative range checks in makechan during runtime.
if size.Type.IsKind(TIDEAL) || maxintval[size.Type.Etype].Cmp(maxintval[TUINT]) <= 0 {
fnname = "makechan"
argtype = types.Types[TINT]
}
n = mkcall1(chanfn(fnname, 1, n.Type), n.Type, init, typename(n.Type), conv(size, argtype))
case OMAKEMAP:
t := n.Type
hmapType := hmap(t)
hint := n.Left
// var h *hmap
var h *Node
if n.Esc == EscNone {
// Allocate hmap on stack.
// var hv hmap
hv := temp(hmapType)
zero := nod(OAS, hv, nil)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
zero = typecheck(zero, ctxStmt)
init.Append(zero)
// h = &hv
h = nod(OADDR, hv, nil)
// Allocate one bucket pointed to by hmap.buckets on stack if hint
// is not larger than BUCKETSIZE. In case hint is larger than
// BUCKETSIZE runtime.makemap will allocate the buckets on the heap.
// Maximum key and elem size is 128 bytes, larger objects
// are stored with an indirection. So max bucket size is 2048+eps.
if !Isconst(hint, CTINT) ||
hint.Val().U.(*Mpint).CmpInt64(BUCKETSIZE) <= 0 {
// In case hint is larger than BUCKETSIZE runtime.makemap
// will allocate the buckets on the heap, see #20184
//
// if hint <= BUCKETSIZE {
// var bv bmap
// b = &bv
// h.buckets = b
// }
nif := nod(OIF, nod(OLE, hint, nodintconst(BUCKETSIZE)), nil)
nif.SetLikely(true)
// var bv bmap
bv := temp(bmap(t))
zero = nod(OAS, bv, nil)
nif.Nbody.Append(zero)
// b = &bv
b := nod(OADDR, bv, nil)
// h.buckets = b
bsym := hmapType.Field(5).Sym // hmap.buckets see reflect.go:hmap
na := nod(OAS, nodSym(ODOT, h, bsym), b)
nif.Nbody.Append(na)
nif = typecheck(nif, ctxStmt)
nif = walkstmt(nif)
init.Append(nif)
}
}
if Isconst(hint, CTINT) && hint.Val().U.(*Mpint).CmpInt64(BUCKETSIZE) <= 0 {
// Handling make(map[any]any) and
// make(map[any]any, hint) where hint <= BUCKETSIZE
// special allows for faster map initialization and
// improves binary size by using calls with fewer arguments.
// For hint <= BUCKETSIZE overLoadFactor(hint, 0) is false
// and no buckets will be allocated by makemap. Therefore,
// no buckets need to be allocated in this code path.
if n.Esc == EscNone {
// Only need to initialize h.hash0 since
// hmap h has been allocated on the stack already.
// h.hash0 = fastrand()
rand := mkcall("fastrand", types.Types[TUINT32], init)
hashsym := hmapType.Field(4).Sym // hmap.hash0 see reflect.go:hmap
a := nod(OAS, nodSym(ODOT, h, hashsym), rand)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
a = typecheck(a, ctxStmt)
a = walkexpr(a, init)
init.Append(a)
n = convnop(h, t)
} else {
// Call runtime.makehmap to allocate an
// hmap on the heap and initialize hmap's hash0 field.
fn := syslook("makemap_small")
fn = substArgTypes(fn, t.Key(), t.Elem())
n = mkcall1(fn, n.Type, init)
}
} else {
if n.Esc != EscNone {
h = nodnil()
}
// Map initialization with a variable or large hint is
// more complicated. We therefore generate a call to
// runtime.makemap to initialize hmap and allocate the
// map buckets.
// When hint fits into int, use makemap instead of
// makemap64, which is faster and shorter on 32 bit platforms.
fnname := "makemap64"
argtype := types.Types[TINT64]
// Type checking guarantees that TIDEAL hint is positive and fits in an int.
// See checkmake call in TMAP case of OMAKE case in OpSwitch in typecheck1 function.
// The case of hint overflow when converting TUINT or TUINTPTR to TINT
// will be handled by the negative range checks in makemap during runtime.
if hint.Type.IsKind(TIDEAL) || maxintval[hint.Type.Etype].Cmp(maxintval[TUINT]) <= 0 {
fnname = "makemap"
argtype = types.Types[TINT]
}
fn := syslook(fnname)
fn = substArgTypes(fn, hmapType, t.Key(), t.Elem())
n = mkcall1(fn, n.Type, init, typename(n.Type), conv(hint, argtype), h)
}
case OMAKESLICE:
l := n.Left
r := n.Right
if r == nil {
r = safeexpr(l, init)
l = r
}
t := n.Type
if t.Elem().NotInHeap() {
yyerror("%v is go:notinheap; heap allocation disallowed", t.Elem())
}
if n.Esc == EscNone {
if !isSmallMakeSlice(n) {
Fatalf("non-small OMAKESLICE with EscNone: %v", n)
}
// var arr [r]T
// n = arr[:l]
i := indexconst(r)
if i < 0 {
Fatalf("walkexpr: invalid index %v", r)
}
// cap is constrained to [0,2^31) or [0,2^63) depending on whether
// we're in 32-bit or 64-bit systems. So it's safe to do:
//
// if uint64(len) > cap {
// if len < 0 { panicmakeslicelen() }
// panicmakeslicecap()
// }
nif := nod(OIF, nod(OGT, conv(l, types.Types[TUINT64]), nodintconst(i)), nil)
niflen := nod(OIF, nod(OLT, l, nodintconst(0)), nil)
niflen.Nbody.Set1(mkcall("panicmakeslicelen", nil, init))
nif.Nbody.Append(niflen, mkcall("panicmakeslicecap", nil, init))
nif = typecheck(nif, ctxStmt)
init.Append(nif)
t = types.NewArray(t.Elem(), i) // [r]T
var_ := temp(t)
a := nod(OAS, var_, nil) // zero temp
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
a = typecheck(a, ctxStmt)
init.Append(a)
r := nod(OSLICE, var_, nil) // arr[:l]
r.SetSliceBounds(nil, l, nil)
r = conv(r, n.Type) // in case n.Type is named.
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
r = typecheck(r, ctxExpr)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
r = walkexpr(r, init)
n = r
} else {
// n escapes; set up a call to makeslice.
// When len and cap can fit into int, use makeslice instead of
// makeslice64, which is faster and shorter on 32 bit platforms.
len, cap := l, r
fnname := "makeslice64"
argtype := types.Types[TINT64]
// Type checking guarantees that TIDEAL len/cap are positive and fit in an int.
// The case of len or cap overflow when converting TUINT or TUINTPTR to TINT
// will be handled by the negative range checks in makeslice during runtime.
if (len.Type.IsKind(TIDEAL) || maxintval[len.Type.Etype].Cmp(maxintval[TUINT]) <= 0) &&
(cap.Type.IsKind(TIDEAL) || maxintval[cap.Type.Etype].Cmp(maxintval[TUINT]) <= 0) {
fnname = "makeslice"
argtype = types.Types[TINT]
}
m := nod(OSLICEHEADER, nil, nil)
m.Type = t
fn := syslook(fnname)
m.Left = mkcall1(fn, types.Types[TUNSAFEPTR], init, typename(t.Elem()), conv(len, argtype), conv(cap, argtype))
m.Left.MarkNonNil()
m.List.Set2(conv(len, types.Types[TINT]), conv(cap, types.Types[TINT]))
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
m = typecheck(m, ctxExpr)
m = walkexpr(m, init)
n = m
}
cmd/compile: optimize make+copy pattern to avoid memclr match: m = make([]T, x); copy(m, s) for pointer free T and x==len(s) rewrite to: m = mallocgc(x*elemsize(T), nil, false); memmove(&m, &s, x*elemsize(T)) otherwise rewrite to: m = makeslicecopy([]T, x, s) This avoids memclear and shading of pointers in the newly created slice before the copy. With this CL "s" is only be allowed to bev a variable and not a more complex expression. This restriction could be lifted in future versions of this optimization when it can be proven that "s" is not referencing "m". Triggers 450 times during make.bash.. Reduces go binary size by ~8 kbyte. name old time/op new time/op delta MakeSliceCopy/mallocmove/Byte 71.1ns ± 1% 65.8ns ± 0% -7.49% (p=0.000 n=10+9) MakeSliceCopy/mallocmove/Int 71.2ns ± 1% 66.0ns ± 0% -7.27% (p=0.000 n=10+8) MakeSliceCopy/mallocmove/Ptr 104ns ± 4% 99ns ± 1% -5.13% (p=0.000 n=10+10) MakeSliceCopy/makecopy/Byte 70.3ns ± 0% 68.0ns ± 0% -3.22% (p=0.000 n=10+9) MakeSliceCopy/makecopy/Int 70.3ns ± 0% 68.5ns ± 1% -2.59% (p=0.000 n=9+10) MakeSliceCopy/makecopy/Ptr 102ns ± 0% 99ns ± 1% -2.97% (p=0.000 n=9+9) MakeSliceCopy/nilappend/Byte 75.4ns ± 0% 74.9ns ± 2% -0.63% (p=0.015 n=9+9) MakeSliceCopy/nilappend/Int 75.6ns ± 0% 76.4ns ± 3% ~ (p=0.245 n=9+10) MakeSliceCopy/nilappend/Ptr 107ns ± 0% 108ns ± 1% +0.93% (p=0.005 n=9+10) Fixes #26252 Change-Id: Iec553dd1fef6ded16197216a472351c8799a8e71 Reviewed-on: https://go-review.googlesource.com/c/go/+/146719 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Martin Möhrmann <moehrmann@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2018-10-23 13:50:07 +02:00
case OMAKESLICECOPY:
if n.Esc == EscNone {
Fatalf("OMAKESLICECOPY with EscNone: %v", n)
}
t := n.Type
if t.Elem().NotInHeap() {
yyerror("%v is go:notinheap; heap allocation disallowed", t.Elem())
cmd/compile: optimize make+copy pattern to avoid memclr match: m = make([]T, x); copy(m, s) for pointer free T and x==len(s) rewrite to: m = mallocgc(x*elemsize(T), nil, false); memmove(&m, &s, x*elemsize(T)) otherwise rewrite to: m = makeslicecopy([]T, x, s) This avoids memclear and shading of pointers in the newly created slice before the copy. With this CL "s" is only be allowed to bev a variable and not a more complex expression. This restriction could be lifted in future versions of this optimization when it can be proven that "s" is not referencing "m". Triggers 450 times during make.bash.. Reduces go binary size by ~8 kbyte. name old time/op new time/op delta MakeSliceCopy/mallocmove/Byte 71.1ns ± 1% 65.8ns ± 0% -7.49% (p=0.000 n=10+9) MakeSliceCopy/mallocmove/Int 71.2ns ± 1% 66.0ns ± 0% -7.27% (p=0.000 n=10+8) MakeSliceCopy/mallocmove/Ptr 104ns ± 4% 99ns ± 1% -5.13% (p=0.000 n=10+10) MakeSliceCopy/makecopy/Byte 70.3ns ± 0% 68.0ns ± 0% -3.22% (p=0.000 n=10+9) MakeSliceCopy/makecopy/Int 70.3ns ± 0% 68.5ns ± 1% -2.59% (p=0.000 n=9+10) MakeSliceCopy/makecopy/Ptr 102ns ± 0% 99ns ± 1% -2.97% (p=0.000 n=9+9) MakeSliceCopy/nilappend/Byte 75.4ns ± 0% 74.9ns ± 2% -0.63% (p=0.015 n=9+9) MakeSliceCopy/nilappend/Int 75.6ns ± 0% 76.4ns ± 3% ~ (p=0.245 n=9+10) MakeSliceCopy/nilappend/Ptr 107ns ± 0% 108ns ± 1% +0.93% (p=0.005 n=9+10) Fixes #26252 Change-Id: Iec553dd1fef6ded16197216a472351c8799a8e71 Reviewed-on: https://go-review.googlesource.com/c/go/+/146719 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Martin Möhrmann <moehrmann@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2018-10-23 13:50:07 +02:00
}
length := conv(n.Left, types.Types[TINT])
copylen := nod(OLEN, n.Right, nil)
copyptr := nod(OSPTR, n.Right, nil)
if !t.Elem().HasPointers() && n.Bounded() {
cmd/compile: optimize make+copy pattern to avoid memclr match: m = make([]T, x); copy(m, s) for pointer free T and x==len(s) rewrite to: m = mallocgc(x*elemsize(T), nil, false); memmove(&m, &s, x*elemsize(T)) otherwise rewrite to: m = makeslicecopy([]T, x, s) This avoids memclear and shading of pointers in the newly created slice before the copy. With this CL "s" is only be allowed to bev a variable and not a more complex expression. This restriction could be lifted in future versions of this optimization when it can be proven that "s" is not referencing "m". Triggers 450 times during make.bash.. Reduces go binary size by ~8 kbyte. name old time/op new time/op delta MakeSliceCopy/mallocmove/Byte 71.1ns ± 1% 65.8ns ± 0% -7.49% (p=0.000 n=10+9) MakeSliceCopy/mallocmove/Int 71.2ns ± 1% 66.0ns ± 0% -7.27% (p=0.000 n=10+8) MakeSliceCopy/mallocmove/Ptr 104ns ± 4% 99ns ± 1% -5.13% (p=0.000 n=10+10) MakeSliceCopy/makecopy/Byte 70.3ns ± 0% 68.0ns ± 0% -3.22% (p=0.000 n=10+9) MakeSliceCopy/makecopy/Int 70.3ns ± 0% 68.5ns ± 1% -2.59% (p=0.000 n=9+10) MakeSliceCopy/makecopy/Ptr 102ns ± 0% 99ns ± 1% -2.97% (p=0.000 n=9+9) MakeSliceCopy/nilappend/Byte 75.4ns ± 0% 74.9ns ± 2% -0.63% (p=0.015 n=9+9) MakeSliceCopy/nilappend/Int 75.6ns ± 0% 76.4ns ± 3% ~ (p=0.245 n=9+10) MakeSliceCopy/nilappend/Ptr 107ns ± 0% 108ns ± 1% +0.93% (p=0.005 n=9+10) Fixes #26252 Change-Id: Iec553dd1fef6ded16197216a472351c8799a8e71 Reviewed-on: https://go-review.googlesource.com/c/go/+/146719 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Martin Möhrmann <moehrmann@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2018-10-23 13:50:07 +02:00
// When len(to)==len(from) and elements have no pointers:
// replace make+copy with runtime.mallocgc+runtime.memmove.
// We do not check for overflow of len(to)*elem.Width here
// since len(from) is an existing checked slice capacity
// with same elem.Width for the from slice.
size := nod(OMUL, conv(length, types.Types[TUINTPTR]), conv(nodintconst(t.Elem().Width), types.Types[TUINTPTR]))
// instantiate mallocgc(size uintptr, typ *byte, needszero bool) unsafe.Pointer
fn := syslook("mallocgc")
sh := nod(OSLICEHEADER, nil, nil)
sh.Left = mkcall1(fn, types.Types[TUNSAFEPTR], init, size, nodnil(), nodbool(false))
sh.Left.MarkNonNil()
sh.List.Set2(length, length)
sh.Type = t
s := temp(t)
r := typecheck(nod(OAS, s, sh), ctxStmt)
r = walkexpr(r, init)
init.Append(r)
// instantiate memmove(to *any, frm *any, size uintptr)
fn = syslook("memmove")
fn = substArgTypes(fn, t.Elem(), t.Elem())
ncopy := mkcall1(fn, nil, init, nod(OSPTR, s, nil), copyptr, size)
ncopy = typecheck(ncopy, ctxStmt)
ncopy = walkexpr(ncopy, init)
init.Append(ncopy)
n = s
} else { // Replace make+copy with runtime.makeslicecopy.
// instantiate makeslicecopy(typ *byte, tolen int, fromlen int, from unsafe.Pointer) unsafe.Pointer
fn := syslook("makeslicecopy")
s := nod(OSLICEHEADER, nil, nil)
s.Left = mkcall1(fn, types.Types[TUNSAFEPTR], init, typename(t.Elem()), length, copylen, conv(copyptr, types.Types[TUNSAFEPTR]))
s.Left.MarkNonNil()
s.List.Set2(length, length)
s.Type = t
n = typecheck(s, ctxExpr)
n = walkexpr(n, init)
}
case ORUNESTR:
a := nodnil()
if n.Esc == EscNone {
t := types.NewArray(types.Types[TUINT8], 4)
a = nod(OADDR, temp(t), nil)
}
// intstring(*[4]byte, rune)
n = mkcall("intstring", n.Type, init, a, conv(n.Left, types.Types[TINT64]))
case OBYTES2STR, ORUNES2STR:
a := nodnil()
if n.Esc == EscNone {
// Create temporary buffer for string on stack.
t := types.NewArray(types.Types[TUINT8], tmpstringbufsize)
a = nod(OADDR, temp(t), nil)
}
if n.Op == ORUNES2STR {
// slicerunetostring(*[32]byte, []rune) string
n = mkcall("slicerunetostring", n.Type, init, a, n.Left)
} else {
// slicebytetostring(*[32]byte, ptr *byte, n int) string
n.Left = cheapexpr(n.Left, init)
ptr, len := n.Left.slicePtrLen()
n = mkcall("slicebytetostring", n.Type, init, a, ptr, len)
}
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
case OBYTES2STRTMP:
n.Left = walkexpr(n.Left, init)
if !instrumenting {
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
// Let the backend handle OBYTES2STRTMP directly
// to avoid a function call to slicebytetostringtmp.
break
}
// slicebytetostringtmp(ptr *byte, n int) string
n.Left = cheapexpr(n.Left, init)
ptr, len := n.Left.slicePtrLen()
n = mkcall("slicebytetostringtmp", n.Type, init, ptr, len)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
case OSTR2BYTES:
s := n.Left
if Isconst(s, CTSTR) {
sc := strlit(s)
// Allocate a [n]byte of the right size.
t := types.NewArray(types.Types[TUINT8], int64(len(sc)))
var a *Node
if n.Esc == EscNone && len(sc) <= int(maxImplicitStackVarSize) {
a = nod(OADDR, temp(t), nil)
} else {
a = callnew(t)
}
p := temp(t.PtrTo()) // *[n]byte
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
init.Append(typecheck(nod(OAS, p, a), ctxStmt))
// Copy from the static string data to the [n]byte.
if len(sc) > 0 {
as := nod(OAS,
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
nod(ODEREF, p, nil),
nod(ODEREF, convnop(nod(OSPTR, s, nil), t.PtrTo()), nil))
as = typecheck(as, ctxStmt)
as = walkstmt(as)
init.Append(as)
}
// Slice the [n]byte to a []byte.
n.Op = OSLICEARR
n.Left = p
n = walkexpr(n, init)
break
}
a := nodnil()
if n.Esc == EscNone {
// Create temporary buffer for slice on stack.
t := types.NewArray(types.Types[TUINT8], tmpstringbufsize)
a = nod(OADDR, temp(t), nil)
}
// stringtoslicebyte(*32[byte], string) []byte
n = mkcall("stringtoslicebyte", n.Type, init, a, conv(s, types.Types[TSTRING]))
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
case OSTR2BYTESTMP:
// []byte(string) conversion that creates a slice
// referring to the actual string bytes.
// This conversion is handled later by the backend and
// is only for use by internal compiler optimizations
// that know that the slice won't be mutated.
// The only such case today is:
// for i, c := range []byte(string)
n.Left = walkexpr(n.Left, init)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
case OSTR2RUNES:
a := nodnil()
if n.Esc == EscNone {
// Create temporary buffer for slice on stack.
t := types.NewArray(types.Types[TINT32], tmpstringbufsize)
a = nod(OADDR, temp(t), nil)
}
// stringtoslicerune(*[32]rune, string) []rune
n = mkcall("stringtoslicerune", n.Type, init, a, conv(n.Left, types.Types[TSTRING]))
case OARRAYLIT, OSLICELIT, OMAPLIT, OSTRUCTLIT, OPTRLIT:
if isStaticCompositeLiteral(n) && !canSSAType(n.Type) {
// n can be directly represented in the read-only data section.
// Make direct reference to the static data. See issue 12841.
vstat := readonlystaticname(n.Type)
fixedlit(inInitFunction, initKindStatic, n, vstat, init)
n = vstat
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
n = typecheck(n, ctxExpr)
break
}
var_ := temp(n.Type)
anylit(n, var_, init)
n = var_
case OSEND:
n1 := n.Right
n1 = assignconv(n1, n.Left.Type.Elem(), "chan send")
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n1 = walkexpr(n1, init)
n1 = nod(OADDR, n1, nil)
n = mkcall1(chanfn("chansend1", 2, n.Left.Type), nil, init, n.Left, n1)
case OCLOSURE:
n = walkclosure(n, init)
case OCALLPART:
n = walkpartialcall(n, init)
}
// Expressions that are constant at run time but not
// considered const by the language spec are not turned into
// constants until walk. For example, if n is y%1 == 0, the
// walk of y%1 may have replaced it by 0.
// Check whether n with its updated args is itself now a constant.
t := n.Type
evconst(n)
if n.Type != t {
Fatalf("evconst changed Type: %v had type %v, now %v", n, t, n.Type)
}
if n.Op == OLITERAL {
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
n = typecheck(n, ctxExpr)
// Emit string symbol now to avoid emitting
// any concurrently during the backend.
if s, ok := n.Val().U.(string); ok {
_ = stringsym(n.Pos, s)
}
}
updateHasCall(n)
if Debug['w'] != 0 && n != nil {
Dump("after walk expr", n)
}
lineno = lno
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
return n
}
// markTypeUsedInInterface marks that type t is converted to an interface.
// This information is used in the linker in dead method elimination.
func markTypeUsedInInterface(t *types.Type) {
typenamesym(t).Linksym().Set(obj.AttrUsedInIface, true)
}
// rtconvfn returns the parameter and result types that will be used by a
// runtime function to convert from type src to type dst. The runtime function
// name can be derived from the names of the returned types.
//
// If no such function is necessary, it returns (Txxx, Txxx).
func rtconvfn(src, dst *types.Type) (param, result types.EType) {
if thearch.SoftFloat {
return Txxx, Txxx
}
switch thearch.LinkArch.Family {
case sys.ARM, sys.MIPS:
if src.IsFloat() {
switch dst.Etype {
case TINT64, TUINT64:
return TFLOAT64, dst.Etype
}
}
if dst.IsFloat() {
switch src.Etype {
case TINT64, TUINT64:
return src.Etype, TFLOAT64
}
}
case sys.I386:
if src.IsFloat() {
switch dst.Etype {
case TINT64, TUINT64:
return TFLOAT64, dst.Etype
case TUINT32, TUINT, TUINTPTR:
return TFLOAT64, TUINT32
}
}
if dst.IsFloat() {
switch src.Etype {
case TINT64, TUINT64:
return src.Etype, TFLOAT64
case TUINT32, TUINT, TUINTPTR:
return TUINT32, TFLOAT64
}
}
}
return Txxx, Txxx
}
// TODO(josharian): combine this with its caller and simplify
cmd/internal/gc: optimize slice + write barrier The code generated for a slice x[i:j] or x[i:j:k] computes the entire new slice (base, len, cap) and then uses it as the evaluation of the slice expression. If the slice is part of an update x = x[i:j] or x = x[i:j:k], there are opportunities to avoid computing some of these fields. For x = x[0:i], we know that only the len is changing; base can be ignored completely, and cap can be left unmodified. For x = x[0:i:j], we know that only len and cap are changing; base can be ignored completely. For x = x[i:i], we know that the resulting cap is zero, and we don't adjust the base during a slice producing a zero-cap result, so again base can be ignored completely. No write to base, no write barrier. The old slice code was trying to work at a Go syntax level, mainly because that was how you wrote code just once instead of once per architecture. Now the compiler is factored a bit better and we can implement slice during code generation but still have one copy of the code. So the new code is working at that lower level. (It must, to update only parts of the result.) This CL by itself: name old mean new mean delta BinaryTree17 5.81s × (0.98,1.03) 5.71s × (0.96,1.05) ~ (p=0.101) Fannkuch11 4.35s × (1.00,1.00) 4.39s × (1.00,1.00) +0.79% (p=0.000) FmtFprintfEmpty 86.0ns × (0.94,1.11) 82.6ns × (0.98,1.04) -3.86% (p=0.048) FmtFprintfString 276ns × (0.98,1.04) 273ns × (0.98,1.02) ~ (p=0.235) FmtFprintfInt 274ns × (0.98,1.06) 270ns × (0.99,1.01) ~ (p=0.119) FmtFprintfIntInt 506ns × (0.99,1.01) 475ns × (0.99,1.01) -6.02% (p=0.000) FmtFprintfPrefixedInt 391ns × (0.99,1.01) 393ns × (1.00,1.01) ~ (p=0.139) FmtFprintfFloat 566ns × (0.99,1.01) 574ns × (1.00,1.01) +1.33% (p=0.001) FmtManyArgs 1.91µs × (0.99,1.01) 1.87µs × (0.99,1.02) -1.83% (p=0.000) GobDecode 15.3ms × (0.99,1.02) 15.0ms × (0.98,1.05) -1.84% (p=0.042) GobEncode 11.5ms × (0.97,1.03) 11.4ms × (0.99,1.03) ~ (p=0.152) Gzip 645ms × (0.99,1.01) 647ms × (0.99,1.01) ~ (p=0.265) Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) +0.90% (p=0.000) HTTPClientServer 90.5µs × (0.97,1.04) 88.5µs × (0.99,1.03) -2.27% (p=0.014) JSONEncode 32.0ms × (0.98,1.03) 29.6ms × (0.98,1.01) -7.51% (p=0.000) JSONDecode 114ms × (0.99,1.01) 104ms × (1.00,1.01) -8.60% (p=0.000) Mandelbrot200 6.04ms × (1.00,1.01) 6.02ms × (1.00,1.00) ~ (p=0.057) GoParse 6.47ms × (0.97,1.05) 6.37ms × (0.97,1.04) ~ (p=0.105) RegexpMatchEasy0_32 171ns × (0.93,1.07) 152ns × (0.99,1.01) -11.09% (p=0.000) RegexpMatchEasy0_1K 550ns × (0.98,1.01) 530ns × (1.00,1.00) -3.78% (p=0.000) RegexpMatchEasy1_32 135ns × (0.99,1.02) 134ns × (0.99,1.01) -1.33% (p=0.002) RegexpMatchEasy1_1K 879ns × (1.00,1.01) 865ns × (1.00,1.00) -1.58% (p=0.000) RegexpMatchMedium_32 243ns × (1.00,1.00) 233ns × (1.00,1.00) -4.30% (p=0.000) RegexpMatchMedium_1K 70.3µs × (1.00,1.00) 69.5µs × (1.00,1.00) -1.13% (p=0.000) RegexpMatchHard_32 3.82µs × (1.00,1.01) 3.74µs × (1.00,1.00) -1.95% (p=0.000) RegexpMatchHard_1K 117µs × (1.00,1.00) 115µs × (1.00,1.00) -1.69% (p=0.000) Revcomp 917ms × (0.97,1.04) 920ms × (0.97,1.04) ~ (p=0.786) Template 114ms × (0.99,1.01) 117ms × (0.99,1.01) +2.58% (p=0.000) TimeParse 622ns × (0.99,1.01) 615ns × (0.99,1.00) -1.06% (p=0.000) TimeFormat 665ns × (0.99,1.01) 654ns × (0.99,1.00) -1.70% (p=0.000) This CL and previous CL (append) combined: name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.71s × (0.96,1.05) ~ (p=0.638) Fannkuch11 4.41s × (0.98,1.03) 4.39s × (1.00,1.00) ~ (p=0.474) FmtFprintfEmpty 92.7ns × (0.91,1.16) 82.6ns × (0.98,1.04) -10.89% (p=0.004) FmtFprintfString 281ns × (0.96,1.08) 273ns × (0.98,1.02) ~ (p=0.078) FmtFprintfInt 288ns × (0.97,1.06) 270ns × (0.99,1.01) -6.37% (p=0.000) FmtFprintfIntInt 493ns × (0.97,1.04) 475ns × (0.99,1.01) -3.53% (p=0.002) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 393ns × (1.00,1.01) -7.07% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 574ns × (1.00,1.01) -4.02% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.87µs × (0.99,1.02) ~ (p=0.305) GobDecode 14.8ms × (0.98,1.03) 15.0ms × (0.98,1.05) ~ (p=0.237) GobEncode 12.3ms × (0.98,1.01) 11.4ms × (0.99,1.03) -6.95% (p=0.000) Gzip 656ms × (0.99,1.05) 647ms × (0.99,1.01) ~ (p=0.101) Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) +0.58% (p=0.001) HTTPClientServer 91.2µs × (0.97,1.04) 88.5µs × (0.99,1.03) -3.02% (p=0.003) JSONEncode 32.6ms × (0.97,1.08) 29.6ms × (0.98,1.01) -9.10% (p=0.000) JSONDecode 114ms × (0.97,1.05) 104ms × (1.00,1.01) -8.74% (p=0.000) Mandelbrot200 6.11ms × (0.98,1.04) 6.02ms × (1.00,1.00) ~ (p=0.090) GoParse 6.66ms × (0.97,1.04) 6.37ms × (0.97,1.04) -4.41% (p=0.000) RegexpMatchEasy0_32 159ns × (0.99,1.00) 152ns × (0.99,1.01) -4.69% (p=0.000) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 530ns × (1.00,1.00) -1.57% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 134ns × (0.99,1.01) -2.91% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 865ns × (1.00,1.00) -0.51% (p=0.012) RegexpMatchMedium_32 252ns × (0.99,1.01) 233ns × (1.00,1.00) -7.85% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 69.5µs × (1.00,1.00) -4.43% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.74µs × (1.00,1.00) -2.74% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 115µs × (1.00,1.00) -2.24% (p=0.000) Revcomp 920ms × (0.97,1.07) 920ms × (0.97,1.04) ~ (p=0.998) Template 129ms × (0.98,1.03) 117ms × (0.99,1.01) -9.79% (p=0.000) TimeParse 619ns × (0.99,1.01) 615ns × (0.99,1.00) -0.57% (p=0.011) TimeFormat 661ns × (0.98,1.04) 654ns × (0.99,1.00) ~ (p=0.223) Change-Id: If054d81ab2c71d8d62cf54b5b1fac2af66b387fc Reviewed-on: https://go-review.googlesource.com/9813 Reviewed-by: David Chase <drchase@google.com> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2015-05-06 12:35:53 -04:00
func reduceSlice(n *Node) *Node {
low, high, max := n.SliceBounds()
if high != nil && high.Op == OLEN && samesafeexpr(n.Left, high.Left) {
cmd/internal/gc: optimize slice + write barrier The code generated for a slice x[i:j] or x[i:j:k] computes the entire new slice (base, len, cap) and then uses it as the evaluation of the slice expression. If the slice is part of an update x = x[i:j] or x = x[i:j:k], there are opportunities to avoid computing some of these fields. For x = x[0:i], we know that only the len is changing; base can be ignored completely, and cap can be left unmodified. For x = x[0:i:j], we know that only len and cap are changing; base can be ignored completely. For x = x[i:i], we know that the resulting cap is zero, and we don't adjust the base during a slice producing a zero-cap result, so again base can be ignored completely. No write to base, no write barrier. The old slice code was trying to work at a Go syntax level, mainly because that was how you wrote code just once instead of once per architecture. Now the compiler is factored a bit better and we can implement slice during code generation but still have one copy of the code. So the new code is working at that lower level. (It must, to update only parts of the result.) This CL by itself: name old mean new mean delta BinaryTree17 5.81s × (0.98,1.03) 5.71s × (0.96,1.05) ~ (p=0.101) Fannkuch11 4.35s × (1.00,1.00) 4.39s × (1.00,1.00) +0.79% (p=0.000) FmtFprintfEmpty 86.0ns × (0.94,1.11) 82.6ns × (0.98,1.04) -3.86% (p=0.048) FmtFprintfString 276ns × (0.98,1.04) 273ns × (0.98,1.02) ~ (p=0.235) FmtFprintfInt 274ns × (0.98,1.06) 270ns × (0.99,1.01) ~ (p=0.119) FmtFprintfIntInt 506ns × (0.99,1.01) 475ns × (0.99,1.01) -6.02% (p=0.000) FmtFprintfPrefixedInt 391ns × (0.99,1.01) 393ns × (1.00,1.01) ~ (p=0.139) FmtFprintfFloat 566ns × (0.99,1.01) 574ns × (1.00,1.01) +1.33% (p=0.001) FmtManyArgs 1.91µs × (0.99,1.01) 1.87µs × (0.99,1.02) -1.83% (p=0.000) GobDecode 15.3ms × (0.99,1.02) 15.0ms × (0.98,1.05) -1.84% (p=0.042) GobEncode 11.5ms × (0.97,1.03) 11.4ms × (0.99,1.03) ~ (p=0.152) Gzip 645ms × (0.99,1.01) 647ms × (0.99,1.01) ~ (p=0.265) Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) +0.90% (p=0.000) HTTPClientServer 90.5µs × (0.97,1.04) 88.5µs × (0.99,1.03) -2.27% (p=0.014) JSONEncode 32.0ms × (0.98,1.03) 29.6ms × (0.98,1.01) -7.51% (p=0.000) JSONDecode 114ms × (0.99,1.01) 104ms × (1.00,1.01) -8.60% (p=0.000) Mandelbrot200 6.04ms × (1.00,1.01) 6.02ms × (1.00,1.00) ~ (p=0.057) GoParse 6.47ms × (0.97,1.05) 6.37ms × (0.97,1.04) ~ (p=0.105) RegexpMatchEasy0_32 171ns × (0.93,1.07) 152ns × (0.99,1.01) -11.09% (p=0.000) RegexpMatchEasy0_1K 550ns × (0.98,1.01) 530ns × (1.00,1.00) -3.78% (p=0.000) RegexpMatchEasy1_32 135ns × (0.99,1.02) 134ns × (0.99,1.01) -1.33% (p=0.002) RegexpMatchEasy1_1K 879ns × (1.00,1.01) 865ns × (1.00,1.00) -1.58% (p=0.000) RegexpMatchMedium_32 243ns × (1.00,1.00) 233ns × (1.00,1.00) -4.30% (p=0.000) RegexpMatchMedium_1K 70.3µs × (1.00,1.00) 69.5µs × (1.00,1.00) -1.13% (p=0.000) RegexpMatchHard_32 3.82µs × (1.00,1.01) 3.74µs × (1.00,1.00) -1.95% (p=0.000) RegexpMatchHard_1K 117µs × (1.00,1.00) 115µs × (1.00,1.00) -1.69% (p=0.000) Revcomp 917ms × (0.97,1.04) 920ms × (0.97,1.04) ~ (p=0.786) Template 114ms × (0.99,1.01) 117ms × (0.99,1.01) +2.58% (p=0.000) TimeParse 622ns × (0.99,1.01) 615ns × (0.99,1.00) -1.06% (p=0.000) TimeFormat 665ns × (0.99,1.01) 654ns × (0.99,1.00) -1.70% (p=0.000) This CL and previous CL (append) combined: name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.71s × (0.96,1.05) ~ (p=0.638) Fannkuch11 4.41s × (0.98,1.03) 4.39s × (1.00,1.00) ~ (p=0.474) FmtFprintfEmpty 92.7ns × (0.91,1.16) 82.6ns × (0.98,1.04) -10.89% (p=0.004) FmtFprintfString 281ns × (0.96,1.08) 273ns × (0.98,1.02) ~ (p=0.078) FmtFprintfInt 288ns × (0.97,1.06) 270ns × (0.99,1.01) -6.37% (p=0.000) FmtFprintfIntInt 493ns × (0.97,1.04) 475ns × (0.99,1.01) -3.53% (p=0.002) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 393ns × (1.00,1.01) -7.07% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 574ns × (1.00,1.01) -4.02% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.87µs × (0.99,1.02) ~ (p=0.305) GobDecode 14.8ms × (0.98,1.03) 15.0ms × (0.98,1.05) ~ (p=0.237) GobEncode 12.3ms × (0.98,1.01) 11.4ms × (0.99,1.03) -6.95% (p=0.000) Gzip 656ms × (0.99,1.05) 647ms × (0.99,1.01) ~ (p=0.101) Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) +0.58% (p=0.001) HTTPClientServer 91.2µs × (0.97,1.04) 88.5µs × (0.99,1.03) -3.02% (p=0.003) JSONEncode 32.6ms × (0.97,1.08) 29.6ms × (0.98,1.01) -9.10% (p=0.000) JSONDecode 114ms × (0.97,1.05) 104ms × (1.00,1.01) -8.74% (p=0.000) Mandelbrot200 6.11ms × (0.98,1.04) 6.02ms × (1.00,1.00) ~ (p=0.090) GoParse 6.66ms × (0.97,1.04) 6.37ms × (0.97,1.04) -4.41% (p=0.000) RegexpMatchEasy0_32 159ns × (0.99,1.00) 152ns × (0.99,1.01) -4.69% (p=0.000) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 530ns × (1.00,1.00) -1.57% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 134ns × (0.99,1.01) -2.91% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 865ns × (1.00,1.00) -0.51% (p=0.012) RegexpMatchMedium_32 252ns × (0.99,1.01) 233ns × (1.00,1.00) -7.85% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 69.5µs × (1.00,1.00) -4.43% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.74µs × (1.00,1.00) -2.74% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 115µs × (1.00,1.00) -2.24% (p=0.000) Revcomp 920ms × (0.97,1.07) 920ms × (0.97,1.04) ~ (p=0.998) Template 129ms × (0.98,1.03) 117ms × (0.99,1.01) -9.79% (p=0.000) TimeParse 619ns × (0.99,1.01) 615ns × (0.99,1.00) -0.57% (p=0.011) TimeFormat 661ns × (0.98,1.04) 654ns × (0.99,1.00) ~ (p=0.223) Change-Id: If054d81ab2c71d8d62cf54b5b1fac2af66b387fc Reviewed-on: https://go-review.googlesource.com/9813 Reviewed-by: David Chase <drchase@google.com> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2015-05-06 12:35:53 -04:00
// Reduce x[i:len(x)] to x[i:].
high = nil
cmd/internal/gc: optimize slice + write barrier The code generated for a slice x[i:j] or x[i:j:k] computes the entire new slice (base, len, cap) and then uses it as the evaluation of the slice expression. If the slice is part of an update x = x[i:j] or x = x[i:j:k], there are opportunities to avoid computing some of these fields. For x = x[0:i], we know that only the len is changing; base can be ignored completely, and cap can be left unmodified. For x = x[0:i:j], we know that only len and cap are changing; base can be ignored completely. For x = x[i:i], we know that the resulting cap is zero, and we don't adjust the base during a slice producing a zero-cap result, so again base can be ignored completely. No write to base, no write barrier. The old slice code was trying to work at a Go syntax level, mainly because that was how you wrote code just once instead of once per architecture. Now the compiler is factored a bit better and we can implement slice during code generation but still have one copy of the code. So the new code is working at that lower level. (It must, to update only parts of the result.) This CL by itself: name old mean new mean delta BinaryTree17 5.81s × (0.98,1.03) 5.71s × (0.96,1.05) ~ (p=0.101) Fannkuch11 4.35s × (1.00,1.00) 4.39s × (1.00,1.00) +0.79% (p=0.000) FmtFprintfEmpty 86.0ns × (0.94,1.11) 82.6ns × (0.98,1.04) -3.86% (p=0.048) FmtFprintfString 276ns × (0.98,1.04) 273ns × (0.98,1.02) ~ (p=0.235) FmtFprintfInt 274ns × (0.98,1.06) 270ns × (0.99,1.01) ~ (p=0.119) FmtFprintfIntInt 506ns × (0.99,1.01) 475ns × (0.99,1.01) -6.02% (p=0.000) FmtFprintfPrefixedInt 391ns × (0.99,1.01) 393ns × (1.00,1.01) ~ (p=0.139) FmtFprintfFloat 566ns × (0.99,1.01) 574ns × (1.00,1.01) +1.33% (p=0.001) FmtManyArgs 1.91µs × (0.99,1.01) 1.87µs × (0.99,1.02) -1.83% (p=0.000) GobDecode 15.3ms × (0.99,1.02) 15.0ms × (0.98,1.05) -1.84% (p=0.042) GobEncode 11.5ms × (0.97,1.03) 11.4ms × (0.99,1.03) ~ (p=0.152) Gzip 645ms × (0.99,1.01) 647ms × (0.99,1.01) ~ (p=0.265) Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) +0.90% (p=0.000) HTTPClientServer 90.5µs × (0.97,1.04) 88.5µs × (0.99,1.03) -2.27% (p=0.014) JSONEncode 32.0ms × (0.98,1.03) 29.6ms × (0.98,1.01) -7.51% (p=0.000) JSONDecode 114ms × (0.99,1.01) 104ms × (1.00,1.01) -8.60% (p=0.000) Mandelbrot200 6.04ms × (1.00,1.01) 6.02ms × (1.00,1.00) ~ (p=0.057) GoParse 6.47ms × (0.97,1.05) 6.37ms × (0.97,1.04) ~ (p=0.105) RegexpMatchEasy0_32 171ns × (0.93,1.07) 152ns × (0.99,1.01) -11.09% (p=0.000) RegexpMatchEasy0_1K 550ns × (0.98,1.01) 530ns × (1.00,1.00) -3.78% (p=0.000) RegexpMatchEasy1_32 135ns × (0.99,1.02) 134ns × (0.99,1.01) -1.33% (p=0.002) RegexpMatchEasy1_1K 879ns × (1.00,1.01) 865ns × (1.00,1.00) -1.58% (p=0.000) RegexpMatchMedium_32 243ns × (1.00,1.00) 233ns × (1.00,1.00) -4.30% (p=0.000) RegexpMatchMedium_1K 70.3µs × (1.00,1.00) 69.5µs × (1.00,1.00) -1.13% (p=0.000) RegexpMatchHard_32 3.82µs × (1.00,1.01) 3.74µs × (1.00,1.00) -1.95% (p=0.000) RegexpMatchHard_1K 117µs × (1.00,1.00) 115µs × (1.00,1.00) -1.69% (p=0.000) Revcomp 917ms × (0.97,1.04) 920ms × (0.97,1.04) ~ (p=0.786) Template 114ms × (0.99,1.01) 117ms × (0.99,1.01) +2.58% (p=0.000) TimeParse 622ns × (0.99,1.01) 615ns × (0.99,1.00) -1.06% (p=0.000) TimeFormat 665ns × (0.99,1.01) 654ns × (0.99,1.00) -1.70% (p=0.000) This CL and previous CL (append) combined: name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.71s × (0.96,1.05) ~ (p=0.638) Fannkuch11 4.41s × (0.98,1.03) 4.39s × (1.00,1.00) ~ (p=0.474) FmtFprintfEmpty 92.7ns × (0.91,1.16) 82.6ns × (0.98,1.04) -10.89% (p=0.004) FmtFprintfString 281ns × (0.96,1.08) 273ns × (0.98,1.02) ~ (p=0.078) FmtFprintfInt 288ns × (0.97,1.06) 270ns × (0.99,1.01) -6.37% (p=0.000) FmtFprintfIntInt 493ns × (0.97,1.04) 475ns × (0.99,1.01) -3.53% (p=0.002) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 393ns × (1.00,1.01) -7.07% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 574ns × (1.00,1.01) -4.02% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.87µs × (0.99,1.02) ~ (p=0.305) GobDecode 14.8ms × (0.98,1.03) 15.0ms × (0.98,1.05) ~ (p=0.237) GobEncode 12.3ms × (0.98,1.01) 11.4ms × (0.99,1.03) -6.95% (p=0.000) Gzip 656ms × (0.99,1.05) 647ms × (0.99,1.01) ~ (p=0.101) Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) +0.58% (p=0.001) HTTPClientServer 91.2µs × (0.97,1.04) 88.5µs × (0.99,1.03) -3.02% (p=0.003) JSONEncode 32.6ms × (0.97,1.08) 29.6ms × (0.98,1.01) -9.10% (p=0.000) JSONDecode 114ms × (0.97,1.05) 104ms × (1.00,1.01) -8.74% (p=0.000) Mandelbrot200 6.11ms × (0.98,1.04) 6.02ms × (1.00,1.00) ~ (p=0.090) GoParse 6.66ms × (0.97,1.04) 6.37ms × (0.97,1.04) -4.41% (p=0.000) RegexpMatchEasy0_32 159ns × (0.99,1.00) 152ns × (0.99,1.01) -4.69% (p=0.000) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 530ns × (1.00,1.00) -1.57% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 134ns × (0.99,1.01) -2.91% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 865ns × (1.00,1.00) -0.51% (p=0.012) RegexpMatchMedium_32 252ns × (0.99,1.01) 233ns × (1.00,1.00) -7.85% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 69.5µs × (1.00,1.00) -4.43% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.74µs × (1.00,1.00) -2.74% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 115µs × (1.00,1.00) -2.24% (p=0.000) Revcomp 920ms × (0.97,1.07) 920ms × (0.97,1.04) ~ (p=0.998) Template 129ms × (0.98,1.03) 117ms × (0.99,1.01) -9.79% (p=0.000) TimeParse 619ns × (0.99,1.01) 615ns × (0.99,1.00) -0.57% (p=0.011) TimeFormat 661ns × (0.98,1.04) 654ns × (0.99,1.00) ~ (p=0.223) Change-Id: If054d81ab2c71d8d62cf54b5b1fac2af66b387fc Reviewed-on: https://go-review.googlesource.com/9813 Reviewed-by: David Chase <drchase@google.com> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2015-05-06 12:35:53 -04:00
}
n.SetSliceBounds(low, high, max)
if (n.Op == OSLICE || n.Op == OSLICESTR) && low == nil && high == nil {
cmd/internal/gc: optimize slice + write barrier The code generated for a slice x[i:j] or x[i:j:k] computes the entire new slice (base, len, cap) and then uses it as the evaluation of the slice expression. If the slice is part of an update x = x[i:j] or x = x[i:j:k], there are opportunities to avoid computing some of these fields. For x = x[0:i], we know that only the len is changing; base can be ignored completely, and cap can be left unmodified. For x = x[0:i:j], we know that only len and cap are changing; base can be ignored completely. For x = x[i:i], we know that the resulting cap is zero, and we don't adjust the base during a slice producing a zero-cap result, so again base can be ignored completely. No write to base, no write barrier. The old slice code was trying to work at a Go syntax level, mainly because that was how you wrote code just once instead of once per architecture. Now the compiler is factored a bit better and we can implement slice during code generation but still have one copy of the code. So the new code is working at that lower level. (It must, to update only parts of the result.) This CL by itself: name old mean new mean delta BinaryTree17 5.81s × (0.98,1.03) 5.71s × (0.96,1.05) ~ (p=0.101) Fannkuch11 4.35s × (1.00,1.00) 4.39s × (1.00,1.00) +0.79% (p=0.000) FmtFprintfEmpty 86.0ns × (0.94,1.11) 82.6ns × (0.98,1.04) -3.86% (p=0.048) FmtFprintfString 276ns × (0.98,1.04) 273ns × (0.98,1.02) ~ (p=0.235) FmtFprintfInt 274ns × (0.98,1.06) 270ns × (0.99,1.01) ~ (p=0.119) FmtFprintfIntInt 506ns × (0.99,1.01) 475ns × (0.99,1.01) -6.02% (p=0.000) FmtFprintfPrefixedInt 391ns × (0.99,1.01) 393ns × (1.00,1.01) ~ (p=0.139) FmtFprintfFloat 566ns × (0.99,1.01) 574ns × (1.00,1.01) +1.33% (p=0.001) FmtManyArgs 1.91µs × (0.99,1.01) 1.87µs × (0.99,1.02) -1.83% (p=0.000) GobDecode 15.3ms × (0.99,1.02) 15.0ms × (0.98,1.05) -1.84% (p=0.042) GobEncode 11.5ms × (0.97,1.03) 11.4ms × (0.99,1.03) ~ (p=0.152) Gzip 645ms × (0.99,1.01) 647ms × (0.99,1.01) ~ (p=0.265) Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) +0.90% (p=0.000) HTTPClientServer 90.5µs × (0.97,1.04) 88.5µs × (0.99,1.03) -2.27% (p=0.014) JSONEncode 32.0ms × (0.98,1.03) 29.6ms × (0.98,1.01) -7.51% (p=0.000) JSONDecode 114ms × (0.99,1.01) 104ms × (1.00,1.01) -8.60% (p=0.000) Mandelbrot200 6.04ms × (1.00,1.01) 6.02ms × (1.00,1.00) ~ (p=0.057) GoParse 6.47ms × (0.97,1.05) 6.37ms × (0.97,1.04) ~ (p=0.105) RegexpMatchEasy0_32 171ns × (0.93,1.07) 152ns × (0.99,1.01) -11.09% (p=0.000) RegexpMatchEasy0_1K 550ns × (0.98,1.01) 530ns × (1.00,1.00) -3.78% (p=0.000) RegexpMatchEasy1_32 135ns × (0.99,1.02) 134ns × (0.99,1.01) -1.33% (p=0.002) RegexpMatchEasy1_1K 879ns × (1.00,1.01) 865ns × (1.00,1.00) -1.58% (p=0.000) RegexpMatchMedium_32 243ns × (1.00,1.00) 233ns × (1.00,1.00) -4.30% (p=0.000) RegexpMatchMedium_1K 70.3µs × (1.00,1.00) 69.5µs × (1.00,1.00) -1.13% (p=0.000) RegexpMatchHard_32 3.82µs × (1.00,1.01) 3.74µs × (1.00,1.00) -1.95% (p=0.000) RegexpMatchHard_1K 117µs × (1.00,1.00) 115µs × (1.00,1.00) -1.69% (p=0.000) Revcomp 917ms × (0.97,1.04) 920ms × (0.97,1.04) ~ (p=0.786) Template 114ms × (0.99,1.01) 117ms × (0.99,1.01) +2.58% (p=0.000) TimeParse 622ns × (0.99,1.01) 615ns × (0.99,1.00) -1.06% (p=0.000) TimeFormat 665ns × (0.99,1.01) 654ns × (0.99,1.00) -1.70% (p=0.000) This CL and previous CL (append) combined: name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.71s × (0.96,1.05) ~ (p=0.638) Fannkuch11 4.41s × (0.98,1.03) 4.39s × (1.00,1.00) ~ (p=0.474) FmtFprintfEmpty 92.7ns × (0.91,1.16) 82.6ns × (0.98,1.04) -10.89% (p=0.004) FmtFprintfString 281ns × (0.96,1.08) 273ns × (0.98,1.02) ~ (p=0.078) FmtFprintfInt 288ns × (0.97,1.06) 270ns × (0.99,1.01) -6.37% (p=0.000) FmtFprintfIntInt 493ns × (0.97,1.04) 475ns × (0.99,1.01) -3.53% (p=0.002) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 393ns × (1.00,1.01) -7.07% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 574ns × (1.00,1.01) -4.02% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.87µs × (0.99,1.02) ~ (p=0.305) GobDecode 14.8ms × (0.98,1.03) 15.0ms × (0.98,1.05) ~ (p=0.237) GobEncode 12.3ms × (0.98,1.01) 11.4ms × (0.99,1.03) -6.95% (p=0.000) Gzip 656ms × (0.99,1.05) 647ms × (0.99,1.01) ~ (p=0.101) Gunzip 142ms × (1.00,1.00) 143ms × (1.00,1.01) +0.58% (p=0.001) HTTPClientServer 91.2µs × (0.97,1.04) 88.5µs × (0.99,1.03) -3.02% (p=0.003) JSONEncode 32.6ms × (0.97,1.08) 29.6ms × (0.98,1.01) -9.10% (p=0.000) JSONDecode 114ms × (0.97,1.05) 104ms × (1.00,1.01) -8.74% (p=0.000) Mandelbrot200 6.11ms × (0.98,1.04) 6.02ms × (1.00,1.00) ~ (p=0.090) GoParse 6.66ms × (0.97,1.04) 6.37ms × (0.97,1.04) -4.41% (p=0.000) RegexpMatchEasy0_32 159ns × (0.99,1.00) 152ns × (0.99,1.01) -4.69% (p=0.000) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 530ns × (1.00,1.00) -1.57% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 134ns × (0.99,1.01) -2.91% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 865ns × (1.00,1.00) -0.51% (p=0.012) RegexpMatchMedium_32 252ns × (0.99,1.01) 233ns × (1.00,1.00) -7.85% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 69.5µs × (1.00,1.00) -4.43% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.74µs × (1.00,1.00) -2.74% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 115µs × (1.00,1.00) -2.24% (p=0.000) Revcomp 920ms × (0.97,1.07) 920ms × (0.97,1.04) ~ (p=0.998) Template 129ms × (0.98,1.03) 117ms × (0.99,1.01) -9.79% (p=0.000) TimeParse 619ns × (0.99,1.01) 615ns × (0.99,1.00) -0.57% (p=0.011) TimeFormat 661ns × (0.98,1.04) 654ns × (0.99,1.00) ~ (p=0.223) Change-Id: If054d81ab2c71d8d62cf54b5b1fac2af66b387fc Reviewed-on: https://go-review.googlesource.com/9813 Reviewed-by: David Chase <drchase@google.com> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2015-05-06 12:35:53 -04:00
// Reduce x[:] to x.
if Debug_slice > 0 {
Warn("slice: omit slice operation")
}
return n.Left
}
return n
}
func ascompatee1(l *Node, r *Node, init *Nodes) *Node {
// convas will turn map assigns into function calls,
// making it impossible for reorder3 to work.
n := nod(OAS, l, r)
if l.Op == OINDEXMAP {
return n
}
return convas(n, init)
}
func ascompatee(op Op, nl, nr []*Node, init *Nodes) []*Node {
// check assign expression list to
// an expression list. called in
// expr-list = expr-list
// ensure order of evaluation for function calls
for i := range nl {
nl[i] = safeexpr(nl[i], init)
}
for i1 := range nr {
nr[i1] = safeexpr(nr[i1], init)
}
var nn []*Node
i := 0
for ; i < len(nl); i++ {
if i >= len(nr) {
break
}
// Do not generate 'x = x' during return. See issue 4014.
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
if op == ORETURN && samesafeexpr(nl[i], nr[i]) {
continue
}
nn = append(nn, ascompatee1(nl[i], nr[i], init))
}
// cannot happen: caller checked that lists had same length
if i < len(nl) || i < len(nr) {
var nln, nrn Nodes
nln.Set(nl)
nrn.Set(nr)
Fatalf("error in shape across %+v %v %+v / %d %d [%s]", nln, op, nrn, len(nl), len(nr), Curfn.funcname())
}
return nn
}
// fncall reports whether assigning an rvalue of type rt to an lvalue l might involve a function call.
func fncall(l *Node, rt *types.Type) bool {
if l.HasCall() || l.Op == OINDEXMAP {
return true
}
if types.Identical(l.Type, rt) {
return false
}
// There might be a conversion required, which might involve a runtime call.
return true
}
// check assign type list to
// an expression list. called in
// expr-list = func()
func ascompatet(nl Nodes, nr *types.Type) []*Node {
if nl.Len() != nr.NumFields() {
Fatalf("ascompatet: assignment count mismatch: %d = %d", nl.Len(), nr.NumFields())
}
var nn, mm Nodes
for i, l := range nl.Slice() {
if l.isBlank() {
continue
}
r := nr.Field(i)
// Any assignment to an lvalue that might cause a function call must be
// deferred until all the returned values have been read.
if fncall(l, r.Type) {
tmp := temp(r.Type)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
tmp = typecheck(tmp, ctxExpr)
a := nod(OAS, l, tmp)
a = convas(a, &mm)
mm.Append(a)
l = tmp
}
res := nod(ORESULT, nil, nil)
res.Xoffset = Ctxt.FixedFrameSize() + r.Offset
res.Type = r.Type
res.SetTypecheck(1)
a := nod(OAS, l, res)
a = convas(a, &nn)
updateHasCall(a)
if a.HasCall() {
Dump("ascompatet ucount", a)
Fatalf("ascompatet: too many function calls evaluating parameters")
}
nn.Append(a)
}
return append(nn.Slice(), mm.Slice()...)
}
// package all the arguments that match a ... T parameter into a []T.
func mkdotargslice(typ *types.Type, args []*Node) *Node {
var n *Node
if len(args) == 0 {
n = nodnil()
n.Type = typ
} else {
n = nod(OCOMPLIT, nil, typenod(typ))
n.List.Append(args...)
cmd/compile: use fixVariadicCall in escape analysis This CL uses fixVariadicCall before escape analyzing function calls. This has a number of benefits, though also some minor obstacles: Most notably, it allows us to remove ODDDARG along with the logic involved in setting it up, manipulating EscHoles, and later copying its escape analysis flags to the actual slice argument. Instead, we uniformly handle all variadic calls the same way. (E.g., issue31573.go is updated because now f() and f(nil...) are handled identically.) It also allows us to simplify handling of builtins and generic function calls. Previously handling of calls was hairy enough to require multiple dispatches on n.Op, whereas now the logic is uniform enough that we can easily handle it with a single dispatch. The downside is handling //go:uintptrescapes is now somewhat clumsy. (It used to be clumsy, but it still is, too.) The proper fix here is probably to stop using escape analysis tags for //go:uintptrescapes and unsafe-uintptr, and have an earlier pass responsible for them. Finally, note that while we now call fixVariadicCall in Escape, we still have to call it in Order, because we don't (yet) run Escape on all compiler-generated functions. In particular, the generated "init" function for initializing package-level variables can contain calls to variadic functions and isn't escape analyzed. Passes toolstash-check -race. Change-Id: I4cdb92a393ac487910aeee58a5cb8c1500eef881 Reviewed-on: https://go-review.googlesource.com/c/go/+/229759 Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
2020-04-21 19:48:02 -07:00
n.SetImplicit(true)
}
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
n = typecheck(n, ctxExpr)
if n.Type == nil {
Fatalf("mkdotargslice: typecheck failed")
}
return n
}
// fixVariadicCall rewrites calls to variadic functions to use an
// explicit ... argument if one is not already present.
func fixVariadicCall(call *Node) {
fntype := call.Left.Type
if !fntype.IsVariadic() || call.IsDDD() {
return
}
vi := fntype.NumParams() - 1
vt := fntype.Params().Field(vi).Type
args := call.List.Slice()
extra := args[vi:]
slice := mkdotargslice(vt, extra)
for i := range extra {
extra[i] = nil // allow GC
}
call.List.Set(append(args[:vi], slice))
call.SetIsDDD(true)
}
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
func walkCall(n *Node, init *Nodes) {
if n.Rlist.Len() != 0 {
return // already walked
}
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
params := n.Left.Type.Params()
args := n.List.Slice()
n.Left = walkexpr(n.Left, init)
walkexprlist(args, init)
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
// If this is a method call, add the receiver at the beginning of the args.
if n.Op == OCALLMETH {
withRecv := make([]*Node, len(args)+1)
withRecv[0] = n.Left.Left
n.Left.Left = nil
copy(withRecv[1:], args)
args = withRecv
}
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
// For any argument whose evaluation might require a function call,
// store that argument into a temporary variable,
// to prevent that calls from clobbering arguments already on the stack.
// When instrumenting, all arguments might require function calls.
var tempAssigns []*Node
for i, arg := range args {
updateHasCall(arg)
// Determine param type.
var t *types.Type
if n.Op == OCALLMETH {
if i == 0 {
t = n.Left.Type.Recv().Type
} else {
t = params.Field(i - 1).Type
}
} else {
t = params.Field(i).Type
}
if instrumenting || fncall(arg, t) {
// make assignment of fncall to tempAt
tmp := temp(t)
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
a := nod(OAS, tmp, arg)
a = convas(a, init)
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
tempAssigns = append(tempAssigns, a)
// replace arg with temp
args[i] = tmp
}
}
n.List.Set(tempAssigns)
n.Rlist.Set(args)
}
// generate code for print
func walkprint(nn *Node, init *Nodes) *Node {
// Hoist all the argument evaluation up before the lock.
walkexprlistcheap(nn.List.Slice(), init)
// For println, add " " between elements and "\n" at the end.
if nn.Op == OPRINTN {
s := nn.List.Slice()
t := make([]*Node, 0, len(s)*2)
for i, n := range s {
if i != 0 {
t = append(t, nodstr(" "))
}
t = append(t, n)
}
t = append(t, nodstr("\n"))
nn.List.Set(t)
}
// Collapse runs of constant strings.
s := nn.List.Slice()
t := make([]*Node, 0, len(s))
for i := 0; i < len(s); {
var strs []string
for i < len(s) && Isconst(s[i], CTSTR) {
strs = append(strs, strlit(s[i]))
i++
}
if len(strs) > 0 {
t = append(t, nodstr(strings.Join(strs, "")))
}
if i < len(s) {
t = append(t, s[i])
i++
}
}
nn.List.Set(t)
calls := []*Node{mkcall("printlock", nil, init)}
for i, n := range nn.List.Slice() {
if n.Op == OLITERAL {
switch n.Val().Ctype() {
case CTRUNE:
n = defaultlit(n, types.Runetype)
case CTINT:
n = defaultlit(n, types.Types[TINT64])
case CTFLT:
n = defaultlit(n, types.Types[TFLOAT64])
}
}
if n.Op != OLITERAL && n.Type != nil && n.Type.Etype == TIDEAL {
n = defaultlit(n, types.Types[TINT64])
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n = defaultlit(n, nil)
nn.List.SetIndex(i, n)
if n.Type == nil || n.Type.Etype == TFORW {
continue
}
var on *Node
switch n.Type.Etype {
case TINTER:
if n.Type.IsEmptyInterface() {
on = syslook("printeface")
} else {
on = syslook("printiface")
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
on = substArgTypes(on, n.Type) // any-1
case TPTR, TCHAN, TMAP, TFUNC, TUNSAFEPTR:
on = syslook("printpointer")
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
on = substArgTypes(on, n.Type) // any-1
case TSLICE:
on = syslook("printslice")
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
on = substArgTypes(on, n.Type) // any-1
case TUINT, TUINT8, TUINT16, TUINT32, TUINT64, TUINTPTR:
if isRuntimePkg(n.Type.Sym.Pkg) && n.Type.Sym.Name == "hex" {
on = syslook("printhex")
} else {
on = syslook("printuint")
}
case TINT, TINT8, TINT16, TINT32, TINT64:
on = syslook("printint")
case TFLOAT32, TFLOAT64:
on = syslook("printfloat")
case TCOMPLEX64, TCOMPLEX128:
on = syslook("printcomplex")
case TBOOL:
on = syslook("printbool")
case TSTRING:
cs := ""
if Isconst(n, CTSTR) {
cs = strlit(n)
}
switch cs {
case " ":
on = syslook("printsp")
case "\n":
on = syslook("printnl")
default:
on = syslook("printstring")
}
default:
badtype(OPRINT, n.Type, nil)
continue
}
r := nod(OCALL, on, nil)
if params := on.Type.Params().FieldSlice(); len(params) > 0 {
t := params[0].Type
if !types.Identical(t, n.Type) {
n = nod(OCONV, n, nil)
n.Type = t
}
r.List.Append(n)
}
calls = append(calls, r)
}
calls = append(calls, mkcall("printunlock", nil, init))
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
typecheckslice(calls, ctxStmt)
walkexprlist(calls, init)
r := nod(OEMPTY, nil, nil)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
r = typecheck(r, ctxStmt)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
r = walkexpr(r, init)
r.Ninit.Set(calls)
return r
}
func callnew(t *types.Type) *Node {
dowidth(t)
cmd/compile: port callnew to ssa conversion This is part of a general effort to shrink walk. In an ideal world, we'd have an SSA op for allocation, but we don't yet have a good mechanism for introducing function calling during SSA compilation. In the meantime, SSA conversion is a better place for it. This also makes it easier to introduce new optimizations; instead of doing the typecheck walk dance, we can simply write what we want the backend to do. I introduced a new opcode in this change because: (a) It avoids a class of bugs involving correctly detecting whether this ONEW is a "before walk" ONEW or an "after walk" ONEW. It also means that using ONEW or ONEWOBJ in the wrong context will generally result in a faster failure. (b) Opcodes are cheap. (c) It provides a better place to put documentation. This change also is also marginally more performant: name old alloc/op new alloc/op delta Template 39.1MB ± 0% 39.0MB ± 0% -0.14% (p=0.008 n=5+5) Unicode 28.4MB ± 0% 28.4MB ± 0% ~ (p=0.421 n=5+5) GoTypes 132MB ± 0% 132MB ± 0% -0.23% (p=0.008 n=5+5) Compiler 608MB ± 0% 607MB ± 0% -0.25% (p=0.008 n=5+5) SSA 2.04GB ± 0% 2.04GB ± 0% -0.01% (p=0.008 n=5+5) Flate 24.4MB ± 0% 24.3MB ± 0% -0.13% (p=0.008 n=5+5) GoParser 29.3MB ± 0% 29.1MB ± 0% -0.54% (p=0.008 n=5+5) Reflect 84.8MB ± 0% 84.7MB ± 0% -0.21% (p=0.008 n=5+5) Tar 36.7MB ± 0% 36.6MB ± 0% -0.10% (p=0.008 n=5+5) XML 48.7MB ± 0% 48.6MB ± 0% -0.24% (p=0.008 n=5+5) [Geo mean] 85.0MB 84.8MB -0.19% name old allocs/op new allocs/op delta Template 383k ± 0% 382k ± 0% -0.26% (p=0.008 n=5+5) Unicode 341k ± 0% 341k ± 0% ~ (p=0.579 n=5+5) GoTypes 1.37M ± 0% 1.36M ± 0% -0.39% (p=0.008 n=5+5) Compiler 5.59M ± 0% 5.56M ± 0% -0.49% (p=0.008 n=5+5) SSA 16.9M ± 0% 16.9M ± 0% -0.03% (p=0.008 n=5+5) Flate 238k ± 0% 238k ± 0% -0.23% (p=0.008 n=5+5) GoParser 306k ± 0% 303k ± 0% -0.93% (p=0.008 n=5+5) Reflect 990k ± 0% 987k ± 0% -0.33% (p=0.008 n=5+5) Tar 356k ± 0% 355k ± 0% -0.20% (p=0.008 n=5+5) XML 444k ± 0% 442k ± 0% -0.45% (p=0.008 n=5+5) [Geo mean] 848k 845k -0.33% Change-Id: I2c36003a7cbf71b53857b7de734852b698f49310 Reviewed-on: https://go-review.googlesource.com/c/go/+/167957 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2019-03-17 07:14:12 -07:00
n := nod(ONEWOBJ, typename(t), nil)
n.Type = types.NewPtr(t)
n.SetTypecheck(1)
n.MarkNonNil()
cmd/compile: port callnew to ssa conversion This is part of a general effort to shrink walk. In an ideal world, we'd have an SSA op for allocation, but we don't yet have a good mechanism for introducing function calling during SSA compilation. In the meantime, SSA conversion is a better place for it. This also makes it easier to introduce new optimizations; instead of doing the typecheck walk dance, we can simply write what we want the backend to do. I introduced a new opcode in this change because: (a) It avoids a class of bugs involving correctly detecting whether this ONEW is a "before walk" ONEW or an "after walk" ONEW. It also means that using ONEW or ONEWOBJ in the wrong context will generally result in a faster failure. (b) Opcodes are cheap. (c) It provides a better place to put documentation. This change also is also marginally more performant: name old alloc/op new alloc/op delta Template 39.1MB ± 0% 39.0MB ± 0% -0.14% (p=0.008 n=5+5) Unicode 28.4MB ± 0% 28.4MB ± 0% ~ (p=0.421 n=5+5) GoTypes 132MB ± 0% 132MB ± 0% -0.23% (p=0.008 n=5+5) Compiler 608MB ± 0% 607MB ± 0% -0.25% (p=0.008 n=5+5) SSA 2.04GB ± 0% 2.04GB ± 0% -0.01% (p=0.008 n=5+5) Flate 24.4MB ± 0% 24.3MB ± 0% -0.13% (p=0.008 n=5+5) GoParser 29.3MB ± 0% 29.1MB ± 0% -0.54% (p=0.008 n=5+5) Reflect 84.8MB ± 0% 84.7MB ± 0% -0.21% (p=0.008 n=5+5) Tar 36.7MB ± 0% 36.6MB ± 0% -0.10% (p=0.008 n=5+5) XML 48.7MB ± 0% 48.6MB ± 0% -0.24% (p=0.008 n=5+5) [Geo mean] 85.0MB 84.8MB -0.19% name old allocs/op new allocs/op delta Template 383k ± 0% 382k ± 0% -0.26% (p=0.008 n=5+5) Unicode 341k ± 0% 341k ± 0% ~ (p=0.579 n=5+5) GoTypes 1.37M ± 0% 1.36M ± 0% -0.39% (p=0.008 n=5+5) Compiler 5.59M ± 0% 5.56M ± 0% -0.49% (p=0.008 n=5+5) SSA 16.9M ± 0% 16.9M ± 0% -0.03% (p=0.008 n=5+5) Flate 238k ± 0% 238k ± 0% -0.23% (p=0.008 n=5+5) GoParser 306k ± 0% 303k ± 0% -0.93% (p=0.008 n=5+5) Reflect 990k ± 0% 987k ± 0% -0.33% (p=0.008 n=5+5) Tar 356k ± 0% 355k ± 0% -0.20% (p=0.008 n=5+5) XML 444k ± 0% 442k ± 0% -0.45% (p=0.008 n=5+5) [Geo mean] 848k 845k -0.33% Change-Id: I2c36003a7cbf71b53857b7de734852b698f49310 Reviewed-on: https://go-review.googlesource.com/c/go/+/167957 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2019-03-17 07:14:12 -07:00
return n
}
// isReflectHeaderDataField reports whether l is an expression p.Data
// where p has type reflect.SliceHeader or reflect.StringHeader.
func isReflectHeaderDataField(l *Node) bool {
if l.Type != types.Types[TUINTPTR] {
return false
}
var tsym *types.Sym
switch l.Op {
case ODOT:
tsym = l.Left.Type.Sym
case ODOTPTR:
tsym = l.Left.Type.Elem().Sym
default:
return false
}
if tsym == nil || l.Sym.Name != "Data" || tsym.Pkg.Path != "reflect" {
return false
}
return tsym.Name == "SliceHeader" || tsym.Name == "StringHeader"
}
func convas(n *Node, init *Nodes) *Node {
if n.Op != OAS {
Fatalf("convas: not OAS %v", n.Op)
}
defer updateHasCall(n)
n.SetTypecheck(1)
if n.Left == nil || n.Right == nil {
return n
}
lt := n.Left.Type
rt := n.Right.Type
if lt == nil || rt == nil {
return n
}
if n.Left.isBlank() {
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Right = defaultlit(n.Right, nil)
return n
}
if !types.Identical(lt, rt) {
n.Right = assignconv(n.Right, lt, "assignment")
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Right = walkexpr(n.Right, init)
}
dowidth(n.Right.Type)
return n
}
// from ascompat[ee]
// a,b = c,d
// simultaneous assignment. there cannot
// be later use of an earlier lvalue.
//
// function calls have been removed.
func reorder3(all []*Node) []*Node {
// If a needed expression may be affected by an
// earlier assignment, make an early copy of that
// expression and use the copy instead.
var early []*Node
var mapinit Nodes
for i, n := range all {
l := n.Left
// Save subexpressions needed on left side.
// Drill through non-dereferences.
for {
if l.Op == ODOT || l.Op == OPAREN {
l = l.Left
continue
}
if l.Op == OINDEX && l.Left.Type.IsArray() {
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
l.Right = reorder3save(l.Right, all, i, &early)
l = l.Left
continue
}
break
}
switch l.Op {
default:
Fatalf("reorder3 unexpected lvalue %#v", l.Op)
case ONAME:
break
case OINDEX, OINDEXMAP:
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
l.Left = reorder3save(l.Left, all, i, &early)
l.Right = reorder3save(l.Right, all, i, &early)
if l.Op == OINDEXMAP {
all[i] = convas(all[i], &mapinit)
}
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
case ODEREF, ODOTPTR:
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
l.Left = reorder3save(l.Left, all, i, &early)
}
// Save expression on right side.
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
all[i].Right = reorder3save(all[i].Right, all, i, &early)
}
early = append(mapinit.Slice(), early...)
return append(early, all...)
}
// if the evaluation of *np would be affected by the
// assignments in all up to but not including the ith assignment,
// copy into a temporary during *early and
// replace *np with that temp.
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
// The result of reorder3save MUST be assigned back to n, e.g.
// n.Left = reorder3save(n.Left, all, i, early)
func reorder3save(n *Node, all []*Node, i int, early *[]*Node) *Node {
if !aliased(n, all, i) {
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
return n
}
q := temp(n.Type)
q = nod(OAS, q, n)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
q = typecheck(q, ctxStmt)
*early = append(*early, q)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
return q.Left
}
// what's the outer value that a write to n affects?
// outer value means containing struct or array.
func outervalue(n *Node) *Node {
for {
switch n.Op {
case OXDOT:
Fatalf("OXDOT in walk")
case ODOT, OPAREN, OCONVNOP:
n = n.Left
continue
case OINDEX:
if n.Left.Type != nil && n.Left.Type.IsArray() {
n = n.Left
continue
}
}
return n
}
}
// Is it possible that the computation of n might be
// affected by writes in as up to but not including the ith element?
func aliased(n *Node, all []*Node, i int) bool {
if n == nil {
return false
}
// Treat all fields of a struct as referring to the whole struct.
// We could do better but we would have to keep track of the fields.
for n.Op == ODOT {
n = n.Left
}
// Look for obvious aliasing: a variable being assigned
// during the all list and appearing in n.
// Also record whether there are any writes to main memory.
// Also record whether there are any writes to variables
// whose addresses have been taken.
memwrite := false
varwrite := false
for _, an := range all[:i] {
a := outervalue(an.Left)
for a.Op == ODOT {
a = a.Left
}
if a.Op != ONAME {
memwrite = true
continue
}
cmd/compile: move Node.Class to flags Put it at position zero, since it is fairly hot. This shrinks gc.Node into a smaller size class on 64 bit systems. name old time/op new time/op delta Template 193ms ± 5% 192ms ± 3% ~ (p=0.353 n=94+93) Unicode 86.1ms ± 5% 85.0ms ± 4% -1.23% (p=0.000 n=95+98) GoTypes 546ms ± 3% 544ms ± 4% -0.40% (p=0.007 n=94+97) Compiler 2.56s ± 3% 2.54s ± 3% -0.67% (p=0.000 n=99+97) SSA 5.13s ± 2% 5.10s ± 3% -0.55% (p=0.000 n=94+98) Flate 122ms ± 6% 121ms ± 4% -0.75% (p=0.002 n=97+95) GoParser 144ms ± 5% 144ms ± 4% ~ (p=0.298 n=98+97) Reflect 348ms ± 4% 349ms ± 4% ~ (p=0.350 n=98+97) Tar 105ms ± 5% 104ms ± 5% ~ (p=0.154 n=96+98) XML 200ms ± 5% 198ms ± 4% -0.71% (p=0.015 n=97+98) [Geo mean] 330ms 328ms -0.52% name old user-time/op new user-time/op delta Template 229ms ±11% 224ms ± 7% -2.16% (p=0.001 n=100+87) Unicode 109ms ± 5% 109ms ± 6% ~ (p=0.897 n=96+91) GoTypes 712ms ± 4% 709ms ± 4% ~ (p=0.085 n=96+98) Compiler 3.41s ± 3% 3.36s ± 3% -1.43% (p=0.000 n=98+98) SSA 7.46s ± 3% 7.31s ± 3% -2.02% (p=0.000 n=100+99) Flate 145ms ± 6% 143ms ± 6% -1.11% (p=0.001 n=99+97) GoParser 177ms ± 5% 176ms ± 5% -0.78% (p=0.018 n=95+95) Reflect 432ms ± 7% 435ms ± 9% ~ (p=0.296 n=100+100) Tar 121ms ± 7% 121ms ± 5% ~ (p=0.072 n=100+95) XML 241ms ± 4% 239ms ± 5% ~ (p=0.085 n=97+99) [Geo mean] 413ms 410ms -0.73% name old alloc/op new alloc/op delta Template 38.4MB ± 0% 37.7MB ± 0% -1.85% (p=0.008 n=5+5) Unicode 30.1MB ± 0% 28.8MB ± 0% -4.09% (p=0.008 n=5+5) GoTypes 112MB ± 0% 110MB ± 0% -1.69% (p=0.008 n=5+5) Compiler 470MB ± 0% 461MB ± 0% -1.91% (p=0.008 n=5+5) SSA 1.13GB ± 0% 1.11GB ± 0% -1.70% (p=0.008 n=5+5) Flate 25.0MB ± 0% 24.6MB ± 0% -1.67% (p=0.008 n=5+5) GoParser 31.6MB ± 0% 31.1MB ± 0% -1.66% (p=0.008 n=5+5) Reflect 77.1MB ± 0% 75.8MB ± 0% -1.69% (p=0.008 n=5+5) Tar 26.3MB ± 0% 25.7MB ± 0% -2.06% (p=0.008 n=5+5) XML 41.9MB ± 0% 41.1MB ± 0% -1.93% (p=0.008 n=5+5) [Geo mean] 73.5MB 72.0MB -2.03% name old allocs/op new allocs/op delta Template 383k ± 0% 383k ± 0% ~ (p=0.690 n=5+5) Unicode 343k ± 0% 343k ± 0% ~ (p=0.841 n=5+5) GoTypes 1.16M ± 0% 1.16M ± 0% ~ (p=0.310 n=5+5) Compiler 4.43M ± 0% 4.42M ± 0% -0.17% (p=0.008 n=5+5) SSA 9.85M ± 0% 9.85M ± 0% ~ (p=0.310 n=5+5) Flate 236k ± 0% 236k ± 1% ~ (p=0.841 n=5+5) GoParser 320k ± 0% 320k ± 0% ~ (p=0.421 n=5+5) Reflect 988k ± 0% 987k ± 0% ~ (p=0.690 n=5+5) Tar 252k ± 0% 251k ± 0% ~ (p=0.095 n=5+5) XML 399k ± 0% 399k ± 0% ~ (p=1.000 n=5+5) [Geo mean] 741k 740k -0.07% Change-Id: I9e952b58a98e30a12494304db9ce50d0a85e459c Reviewed-on: https://go-review.googlesource.com/41797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com>
2017-04-25 18:14:12 -07:00
switch n.Class() {
default:
varwrite = true
continue
case PAUTO, PPARAM, PPARAMOUT:
if n.Name.Addrtaken() {
varwrite = true
continue
}
if vmatch2(a, n) {
// Direct hit.
return true
}
}
}
// The variables being written do not appear in n.
// However, n might refer to computed addresses
// that are being written.
// If no computed addresses are affected by the writes, no aliasing.
if !memwrite && !varwrite {
return false
}
// If n does not refer to computed addresses
// (that is, if n only refers to variables whose addresses
// have not been taken), no aliasing.
if varexpr(n) {
return false
}
// Otherwise, both the writes and n refer to computed memory addresses.
// Assume that they might conflict.
return true
}
// does the evaluation of n only refer to variables
// whose addresses have not been taken?
// (and no other memory)
func varexpr(n *Node) bool {
if n == nil {
return true
}
switch n.Op {
case OLITERAL:
return true
case ONAME:
cmd/compile: move Node.Class to flags Put it at position zero, since it is fairly hot. This shrinks gc.Node into a smaller size class on 64 bit systems. name old time/op new time/op delta Template 193ms ± 5% 192ms ± 3% ~ (p=0.353 n=94+93) Unicode 86.1ms ± 5% 85.0ms ± 4% -1.23% (p=0.000 n=95+98) GoTypes 546ms ± 3% 544ms ± 4% -0.40% (p=0.007 n=94+97) Compiler 2.56s ± 3% 2.54s ± 3% -0.67% (p=0.000 n=99+97) SSA 5.13s ± 2% 5.10s ± 3% -0.55% (p=0.000 n=94+98) Flate 122ms ± 6% 121ms ± 4% -0.75% (p=0.002 n=97+95) GoParser 144ms ± 5% 144ms ± 4% ~ (p=0.298 n=98+97) Reflect 348ms ± 4% 349ms ± 4% ~ (p=0.350 n=98+97) Tar 105ms ± 5% 104ms ± 5% ~ (p=0.154 n=96+98) XML 200ms ± 5% 198ms ± 4% -0.71% (p=0.015 n=97+98) [Geo mean] 330ms 328ms -0.52% name old user-time/op new user-time/op delta Template 229ms ±11% 224ms ± 7% -2.16% (p=0.001 n=100+87) Unicode 109ms ± 5% 109ms ± 6% ~ (p=0.897 n=96+91) GoTypes 712ms ± 4% 709ms ± 4% ~ (p=0.085 n=96+98) Compiler 3.41s ± 3% 3.36s ± 3% -1.43% (p=0.000 n=98+98) SSA 7.46s ± 3% 7.31s ± 3% -2.02% (p=0.000 n=100+99) Flate 145ms ± 6% 143ms ± 6% -1.11% (p=0.001 n=99+97) GoParser 177ms ± 5% 176ms ± 5% -0.78% (p=0.018 n=95+95) Reflect 432ms ± 7% 435ms ± 9% ~ (p=0.296 n=100+100) Tar 121ms ± 7% 121ms ± 5% ~ (p=0.072 n=100+95) XML 241ms ± 4% 239ms ± 5% ~ (p=0.085 n=97+99) [Geo mean] 413ms 410ms -0.73% name old alloc/op new alloc/op delta Template 38.4MB ± 0% 37.7MB ± 0% -1.85% (p=0.008 n=5+5) Unicode 30.1MB ± 0% 28.8MB ± 0% -4.09% (p=0.008 n=5+5) GoTypes 112MB ± 0% 110MB ± 0% -1.69% (p=0.008 n=5+5) Compiler 470MB ± 0% 461MB ± 0% -1.91% (p=0.008 n=5+5) SSA 1.13GB ± 0% 1.11GB ± 0% -1.70% (p=0.008 n=5+5) Flate 25.0MB ± 0% 24.6MB ± 0% -1.67% (p=0.008 n=5+5) GoParser 31.6MB ± 0% 31.1MB ± 0% -1.66% (p=0.008 n=5+5) Reflect 77.1MB ± 0% 75.8MB ± 0% -1.69% (p=0.008 n=5+5) Tar 26.3MB ± 0% 25.7MB ± 0% -2.06% (p=0.008 n=5+5) XML 41.9MB ± 0% 41.1MB ± 0% -1.93% (p=0.008 n=5+5) [Geo mean] 73.5MB 72.0MB -2.03% name old allocs/op new allocs/op delta Template 383k ± 0% 383k ± 0% ~ (p=0.690 n=5+5) Unicode 343k ± 0% 343k ± 0% ~ (p=0.841 n=5+5) GoTypes 1.16M ± 0% 1.16M ± 0% ~ (p=0.310 n=5+5) Compiler 4.43M ± 0% 4.42M ± 0% -0.17% (p=0.008 n=5+5) SSA 9.85M ± 0% 9.85M ± 0% ~ (p=0.310 n=5+5) Flate 236k ± 0% 236k ± 1% ~ (p=0.841 n=5+5) GoParser 320k ± 0% 320k ± 0% ~ (p=0.421 n=5+5) Reflect 988k ± 0% 987k ± 0% ~ (p=0.690 n=5+5) Tar 252k ± 0% 251k ± 0% ~ (p=0.095 n=5+5) XML 399k ± 0% 399k ± 0% ~ (p=1.000 n=5+5) [Geo mean] 741k 740k -0.07% Change-Id: I9e952b58a98e30a12494304db9ce50d0a85e459c Reviewed-on: https://go-review.googlesource.com/41797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com>
2017-04-25 18:14:12 -07:00
switch n.Class() {
case PAUTO, PPARAM, PPARAMOUT:
if !n.Name.Addrtaken() {
return true
}
}
return false
case OADD,
OSUB,
OOR,
OXOR,
OMUL,
ODIV,
OMOD,
OLSH,
ORSH,
OAND,
OANDNOT,
OPLUS,
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
ONEG,
OBITNOT,
OPAREN,
OANDAND,
OOROR,
OCONV,
OCONVNOP,
OCONVIFACE,
ODOTTYPE:
return varexpr(n.Left) && varexpr(n.Right)
case ODOT: // but not ODOTPTR
// Should have been handled in aliased.
Fatalf("varexpr unexpected ODOT")
}
// Be conservative.
return false
}
// is the name l mentioned in r?
func vmatch2(l *Node, r *Node) bool {
if r == nil {
return false
}
switch r.Op {
// match each right given left
case ONAME:
return l == r
case OLITERAL:
return false
}
if vmatch2(l, r.Left) {
return true
}
if vmatch2(l, r.Right) {
return true
}
for _, n := range r.List.Slice() {
if vmatch2(l, n) {
return true
}
}
return false
}
// is any name mentioned in l also mentioned in r?
// called by sinit.go
func vmatch1(l *Node, r *Node) bool {
// isolate all left sides
if l == nil || r == nil {
return false
}
switch l.Op {
case ONAME:
cmd/compile: move Node.Class to flags Put it at position zero, since it is fairly hot. This shrinks gc.Node into a smaller size class on 64 bit systems. name old time/op new time/op delta Template 193ms ± 5% 192ms ± 3% ~ (p=0.353 n=94+93) Unicode 86.1ms ± 5% 85.0ms ± 4% -1.23% (p=0.000 n=95+98) GoTypes 546ms ± 3% 544ms ± 4% -0.40% (p=0.007 n=94+97) Compiler 2.56s ± 3% 2.54s ± 3% -0.67% (p=0.000 n=99+97) SSA 5.13s ± 2% 5.10s ± 3% -0.55% (p=0.000 n=94+98) Flate 122ms ± 6% 121ms ± 4% -0.75% (p=0.002 n=97+95) GoParser 144ms ± 5% 144ms ± 4% ~ (p=0.298 n=98+97) Reflect 348ms ± 4% 349ms ± 4% ~ (p=0.350 n=98+97) Tar 105ms ± 5% 104ms ± 5% ~ (p=0.154 n=96+98) XML 200ms ± 5% 198ms ± 4% -0.71% (p=0.015 n=97+98) [Geo mean] 330ms 328ms -0.52% name old user-time/op new user-time/op delta Template 229ms ±11% 224ms ± 7% -2.16% (p=0.001 n=100+87) Unicode 109ms ± 5% 109ms ± 6% ~ (p=0.897 n=96+91) GoTypes 712ms ± 4% 709ms ± 4% ~ (p=0.085 n=96+98) Compiler 3.41s ± 3% 3.36s ± 3% -1.43% (p=0.000 n=98+98) SSA 7.46s ± 3% 7.31s ± 3% -2.02% (p=0.000 n=100+99) Flate 145ms ± 6% 143ms ± 6% -1.11% (p=0.001 n=99+97) GoParser 177ms ± 5% 176ms ± 5% -0.78% (p=0.018 n=95+95) Reflect 432ms ± 7% 435ms ± 9% ~ (p=0.296 n=100+100) Tar 121ms ± 7% 121ms ± 5% ~ (p=0.072 n=100+95) XML 241ms ± 4% 239ms ± 5% ~ (p=0.085 n=97+99) [Geo mean] 413ms 410ms -0.73% name old alloc/op new alloc/op delta Template 38.4MB ± 0% 37.7MB ± 0% -1.85% (p=0.008 n=5+5) Unicode 30.1MB ± 0% 28.8MB ± 0% -4.09% (p=0.008 n=5+5) GoTypes 112MB ± 0% 110MB ± 0% -1.69% (p=0.008 n=5+5) Compiler 470MB ± 0% 461MB ± 0% -1.91% (p=0.008 n=5+5) SSA 1.13GB ± 0% 1.11GB ± 0% -1.70% (p=0.008 n=5+5) Flate 25.0MB ± 0% 24.6MB ± 0% -1.67% (p=0.008 n=5+5) GoParser 31.6MB ± 0% 31.1MB ± 0% -1.66% (p=0.008 n=5+5) Reflect 77.1MB ± 0% 75.8MB ± 0% -1.69% (p=0.008 n=5+5) Tar 26.3MB ± 0% 25.7MB ± 0% -2.06% (p=0.008 n=5+5) XML 41.9MB ± 0% 41.1MB ± 0% -1.93% (p=0.008 n=5+5) [Geo mean] 73.5MB 72.0MB -2.03% name old allocs/op new allocs/op delta Template 383k ± 0% 383k ± 0% ~ (p=0.690 n=5+5) Unicode 343k ± 0% 343k ± 0% ~ (p=0.841 n=5+5) GoTypes 1.16M ± 0% 1.16M ± 0% ~ (p=0.310 n=5+5) Compiler 4.43M ± 0% 4.42M ± 0% -0.17% (p=0.008 n=5+5) SSA 9.85M ± 0% 9.85M ± 0% ~ (p=0.310 n=5+5) Flate 236k ± 0% 236k ± 1% ~ (p=0.841 n=5+5) GoParser 320k ± 0% 320k ± 0% ~ (p=0.421 n=5+5) Reflect 988k ± 0% 987k ± 0% ~ (p=0.690 n=5+5) Tar 252k ± 0% 251k ± 0% ~ (p=0.095 n=5+5) XML 399k ± 0% 399k ± 0% ~ (p=1.000 n=5+5) [Geo mean] 741k 740k -0.07% Change-Id: I9e952b58a98e30a12494304db9ce50d0a85e459c Reviewed-on: https://go-review.googlesource.com/41797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com>
2017-04-25 18:14:12 -07:00
switch l.Class() {
case PPARAM, PAUTO:
break
default:
// assignment to non-stack variable must be
// delayed if right has function calls.
if r.HasCall() {
return true
}
}
return vmatch2(l, r)
case OLITERAL:
return false
}
if vmatch1(l.Left, r) {
return true
}
if vmatch1(l.Right, r) {
return true
}
for _, n := range l.List.Slice() {
if vmatch1(n, r) {
return true
}
}
return false
}
// paramstoheap returns code to allocate memory for heap-escaped parameters
// and to copy non-result parameters' values from the stack.
func paramstoheap(params *types.Type) []*Node {
var nn []*Node
for _, t := range params.Fields().Slice() {
v := asNode(t.Nname)
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
if v != nil && v.Sym != nil && strings.HasPrefix(v.Sym.Name, "~r") { // unnamed result
v = nil
}
if v == nil {
continue
}
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
if stackcopy := v.Name.Param.Stackcopy; stackcopy != nil {
nn = append(nn, walkstmt(nod(ODCL, v, nil)))
cmd/compile: move Node.Class to flags Put it at position zero, since it is fairly hot. This shrinks gc.Node into a smaller size class on 64 bit systems. name old time/op new time/op delta Template 193ms ± 5% 192ms ± 3% ~ (p=0.353 n=94+93) Unicode 86.1ms ± 5% 85.0ms ± 4% -1.23% (p=0.000 n=95+98) GoTypes 546ms ± 3% 544ms ± 4% -0.40% (p=0.007 n=94+97) Compiler 2.56s ± 3% 2.54s ± 3% -0.67% (p=0.000 n=99+97) SSA 5.13s ± 2% 5.10s ± 3% -0.55% (p=0.000 n=94+98) Flate 122ms ± 6% 121ms ± 4% -0.75% (p=0.002 n=97+95) GoParser 144ms ± 5% 144ms ± 4% ~ (p=0.298 n=98+97) Reflect 348ms ± 4% 349ms ± 4% ~ (p=0.350 n=98+97) Tar 105ms ± 5% 104ms ± 5% ~ (p=0.154 n=96+98) XML 200ms ± 5% 198ms ± 4% -0.71% (p=0.015 n=97+98) [Geo mean] 330ms 328ms -0.52% name old user-time/op new user-time/op delta Template 229ms ±11% 224ms ± 7% -2.16% (p=0.001 n=100+87) Unicode 109ms ± 5% 109ms ± 6% ~ (p=0.897 n=96+91) GoTypes 712ms ± 4% 709ms ± 4% ~ (p=0.085 n=96+98) Compiler 3.41s ± 3% 3.36s ± 3% -1.43% (p=0.000 n=98+98) SSA 7.46s ± 3% 7.31s ± 3% -2.02% (p=0.000 n=100+99) Flate 145ms ± 6% 143ms ± 6% -1.11% (p=0.001 n=99+97) GoParser 177ms ± 5% 176ms ± 5% -0.78% (p=0.018 n=95+95) Reflect 432ms ± 7% 435ms ± 9% ~ (p=0.296 n=100+100) Tar 121ms ± 7% 121ms ± 5% ~ (p=0.072 n=100+95) XML 241ms ± 4% 239ms ± 5% ~ (p=0.085 n=97+99) [Geo mean] 413ms 410ms -0.73% name old alloc/op new alloc/op delta Template 38.4MB ± 0% 37.7MB ± 0% -1.85% (p=0.008 n=5+5) Unicode 30.1MB ± 0% 28.8MB ± 0% -4.09% (p=0.008 n=5+5) GoTypes 112MB ± 0% 110MB ± 0% -1.69% (p=0.008 n=5+5) Compiler 470MB ± 0% 461MB ± 0% -1.91% (p=0.008 n=5+5) SSA 1.13GB ± 0% 1.11GB ± 0% -1.70% (p=0.008 n=5+5) Flate 25.0MB ± 0% 24.6MB ± 0% -1.67% (p=0.008 n=5+5) GoParser 31.6MB ± 0% 31.1MB ± 0% -1.66% (p=0.008 n=5+5) Reflect 77.1MB ± 0% 75.8MB ± 0% -1.69% (p=0.008 n=5+5) Tar 26.3MB ± 0% 25.7MB ± 0% -2.06% (p=0.008 n=5+5) XML 41.9MB ± 0% 41.1MB ± 0% -1.93% (p=0.008 n=5+5) [Geo mean] 73.5MB 72.0MB -2.03% name old allocs/op new allocs/op delta Template 383k ± 0% 383k ± 0% ~ (p=0.690 n=5+5) Unicode 343k ± 0% 343k ± 0% ~ (p=0.841 n=5+5) GoTypes 1.16M ± 0% 1.16M ± 0% ~ (p=0.310 n=5+5) Compiler 4.43M ± 0% 4.42M ± 0% -0.17% (p=0.008 n=5+5) SSA 9.85M ± 0% 9.85M ± 0% ~ (p=0.310 n=5+5) Flate 236k ± 0% 236k ± 1% ~ (p=0.841 n=5+5) GoParser 320k ± 0% 320k ± 0% ~ (p=0.421 n=5+5) Reflect 988k ± 0% 987k ± 0% ~ (p=0.690 n=5+5) Tar 252k ± 0% 251k ± 0% ~ (p=0.095 n=5+5) XML 399k ± 0% 399k ± 0% ~ (p=1.000 n=5+5) [Geo mean] 741k 740k -0.07% Change-Id: I9e952b58a98e30a12494304db9ce50d0a85e459c Reviewed-on: https://go-review.googlesource.com/41797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com>
2017-04-25 18:14:12 -07:00
if stackcopy.Class() == PPARAM {
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
nn = append(nn, walkstmt(typecheck(nod(OAS, v, stackcopy), ctxStmt)))
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
}
}
}
return nn
}
// zeroResults zeros the return values at the start of the function.
// We need to do this very early in the function. Defer might stop a
// panic and show the return values as they exist at the time of
// panic. For precise stacks, the garbage collector assumes results
// are always live, so we need to zero them before any allocations,
// even allocations to move params/results to the heap.
// The generated code is added to Curfn's Enter list.
func zeroResults() {
for _, f := range Curfn.Type.Results().Fields().Slice() {
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
v := asNode(f.Nname)
if v != nil && v.Name.Param.Heapaddr != nil {
// The local which points to the return value is the
// thing that needs zeroing. This is already handled
// by a Needzero annotation in plive.go:livenessepilogue.
continue
}
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
if v.isParamHeapCopy() {
// TODO(josharian/khr): Investigate whether we can switch to "continue" here,
// and document more in either case.
// In the review of CL 114797, Keith wrote (roughly):
// I don't think the zeroing below matters.
// The stack return value will never be marked as live anywhere in the function.
// It is not written to until deferreturn returns.
v = v.Name.Param.Stackcopy
}
// Zero the stack location containing f.
cmd/compile: move argument stack construction to SSA generation The goal of this change is to move work from walk to SSA, and simplify things along the way. This is hard to accomplish cleanly with small incremental changes, so this large commit message aims to provide a roadmap to the diff. High level description: Prior to this change, walk was responsible for constructing (most of) the stack for function calls. ascompatte gathered variadic arguments into a slice. It also rewrote n.List from a list of arguments to a list of assignments to stack slots. ascompatte was called multiple times to handle the receiver in a method call. reorder1 then introduced temporaries into n.List as needed to avoid smashing the stack. adjustargs then made extra stack space for go/defer args as needed. Node to SSA construction evaluated all the statements in n.List, and issued the function call, assuming that the stack was correctly constructed. Intrinsic calls had to dig around inside n.List to extract the arguments, since intrinsics don't use the stack to make function calls. This change moves stack construction to the SSA construction phase. ascompatte, now called walkParams, does all the work that ascompatte and reorder1 did. It handles variadic arguments, inserts the method receiver if needed, and allocates temporaries. It does not, however, make any assignments to stack slots. Instead, it moves the function arguments to n.Rlist, leaving assignments to temporaries in n.List. (It would be better to use Ninit instead of List; future work.) During SSA construction, after doing all the temporary assignments in n.List, the function arguments are assigned to stack slots by constructing the appropriate SSA Value, using (*state).storeArg. SSA construction also now handles adjustments for go/defer args. This change also simplifies intrinsic calls, since we no longer need to undo walk's work. Along the way, we simplify nodarg by pushing the fp==1 case to its callers, where it fits nicely. Generated code differences: There were a few optimizations applied along the way, the old way. f(g()) was rewritten to do a block copy of function results to function arguments. And reorder1 avoided introducing the final "save the stack" temporary in n.List. The f(g()) block copy optimization never actually triggered; the order pass rewrote away g(), so that has been removed. SSA optimizations mostly obviated the need for reorder1's optimization of avoiding the final temporary. The exception was when the temporary's type was not SSA-able; in that case, we got a Move into an autotmp and then an immediate Move onto the stack, with the autotmp never read or used again. This change introduces a new rewrite rule to detect such pointless double Moves and collapse them into a single Move. This is actually more powerful than the original optimization, since the original optimization relied on the imprecise Node.HasCall calculation. The other significant difference in the generated code is that the stack is now constructed completely in SP-offset order. Prior to this change, the stack was constructed somewhat haphazardly: first the final argument that Node.HasCall deemed to require a temporary, then other arguments, then the method receiver, then the defer/go args. SP-offset is probably a good default order. See future work. There are a few minor object file size changes as a result of this change. I investigated some regressions in early versions of this change. One regression (in archive/tar) was the addition of a single CMPQ instruction, which would be eliminated were this TODO from flagalloc to be done: // TODO: Remove original instructions if they are never used. One regression (in text/template) was an ADDQconstmodify that is now a regular MOVQLoad+ADDQconst+MOVQStore, due to an unlucky change in the order in which arguments are written. The argument change order can also now be luckier, so this appears to be a wash. All in all, though there will be minor winners and losers, this change appears to be performance neutral. Future work: Move loading the result of function calls to SSA construction; eliminate OINDREGSP. Consider pushing stack construction deeper into SSA world, perhaps in an arch-specific pass. Among other benefits, this would make it easier to transition to a new calling convention. This would require rethinking the handling of stack conflicts and is non-trivial. Figure out some clean way to indicate that stack construction Stores/Moves do not alias each other, so that subsequent passes may do things like CSE+tighten shared stack setup, do DSE using non-first Stores, etc. This would allow us to eliminate the minor text/template regression. Possibly make assignments to stack slots not treated as statements by DWARF. Compiler benchmarks: name old time/op new time/op delta Template 182ms ± 2% 179ms ± 2% -1.69% (p=0.000 n=47+48) Unicode 86.3ms ± 5% 85.1ms ± 4% -1.36% (p=0.001 n=50+50) GoTypes 646ms ± 1% 642ms ± 1% -0.63% (p=0.000 n=49+48) Compiler 2.89s ± 1% 2.86s ± 2% -1.36% (p=0.000 n=48+50) SSA 8.47s ± 1% 8.37s ± 2% -1.22% (p=0.000 n=47+50) Flate 122ms ± 2% 121ms ± 2% -0.66% (p=0.000 n=47+45) GoParser 147ms ± 2% 146ms ± 2% -0.53% (p=0.006 n=46+49) Reflect 406ms ± 2% 403ms ± 2% -0.76% (p=0.000 n=48+43) Tar 162ms ± 3% 162ms ± 4% ~ (p=0.191 n=46+50) XML 223ms ± 2% 222ms ± 2% -0.37% (p=0.031 n=45+49) [Geo mean] 382ms 378ms -0.89% name old user-time/op new user-time/op delta Template 219ms ± 3% 216ms ± 3% -1.56% (p=0.000 n=50+48) Unicode 109ms ± 6% 109ms ± 5% ~ (p=0.190 n=50+49) GoTypes 836ms ± 2% 828ms ± 2% -0.96% (p=0.000 n=49+48) Compiler 3.87s ± 2% 3.80s ± 1% -1.81% (p=0.000 n=49+46) SSA 12.0s ± 1% 11.8s ± 1% -2.01% (p=0.000 n=48+50) Flate 142ms ± 3% 141ms ± 3% -0.85% (p=0.003 n=50+48) GoParser 178ms ± 4% 175ms ± 4% -1.66% (p=0.000 n=48+46) Reflect 520ms ± 2% 512ms ± 2% -1.44% (p=0.000 n=45+48) Tar 200ms ± 3% 198ms ± 4% -0.61% (p=0.037 n=47+50) XML 277ms ± 3% 275ms ± 3% -0.85% (p=0.000 n=49+48) [Geo mean] 482ms 476ms -1.23% name old alloc/op new alloc/op delta Template 36.1MB ± 0% 35.3MB ± 0% -2.18% (p=0.008 n=5+5) Unicode 29.8MB ± 0% 29.3MB ± 0% -1.58% (p=0.008 n=5+5) GoTypes 125MB ± 0% 123MB ± 0% -2.13% (p=0.008 n=5+5) Compiler 531MB ± 0% 513MB ± 0% -3.40% (p=0.008 n=5+5) SSA 2.00GB ± 0% 1.93GB ± 0% -3.34% (p=0.008 n=5+5) Flate 24.5MB ± 0% 24.3MB ± 0% -1.18% (p=0.008 n=5+5) GoParser 29.4MB ± 0% 28.7MB ± 0% -2.34% (p=0.008 n=5+5) Reflect 87.1MB ± 0% 86.0MB ± 0% -1.33% (p=0.008 n=5+5) Tar 35.3MB ± 0% 34.8MB ± 0% -1.44% (p=0.008 n=5+5) XML 47.9MB ± 0% 47.1MB ± 0% -1.86% (p=0.008 n=5+5) [Geo mean] 82.8MB 81.1MB -2.08% name old allocs/op new allocs/op delta Template 352k ± 0% 347k ± 0% -1.32% (p=0.008 n=5+5) Unicode 342k ± 0% 339k ± 0% -0.66% (p=0.008 n=5+5) GoTypes 1.29M ± 0% 1.27M ± 0% -1.30% (p=0.008 n=5+5) Compiler 4.98M ± 0% 4.87M ± 0% -2.14% (p=0.008 n=5+5) SSA 15.7M ± 0% 15.2M ± 0% -2.86% (p=0.008 n=5+5) Flate 233k ± 0% 231k ± 0% -0.83% (p=0.008 n=5+5) GoParser 296k ± 0% 291k ± 0% -1.54% (p=0.016 n=5+4) Reflect 1.05M ± 0% 1.04M ± 0% -0.65% (p=0.008 n=5+5) Tar 343k ± 0% 339k ± 0% -0.97% (p=0.008 n=5+5) XML 432k ± 0% 426k ± 0% -1.19% (p=0.008 n=5+5) [Geo mean] 815k 804k -1.35% name old object-bytes new object-bytes delta Template 505kB ± 0% 505kB ± 0% -0.01% (p=0.008 n=5+5) Unicode 224kB ± 0% 224kB ± 0% ~ (all equal) GoTypes 1.82MB ± 0% 1.83MB ± 0% +0.06% (p=0.008 n=5+5) Flate 324kB ± 0% 324kB ± 0% +0.00% (p=0.008 n=5+5) GoParser 402kB ± 0% 402kB ± 0% +0.04% (p=0.008 n=5+5) Reflect 1.39MB ± 0% 1.39MB ± 0% -0.01% (p=0.008 n=5+5) Tar 449kB ± 0% 449kB ± 0% -0.02% (p=0.008 n=5+5) XML 598kB ± 0% 597kB ± 0% -0.05% (p=0.008 n=5+5) Change-Id: Ifc9d5c1bd01f90171414b8fb18ffe2290d271143 Reviewed-on: https://go-review.googlesource.com/c/114797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-05-06 12:58:53 -07:00
Curfn.Func.Enter.Append(nodl(Curfn.Pos, OAS, v, nil))
}
}
// returnsfromheap returns code to copy values for heap-escaped parameters
// back to the stack.
func returnsfromheap(params *types.Type) []*Node {
var nn []*Node
for _, t := range params.Fields().Slice() {
v := asNode(t.Nname)
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
if v == nil {
continue
}
cmd/compile: move Node.Class to flags Put it at position zero, since it is fairly hot. This shrinks gc.Node into a smaller size class on 64 bit systems. name old time/op new time/op delta Template 193ms ± 5% 192ms ± 3% ~ (p=0.353 n=94+93) Unicode 86.1ms ± 5% 85.0ms ± 4% -1.23% (p=0.000 n=95+98) GoTypes 546ms ± 3% 544ms ± 4% -0.40% (p=0.007 n=94+97) Compiler 2.56s ± 3% 2.54s ± 3% -0.67% (p=0.000 n=99+97) SSA 5.13s ± 2% 5.10s ± 3% -0.55% (p=0.000 n=94+98) Flate 122ms ± 6% 121ms ± 4% -0.75% (p=0.002 n=97+95) GoParser 144ms ± 5% 144ms ± 4% ~ (p=0.298 n=98+97) Reflect 348ms ± 4% 349ms ± 4% ~ (p=0.350 n=98+97) Tar 105ms ± 5% 104ms ± 5% ~ (p=0.154 n=96+98) XML 200ms ± 5% 198ms ± 4% -0.71% (p=0.015 n=97+98) [Geo mean] 330ms 328ms -0.52% name old user-time/op new user-time/op delta Template 229ms ±11% 224ms ± 7% -2.16% (p=0.001 n=100+87) Unicode 109ms ± 5% 109ms ± 6% ~ (p=0.897 n=96+91) GoTypes 712ms ± 4% 709ms ± 4% ~ (p=0.085 n=96+98) Compiler 3.41s ± 3% 3.36s ± 3% -1.43% (p=0.000 n=98+98) SSA 7.46s ± 3% 7.31s ± 3% -2.02% (p=0.000 n=100+99) Flate 145ms ± 6% 143ms ± 6% -1.11% (p=0.001 n=99+97) GoParser 177ms ± 5% 176ms ± 5% -0.78% (p=0.018 n=95+95) Reflect 432ms ± 7% 435ms ± 9% ~ (p=0.296 n=100+100) Tar 121ms ± 7% 121ms ± 5% ~ (p=0.072 n=100+95) XML 241ms ± 4% 239ms ± 5% ~ (p=0.085 n=97+99) [Geo mean] 413ms 410ms -0.73% name old alloc/op new alloc/op delta Template 38.4MB ± 0% 37.7MB ± 0% -1.85% (p=0.008 n=5+5) Unicode 30.1MB ± 0% 28.8MB ± 0% -4.09% (p=0.008 n=5+5) GoTypes 112MB ± 0% 110MB ± 0% -1.69% (p=0.008 n=5+5) Compiler 470MB ± 0% 461MB ± 0% -1.91% (p=0.008 n=5+5) SSA 1.13GB ± 0% 1.11GB ± 0% -1.70% (p=0.008 n=5+5) Flate 25.0MB ± 0% 24.6MB ± 0% -1.67% (p=0.008 n=5+5) GoParser 31.6MB ± 0% 31.1MB ± 0% -1.66% (p=0.008 n=5+5) Reflect 77.1MB ± 0% 75.8MB ± 0% -1.69% (p=0.008 n=5+5) Tar 26.3MB ± 0% 25.7MB ± 0% -2.06% (p=0.008 n=5+5) XML 41.9MB ± 0% 41.1MB ± 0% -1.93% (p=0.008 n=5+5) [Geo mean] 73.5MB 72.0MB -2.03% name old allocs/op new allocs/op delta Template 383k ± 0% 383k ± 0% ~ (p=0.690 n=5+5) Unicode 343k ± 0% 343k ± 0% ~ (p=0.841 n=5+5) GoTypes 1.16M ± 0% 1.16M ± 0% ~ (p=0.310 n=5+5) Compiler 4.43M ± 0% 4.42M ± 0% -0.17% (p=0.008 n=5+5) SSA 9.85M ± 0% 9.85M ± 0% ~ (p=0.310 n=5+5) Flate 236k ± 0% 236k ± 1% ~ (p=0.841 n=5+5) GoParser 320k ± 0% 320k ± 0% ~ (p=0.421 n=5+5) Reflect 988k ± 0% 987k ± 0% ~ (p=0.690 n=5+5) Tar 252k ± 0% 251k ± 0% ~ (p=0.095 n=5+5) XML 399k ± 0% 399k ± 0% ~ (p=1.000 n=5+5) [Geo mean] 741k 740k -0.07% Change-Id: I9e952b58a98e30a12494304db9ce50d0a85e459c Reviewed-on: https://go-review.googlesource.com/41797 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com>
2017-04-25 18:14:12 -07:00
if stackcopy := v.Name.Param.Stackcopy; stackcopy != nil && stackcopy.Class() == PPARAMOUT {
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
nn = append(nn, walkstmt(typecheck(nod(OAS, stackcopy, v), ctxStmt)))
cmd/compile: fix liveness computation for heap-escaped parameters The liveness computation of parameters generally was never correct, but forcing all parameters to be live throughout the function covered up that problem. The new SSA back end is too clever: even though it currently keeps the parameter values live throughout the function, it may find optimizations that mean the current values are not written back to the original parameter stack slots immediately or ever (for example if a parameter is set to nil, SSA constant propagation may replace all later uses of the parameter with a constant nil, eliminating the need to write the nil value back to the stack slot), so the liveness code must now track the actual operations on the stack slots, exposing these problems. One small problem in the handling of arguments is that nodarg can return ONAME PPARAM nodes with adjusted offsets, so that there are actually multiple *Node pointers for the same parameter in the instruction stream. This might be possible to correct, but not in this CL. For now, we fix this by using n.Orig instead of n when considering PPARAM and PPARAMOUT nodes. The major problem in the handling of arguments is general confusion in the liveness code about the meaning of PPARAM|PHEAP and PPARAMOUT|PHEAP nodes, especially as contrasted with PAUTO|PHEAP. The difference between these two is that when a local variable "moves" to the heap, it's really just allocated there to start with; in contrast, when an argument moves to the heap, the actual data has to be copied there from the stack at the beginning of the function, and when a result "moves" to the heap the value in the heap has to be copied back to the stack when the function returns This general confusion is also present in the SSA back end. The PHEAP bit worked decently when I first introduced it 7 years ago (!) in 391425ae. The back end did nothing sophisticated, and in particular there was no analysis at all: no escape analysis, no liveness analysis, and certainly no SSA back end. But the complications caused in the various downstream consumers suggest that this should be a detail kept mainly in the front end. This CL therefore eliminates both the PHEAP bit and even the idea of "heap variables" from the back ends. First, it replaces the PPARAM|PHEAP, PPARAMOUT|PHEAP, and PAUTO|PHEAP variable classes with the single PAUTOHEAP, a pseudo-class indicating a variable maintained on the heap and available by indirecting a local variable kept on the stack (a plain PAUTO). Second, walkexpr replaces all references to PAUTOHEAP variables with indirections of the corresponding PAUTO variable. The back ends and the liveness code now just see plain indirected variables. This may actually produce better code, but the real goal here is to eliminate these little-used and somewhat suspect code paths in the back end analyses. The OPARAM node type goes away too. A followup CL will do the same to PPARAMREF. I'm not sure that the back ends (SSA in particular) are handling those right either, and with the framework established in this CL that change is trivial and the result clearly more correct. Fixes #15747. Change-Id: I2770b1ce3cbc93981bfc7166be66a9da12013d74 Reviewed-on: https://go-review.googlesource.com/23393 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Russ Cox <rsc@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2016-05-25 01:33:24 -04:00
}
}
return nn
}
// heapmoves generates code to handle migrating heap-escaped parameters
// between the stack and the heap. The generated code is added to Curfn's
// Enter and Exit lists.
func heapmoves() {
lno := lineno
lineno = Curfn.Pos
nn := paramstoheap(Curfn.Type.Recvs())
nn = append(nn, paramstoheap(Curfn.Type.Params())...)
nn = append(nn, paramstoheap(Curfn.Type.Results())...)
Curfn.Func.Enter.Append(nn...)
lineno = Curfn.Func.Endlineno
Curfn.Func.Exit.Append(returnsfromheap(Curfn.Type.Results())...)
lineno = lno
}
func vmkcall(fn *Node, t *types.Type, init *Nodes, va []*Node) *Node {
if fn.Type == nil || fn.Type.Etype != TFUNC {
Fatalf("mkcall %v %v", fn, fn.Type)
}
n := fn.Type.NumParams()
if n != len(va) {
Fatalf("vmkcall %v needs %v args got %v", fn, n, len(va))
}
r := nod(OCALL, fn, nil)
r.List.Set(va)
if fn.Type.NumResults() > 0 {
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
r = typecheck(r, ctxExpr|ctxMultiOK)
} else {
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
r = typecheck(r, ctxStmt)
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
r = walkexpr(r, init)
r.Type = t
return r
}
func mkcall(name string, t *types.Type, init *Nodes, args ...*Node) *Node {
return vmkcall(syslook(name), t, init, args)
}
func mkcall1(fn *Node, t *types.Type, init *Nodes, args ...*Node) *Node {
return vmkcall(fn, t, init, args)
}
func conv(n *Node, t *types.Type) *Node {
if types.Identical(n.Type, t) {
return n
}
n = nod(OCONV, n, nil)
n.Type = t
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
n = typecheck(n, ctxExpr)
return n
}
// convnop converts node n to type t using the OCONVNOP op
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
// and typechecks the result with ctxExpr.
func convnop(n *Node, t *types.Type) *Node {
if types.Identical(n.Type, t) {
return n
}
n = nod(OCONVNOP, n, nil)
n.Type = t
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
n = typecheck(n, ctxExpr)
return n
}
// byteindex converts n, which is byte-sized, to an int used to index into an array.
// We cannot use conv, because we allow converting bool to int here,
// which is forbidden in user code.
func byteindex(n *Node) *Node {
// We cannot convert from bool to int directly.
// While converting from int8 to int is possible, it would yield
// the wrong result for negative values.
// Reinterpreting the value as an unsigned byte solves both cases.
if !types.Identical(n.Type, types.Types[TUINT8]) {
n = nod(OCONV, n, nil)
n.Type = types.Types[TUINT8]
n.SetTypecheck(1)
}
n = nod(OCONV, n, nil)
n.Type = types.Types[TINT]
n.SetTypecheck(1)
return n
}
func chanfn(name string, n int, t *types.Type) *Node {
if !t.IsChan() {
Fatalf("chanfn %v", t)
}
fn := syslook(name)
switch n {
default:
Fatalf("chanfn %d", n)
case 1:
fn = substArgTypes(fn, t.Elem())
case 2:
fn = substArgTypes(fn, t.Elem(), t.Elem())
}
return fn
}
func mapfn(name string, t *types.Type) *Node {
if !t.IsMap() {
Fatalf("mapfn %v", t)
}
fn := syslook(name)
fn = substArgTypes(fn, t.Key(), t.Elem(), t.Key(), t.Elem())
return fn
}
func mapfndel(name string, t *types.Type) *Node {
if !t.IsMap() {
Fatalf("mapfn %v", t)
}
fn := syslook(name)
fn = substArgTypes(fn, t.Key(), t.Elem(), t.Key())
return fn
}
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
const (
mapslow = iota
mapfast32
mapfast32ptr
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
mapfast64
mapfast64ptr
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
mapfaststr
nmapfast
)
type mapnames [nmapfast]string
func mkmapnames(base string, ptr string) mapnames {
return mapnames{base, base + "_fast32", base + "_fast32" + ptr, base + "_fast64", base + "_fast64" + ptr, base + "_faststr"}
}
var mapaccess1 = mkmapnames("mapaccess1", "")
var mapaccess2 = mkmapnames("mapaccess2", "")
var mapassign = mkmapnames("mapassign", "ptr")
var mapdelete = mkmapnames("mapdelete", "")
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
func mapfast(t *types.Type) int {
// Check runtime/map.go:maxElemSize before changing.
if t.Elem().Width > 128 {
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
return mapslow
runtime: add mapassign_fast* Add benchmarks for map assignment with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapAssignInt32_255-8 24.7ns ± 3% 17.4ns ± 2% -29.75% (p=0.000 n=10+10) MapAssignInt32_64k-8 45.5ns ± 4% 37.6ns ± 4% -17.18% (p=0.000 n=10+10) MapAssignInt64_255-8 26.0ns ± 3% 17.9ns ± 4% -31.03% (p=0.000 n=10+10) MapAssignInt64_64k-8 46.9ns ± 5% 38.7ns ± 2% -17.53% (p=0.000 n=9+10) MapAssignStr_255-8 47.8ns ± 3% 24.8ns ± 4% -48.01% (p=0.000 n=10+10) MapAssignStr_64k-8 83.0ns ± 3% 51.9ns ± 3% -37.45% (p=0.000 n=10+9) name old time/op new time/op delta BinaryTree17-8 3.11s ±19% 2.78s ± 3% ~ (p=0.095 n=5+5) Fannkuch11-8 3.26s ± 1% 3.21s ± 2% ~ (p=0.056 n=5+5) FmtFprintfEmpty-8 50.3ns ± 1% 50.8ns ± 2% ~ (p=0.246 n=5+5) FmtFprintfString-8 82.7ns ± 4% 80.1ns ± 5% ~ (p=0.238 n=5+5) FmtFprintfInt-8 82.6ns ± 2% 81.9ns ± 3% ~ (p=0.508 n=5+5) FmtFprintfIntInt-8 124ns ± 4% 121ns ± 3% ~ (p=0.111 n=5+5) FmtFprintfPrefixedInt-8 158ns ± 6% 160ns ± 2% ~ (p=0.341 n=5+5) FmtFprintfFloat-8 249ns ± 2% 245ns ± 2% ~ (p=0.095 n=5+5) FmtManyArgs-8 513ns ± 2% 519ns ± 3% ~ (p=0.151 n=5+5) GobDecode-8 7.48ms ±12% 7.11ms ± 2% ~ (p=0.222 n=5+5) GobEncode-8 6.25ms ± 1% 6.03ms ± 2% -3.56% (p=0.008 n=5+5) Gzip-8 252ms ± 4% 252ms ± 4% ~ (p=1.000 n=5+5) Gunzip-8 38.4ms ± 3% 38.6ms ± 2% ~ (p=0.690 n=5+5) HTTPClientServer-8 76.9µs ±41% 66.4µs ± 6% ~ (p=0.310 n=5+5) JSONEncode-8 16.5ms ± 3% 16.7ms ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 54.6ms ± 1% 54.3ms ± 2% ~ (p=0.548 n=5+5) Mandelbrot200-8 4.45ms ± 3% 4.47ms ± 1% ~ (p=0.841 n=5+5) GoParse-8 3.43ms ± 1% 3.32ms ± 2% -3.28% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 88.2ns ± 3% 89.4ns ± 2% ~ (p=0.333 n=5+5) RegexpMatchEasy0_1K-8 205ns ± 1% 206ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchEasy1_32-8 85.1ns ± 1% 85.5ns ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 365ns ± 1% 371ns ± 9% ~ (p=1.000 n=5+5) RegexpMatchMedium_32-8 129ns ± 2% 128ns ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 39.8µs ± 0% 39.7µs ± 4% ~ (p=0.730 n=4+5) RegexpMatchHard_32-8 1.99µs ± 3% 2.05µs ±16% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 59.3µs ± 1% 60.3µs ± 7% ~ (p=1.000 n=5+5) Revcomp-8 1.36s ±63% 0.52s ± 5% ~ (p=0.095 n=5+5) Template-8 62.6ms ±14% 60.5ms ± 5% ~ (p=0.690 n=5+5) TimeParse-8 330ns ± 2% 324ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 350ns ± 3% 340ns ± 1% -2.86% (p=0.008 n=5+5) name old speed new speed delta GobDecode-8 103MB/s ±11% 108MB/s ± 2% ~ (p=0.222 n=5+5) GobEncode-8 123MB/s ± 1% 127MB/s ± 2% +3.71% (p=0.008 n=5+5) Gzip-8 77.1MB/s ± 4% 76.9MB/s ± 3% ~ (p=1.000 n=5+5) Gunzip-8 505MB/s ± 3% 503MB/s ± 2% ~ (p=0.690 n=5+5) JSONEncode-8 118MB/s ± 3% 116MB/s ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 35.5MB/s ± 1% 35.8MB/s ± 2% ~ (p=0.397 n=5+5) GoParse-8 16.9MB/s ± 1% 17.4MB/s ± 2% +3.45% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 363MB/s ± 3% 358MB/s ± 2% ~ (p=0.421 n=5+5) RegexpMatchEasy0_1K-8 4.98GB/s ± 1% 4.97GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 376MB/s ± 1% 375MB/s ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 2.80GB/s ± 1% 2.76GB/s ± 9% ~ (p=0.841 n=5+5) RegexpMatchMedium_32-8 7.73MB/s ± 1% 7.76MB/s ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 25.8MB/s ± 0% 25.8MB/s ± 4% ~ (p=0.651 n=4+5) RegexpMatchHard_32-8 16.1MB/s ± 3% 15.7MB/s ±14% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 17.3MB/s ± 1% 17.0MB/s ± 7% ~ (p=0.984 n=5+5) Revcomp-8 273MB/s ±83% 488MB/s ± 5% ~ (p=0.095 n=5+5) Template-8 31.1MB/s ±13% 32.1MB/s ± 5% ~ (p=0.690 n=5+5) Updates #19495 Change-Id: I116e9a2a4594769318b22d736464de8a98499909 Reviewed-on: https://go-review.googlesource.com/38091 Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-12 14:47:59 -07:00
}
switch algtype(t.Key()) {
case AMEM32:
if !t.Key().HasPointers() {
return mapfast32
}
if Widthptr == 4 {
return mapfast32ptr
}
Fatalf("small pointer %v", t.Key())
runtime: add mapassign_fast* Add benchmarks for map assignment with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapAssignInt32_255-8 24.7ns ± 3% 17.4ns ± 2% -29.75% (p=0.000 n=10+10) MapAssignInt32_64k-8 45.5ns ± 4% 37.6ns ± 4% -17.18% (p=0.000 n=10+10) MapAssignInt64_255-8 26.0ns ± 3% 17.9ns ± 4% -31.03% (p=0.000 n=10+10) MapAssignInt64_64k-8 46.9ns ± 5% 38.7ns ± 2% -17.53% (p=0.000 n=9+10) MapAssignStr_255-8 47.8ns ± 3% 24.8ns ± 4% -48.01% (p=0.000 n=10+10) MapAssignStr_64k-8 83.0ns ± 3% 51.9ns ± 3% -37.45% (p=0.000 n=10+9) name old time/op new time/op delta BinaryTree17-8 3.11s ±19% 2.78s ± 3% ~ (p=0.095 n=5+5) Fannkuch11-8 3.26s ± 1% 3.21s ± 2% ~ (p=0.056 n=5+5) FmtFprintfEmpty-8 50.3ns ± 1% 50.8ns ± 2% ~ (p=0.246 n=5+5) FmtFprintfString-8 82.7ns ± 4% 80.1ns ± 5% ~ (p=0.238 n=5+5) FmtFprintfInt-8 82.6ns ± 2% 81.9ns ± 3% ~ (p=0.508 n=5+5) FmtFprintfIntInt-8 124ns ± 4% 121ns ± 3% ~ (p=0.111 n=5+5) FmtFprintfPrefixedInt-8 158ns ± 6% 160ns ± 2% ~ (p=0.341 n=5+5) FmtFprintfFloat-8 249ns ± 2% 245ns ± 2% ~ (p=0.095 n=5+5) FmtManyArgs-8 513ns ± 2% 519ns ± 3% ~ (p=0.151 n=5+5) GobDecode-8 7.48ms ±12% 7.11ms ± 2% ~ (p=0.222 n=5+5) GobEncode-8 6.25ms ± 1% 6.03ms ± 2% -3.56% (p=0.008 n=5+5) Gzip-8 252ms ± 4% 252ms ± 4% ~ (p=1.000 n=5+5) Gunzip-8 38.4ms ± 3% 38.6ms ± 2% ~ (p=0.690 n=5+5) HTTPClientServer-8 76.9µs ±41% 66.4µs ± 6% ~ (p=0.310 n=5+5) JSONEncode-8 16.5ms ± 3% 16.7ms ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 54.6ms ± 1% 54.3ms ± 2% ~ (p=0.548 n=5+5) Mandelbrot200-8 4.45ms ± 3% 4.47ms ± 1% ~ (p=0.841 n=5+5) GoParse-8 3.43ms ± 1% 3.32ms ± 2% -3.28% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 88.2ns ± 3% 89.4ns ± 2% ~ (p=0.333 n=5+5) RegexpMatchEasy0_1K-8 205ns ± 1% 206ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchEasy1_32-8 85.1ns ± 1% 85.5ns ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 365ns ± 1% 371ns ± 9% ~ (p=1.000 n=5+5) RegexpMatchMedium_32-8 129ns ± 2% 128ns ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 39.8µs ± 0% 39.7µs ± 4% ~ (p=0.730 n=4+5) RegexpMatchHard_32-8 1.99µs ± 3% 2.05µs ±16% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 59.3µs ± 1% 60.3µs ± 7% ~ (p=1.000 n=5+5) Revcomp-8 1.36s ±63% 0.52s ± 5% ~ (p=0.095 n=5+5) Template-8 62.6ms ±14% 60.5ms ± 5% ~ (p=0.690 n=5+5) TimeParse-8 330ns ± 2% 324ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 350ns ± 3% 340ns ± 1% -2.86% (p=0.008 n=5+5) name old speed new speed delta GobDecode-8 103MB/s ±11% 108MB/s ± 2% ~ (p=0.222 n=5+5) GobEncode-8 123MB/s ± 1% 127MB/s ± 2% +3.71% (p=0.008 n=5+5) Gzip-8 77.1MB/s ± 4% 76.9MB/s ± 3% ~ (p=1.000 n=5+5) Gunzip-8 505MB/s ± 3% 503MB/s ± 2% ~ (p=0.690 n=5+5) JSONEncode-8 118MB/s ± 3% 116MB/s ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 35.5MB/s ± 1% 35.8MB/s ± 2% ~ (p=0.397 n=5+5) GoParse-8 16.9MB/s ± 1% 17.4MB/s ± 2% +3.45% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 363MB/s ± 3% 358MB/s ± 2% ~ (p=0.421 n=5+5) RegexpMatchEasy0_1K-8 4.98GB/s ± 1% 4.97GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 376MB/s ± 1% 375MB/s ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 2.80GB/s ± 1% 2.76GB/s ± 9% ~ (p=0.841 n=5+5) RegexpMatchMedium_32-8 7.73MB/s ± 1% 7.76MB/s ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 25.8MB/s ± 0% 25.8MB/s ± 4% ~ (p=0.651 n=4+5) RegexpMatchHard_32-8 16.1MB/s ± 3% 15.7MB/s ±14% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 17.3MB/s ± 1% 17.0MB/s ± 7% ~ (p=0.984 n=5+5) Revcomp-8 273MB/s ±83% 488MB/s ± 5% ~ (p=0.095 n=5+5) Template-8 31.1MB/s ±13% 32.1MB/s ± 5% ~ (p=0.690 n=5+5) Updates #19495 Change-Id: I116e9a2a4594769318b22d736464de8a98499909 Reviewed-on: https://go-review.googlesource.com/38091 Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-12 14:47:59 -07:00
case AMEM64:
if !t.Key().HasPointers() {
return mapfast64
}
if Widthptr == 8 {
return mapfast64ptr
}
// Two-word object, at least one of which is a pointer.
// Use the slow path.
runtime: add mapassign_fast* Add benchmarks for map assignment with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapAssignInt32_255-8 24.7ns ± 3% 17.4ns ± 2% -29.75% (p=0.000 n=10+10) MapAssignInt32_64k-8 45.5ns ± 4% 37.6ns ± 4% -17.18% (p=0.000 n=10+10) MapAssignInt64_255-8 26.0ns ± 3% 17.9ns ± 4% -31.03% (p=0.000 n=10+10) MapAssignInt64_64k-8 46.9ns ± 5% 38.7ns ± 2% -17.53% (p=0.000 n=9+10) MapAssignStr_255-8 47.8ns ± 3% 24.8ns ± 4% -48.01% (p=0.000 n=10+10) MapAssignStr_64k-8 83.0ns ± 3% 51.9ns ± 3% -37.45% (p=0.000 n=10+9) name old time/op new time/op delta BinaryTree17-8 3.11s ±19% 2.78s ± 3% ~ (p=0.095 n=5+5) Fannkuch11-8 3.26s ± 1% 3.21s ± 2% ~ (p=0.056 n=5+5) FmtFprintfEmpty-8 50.3ns ± 1% 50.8ns ± 2% ~ (p=0.246 n=5+5) FmtFprintfString-8 82.7ns ± 4% 80.1ns ± 5% ~ (p=0.238 n=5+5) FmtFprintfInt-8 82.6ns ± 2% 81.9ns ± 3% ~ (p=0.508 n=5+5) FmtFprintfIntInt-8 124ns ± 4% 121ns ± 3% ~ (p=0.111 n=5+5) FmtFprintfPrefixedInt-8 158ns ± 6% 160ns ± 2% ~ (p=0.341 n=5+5) FmtFprintfFloat-8 249ns ± 2% 245ns ± 2% ~ (p=0.095 n=5+5) FmtManyArgs-8 513ns ± 2% 519ns ± 3% ~ (p=0.151 n=5+5) GobDecode-8 7.48ms ±12% 7.11ms ± 2% ~ (p=0.222 n=5+5) GobEncode-8 6.25ms ± 1% 6.03ms ± 2% -3.56% (p=0.008 n=5+5) Gzip-8 252ms ± 4% 252ms ± 4% ~ (p=1.000 n=5+5) Gunzip-8 38.4ms ± 3% 38.6ms ± 2% ~ (p=0.690 n=5+5) HTTPClientServer-8 76.9µs ±41% 66.4µs ± 6% ~ (p=0.310 n=5+5) JSONEncode-8 16.5ms ± 3% 16.7ms ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 54.6ms ± 1% 54.3ms ± 2% ~ (p=0.548 n=5+5) Mandelbrot200-8 4.45ms ± 3% 4.47ms ± 1% ~ (p=0.841 n=5+5) GoParse-8 3.43ms ± 1% 3.32ms ± 2% -3.28% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 88.2ns ± 3% 89.4ns ± 2% ~ (p=0.333 n=5+5) RegexpMatchEasy0_1K-8 205ns ± 1% 206ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchEasy1_32-8 85.1ns ± 1% 85.5ns ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 365ns ± 1% 371ns ± 9% ~ (p=1.000 n=5+5) RegexpMatchMedium_32-8 129ns ± 2% 128ns ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 39.8µs ± 0% 39.7µs ± 4% ~ (p=0.730 n=4+5) RegexpMatchHard_32-8 1.99µs ± 3% 2.05µs ±16% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 59.3µs ± 1% 60.3µs ± 7% ~ (p=1.000 n=5+5) Revcomp-8 1.36s ±63% 0.52s ± 5% ~ (p=0.095 n=5+5) Template-8 62.6ms ±14% 60.5ms ± 5% ~ (p=0.690 n=5+5) TimeParse-8 330ns ± 2% 324ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 350ns ± 3% 340ns ± 1% -2.86% (p=0.008 n=5+5) name old speed new speed delta GobDecode-8 103MB/s ±11% 108MB/s ± 2% ~ (p=0.222 n=5+5) GobEncode-8 123MB/s ± 1% 127MB/s ± 2% +3.71% (p=0.008 n=5+5) Gzip-8 77.1MB/s ± 4% 76.9MB/s ± 3% ~ (p=1.000 n=5+5) Gunzip-8 505MB/s ± 3% 503MB/s ± 2% ~ (p=0.690 n=5+5) JSONEncode-8 118MB/s ± 3% 116MB/s ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 35.5MB/s ± 1% 35.8MB/s ± 2% ~ (p=0.397 n=5+5) GoParse-8 16.9MB/s ± 1% 17.4MB/s ± 2% +3.45% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 363MB/s ± 3% 358MB/s ± 2% ~ (p=0.421 n=5+5) RegexpMatchEasy0_1K-8 4.98GB/s ± 1% 4.97GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 376MB/s ± 1% 375MB/s ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 2.80GB/s ± 1% 2.76GB/s ± 9% ~ (p=0.841 n=5+5) RegexpMatchMedium_32-8 7.73MB/s ± 1% 7.76MB/s ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 25.8MB/s ± 0% 25.8MB/s ± 4% ~ (p=0.651 n=4+5) RegexpMatchHard_32-8 16.1MB/s ± 3% 15.7MB/s ±14% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 17.3MB/s ± 1% 17.0MB/s ± 7% ~ (p=0.984 n=5+5) Revcomp-8 273MB/s ±83% 488MB/s ± 5% ~ (p=0.095 n=5+5) Template-8 31.1MB/s ±13% 32.1MB/s ± 5% ~ (p=0.690 n=5+5) Updates #19495 Change-Id: I116e9a2a4594769318b22d736464de8a98499909 Reviewed-on: https://go-review.googlesource.com/38091 Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-12 14:47:59 -07:00
case ASTRING:
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
return mapfaststr
runtime: add mapassign_fast* Add benchmarks for map assignment with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapAssignInt32_255-8 24.7ns ± 3% 17.4ns ± 2% -29.75% (p=0.000 n=10+10) MapAssignInt32_64k-8 45.5ns ± 4% 37.6ns ± 4% -17.18% (p=0.000 n=10+10) MapAssignInt64_255-8 26.0ns ± 3% 17.9ns ± 4% -31.03% (p=0.000 n=10+10) MapAssignInt64_64k-8 46.9ns ± 5% 38.7ns ± 2% -17.53% (p=0.000 n=9+10) MapAssignStr_255-8 47.8ns ± 3% 24.8ns ± 4% -48.01% (p=0.000 n=10+10) MapAssignStr_64k-8 83.0ns ± 3% 51.9ns ± 3% -37.45% (p=0.000 n=10+9) name old time/op new time/op delta BinaryTree17-8 3.11s ±19% 2.78s ± 3% ~ (p=0.095 n=5+5) Fannkuch11-8 3.26s ± 1% 3.21s ± 2% ~ (p=0.056 n=5+5) FmtFprintfEmpty-8 50.3ns ± 1% 50.8ns ± 2% ~ (p=0.246 n=5+5) FmtFprintfString-8 82.7ns ± 4% 80.1ns ± 5% ~ (p=0.238 n=5+5) FmtFprintfInt-8 82.6ns ± 2% 81.9ns ± 3% ~ (p=0.508 n=5+5) FmtFprintfIntInt-8 124ns ± 4% 121ns ± 3% ~ (p=0.111 n=5+5) FmtFprintfPrefixedInt-8 158ns ± 6% 160ns ± 2% ~ (p=0.341 n=5+5) FmtFprintfFloat-8 249ns ± 2% 245ns ± 2% ~ (p=0.095 n=5+5) FmtManyArgs-8 513ns ± 2% 519ns ± 3% ~ (p=0.151 n=5+5) GobDecode-8 7.48ms ±12% 7.11ms ± 2% ~ (p=0.222 n=5+5) GobEncode-8 6.25ms ± 1% 6.03ms ± 2% -3.56% (p=0.008 n=5+5) Gzip-8 252ms ± 4% 252ms ± 4% ~ (p=1.000 n=5+5) Gunzip-8 38.4ms ± 3% 38.6ms ± 2% ~ (p=0.690 n=5+5) HTTPClientServer-8 76.9µs ±41% 66.4µs ± 6% ~ (p=0.310 n=5+5) JSONEncode-8 16.5ms ± 3% 16.7ms ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 54.6ms ± 1% 54.3ms ± 2% ~ (p=0.548 n=5+5) Mandelbrot200-8 4.45ms ± 3% 4.47ms ± 1% ~ (p=0.841 n=5+5) GoParse-8 3.43ms ± 1% 3.32ms ± 2% -3.28% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 88.2ns ± 3% 89.4ns ± 2% ~ (p=0.333 n=5+5) RegexpMatchEasy0_1K-8 205ns ± 1% 206ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchEasy1_32-8 85.1ns ± 1% 85.5ns ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 365ns ± 1% 371ns ± 9% ~ (p=1.000 n=5+5) RegexpMatchMedium_32-8 129ns ± 2% 128ns ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 39.8µs ± 0% 39.7µs ± 4% ~ (p=0.730 n=4+5) RegexpMatchHard_32-8 1.99µs ± 3% 2.05µs ±16% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 59.3µs ± 1% 60.3µs ± 7% ~ (p=1.000 n=5+5) Revcomp-8 1.36s ±63% 0.52s ± 5% ~ (p=0.095 n=5+5) Template-8 62.6ms ±14% 60.5ms ± 5% ~ (p=0.690 n=5+5) TimeParse-8 330ns ± 2% 324ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 350ns ± 3% 340ns ± 1% -2.86% (p=0.008 n=5+5) name old speed new speed delta GobDecode-8 103MB/s ±11% 108MB/s ± 2% ~ (p=0.222 n=5+5) GobEncode-8 123MB/s ± 1% 127MB/s ± 2% +3.71% (p=0.008 n=5+5) Gzip-8 77.1MB/s ± 4% 76.9MB/s ± 3% ~ (p=1.000 n=5+5) Gunzip-8 505MB/s ± 3% 503MB/s ± 2% ~ (p=0.690 n=5+5) JSONEncode-8 118MB/s ± 3% 116MB/s ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 35.5MB/s ± 1% 35.8MB/s ± 2% ~ (p=0.397 n=5+5) GoParse-8 16.9MB/s ± 1% 17.4MB/s ± 2% +3.45% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 363MB/s ± 3% 358MB/s ± 2% ~ (p=0.421 n=5+5) RegexpMatchEasy0_1K-8 4.98GB/s ± 1% 4.97GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 376MB/s ± 1% 375MB/s ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 2.80GB/s ± 1% 2.76GB/s ± 9% ~ (p=0.841 n=5+5) RegexpMatchMedium_32-8 7.73MB/s ± 1% 7.76MB/s ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 25.8MB/s ± 0% 25.8MB/s ± 4% ~ (p=0.651 n=4+5) RegexpMatchHard_32-8 16.1MB/s ± 3% 15.7MB/s ±14% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 17.3MB/s ± 1% 17.0MB/s ± 7% ~ (p=0.984 n=5+5) Revcomp-8 273MB/s ±83% 488MB/s ± 5% ~ (p=0.095 n=5+5) Template-8 31.1MB/s ±13% 32.1MB/s ± 5% ~ (p=0.690 n=5+5) Updates #19495 Change-Id: I116e9a2a4594769318b22d736464de8a98499909 Reviewed-on: https://go-review.googlesource.com/38091 Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-12 14:47:59 -07:00
}
runtime: add mapdelete_fast* Add benchmarks for map delete with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapDelete/Int32/1-8 151ns ± 8% 99ns ± 3% -34.39% (p=0.008 n=5+5) MapDelete/Int32/2-8 128ns ± 2% 111ns ±15% -13.40% (p=0.040 n=5+5) MapDelete/Int32/4-8 128ns ± 5% 114ns ± 2% -10.82% (p=0.008 n=5+5) MapDelete/Int64/1-8 144ns ± 0% 104ns ± 3% -27.53% (p=0.016 n=4+5) MapDelete/Int64/2-8 153ns ± 1% 126ns ± 3% -17.17% (p=0.008 n=5+5) MapDelete/Int64/4-8 178ns ± 3% 136ns ± 2% -23.60% (p=0.008 n=5+5) MapDelete/Str/1-8 187ns ± 3% 171ns ± 3% -8.54% (p=0.008 n=5+5) MapDelete/Str/2-8 221ns ± 3% 206ns ± 4% -7.18% (p=0.016 n=5+4) MapDelete/Str/4-8 256ns ± 5% 232ns ± 2% -9.36% (p=0.016 n=4+5) name old time/op new time/op delta BinaryTree17-8 2.78s ± 7% 2.70s ± 1% ~ (p=0.151 n=5+5) Fannkuch11-8 3.21s ± 2% 3.19s ± 1% ~ (p=0.310 n=5+5) FmtFprintfEmpty-8 49.1ns ± 3% 50.2ns ± 2% ~ (p=0.095 n=5+5) FmtFprintfString-8 78.6ns ± 4% 80.2ns ± 5% ~ (p=0.460 n=5+5) FmtFprintfInt-8 79.7ns ± 1% 81.0ns ± 3% ~ (p=0.103 n=5+5) FmtFprintfIntInt-8 117ns ± 2% 119ns ± 0% ~ (p=0.079 n=5+4) FmtFprintfPrefixedInt-8 153ns ± 1% 146ns ± 3% -4.19% (p=0.024 n=5+5) FmtFprintfFloat-8 239ns ± 1% 237ns ± 1% ~ (p=0.246 n=5+5) FmtManyArgs-8 506ns ± 2% 509ns ± 2% ~ (p=0.238 n=5+5) GobDecode-8 7.06ms ± 4% 6.86ms ± 1% ~ (p=0.222 n=5+5) GobEncode-8 6.01ms ± 5% 5.87ms ± 2% ~ (p=0.222 n=5+5) Gzip-8 246ms ± 4% 236ms ± 1% -4.12% (p=0.008 n=5+5) Gunzip-8 37.7ms ± 4% 37.3ms ± 1% ~ (p=0.841 n=5+5) HTTPClientServer-8 64.9µs ± 1% 64.4µs ± 0% -0.80% (p=0.032 n=5+4) JSONEncode-8 16.0ms ± 2% 16.2ms ±11% ~ (p=0.548 n=5+5) JSONDecode-8 53.2ms ± 2% 53.1ms ± 4% ~ (p=1.000 n=5+5) Mandelbrot200-8 4.33ms ± 2% 4.32ms ± 2% ~ (p=0.841 n=5+5) GoParse-8 3.24ms ± 2% 3.27ms ± 4% ~ (p=0.690 n=5+5) RegexpMatchEasy0_32-8 86.2ns ± 1% 85.2ns ± 3% ~ (p=0.286 n=5+5) RegexpMatchEasy0_1K-8 198ns ± 2% 199ns ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_32-8 82.6ns ± 2% 81.8ns ± 1% ~ (p=0.294 n=5+5) RegexpMatchEasy1_1K-8 359ns ± 2% 354ns ± 1% -1.39% (p=0.048 n=5+5) RegexpMatchMedium_32-8 123ns ± 2% 123ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchMedium_1K-8 38.2µs ± 2% 38.6µs ± 8% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 1.92µs ± 2% 1.91µs ± 5% ~ (p=0.460 n=5+5) RegexpMatchHard_1K-8 57.6µs ± 1% 57.0µs ± 2% ~ (p=0.310 n=5+5) Revcomp-8 483ms ± 7% 441ms ± 1% -8.79% (p=0.016 n=5+4) Template-8 58.0ms ± 1% 58.2ms ± 7% ~ (p=0.310 n=5+5) TimeParse-8 324ns ± 6% 312ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 330ns ± 1% 329ns ± 1% ~ (p=0.968 n=5+5) name old speed new speed delta GobDecode-8 109MB/s ± 4% 112MB/s ± 1% ~ (p=0.222 n=5+5) GobEncode-8 128MB/s ± 5% 131MB/s ± 2% ~ (p=0.222 n=5+5) Gzip-8 78.9MB/s ± 4% 82.3MB/s ± 1% +4.25% (p=0.008 n=5+5) Gunzip-8 514MB/s ± 4% 521MB/s ± 1% ~ (p=0.841 n=5+5) JSONEncode-8 121MB/s ± 2% 120MB/s ±10% ~ (p=0.548 n=5+5) JSONDecode-8 36.5MB/s ± 2% 36.6MB/s ± 4% ~ (p=1.000 n=5+5) GoParse-8 17.9MB/s ± 2% 17.7MB/s ± 4% ~ (p=0.730 n=5+5) RegexpMatchEasy0_32-8 371MB/s ± 1% 375MB/s ± 3% ~ (p=0.310 n=5+5) RegexpMatchEasy0_1K-8 5.15GB/s ± 1% 5.13GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 387MB/s ± 2% 391MB/s ± 1% ~ (p=0.310 n=5+5) RegexpMatchEasy1_1K-8 2.85GB/s ± 2% 2.89GB/s ± 1% ~ (p=0.056 n=5+5) RegexpMatchMedium_32-8 8.07MB/s ± 2% 8.06MB/s ± 1% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 26.8MB/s ± 2% 26.6MB/s ± 7% ~ (p=0.690 n=5+5) RegexpMatchHard_32-8 16.7MB/s ± 2% 16.7MB/s ± 5% ~ (p=0.421 n=5+5) RegexpMatchHard_1K-8 17.8MB/s ± 1% 18.0MB/s ± 2% ~ (p=0.310 n=5+5) Revcomp-8 527MB/s ± 6% 577MB/s ± 1% +9.44% (p=0.016 n=5+4) Template-8 33.5MB/s ± 1% 33.4MB/s ± 7% ~ (p=0.310 n=5+5) Updates #19495 Change-Id: Ib9ece1690813d9b4788455db43d30891e2138df5 Reviewed-on: https://go-review.googlesource.com/38172 Reviewed-by: Hugues Bruant <hugues.bruant@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-14 11:11:28 -07:00
return mapslow
runtime: add mapassign_fast* Add benchmarks for map assignment with int32/int64/string key Benchmark results on darwin/amd64 name old time/op new time/op delta MapAssignInt32_255-8 24.7ns ± 3% 17.4ns ± 2% -29.75% (p=0.000 n=10+10) MapAssignInt32_64k-8 45.5ns ± 4% 37.6ns ± 4% -17.18% (p=0.000 n=10+10) MapAssignInt64_255-8 26.0ns ± 3% 17.9ns ± 4% -31.03% (p=0.000 n=10+10) MapAssignInt64_64k-8 46.9ns ± 5% 38.7ns ± 2% -17.53% (p=0.000 n=9+10) MapAssignStr_255-8 47.8ns ± 3% 24.8ns ± 4% -48.01% (p=0.000 n=10+10) MapAssignStr_64k-8 83.0ns ± 3% 51.9ns ± 3% -37.45% (p=0.000 n=10+9) name old time/op new time/op delta BinaryTree17-8 3.11s ±19% 2.78s ± 3% ~ (p=0.095 n=5+5) Fannkuch11-8 3.26s ± 1% 3.21s ± 2% ~ (p=0.056 n=5+5) FmtFprintfEmpty-8 50.3ns ± 1% 50.8ns ± 2% ~ (p=0.246 n=5+5) FmtFprintfString-8 82.7ns ± 4% 80.1ns ± 5% ~ (p=0.238 n=5+5) FmtFprintfInt-8 82.6ns ± 2% 81.9ns ± 3% ~ (p=0.508 n=5+5) FmtFprintfIntInt-8 124ns ± 4% 121ns ± 3% ~ (p=0.111 n=5+5) FmtFprintfPrefixedInt-8 158ns ± 6% 160ns ± 2% ~ (p=0.341 n=5+5) FmtFprintfFloat-8 249ns ± 2% 245ns ± 2% ~ (p=0.095 n=5+5) FmtManyArgs-8 513ns ± 2% 519ns ± 3% ~ (p=0.151 n=5+5) GobDecode-8 7.48ms ±12% 7.11ms ± 2% ~ (p=0.222 n=5+5) GobEncode-8 6.25ms ± 1% 6.03ms ± 2% -3.56% (p=0.008 n=5+5) Gzip-8 252ms ± 4% 252ms ± 4% ~ (p=1.000 n=5+5) Gunzip-8 38.4ms ± 3% 38.6ms ± 2% ~ (p=0.690 n=5+5) HTTPClientServer-8 76.9µs ±41% 66.4µs ± 6% ~ (p=0.310 n=5+5) JSONEncode-8 16.5ms ± 3% 16.7ms ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 54.6ms ± 1% 54.3ms ± 2% ~ (p=0.548 n=5+5) Mandelbrot200-8 4.45ms ± 3% 4.47ms ± 1% ~ (p=0.841 n=5+5) GoParse-8 3.43ms ± 1% 3.32ms ± 2% -3.28% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 88.2ns ± 3% 89.4ns ± 2% ~ (p=0.333 n=5+5) RegexpMatchEasy0_1K-8 205ns ± 1% 206ns ± 1% ~ (p=0.905 n=5+5) RegexpMatchEasy1_32-8 85.1ns ± 1% 85.5ns ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 365ns ± 1% 371ns ± 9% ~ (p=1.000 n=5+5) RegexpMatchMedium_32-8 129ns ± 2% 128ns ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 39.8µs ± 0% 39.7µs ± 4% ~ (p=0.730 n=4+5) RegexpMatchHard_32-8 1.99µs ± 3% 2.05µs ±16% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 59.3µs ± 1% 60.3µs ± 7% ~ (p=1.000 n=5+5) Revcomp-8 1.36s ±63% 0.52s ± 5% ~ (p=0.095 n=5+5) Template-8 62.6ms ±14% 60.5ms ± 5% ~ (p=0.690 n=5+5) TimeParse-8 330ns ± 2% 324ns ± 2% ~ (p=0.087 n=5+5) TimeFormat-8 350ns ± 3% 340ns ± 1% -2.86% (p=0.008 n=5+5) name old speed new speed delta GobDecode-8 103MB/s ±11% 108MB/s ± 2% ~ (p=0.222 n=5+5) GobEncode-8 123MB/s ± 1% 127MB/s ± 2% +3.71% (p=0.008 n=5+5) Gzip-8 77.1MB/s ± 4% 76.9MB/s ± 3% ~ (p=1.000 n=5+5) Gunzip-8 505MB/s ± 3% 503MB/s ± 2% ~ (p=0.690 n=5+5) JSONEncode-8 118MB/s ± 3% 116MB/s ± 3% ~ (p=0.421 n=5+5) JSONDecode-8 35.5MB/s ± 1% 35.8MB/s ± 2% ~ (p=0.397 n=5+5) GoParse-8 16.9MB/s ± 1% 17.4MB/s ± 2% +3.45% (p=0.008 n=5+5) RegexpMatchEasy0_32-8 363MB/s ± 3% 358MB/s ± 2% ~ (p=0.421 n=5+5) RegexpMatchEasy0_1K-8 4.98GB/s ± 1% 4.97GB/s ± 1% ~ (p=0.548 n=5+5) RegexpMatchEasy1_32-8 376MB/s ± 1% 375MB/s ± 5% ~ (p=0.690 n=5+5) RegexpMatchEasy1_1K-8 2.80GB/s ± 1% 2.76GB/s ± 9% ~ (p=0.841 n=5+5) RegexpMatchMedium_32-8 7.73MB/s ± 1% 7.76MB/s ± 3% ~ (p=0.730 n=5+5) RegexpMatchMedium_1K-8 25.8MB/s ± 0% 25.8MB/s ± 4% ~ (p=0.651 n=4+5) RegexpMatchHard_32-8 16.1MB/s ± 3% 15.7MB/s ±14% ~ (p=0.794 n=5+5) RegexpMatchHard_1K-8 17.3MB/s ± 1% 17.0MB/s ± 7% ~ (p=0.984 n=5+5) Revcomp-8 273MB/s ±83% 488MB/s ± 5% ~ (p=0.095 n=5+5) Template-8 31.1MB/s ±13% 32.1MB/s ± 5% ~ (p=0.690 n=5+5) Updates #19495 Change-Id: I116e9a2a4594769318b22d736464de8a98499909 Reviewed-on: https://go-review.googlesource.com/38091 Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-03-12 14:47:59 -07:00
}
func writebarrierfn(name string, l *types.Type, r *types.Type) *Node {
fn := syslook(name)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
fn = substArgTypes(fn, l, r)
return fn
}
func addstr(n *Node, init *Nodes) *Node {
// order.expr rewrote OADDSTR to have a list of strings.
c := n.List.Len()
if c < 2 {
Fatalf("addstr count %d too small", c)
}
buf := nodnil()
if n.Esc == EscNone {
sz := int64(0)
for _, n1 := range n.List.Slice() {
if n1.Op == OLITERAL {
sz += int64(len(strlit(n1)))
}
}
// Don't allocate the buffer if the result won't fit.
if sz < tmpstringbufsize {
// Create temporary buffer for result string on stack.
t := types.NewArray(types.Types[TUINT8], tmpstringbufsize)
buf = nod(OADDR, temp(t), nil)
}
}
// build list of string arguments
args := []*Node{buf}
for _, n2 := range n.List.Slice() {
args = append(args, conv(n2, types.Types[TSTRING]))
}
var fn string
if c <= 5 {
// small numbers of strings use direct runtime helpers.
// note: order.expr knows this cutoff too.
fn = fmt.Sprintf("concatstring%d", c)
} else {
// large numbers of strings are passed to the runtime as a slice.
fn = "concatstrings"
t := types.NewSlice(types.Types[TSTRING])
slice := nod(OCOMPLIT, nil, typenod(t))
if prealloc[n] != nil {
prealloc[slice] = prealloc[n]
}
slice.List.Set(args[1:]) // skip buf arg
args = []*Node{buf, slice}
slice.Esc = EscNone
}
cat := syslook(fn)
r := nod(OCALL, cat, nil)
r.List.Set(args)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
r = typecheck(r, ctxExpr)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
r = walkexpr(r, init)
r.Type = n.Type
return r
}
func walkAppendArgs(n *Node, init *Nodes) {
walkexprlistsafe(n.List.Slice(), init)
// walkexprlistsafe will leave OINDEX (s[n]) alone if both s
// and n are name or literal, but those may index the slice we're
// modifying here. Fix explicitly.
ls := n.List.Slice()
for i1, n1 := range ls {
ls[i1] = cheapexpr(n1, init)
}
}
// expand append(l1, l2...) to
// init {
// s := l1
// n := len(s) + len(l2)
// // Compare as uint so growslice can panic on overflow.
// if uint(n) > uint(cap(s)) {
// s = growslice(s, n)
// }
// s = s[:n]
// memmove(&s[len(l1)], &l2[0], len(l2)*sizeof(T))
// }
// s
//
// l2 is allowed to be a string.
func appendslice(n *Node, init *Nodes) *Node {
walkAppendArgs(n, init)
l1 := n.List.First()
l2 := n.List.Second()
l2 = cheapexpr(l2, init)
n.List.SetSecond(l2)
var nodes Nodes
// var s []T
s := temp(l1.Type)
nodes.Append(nod(OAS, s, l1)) // s = l1
elemtype := s.Type.Elem()
// n := len(s) + len(l2)
nn := temp(types.Types[TINT])
nodes.Append(nod(OAS, nn, nod(OADD, nod(OLEN, s, nil), nod(OLEN, l2, nil))))
// if uint(n) > uint(cap(s))
nif := nod(OIF, nil, nil)
nuint := conv(nn, types.Types[TUINT])
scapuint := conv(nod(OCAP, s, nil), types.Types[TUINT])
nif.Left = nod(OGT, nuint, scapuint)
// instantiate growslice(typ *type, []any, int) []any
fn := syslook("growslice")
fn = substArgTypes(fn, elemtype, elemtype)
// s = growslice(T, s, n)
nif.Nbody.Set1(nod(OAS, s, mkcall1(fn, s.Type, &nif.Ninit, typename(elemtype), s, nn)))
nodes.Append(nif)
// s = s[:n]
nt := nod(OSLICE, s, nil)
nt.SetSliceBounds(nil, nn, nil)
nt.SetBounded(true)
nodes.Append(nod(OAS, s, nt))
var ncopy *Node
if elemtype.HasPointers() {
// copy(s[len(l1):], l2)
nptr1 := nod(OSLICE, s, nil)
nptr1.Type = s.Type
nptr1.SetSliceBounds(nod(OLEN, l1, nil), nil, nil)
nptr1 = cheapexpr(nptr1, &nodes)
nptr2 := l2
Curfn.Func.setWBPos(n.Pos)
// instantiate typedslicecopy(typ *type, dstPtr *any, dstLen int, srcPtr *any, srcLen int) int
fn := syslook("typedslicecopy")
fn = substArgTypes(fn, l1.Type.Elem(), l2.Type.Elem())
ptr1, len1 := nptr1.slicePtrLen()
ptr2, len2 := nptr2.slicePtrLen()
ncopy = mkcall1(fn, types.Types[TINT], &nodes, typename(elemtype), ptr1, len1, ptr2, len2)
} else if instrumenting && !compiling_runtime {
// rely on runtime to instrument copy.
// copy(s[len(l1):], l2)
nptr1 := nod(OSLICE, s, nil)
nptr1.Type = s.Type
nptr1.SetSliceBounds(nod(OLEN, l1, nil), nil, nil)
nptr1 = cheapexpr(nptr1, &nodes)
nptr2 := l2
if l2.Type.IsString() {
// instantiate func slicestringcopy(toPtr *byte, toLen int, fr string) int
fn := syslook("slicestringcopy")
ptr, len := nptr1.slicePtrLen()
str := nod(OCONVNOP, nptr2, nil)
str.Type = types.Types[TSTRING]
ncopy = mkcall1(fn, types.Types[TINT], &nodes, ptr, len, str)
} else {
// instantiate func slicecopy(to any, fr any, wid uintptr) int
fn := syslook("slicecopy")
fn = substArgTypes(fn, l1.Type.Elem(), l2.Type.Elem())
ptr1, len1 := nptr1.slicePtrLen()
ptr2, len2 := nptr2.slicePtrLen()
ncopy = mkcall1(fn, types.Types[TINT], &nodes, ptr1, len1, ptr2, len2, nodintconst(elemtype.Width))
}
} else {
// memmove(&s[len(l1)], &l2[0], len(l2)*sizeof(T))
nptr1 := nod(OINDEX, s, nod(OLEN, l1, nil))
cmd/compile: pack bool fields in Node, Name, Func and Type structs to bitsets This reduces compiler memory usage by up to 4% - see compilebench results below. name old time/op new time/op delta Template 245ms ± 4% 241ms ± 2% -1.88% (p=0.029 n=10+10) Unicode 126ms ± 3% 124ms ± 3% ~ (p=0.105 n=10+10) GoTypes 805ms ± 2% 813ms ± 3% ~ (p=0.515 n=8+10) Compiler 3.95s ± 2% 3.83s ± 1% -2.96% (p=0.000 n=9+10) MakeBash 47.4s ± 4% 46.6s ± 1% -1.59% (p=0.028 n=9+10) name old user-ns/op new user-ns/op delta Template 324M ± 5% 326M ± 3% ~ (p=0.935 n=10+10) Unicode 186M ± 5% 178M ±10% ~ (p=0.067 n=9+10) GoTypes 1.08G ± 7% 1.09G ± 4% ~ (p=0.956 n=10+10) Compiler 5.34G ± 4% 5.31G ± 1% ~ (p=0.501 n=10+8) name old alloc/op new alloc/op delta Template 41.0MB ± 0% 39.8MB ± 0% -3.03% (p=0.000 n=10+10) Unicode 32.3MB ± 0% 31.0MB ± 0% -4.13% (p=0.000 n=10+10) GoTypes 119MB ± 0% 116MB ± 0% -2.39% (p=0.000 n=10+10) Compiler 499MB ± 0% 487MB ± 0% -2.48% (p=0.000 n=10+10) name old allocs/op new allocs/op delta Template 380k ± 1% 379k ± 1% ~ (p=0.436 n=10+10) Unicode 324k ± 1% 324k ± 0% ~ (p=0.853 n=10+10) GoTypes 1.15M ± 0% 1.15M ± 0% ~ (p=0.481 n=10+10) Compiler 4.41M ± 0% 4.41M ± 0% -0.12% (p=0.007 n=10+10) name old text-bytes new text-bytes delta HelloSize 623k ± 0% 623k ± 0% ~ (all equal) CmdGoSize 6.64M ± 0% 6.64M ± 0% ~ (all equal) name old data-bytes new data-bytes delta HelloSize 5.81k ± 0% 5.81k ± 0% ~ (all equal) CmdGoSize 238k ± 0% 238k ± 0% ~ (all equal) name old bss-bytes new bss-bytes delta HelloSize 134k ± 0% 134k ± 0% ~ (all equal) CmdGoSize 152k ± 0% 152k ± 0% ~ (all equal) name old exe-bytes new exe-bytes delta HelloSize 967k ± 0% 967k ± 0% ~ (all equal) CmdGoSize 10.2M ± 0% 10.2M ± 0% ~ (all equal) Change-Id: I1f40af738254892bd6c8ba2eb43390b175753d52 Reviewed-on: https://go-review.googlesource.com/37445 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-02-27 19:56:38 +02:00
nptr1.SetBounded(true)
nptr1 = nod(OADDR, nptr1, nil)
nptr2 := nod(OSPTR, l2, nil)
nwid := cheapexpr(conv(nod(OLEN, l2, nil), types.Types[TUINTPTR]), &nodes)
nwid = nod(OMUL, nwid, nodintconst(elemtype.Width))
// instantiate func memmove(to *any, frm *any, length uintptr)
fn := syslook("memmove")
fn = substArgTypes(fn, elemtype, elemtype)
ncopy = mkcall1(fn, nil, &nodes, nptr1, nptr2, nwid)
}
ln := append(nodes.Slice(), ncopy)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
typecheckslice(ln, ctxStmt)
walkstmtlist(ln)
init.Append(ln...)
return s
}
// isAppendOfMake reports whether n is of the form append(x , make([]T, y)...).
// isAppendOfMake assumes n has already been typechecked.
func isAppendOfMake(n *Node) bool {
if Debug['N'] != 0 || instrumenting {
return false
}
if n.Typecheck() == 0 {
Fatalf("missing typecheck: %+v", n)
}
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
if n.Op != OAPPEND || !n.IsDDD() || n.List.Len() != 2 {
return false
}
second := n.List.Second()
if second.Op != OMAKESLICE || second.Right != nil {
return false
}
// y must be either an integer constant or the largest possible positive value
// of variable y needs to fit into an uint.
// typecheck made sure that constant arguments to make are not negative and fit into an int.
// The care of overflow of the len argument to make will be handled by an explicit check of int(len) < 0 during runtime.
y := second.Left
if !Isconst(y, CTINT) && maxintval[y.Type.Etype].Cmp(maxintval[TUINT]) > 0 {
return false
}
return true
}
// extendslice rewrites append(l1, make([]T, l2)...) to
// init {
// if l2 >= 0 { // Empty if block here for more meaningful node.SetLikely(true)
// } else {
// panicmakeslicelen()
// }
// s := l1
// n := len(s) + l2
// // Compare n and s as uint so growslice can panic on overflow of len(s) + l2.
// // cap is a positive int and n can become negative when len(s) + l2
// // overflows int. Interpreting n when negative as uint makes it larger
// // than cap(s). growslice will check the int n arg and panic if n is
// // negative. This prevents the overflow from being undetected.
// if uint(n) > uint(cap(s)) {
// s = growslice(T, s, n)
// }
// s = s[:n]
// lptr := &l1[0]
// sptr := &s[0]
// if lptr == sptr || !T.HasPointers() {
// // growslice did not clear the whole underlying array (or did not get called)
// hp := &s[len(l1)]
// hn := l2 * sizeof(T)
// memclr(hp, hn)
// }
// }
// s
func extendslice(n *Node, init *Nodes) *Node {
// isAppendOfMake made sure all possible positive values of l2 fit into an uint.
// The case of l2 overflow when converting from e.g. uint to int is handled by an explicit
// check of l2 < 0 at runtime which is generated below.
l2 := conv(n.List.Second().Left, types.Types[TINT])
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
l2 = typecheck(l2, ctxExpr)
n.List.SetSecond(l2) // walkAppendArgs expects l2 in n.List.Second().
walkAppendArgs(n, init)
l1 := n.List.First()
l2 = n.List.Second() // re-read l2, as it may have been updated by walkAppendArgs
var nodes []*Node
// if l2 >= 0 (likely happens), do nothing
nifneg := nod(OIF, nod(OGE, l2, nodintconst(0)), nil)
nifneg.SetLikely(true)
// else panicmakeslicelen()
nifneg.Rlist.Set1(mkcall("panicmakeslicelen", nil, init))
nodes = append(nodes, nifneg)
// s := l1
s := temp(l1.Type)
nodes = append(nodes, nod(OAS, s, l1))
elemtype := s.Type.Elem()
// n := len(s) + l2
nn := temp(types.Types[TINT])
nodes = append(nodes, nod(OAS, nn, nod(OADD, nod(OLEN, s, nil), l2)))
// if uint(n) > uint(cap(s))
nuint := conv(nn, types.Types[TUINT])
capuint := conv(nod(OCAP, s, nil), types.Types[TUINT])
nif := nod(OIF, nod(OGT, nuint, capuint), nil)
// instantiate growslice(typ *type, old []any, newcap int) []any
fn := syslook("growslice")
fn = substArgTypes(fn, elemtype, elemtype)
// s = growslice(T, s, n)
nif.Nbody.Set1(nod(OAS, s, mkcall1(fn, s.Type, &nif.Ninit, typename(elemtype), s, nn)))
nodes = append(nodes, nif)
// s = s[:n]
nt := nod(OSLICE, s, nil)
nt.SetSliceBounds(nil, nn, nil)
nt.SetBounded(true)
nodes = append(nodes, nod(OAS, s, nt))
// lptr := &l1[0]
l1ptr := temp(l1.Type.Elem().PtrTo())
tmp := nod(OSPTR, l1, nil)
nodes = append(nodes, nod(OAS, l1ptr, tmp))
// sptr := &s[0]
sptr := temp(elemtype.PtrTo())
tmp = nod(OSPTR, s, nil)
nodes = append(nodes, nod(OAS, sptr, tmp))
// hp := &s[len(l1)]
hp := nod(OINDEX, s, nod(OLEN, l1, nil))
hp.SetBounded(true)
hp = nod(OADDR, hp, nil)
hp = convnop(hp, types.Types[TUNSAFEPTR])
// hn := l2 * sizeof(elem(s))
hn := nod(OMUL, l2, nodintconst(elemtype.Width))
hn = conv(hn, types.Types[TUINTPTR])
clrname := "memclrNoHeapPointers"
hasPointers := elemtype.HasPointers()
if hasPointers {
clrname = "memclrHasPointers"
Curfn.Func.setWBPos(n.Pos)
}
var clr Nodes
clrfn := mkcall(clrname, nil, &clr, hp, hn)
clr.Append(clrfn)
if hasPointers {
// if l1ptr == sptr
nifclr := nod(OIF, nod(OEQ, l1ptr, sptr), nil)
nifclr.Nbody = clr
nodes = append(nodes, nifclr)
} else {
nodes = append(nodes, clr.Slice()...)
}
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
typecheckslice(nodes, ctxStmt)
walkstmtlist(nodes)
init.Append(nodes...)
return s
}
cmd/internal/gc: optimize append + write barrier The code generated for x = append(x, v) is roughly: t := x if len(t)+1 > cap(t) { t = grow(t) } t[len(t)] = v len(t)++ x = t We used to generate this code as Go pseudocode during walk. Generate it instead as actual instructions during gen. Doing so lets us apply a few optimizations. The most important is that when, as in the above example, the source slice and the destination slice are the same, the code can instead do: t := x if len(t)+1 > cap(t) { t = grow(t) x = {base(t), len(t)+1, cap(t)} } else { len(x)++ } t[len(t)] = v That is, in the fast path that does not reallocate the array, only the updated length needs to be written back to x, not the array pointer and not the capacity. This is more like what you'd write by hand in C. It's faster in general, since the fast path elides two of the three stores, but it's especially faster when the form of x is such that the base pointer write would turn into a write barrier. No write, no barrier. name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.81s × (0.98,1.03) +2.35% (p=0.023) Fannkuch11 4.41s × (0.98,1.03) 4.35s × (1.00,1.00) ~ (p=0.090) FmtFprintfEmpty 92.7ns × (0.91,1.16) 86.0ns × (0.94,1.11) -7.31% (p=0.038) FmtFprintfString 281ns × (0.96,1.08) 276ns × (0.98,1.04) ~ (p=0.219) FmtFprintfInt 288ns × (0.97,1.06) 274ns × (0.98,1.06) -4.94% (p=0.002) FmtFprintfIntInt 493ns × (0.97,1.04) 506ns × (0.99,1.01) +2.65% (p=0.009) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 391ns × (0.99,1.01) -7.52% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 566ns × (0.99,1.01) -5.27% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.91µs × (0.99,1.01) ~ (p=0.231) GobDecode 14.8ms × (0.98,1.03) 15.3ms × (0.99,1.02) +3.01% (p=0.000) GobEncode 12.3ms × (0.98,1.01) 11.5ms × (0.97,1.03) -5.93% (p=0.000) Gzip 656ms × (0.99,1.05) 645ms × (0.99,1.01) ~ (p=0.055) Gunzip 142ms × (1.00,1.00) 142ms × (1.00,1.00) -0.32% (p=0.034) HTTPClientServer 91.2µs × (0.97,1.04) 90.5µs × (0.97,1.04) ~ (p=0.468) JSONEncode 32.6ms × (0.97,1.08) 32.0ms × (0.98,1.03) ~ (p=0.190) JSONDecode 114ms × (0.97,1.05) 114ms × (0.99,1.01) ~ (p=0.887) Mandelbrot200 6.11ms × (0.98,1.04) 6.04ms × (1.00,1.01) ~ (p=0.167) GoParse 6.66ms × (0.97,1.04) 6.47ms × (0.97,1.05) -2.81% (p=0.014) RegexpMatchEasy0_32 159ns × (0.99,1.00) 171ns × (0.93,1.07) +7.19% (p=0.002) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 550ns × (0.98,1.01) +2.30% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 135ns × (0.99,1.02) -1.60% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 879ns × (1.00,1.01) +1.08% (p=0.000) RegexpMatchMedium_32 252ns × (0.99,1.01) 243ns × (1.00,1.00) -3.71% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 70.3µs × (1.00,1.00) -3.34% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.82µs × (1.00,1.01) -0.81% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 117µs × (1.00,1.00) -0.56% (p=0.000) Revcomp 920ms × (0.97,1.07) 917ms × (0.97,1.04) ~ (p=0.808) Template 129ms × (0.98,1.03) 114ms × (0.99,1.01) -12.06% (p=0.000) TimeParse 619ns × (0.99,1.01) 622ns × (0.99,1.01) ~ (p=0.062) TimeFormat 661ns × (0.98,1.04) 665ns × (0.99,1.01) ~ (p=0.524) See next CL for combination with a similar optimization for slice. The benchmarks that are slower in this CL are still faster overall with the combination of the two. Change-Id: I2a7421658091b2488c64741b4db15ab6c3b4cb7e Reviewed-on: https://go-review.googlesource.com/9812 Reviewed-by: David Chase <drchase@google.com>
2015-05-06 12:34:30 -04:00
// Rewrite append(src, x, y, z) so that any side effects in
// x, y, z (including runtime panics) are evaluated in
// initialization statements before the append.
// For normal code generation, stop there and leave the
// rest to cgen_append.
//
// For race detector, expand append(src, a [, b]* ) to
//
// init {
// s := src
// const argc = len(args) - 1
// if cap(s) - len(s) < argc {
// s = growslice(s, len(s)+argc)
// }
// n := len(s)
// s = s[:n+argc]
// s[n] = a
// s[n+1] = b
// ...
// }
// s
func walkappend(n *Node, init *Nodes, dst *Node) *Node {
if !samesafeexpr(dst, n.List.First()) {
n.List.SetFirst(safeexpr(n.List.First(), init))
n.List.SetFirst(walkexpr(n.List.First(), init))
cmd/internal/gc: optimize append + write barrier The code generated for x = append(x, v) is roughly: t := x if len(t)+1 > cap(t) { t = grow(t) } t[len(t)] = v len(t)++ x = t We used to generate this code as Go pseudocode during walk. Generate it instead as actual instructions during gen. Doing so lets us apply a few optimizations. The most important is that when, as in the above example, the source slice and the destination slice are the same, the code can instead do: t := x if len(t)+1 > cap(t) { t = grow(t) x = {base(t), len(t)+1, cap(t)} } else { len(x)++ } t[len(t)] = v That is, in the fast path that does not reallocate the array, only the updated length needs to be written back to x, not the array pointer and not the capacity. This is more like what you'd write by hand in C. It's faster in general, since the fast path elides two of the three stores, but it's especially faster when the form of x is such that the base pointer write would turn into a write barrier. No write, no barrier. name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.81s × (0.98,1.03) +2.35% (p=0.023) Fannkuch11 4.41s × (0.98,1.03) 4.35s × (1.00,1.00) ~ (p=0.090) FmtFprintfEmpty 92.7ns × (0.91,1.16) 86.0ns × (0.94,1.11) -7.31% (p=0.038) FmtFprintfString 281ns × (0.96,1.08) 276ns × (0.98,1.04) ~ (p=0.219) FmtFprintfInt 288ns × (0.97,1.06) 274ns × (0.98,1.06) -4.94% (p=0.002) FmtFprintfIntInt 493ns × (0.97,1.04) 506ns × (0.99,1.01) +2.65% (p=0.009) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 391ns × (0.99,1.01) -7.52% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 566ns × (0.99,1.01) -5.27% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.91µs × (0.99,1.01) ~ (p=0.231) GobDecode 14.8ms × (0.98,1.03) 15.3ms × (0.99,1.02) +3.01% (p=0.000) GobEncode 12.3ms × (0.98,1.01) 11.5ms × (0.97,1.03) -5.93% (p=0.000) Gzip 656ms × (0.99,1.05) 645ms × (0.99,1.01) ~ (p=0.055) Gunzip 142ms × (1.00,1.00) 142ms × (1.00,1.00) -0.32% (p=0.034) HTTPClientServer 91.2µs × (0.97,1.04) 90.5µs × (0.97,1.04) ~ (p=0.468) JSONEncode 32.6ms × (0.97,1.08) 32.0ms × (0.98,1.03) ~ (p=0.190) JSONDecode 114ms × (0.97,1.05) 114ms × (0.99,1.01) ~ (p=0.887) Mandelbrot200 6.11ms × (0.98,1.04) 6.04ms × (1.00,1.01) ~ (p=0.167) GoParse 6.66ms × (0.97,1.04) 6.47ms × (0.97,1.05) -2.81% (p=0.014) RegexpMatchEasy0_32 159ns × (0.99,1.00) 171ns × (0.93,1.07) +7.19% (p=0.002) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 550ns × (0.98,1.01) +2.30% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 135ns × (0.99,1.02) -1.60% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 879ns × (1.00,1.01) +1.08% (p=0.000) RegexpMatchMedium_32 252ns × (0.99,1.01) 243ns × (1.00,1.00) -3.71% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 70.3µs × (1.00,1.00) -3.34% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.82µs × (1.00,1.01) -0.81% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 117µs × (1.00,1.00) -0.56% (p=0.000) Revcomp 920ms × (0.97,1.07) 917ms × (0.97,1.04) ~ (p=0.808) Template 129ms × (0.98,1.03) 114ms × (0.99,1.01) -12.06% (p=0.000) TimeParse 619ns × (0.99,1.01) 622ns × (0.99,1.01) ~ (p=0.062) TimeFormat 661ns × (0.98,1.04) 665ns × (0.99,1.01) ~ (p=0.524) See next CL for combination with a similar optimization for slice. The benchmarks that are slower in this CL are still faster overall with the combination of the two. Change-Id: I2a7421658091b2488c64741b4db15ab6c3b4cb7e Reviewed-on: https://go-review.googlesource.com/9812 Reviewed-by: David Chase <drchase@google.com>
2015-05-06 12:34:30 -04:00
}
walkexprlistsafe(n.List.Slice()[1:], init)
nsrc := n.List.First()
// walkexprlistsafe will leave OINDEX (s[n]) alone if both s
// and n are name or literal, but those may index the slice we're
// modifying here. Fix explicitly.
cmd/internal/gc: optimize append + write barrier The code generated for x = append(x, v) is roughly: t := x if len(t)+1 > cap(t) { t = grow(t) } t[len(t)] = v len(t)++ x = t We used to generate this code as Go pseudocode during walk. Generate it instead as actual instructions during gen. Doing so lets us apply a few optimizations. The most important is that when, as in the above example, the source slice and the destination slice are the same, the code can instead do: t := x if len(t)+1 > cap(t) { t = grow(t) x = {base(t), len(t)+1, cap(t)} } else { len(x)++ } t[len(t)] = v That is, in the fast path that does not reallocate the array, only the updated length needs to be written back to x, not the array pointer and not the capacity. This is more like what you'd write by hand in C. It's faster in general, since the fast path elides two of the three stores, but it's especially faster when the form of x is such that the base pointer write would turn into a write barrier. No write, no barrier. name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.81s × (0.98,1.03) +2.35% (p=0.023) Fannkuch11 4.41s × (0.98,1.03) 4.35s × (1.00,1.00) ~ (p=0.090) FmtFprintfEmpty 92.7ns × (0.91,1.16) 86.0ns × (0.94,1.11) -7.31% (p=0.038) FmtFprintfString 281ns × (0.96,1.08) 276ns × (0.98,1.04) ~ (p=0.219) FmtFprintfInt 288ns × (0.97,1.06) 274ns × (0.98,1.06) -4.94% (p=0.002) FmtFprintfIntInt 493ns × (0.97,1.04) 506ns × (0.99,1.01) +2.65% (p=0.009) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 391ns × (0.99,1.01) -7.52% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 566ns × (0.99,1.01) -5.27% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.91µs × (0.99,1.01) ~ (p=0.231) GobDecode 14.8ms × (0.98,1.03) 15.3ms × (0.99,1.02) +3.01% (p=0.000) GobEncode 12.3ms × (0.98,1.01) 11.5ms × (0.97,1.03) -5.93% (p=0.000) Gzip 656ms × (0.99,1.05) 645ms × (0.99,1.01) ~ (p=0.055) Gunzip 142ms × (1.00,1.00) 142ms × (1.00,1.00) -0.32% (p=0.034) HTTPClientServer 91.2µs × (0.97,1.04) 90.5µs × (0.97,1.04) ~ (p=0.468) JSONEncode 32.6ms × (0.97,1.08) 32.0ms × (0.98,1.03) ~ (p=0.190) JSONDecode 114ms × (0.97,1.05) 114ms × (0.99,1.01) ~ (p=0.887) Mandelbrot200 6.11ms × (0.98,1.04) 6.04ms × (1.00,1.01) ~ (p=0.167) GoParse 6.66ms × (0.97,1.04) 6.47ms × (0.97,1.05) -2.81% (p=0.014) RegexpMatchEasy0_32 159ns × (0.99,1.00) 171ns × (0.93,1.07) +7.19% (p=0.002) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 550ns × (0.98,1.01) +2.30% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 135ns × (0.99,1.02) -1.60% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 879ns × (1.00,1.01) +1.08% (p=0.000) RegexpMatchMedium_32 252ns × (0.99,1.01) 243ns × (1.00,1.00) -3.71% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 70.3µs × (1.00,1.00) -3.34% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.82µs × (1.00,1.01) -0.81% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 117µs × (1.00,1.00) -0.56% (p=0.000) Revcomp 920ms × (0.97,1.07) 917ms × (0.97,1.04) ~ (p=0.808) Template 129ms × (0.98,1.03) 114ms × (0.99,1.01) -12.06% (p=0.000) TimeParse 619ns × (0.99,1.01) 622ns × (0.99,1.01) ~ (p=0.062) TimeFormat 661ns × (0.98,1.04) 665ns × (0.99,1.01) ~ (p=0.524) See next CL for combination with a similar optimization for slice. The benchmarks that are slower in this CL are still faster overall with the combination of the two. Change-Id: I2a7421658091b2488c64741b4db15ab6c3b4cb7e Reviewed-on: https://go-review.googlesource.com/9812 Reviewed-by: David Chase <drchase@google.com>
2015-05-06 12:34:30 -04:00
// Using cheapexpr also makes sure that the evaluation
// of all arguments (and especially any panics) happen
// before we begin to modify the slice in a visible way.
ls := n.List.Slice()[1:]
for i, n := range ls {
n = cheapexpr(n, init)
if !types.Identical(n.Type, nsrc.Type.Elem()) {
n = assignconv(n, nsrc.Type.Elem(), "append")
n = walkexpr(n, init)
}
ls[i] = n
}
argc := n.List.Len() - 1
if argc < 1 {
return nsrc
}
cmd/internal/gc: optimize append + write barrier The code generated for x = append(x, v) is roughly: t := x if len(t)+1 > cap(t) { t = grow(t) } t[len(t)] = v len(t)++ x = t We used to generate this code as Go pseudocode during walk. Generate it instead as actual instructions during gen. Doing so lets us apply a few optimizations. The most important is that when, as in the above example, the source slice and the destination slice are the same, the code can instead do: t := x if len(t)+1 > cap(t) { t = grow(t) x = {base(t), len(t)+1, cap(t)} } else { len(x)++ } t[len(t)] = v That is, in the fast path that does not reallocate the array, only the updated length needs to be written back to x, not the array pointer and not the capacity. This is more like what you'd write by hand in C. It's faster in general, since the fast path elides two of the three stores, but it's especially faster when the form of x is such that the base pointer write would turn into a write barrier. No write, no barrier. name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.81s × (0.98,1.03) +2.35% (p=0.023) Fannkuch11 4.41s × (0.98,1.03) 4.35s × (1.00,1.00) ~ (p=0.090) FmtFprintfEmpty 92.7ns × (0.91,1.16) 86.0ns × (0.94,1.11) -7.31% (p=0.038) FmtFprintfString 281ns × (0.96,1.08) 276ns × (0.98,1.04) ~ (p=0.219) FmtFprintfInt 288ns × (0.97,1.06) 274ns × (0.98,1.06) -4.94% (p=0.002) FmtFprintfIntInt 493ns × (0.97,1.04) 506ns × (0.99,1.01) +2.65% (p=0.009) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 391ns × (0.99,1.01) -7.52% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 566ns × (0.99,1.01) -5.27% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.91µs × (0.99,1.01) ~ (p=0.231) GobDecode 14.8ms × (0.98,1.03) 15.3ms × (0.99,1.02) +3.01% (p=0.000) GobEncode 12.3ms × (0.98,1.01) 11.5ms × (0.97,1.03) -5.93% (p=0.000) Gzip 656ms × (0.99,1.05) 645ms × (0.99,1.01) ~ (p=0.055) Gunzip 142ms × (1.00,1.00) 142ms × (1.00,1.00) -0.32% (p=0.034) HTTPClientServer 91.2µs × (0.97,1.04) 90.5µs × (0.97,1.04) ~ (p=0.468) JSONEncode 32.6ms × (0.97,1.08) 32.0ms × (0.98,1.03) ~ (p=0.190) JSONDecode 114ms × (0.97,1.05) 114ms × (0.99,1.01) ~ (p=0.887) Mandelbrot200 6.11ms × (0.98,1.04) 6.04ms × (1.00,1.01) ~ (p=0.167) GoParse 6.66ms × (0.97,1.04) 6.47ms × (0.97,1.05) -2.81% (p=0.014) RegexpMatchEasy0_32 159ns × (0.99,1.00) 171ns × (0.93,1.07) +7.19% (p=0.002) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 550ns × (0.98,1.01) +2.30% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 135ns × (0.99,1.02) -1.60% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 879ns × (1.00,1.01) +1.08% (p=0.000) RegexpMatchMedium_32 252ns × (0.99,1.01) 243ns × (1.00,1.00) -3.71% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 70.3µs × (1.00,1.00) -3.34% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.82µs × (1.00,1.01) -0.81% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 117µs × (1.00,1.00) -0.56% (p=0.000) Revcomp 920ms × (0.97,1.07) 917ms × (0.97,1.04) ~ (p=0.808) Template 129ms × (0.98,1.03) 114ms × (0.99,1.01) -12.06% (p=0.000) TimeParse 619ns × (0.99,1.01) 622ns × (0.99,1.01) ~ (p=0.062) TimeFormat 661ns × (0.98,1.04) 665ns × (0.99,1.01) ~ (p=0.524) See next CL for combination with a similar optimization for slice. The benchmarks that are slower in this CL are still faster overall with the combination of the two. Change-Id: I2a7421658091b2488c64741b4db15ab6c3b4cb7e Reviewed-on: https://go-review.googlesource.com/9812 Reviewed-by: David Chase <drchase@google.com>
2015-05-06 12:34:30 -04:00
// General case, with no function calls left as arguments.
// Leave for gen, except that instrumentation requires old form.
if !instrumenting || compiling_runtime {
cmd/internal/gc: optimize append + write barrier The code generated for x = append(x, v) is roughly: t := x if len(t)+1 > cap(t) { t = grow(t) } t[len(t)] = v len(t)++ x = t We used to generate this code as Go pseudocode during walk. Generate it instead as actual instructions during gen. Doing so lets us apply a few optimizations. The most important is that when, as in the above example, the source slice and the destination slice are the same, the code can instead do: t := x if len(t)+1 > cap(t) { t = grow(t) x = {base(t), len(t)+1, cap(t)} } else { len(x)++ } t[len(t)] = v That is, in the fast path that does not reallocate the array, only the updated length needs to be written back to x, not the array pointer and not the capacity. This is more like what you'd write by hand in C. It's faster in general, since the fast path elides two of the three stores, but it's especially faster when the form of x is such that the base pointer write would turn into a write barrier. No write, no barrier. name old mean new mean delta BinaryTree17 5.68s × (0.97,1.04) 5.81s × (0.98,1.03) +2.35% (p=0.023) Fannkuch11 4.41s × (0.98,1.03) 4.35s × (1.00,1.00) ~ (p=0.090) FmtFprintfEmpty 92.7ns × (0.91,1.16) 86.0ns × (0.94,1.11) -7.31% (p=0.038) FmtFprintfString 281ns × (0.96,1.08) 276ns × (0.98,1.04) ~ (p=0.219) FmtFprintfInt 288ns × (0.97,1.06) 274ns × (0.98,1.06) -4.94% (p=0.002) FmtFprintfIntInt 493ns × (0.97,1.04) 506ns × (0.99,1.01) +2.65% (p=0.009) FmtFprintfPrefixedInt 423ns × (0.97,1.04) 391ns × (0.99,1.01) -7.52% (p=0.000) FmtFprintfFloat 598ns × (0.99,1.01) 566ns × (0.99,1.01) -5.27% (p=0.000) FmtManyArgs 1.89µs × (0.98,1.05) 1.91µs × (0.99,1.01) ~ (p=0.231) GobDecode 14.8ms × (0.98,1.03) 15.3ms × (0.99,1.02) +3.01% (p=0.000) GobEncode 12.3ms × (0.98,1.01) 11.5ms × (0.97,1.03) -5.93% (p=0.000) Gzip 656ms × (0.99,1.05) 645ms × (0.99,1.01) ~ (p=0.055) Gunzip 142ms × (1.00,1.00) 142ms × (1.00,1.00) -0.32% (p=0.034) HTTPClientServer 91.2µs × (0.97,1.04) 90.5µs × (0.97,1.04) ~ (p=0.468) JSONEncode 32.6ms × (0.97,1.08) 32.0ms × (0.98,1.03) ~ (p=0.190) JSONDecode 114ms × (0.97,1.05) 114ms × (0.99,1.01) ~ (p=0.887) Mandelbrot200 6.11ms × (0.98,1.04) 6.04ms × (1.00,1.01) ~ (p=0.167) GoParse 6.66ms × (0.97,1.04) 6.47ms × (0.97,1.05) -2.81% (p=0.014) RegexpMatchEasy0_32 159ns × (0.99,1.00) 171ns × (0.93,1.07) +7.19% (p=0.002) RegexpMatchEasy0_1K 538ns × (1.00,1.01) 550ns × (0.98,1.01) +2.30% (p=0.000) RegexpMatchEasy1_32 138ns × (1.00,1.00) 135ns × (0.99,1.02) -1.60% (p=0.000) RegexpMatchEasy1_1K 869ns × (0.99,1.01) 879ns × (1.00,1.01) +1.08% (p=0.000) RegexpMatchMedium_32 252ns × (0.99,1.01) 243ns × (1.00,1.00) -3.71% (p=0.000) RegexpMatchMedium_1K 72.7µs × (1.00,1.00) 70.3µs × (1.00,1.00) -3.34% (p=0.000) RegexpMatchHard_32 3.85µs × (1.00,1.00) 3.82µs × (1.00,1.01) -0.81% (p=0.000) RegexpMatchHard_1K 118µs × (1.00,1.00) 117µs × (1.00,1.00) -0.56% (p=0.000) Revcomp 920ms × (0.97,1.07) 917ms × (0.97,1.04) ~ (p=0.808) Template 129ms × (0.98,1.03) 114ms × (0.99,1.01) -12.06% (p=0.000) TimeParse 619ns × (0.99,1.01) 622ns × (0.99,1.01) ~ (p=0.062) TimeFormat 661ns × (0.98,1.04) 665ns × (0.99,1.01) ~ (p=0.524) See next CL for combination with a similar optimization for slice. The benchmarks that are slower in this CL are still faster overall with the combination of the two. Change-Id: I2a7421658091b2488c64741b4db15ab6c3b4cb7e Reviewed-on: https://go-review.googlesource.com/9812 Reviewed-by: David Chase <drchase@google.com>
2015-05-06 12:34:30 -04:00
return n
}
var l []*Node
ns := temp(nsrc.Type)
l = append(l, nod(OAS, ns, nsrc)) // s = src
na := nodintconst(int64(argc)) // const argc
nx := nod(OIF, nil, nil) // if cap(s) - len(s) < argc
nx.Left = nod(OLT, nod(OSUB, nod(OCAP, ns, nil), nod(OLEN, ns, nil)), na)
fn := syslook("growslice") // growslice(<type>, old []T, mincap int) (ret []T)
fn = substArgTypes(fn, ns.Type.Elem(), ns.Type.Elem())
nx.Nbody.Set1(nod(OAS, ns,
mkcall1(fn, ns.Type, &nx.Ninit, typename(ns.Type.Elem()), ns,
nod(OADD, nod(OLEN, ns, nil), na))))
l = append(l, nx)
nn := temp(types.Types[TINT])
l = append(l, nod(OAS, nn, nod(OLEN, ns, nil))) // n = len(s)
nx = nod(OSLICE, ns, nil) // ...s[:n+argc]
nx.SetSliceBounds(nil, nod(OADD, nn, na), nil)
nx.SetBounded(true)
l = append(l, nod(OAS, ns, nx)) // s = s[:n+argc]
ls = n.List.Slice()[1:]
for i, n := range ls {
nx = nod(OINDEX, ns, nn) // s[n] ...
cmd/compile: pack bool fields in Node, Name, Func and Type structs to bitsets This reduces compiler memory usage by up to 4% - see compilebench results below. name old time/op new time/op delta Template 245ms ± 4% 241ms ± 2% -1.88% (p=0.029 n=10+10) Unicode 126ms ± 3% 124ms ± 3% ~ (p=0.105 n=10+10) GoTypes 805ms ± 2% 813ms ± 3% ~ (p=0.515 n=8+10) Compiler 3.95s ± 2% 3.83s ± 1% -2.96% (p=0.000 n=9+10) MakeBash 47.4s ± 4% 46.6s ± 1% -1.59% (p=0.028 n=9+10) name old user-ns/op new user-ns/op delta Template 324M ± 5% 326M ± 3% ~ (p=0.935 n=10+10) Unicode 186M ± 5% 178M ±10% ~ (p=0.067 n=9+10) GoTypes 1.08G ± 7% 1.09G ± 4% ~ (p=0.956 n=10+10) Compiler 5.34G ± 4% 5.31G ± 1% ~ (p=0.501 n=10+8) name old alloc/op new alloc/op delta Template 41.0MB ± 0% 39.8MB ± 0% -3.03% (p=0.000 n=10+10) Unicode 32.3MB ± 0% 31.0MB ± 0% -4.13% (p=0.000 n=10+10) GoTypes 119MB ± 0% 116MB ± 0% -2.39% (p=0.000 n=10+10) Compiler 499MB ± 0% 487MB ± 0% -2.48% (p=0.000 n=10+10) name old allocs/op new allocs/op delta Template 380k ± 1% 379k ± 1% ~ (p=0.436 n=10+10) Unicode 324k ± 1% 324k ± 0% ~ (p=0.853 n=10+10) GoTypes 1.15M ± 0% 1.15M ± 0% ~ (p=0.481 n=10+10) Compiler 4.41M ± 0% 4.41M ± 0% -0.12% (p=0.007 n=10+10) name old text-bytes new text-bytes delta HelloSize 623k ± 0% 623k ± 0% ~ (all equal) CmdGoSize 6.64M ± 0% 6.64M ± 0% ~ (all equal) name old data-bytes new data-bytes delta HelloSize 5.81k ± 0% 5.81k ± 0% ~ (all equal) CmdGoSize 238k ± 0% 238k ± 0% ~ (all equal) name old bss-bytes new bss-bytes delta HelloSize 134k ± 0% 134k ± 0% ~ (all equal) CmdGoSize 152k ± 0% 152k ± 0% ~ (all equal) name old exe-bytes new exe-bytes delta HelloSize 967k ± 0% 967k ± 0% ~ (all equal) CmdGoSize 10.2M ± 0% 10.2M ± 0% ~ (all equal) Change-Id: I1f40af738254892bd6c8ba2eb43390b175753d52 Reviewed-on: https://go-review.googlesource.com/37445 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2017-02-27 19:56:38 +02:00
nx.SetBounded(true)
l = append(l, nod(OAS, nx, n)) // s[n] = arg
if i+1 < len(ls) {
l = append(l, nod(OAS, nn, nod(OADD, nn, nodintconst(1)))) // n = n + 1
}
}
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
typecheckslice(l, ctxStmt)
walkstmtlist(l)
init.Append(l...)
return ns
}
// Lower copy(a, b) to a memmove call or a runtime call.
//
// init {
// n := len(a)
// if n > len(b) { n = len(b) }
// if a.ptr != b.ptr { memmove(a.ptr, b.ptr, n*sizeof(elem(a))) }
// }
// n;
//
// Also works if b is a string.
//
func copyany(n *Node, init *Nodes, runtimecall bool) *Node {
if n.Left.Type.Elem().HasPointers() {
Curfn.Func.setWBPos(n.Pos)
fn := writebarrierfn("typedslicecopy", n.Left.Type.Elem(), n.Right.Type.Elem())
n.Left = cheapexpr(n.Left, init)
ptrL, lenL := n.Left.slicePtrLen()
n.Right = cheapexpr(n.Right, init)
ptrR, lenR := n.Right.slicePtrLen()
return mkcall1(fn, n.Type, init, typename(n.Left.Type.Elem()), ptrL, lenL, ptrR, lenR)
}
if runtimecall {
if n.Right.Type.IsString() {
fn := syslook("slicestringcopy")
n.Left = cheapexpr(n.Left, init)
ptr, len := n.Left.slicePtrLen()
str := nod(OCONVNOP, n.Right, nil)
str.Type = types.Types[TSTRING]
return mkcall1(fn, n.Type, init, ptr, len, str)
}
fn := syslook("slicecopy")
fn = substArgTypes(fn, n.Left.Type.Elem(), n.Right.Type.Elem())
n.Left = cheapexpr(n.Left, init)
ptrL, lenL := n.Left.slicePtrLen()
n.Right = cheapexpr(n.Right, init)
ptrR, lenR := n.Right.slicePtrLen()
return mkcall1(fn, n.Type, init, ptrL, lenL, ptrR, lenR, nodintconst(n.Left.Type.Elem().Width))
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n.Left = walkexpr(n.Left, init)
n.Right = walkexpr(n.Right, init)
nl := temp(n.Left.Type)
nr := temp(n.Right.Type)
var l []*Node
l = append(l, nod(OAS, nl, n.Left))
l = append(l, nod(OAS, nr, n.Right))
nfrm := nod(OSPTR, nr, nil)
nto := nod(OSPTR, nl, nil)
nlen := temp(types.Types[TINT])
// n = len(to)
l = append(l, nod(OAS, nlen, nod(OLEN, nl, nil)))
// if n > len(frm) { n = len(frm) }
nif := nod(OIF, nil, nil)
nif.Left = nod(OGT, nlen, nod(OLEN, nr, nil))
nif.Nbody.Append(nod(OAS, nlen, nod(OLEN, nr, nil)))
l = append(l, nif)
// if to.ptr != frm.ptr { memmove( ... ) }
ne := nod(OIF, nod(ONE, nto, nfrm), nil)
ne.SetLikely(true)
l = append(l, ne)
fn := syslook("memmove")
fn = substArgTypes(fn, nl.Type.Elem(), nl.Type.Elem())
nwid := temp(types.Types[TUINTPTR])
setwid := nod(OAS, nwid, conv(nlen, types.Types[TUINTPTR]))
ne.Nbody.Append(setwid)
nwid = nod(OMUL, nwid, nodintconst(nl.Type.Elem().Width))
call := mkcall1(fn, nil, init, nto, nfrm, nwid)
ne.Nbody.Append(call)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
typecheckslice(l, ctxStmt)
walkstmtlist(l)
init.Append(l...)
return nlen
}
func eqfor(t *types.Type) (n *Node, needsize bool) {
// Should only arrive here with large memory or
// a struct/array containing a non-memory field/element.
// Small memory is handled inline, and single non-memory
// is handled by walkcompare.
switch a, _ := algtype1(t); a {
case AMEM:
n := syslook("memequal")
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
n = substArgTypes(n, t, t)
return n, true
case ASPECIAL:
sym := typesymprefix(".eq", t)
n := newname(sym)
setNodeNameFunc(n)
n.Type = functype(nil, []*Node{
anonfield(types.NewPtr(t)),
anonfield(types.NewPtr(t)),
}, []*Node{
anonfield(types.Types[TBOOL]),
})
return n, false
}
Fatalf("eqfor %v", t)
return nil, false
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
// The result of walkcompare MUST be assigned back to n, e.g.
// n.Left = walkcompare(n.Left, init)
func walkcompare(n *Node, init *Nodes) *Node {
if n.Left.Type.IsInterface() && n.Right.Type.IsInterface() && n.Left.Op != OLITERAL && n.Right.Op != OLITERAL {
return walkcompareInterface(n, init)
}
if n.Left.Type.IsString() && n.Right.Type.IsString() {
return walkcompareString(n, init)
}
n.Left = walkexpr(n.Left, init)
n.Right = walkexpr(n.Right, init)
// Given mixed interface/concrete comparison,
// rewrite into types-equal && data-equal.
// This is efficient, avoids allocations, and avoids runtime calls.
if n.Left.Type.IsInterface() != n.Right.Type.IsInterface() {
// Preserve side-effects in case of short-circuiting; see #32187.
l := cheapexpr(n.Left, init)
r := cheapexpr(n.Right, init)
// Swap so that l is the interface value and r is the concrete value.
if n.Right.Type.IsInterface() {
l, r = r, l
}
// Handle both == and !=.
eq := n.Op
andor := OOROR
if eq == OEQ {
andor = OANDAND
}
// Check for types equal.
// For empty interface, this is:
// l.tab == type(r)
// For non-empty interface, this is:
// l.tab != nil && l.tab._type == type(r)
var eqtype *Node
tab := nod(OITAB, l, nil)
rtyp := typename(r.Type)
if l.Type.IsEmptyInterface() {
tab.Type = types.NewPtr(types.Types[TUINT8])
tab.SetTypecheck(1)
eqtype = nod(eq, tab, rtyp)
} else {
nonnil := nod(brcom(eq), nodnil(), tab)
match := nod(eq, itabType(tab), rtyp)
eqtype = nod(andor, nonnil, match)
}
// Check for data equal.
eqdata := nod(eq, ifaceData(n.Pos, l, r.Type), r)
// Put it all together.
expr := nod(andor, eqtype, eqdata)
n = finishcompare(n, expr, init)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
return n
}
// Must be comparison of array or struct.
// Otherwise back end handles it.
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
// While we're here, decide whether to
// inline or call an eq alg.
t := n.Left.Type
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
var inline bool
maxcmpsize := int64(4)
unalignedLoad := canMergeLoads()
if unalignedLoad {
// Keep this low enough to generate less code than a function call.
maxcmpsize = 2 * int64(thearch.LinkArch.RegSize)
}
switch t.Etype {
default:
if Debug_libfuzzer != 0 && t.IsInteger() {
n.Left = cheapexpr(n.Left, init)
n.Right = cheapexpr(n.Right, init)
// If exactly one comparison operand is
// constant, invoke the constcmp functions
// instead, and arrange for the constant
// operand to be the first argument.
l, r := n.Left, n.Right
if r.Op == OLITERAL {
l, r = r, l
}
constcmp := l.Op == OLITERAL && r.Op != OLITERAL
var fn string
var paramType *types.Type
switch t.Size() {
case 1:
fn = "libfuzzerTraceCmp1"
if constcmp {
fn = "libfuzzerTraceConstCmp1"
}
paramType = types.Types[TUINT8]
case 2:
fn = "libfuzzerTraceCmp2"
if constcmp {
fn = "libfuzzerTraceConstCmp2"
}
paramType = types.Types[TUINT16]
case 4:
fn = "libfuzzerTraceCmp4"
if constcmp {
fn = "libfuzzerTraceConstCmp4"
}
paramType = types.Types[TUINT32]
case 8:
fn = "libfuzzerTraceCmp8"
if constcmp {
fn = "libfuzzerTraceConstCmp8"
}
paramType = types.Types[TUINT64]
default:
Fatalf("unexpected integer size %d for %v", t.Size(), t)
}
init.Append(mkcall(fn, nil, init, tracecmpArg(l, paramType, init), tracecmpArg(r, paramType, init)))
}
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
return n
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
case TARRAY:
// We can compare several elements at once with 2/4/8 byte integer compares
inline = t.NumElem() <= 1 || (issimple[t.Elem().Etype] && (t.NumElem() <= 4 || t.Elem().Width*t.NumElem() <= maxcmpsize))
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
case TSTRUCT:
inline = t.NumComponents(types.IgnoreBlankFields) <= 4
}
cmpl := n.Left
for cmpl != nil && cmpl.Op == OCONVNOP {
cmpl = cmpl.Left
}
cmpr := n.Right
for cmpr != nil && cmpr.Op == OCONVNOP {
cmpr = cmpr.Left
}
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
// Chose not to inline. Call equality function directly.
if !inline {
// eq algs take pointers; cmpl and cmpr must be addressable
if !islvalue(cmpl) || !islvalue(cmpr) {
Fatalf("arguments of comparison must be lvalues - %v %v", cmpl, cmpr)
}
fn, needsize := eqfor(t)
call := nod(OCALL, fn, nil)
call.List.Append(nod(OADDR, cmpl, nil))
call.List.Append(nod(OADDR, cmpr, nil))
if needsize {
call.List.Append(nodintconst(t.Width))
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
}
res := call
if n.Op != OEQ {
res = nod(ONOT, res, nil)
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
}
n = finishcompare(n, res, init)
return n
}
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
// inline: build boolean expression comparing element by element
andor := OANDAND
if n.Op == ONE {
andor = OOROR
}
var expr *Node
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
compare := func(el, er *Node) {
a := nod(n.Op, el, er)
if expr == nil {
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
expr = a
} else {
expr = nod(andor, expr, a)
}
}
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
cmpl = safeexpr(cmpl, init)
cmpr = safeexpr(cmpr, init)
if t.IsStruct() {
for _, f := range t.Fields().Slice() {
sym := f.Sym
if sym.IsBlank() {
continue
}
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
compare(
nodSym(OXDOT, cmpl, sym),
nodSym(OXDOT, cmpr, sym),
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
)
}
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
} else {
step := int64(1)
remains := t.NumElem() * t.Elem().Width
combine64bit := unalignedLoad && Widthreg == 8 && t.Elem().Width <= 4 && t.Elem().IsInteger()
combine32bit := unalignedLoad && t.Elem().Width <= 2 && t.Elem().IsInteger()
combine16bit := unalignedLoad && t.Elem().Width == 1 && t.Elem().IsInteger()
for i := int64(0); remains > 0; {
var convType *types.Type
switch {
case remains >= 8 && combine64bit:
convType = types.Types[TINT64]
step = 8 / t.Elem().Width
case remains >= 4 && combine32bit:
convType = types.Types[TUINT32]
step = 4 / t.Elem().Width
case remains >= 2 && combine16bit:
convType = types.Types[TUINT16]
step = 2 / t.Elem().Width
default:
step = 1
}
if step == 1 {
compare(
nod(OINDEX, cmpl, nodintconst(i)),
nod(OINDEX, cmpr, nodintconst(i)),
)
i++
remains -= t.Elem().Width
} else {
elemType := t.Elem().ToUnsigned()
cmplw := nod(OINDEX, cmpl, nodintconst(i))
cmplw = conv(cmplw, elemType) // convert to unsigned
cmplw = conv(cmplw, convType) // widen
cmprw := nod(OINDEX, cmpr, nodintconst(i))
cmprw = conv(cmprw, elemType)
cmprw = conv(cmprw, convType)
// For code like this: uint32(s[0]) | uint32(s[1])<<8 | uint32(s[2])<<16 ...
// ssa will generate a single large load.
for offset := int64(1); offset < step; offset++ {
lb := nod(OINDEX, cmpl, nodintconst(i+offset))
lb = conv(lb, elemType)
lb = conv(lb, convType)
lb = nod(OLSH, lb, nodintconst(8*t.Elem().Width*offset))
cmplw = nod(OOR, cmplw, lb)
rb := nod(OINDEX, cmpr, nodintconst(i+offset))
rb = conv(rb, elemType)
rb = conv(rb, convType)
rb = nod(OLSH, rb, nodintconst(8*t.Elem().Width*offset))
cmprw = nod(OOR, cmprw, rb)
}
compare(cmplw, cmprw)
i += step
remains -= step * t.Elem().Width
}
}
}
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
if expr == nil {
expr = nodbool(n.Op == OEQ)
// We still need to use cmpl and cmpr, in case they contain
// an expression which might panic. See issue 23837.
t := temp(cmpl.Type)
a1 := nod(OAS, t, cmpl)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
a1 = typecheck(a1, ctxStmt)
a2 := nod(OAS, t, cmpr)
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
a2 = typecheck(a2, ctxStmt)
init.Append(a1, a2)
}
cmd/compile: when inlining ==, don’t take the address of the values This CL reworks walkcompare for clarity and concision. It also makes one significant functional change. (The functional change is hard to separate cleanly from the cleanup, so I just did them together.) When inlining and unrolling an equality comparison for a small struct or array, compare the elements like: a[0] == b[0] && a[1] == b[1] rather than pa := &a pb := &b pa[0] == pb[0] && pa[1] == pb[1] The result is the same, but taking the address and working through the indirect forces the backends to generate less efficient code. This is only an improvement with the SSA backend. However, every port but s390x now has a working SSA backend, and switching to the SSA backend by default everywhere is a priority for Go 1.8. It thus seems reasonable to start to prioritize SSA performance over the old backend. Updates #15303 Sample code: type T struct { a, b int8 } func g(a T) bool { return a == T{1, 2} } SSA before: "".g t=1 size=80 args=0x10 locals=0x8 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $8-16 0x0000 00000 (badeq.go:7) SUBQ $8, SP 0x0004 00004 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0004 00004 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0004 00004 (badeq.go:8) MOVBLZX "".a+16(FP), AX 0x0009 00009 (badeq.go:8) MOVB AL, "".autotmp_0+6(SP) 0x000d 00013 (badeq.go:8) MOVBLZX "".a+17(FP), AX 0x0012 00018 (badeq.go:8) MOVB AL, "".autotmp_0+7(SP) 0x0016 00022 (badeq.go:8) MOVB $0, "".autotmp_1+4(SP) 0x001b 00027 (badeq.go:8) MOVB $1, "".autotmp_1+4(SP) 0x0020 00032 (badeq.go:8) MOVB $2, "".autotmp_1+5(SP) 0x0025 00037 (badeq.go:8) MOVBLZX "".autotmp_0+6(SP), AX 0x002a 00042 (badeq.go:8) MOVBLZX "".autotmp_1+4(SP), CX 0x002f 00047 (badeq.go:8) CMPB AL, CL 0x0031 00049 (badeq.go:8) JNE 70 0x0033 00051 (badeq.go:8) MOVBLZX "".autotmp_0+7(SP), AX 0x0038 00056 (badeq.go:8) CMPB AL, $2 0x003a 00058 (badeq.go:8) SETEQ AL 0x003d 00061 (badeq.go:8) MOVB AL, "".~r1+24(FP) 0x0041 00065 (badeq.go:8) ADDQ $8, SP 0x0045 00069 (badeq.go:8) RET 0x0046 00070 (badeq.go:8) MOVB $0, AL 0x0048 00072 (badeq.go:8) JMP 61 SSA after: "".g t=1 size=32 args=0x10 locals=0x0 0x0000 00000 (badeq.go:7) TEXT "".g(SB), $0-16 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) NOP 0x0000 00000 (badeq.go:7) FUNCDATA $0, gclocals·23e8278e2b69a3a75fa59b23c49ed6ad(SB) 0x0000 00000 (badeq.go:7) FUNCDATA $1, gclocals·33cdeccccebe80329f1fdbee7f5874cb(SB) 0x0000 00000 (badeq.go:8) MOVBLZX "".a+8(FP), AX 0x0005 00005 (badeq.go:8) CMPB AL, $1 0x0007 00007 (badeq.go:8) JNE 25 0x0009 00009 (badeq.go:8) MOVBLZX "".a+9(FP), CX 0x000e 00014 (badeq.go:8) CMPB CL, $2 0x0011 00017 (badeq.go:8) SETEQ AL 0x0014 00020 (badeq.go:8) MOVB AL, "".~r1+16(FP) 0x0018 00024 (badeq.go:8) RET 0x0019 00025 (badeq.go:8) MOVB $0, AL 0x001b 00027 (badeq.go:8) JMP 20 Change-Id: I120185d58012b7bbcdb1ec01225b5b08d0855d86 Reviewed-on: https://go-review.googlesource.com/22277 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2016-04-15 15:49:00 -07:00
n = finishcompare(n, expr, init)
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
return n
}
func tracecmpArg(n *Node, t *types.Type, init *Nodes) *Node {
// Ugly hack to avoid "constant -1 overflows uintptr" errors, etc.
if n.Op == OLITERAL && n.Type.IsSigned() && n.Int64() < 0 {
n = copyexpr(n, n.Type, init)
}
return conv(n, t)
}
func walkcompareInterface(n *Node, init *Nodes) *Node {
n.Right = cheapexpr(n.Right, init)
n.Left = cheapexpr(n.Left, init)
eqtab, eqdata := eqinterface(n.Left, n.Right)
var cmp *Node
if n.Op == OEQ {
cmp = nod(OANDAND, eqtab, eqdata)
} else {
eqtab.Op = ONE
cmp = nod(OOROR, eqtab, nod(ONOT, eqdata, nil))
}
return finishcompare(n, cmp, init)
}
func walkcompareString(n *Node, init *Nodes) *Node {
// Rewrite comparisons to short constant strings as length+byte-wise comparisons.
var cs, ncs *Node // const string, non-const string
switch {
case Isconst(n.Left, CTSTR) && Isconst(n.Right, CTSTR):
// ignore; will be constant evaluated
case Isconst(n.Left, CTSTR):
cs = n.Left
ncs = n.Right
case Isconst(n.Right, CTSTR):
cs = n.Right
ncs = n.Left
}
if cs != nil {
cmp := n.Op
// Our comparison below assumes that the non-constant string
// is on the left hand side, so rewrite "" cmp x to x cmp "".
// See issue 24817.
if Isconst(n.Left, CTSTR) {
cmp = brrev(cmp)
}
// maxRewriteLen was chosen empirically.
// It is the value that minimizes cmd/go file size
// across most architectures.
// See the commit description for CL 26758 for details.
maxRewriteLen := 6
// Some architectures can load unaligned byte sequence as 1 word.
// So we can cover longer strings with the same amount of code.
canCombineLoads := canMergeLoads()
combine64bit := false
if canCombineLoads {
// Keep this low enough to generate less code than a function call.
maxRewriteLen = 2 * thearch.LinkArch.RegSize
combine64bit = thearch.LinkArch.RegSize >= 8
}
var and Op
switch cmp {
case OEQ:
and = OANDAND
case ONE:
and = OOROR
default:
// Don't do byte-wise comparisons for <, <=, etc.
// They're fairly complicated.
// Length-only checks are ok, though.
maxRewriteLen = 0
}
if s := strlit(cs); len(s) <= maxRewriteLen {
if len(s) > 0 {
ncs = safeexpr(ncs, init)
}
r := nod(cmp, nod(OLEN, ncs, nil), nodintconst(int64(len(s))))
remains := len(s)
for i := 0; remains > 0; {
if remains == 1 || !canCombineLoads {
cb := nodintconst(int64(s[i]))
ncb := nod(OINDEX, ncs, nodintconst(int64(i)))
r = nod(and, r, nod(cmp, ncb, cb))
remains--
i++
continue
}
var step int
var convType *types.Type
switch {
case remains >= 8 && combine64bit:
convType = types.Types[TINT64]
step = 8
case remains >= 4:
convType = types.Types[TUINT32]
step = 4
case remains >= 2:
convType = types.Types[TUINT16]
step = 2
}
ncsubstr := nod(OINDEX, ncs, nodintconst(int64(i)))
ncsubstr = conv(ncsubstr, convType)
csubstr := int64(s[i])
// Calculate large constant from bytes as sequence of shifts and ors.
// Like this: uint32(s[0]) | uint32(s[1])<<8 | uint32(s[2])<<16 ...
// ssa will combine this into a single large load.
for offset := 1; offset < step; offset++ {
b := nod(OINDEX, ncs, nodintconst(int64(i+offset)))
b = conv(b, convType)
b = nod(OLSH, b, nodintconst(int64(8*offset)))
ncsubstr = nod(OOR, ncsubstr, b)
csubstr |= int64(s[i+offset]) << uint8(8*offset)
}
csubstrPart := nodintconst(csubstr)
// Compare "step" bytes as once
r = nod(and, r, nod(cmp, csubstrPart, ncsubstr))
remains -= step
i += step
}
return finishcompare(n, r, init)
}
}
var r *Node
if n.Op == OEQ || n.Op == ONE {
// prepare for rewrite below
n.Left = cheapexpr(n.Left, init)
n.Right = cheapexpr(n.Right, init)
eqlen, eqmem := eqstring(n.Left, n.Right)
// quick check of len before full compare for == or !=.
// memequal then tests equality up to length len.
if n.Op == OEQ {
// len(left) == len(right) && memequal(left, right, len)
r = nod(OANDAND, eqlen, eqmem)
} else {
// len(left) != len(right) || !memequal(left, right, len)
eqlen.Op = ONE
r = nod(OOROR, eqlen, nod(ONOT, eqmem, nil))
}
} else {
// sys_cmpstring(s1, s2) :: 0
r = mkcall("cmpstring", types.Types[TINT], init, conv(n.Left, types.Types[TSTRING]), conv(n.Right, types.Types[TSTRING]))
r = nod(n.Op, r, nodintconst(0))
}
return finishcompare(n, r, init)
}
// The result of finishcompare MUST be assigned back to n, e.g.
cmd/compile: reduce use of **Node parameters Escape analysis has a hard time with tree-like structures (see #13493 and #14858). This is unlikely to change. As a result, when invoking a function that accepts a **Node parameter, we usually allocate a *Node on the heap. This happens a whole lot. This CL changes functions from taking a **Node to acting more like append: It both modifies the input and returns a replacement for it. Because of the cascading nature of escape analysis, in order to get the benefits, I had to modify almost all such functions. The remaining functions are in racewalk and the backend. I would be happy to update them as well in a separate CL. This CL was created by manually updating the function signatures and the directly impacted bits of code. The callsites were then automatically updated using a bespoke script: https://gist.github.com/josharian/046b1be7aceae244de39 For ease of reviewing and future understanding, this CL is also broken down into four CLs, mailed separately, which show the manual and the automated changes separately. They are CLs 20990, 20991, 20992, and 20993. Passes toolstash -cmp. name old time/op new time/op delta Template 335ms ± 5% 324ms ± 5% -3.35% (p=0.000 n=23+24) Unicode 176ms ± 9% 165ms ± 6% -6.12% (p=0.000 n=23+24) GoTypes 1.10s ± 4% 1.07s ± 2% -2.77% (p=0.000 n=24+24) Compiler 5.31s ± 3% 5.15s ± 3% -2.95% (p=0.000 n=24+24) MakeBash 41.6s ± 1% 41.7s ± 2% ~ (p=0.586 n=23+23) name old alloc/op new alloc/op delta Template 63.3MB ± 0% 62.4MB ± 0% -1.36% (p=0.000 n=25+23) Unicode 42.4MB ± 0% 41.6MB ± 0% -1.99% (p=0.000 n=24+25) GoTypes 220MB ± 0% 217MB ± 0% -1.11% (p=0.000 n=25+25) Compiler 994MB ± 0% 973MB ± 0% -2.08% (p=0.000 n=24+25) name old allocs/op new allocs/op delta Template 681k ± 0% 574k ± 0% -15.71% (p=0.000 n=24+25) Unicode 518k ± 0% 413k ± 0% -20.34% (p=0.000 n=25+24) GoTypes 2.08M ± 0% 1.78M ± 0% -14.62% (p=0.000 n=25+25) Compiler 9.26M ± 0% 7.64M ± 0% -17.48% (p=0.000 n=25+25) name old text-bytes new text-bytes delta HelloSize 578k ± 0% 578k ± 0% ~ (all samples are equal) CmdGoSize 6.46M ± 0% 6.46M ± 0% ~ (all samples are equal) name old data-bytes new data-bytes delta HelloSize 128k ± 0% 128k ± 0% ~ (all samples are equal) CmdGoSize 281k ± 0% 281k ± 0% ~ (all samples are equal) name old exe-bytes new exe-bytes delta HelloSize 921k ± 0% 921k ± 0% ~ (all samples are equal) CmdGoSize 9.86M ± 0% 9.86M ± 0% ~ (all samples are equal) Change-Id: I277d95bd56d51c166ef7f560647aeaa092f3f475 Reviewed-on: https://go-review.googlesource.com/20959 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
2016-03-20 08:03:31 -07:00
// n.Left = finishcompare(n.Left, x, r, init)
func finishcompare(n, r *Node, init *Nodes) *Node {
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
r = typecheck(r, ctxExpr)
r = conv(r, n.Type)
r = walkexpr(r, init)
return r
}
// return 1 if integer n must be in range [0, max), 0 otherwise
func bounded(n *Node, max int64) bool {
if n.Type == nil || !n.Type.IsInteger() {
return false
}
sign := n.Type.IsSigned()
bits := int32(8 * n.Type.Width)
if smallintconst(n) {
v := n.Int64()
return 0 <= v && v < max
}
switch n.Op {
case OAND:
v := int64(-1)
if smallintconst(n.Left) {
v = n.Left.Int64()
} else if smallintconst(n.Right) {
v = n.Right.Int64()
}
if 0 <= v && v < max {
return true
}
case OMOD:
if !sign && smallintconst(n.Right) {
v := n.Right.Int64()
if 0 <= v && v <= max {
return true
}
}
case ODIV:
if !sign && smallintconst(n.Right) {
v := n.Right.Int64()
for bits > 0 && v >= 2 {
bits--
v >>= 1
}
}
case ORSH:
if !sign && smallintconst(n.Right) {
v := n.Right.Int64()
if v > int64(bits) {
return true
}
bits -= int32(v)
}
}
if !sign && bits <= 62 && 1<<uint(bits) <= max {
return true
}
return false
}
// usemethod checks interface method calls for uses of reflect.Type.Method.
func usemethod(n *Node) {
t := n.Left.Type
// Looking for either of:
// Method(int) reflect.Method
// MethodByName(string) (reflect.Method, bool)
//
// TODO(crawshaw): improve precision of match by working out
// how to check the method name.
if n := t.NumParams(); n != 1 {
return
}
if n := t.NumResults(); n != 1 && n != 2 {
return
}
p0 := t.Params().Field(0)
res0 := t.Results().Field(0)
var res1 *types.Field
if t.NumResults() == 2 {
res1 = t.Results().Field(1)
}
if res1 == nil {
if p0.Type.Etype != TINT {
return
}
} else {
if !p0.Type.IsString() {
return
}
if !res1.Type.IsBoolean() {
return
}
}
// Note: Don't rely on res0.Type.String() since its formatting depends on multiple factors
// (including global variables such as numImports - was issue #19028).
// Also need to check for reflect package itself (see Issue #38515).
if s := res0.Type.Sym; s != nil && s.Name == "Method" && isReflectPkg(s.Pkg) {
Curfn.Func.SetReflectMethod(true)
}
}
func usefield(n *Node) {
if objabi.Fieldtrack_enabled == 0 {
return
}
switch n.Op {
default:
Fatalf("usefield %v", n.Op)
case ODOT, ODOTPTR:
break
}
if n.Sym == nil {
// No field name. This DOTPTR was built by the compiler for access
// to runtime data structures. Ignore.
return
}
t := n.Left.Type
if t.IsPtr() {
t = t.Elem()
}
field := dotField[typeSymKey{t.Orig, n.Sym}]
if field == nil {
Fatalf("usefield %v %v without paramfld", n.Left.Type, n.Sym)
}
if !strings.Contains(field.Note, "go:\"track\"") {
return
}
outer := n.Left.Type
if outer.IsPtr() {
outer = outer.Elem()
}
if outer.Sym == nil {
yyerror("tracked field must be in named struct type")
}
if !types.IsExported(field.Sym.Name) {
yyerror("tracked field must be exported (upper case)")
}
sym := tracksym(outer, field)
if Curfn.Func.FieldTrack == nil {
Curfn.Func.FieldTrack = make(map[*types.Sym]struct{})
}
Curfn.Func.FieldTrack[sym] = struct{}{}
}
func candiscardlist(l Nodes) bool {
for _, n := range l.Slice() {
if !candiscard(n) {
return false
}
}
return true
}
func candiscard(n *Node) bool {
if n == nil {
return true
}
switch n.Op {
default:
return false
// Discardable as long as the subpieces are.
case ONAME,
ONONAME,
OTYPE,
OPACK,
OLITERAL,
OADD,
OSUB,
OOR,
OXOR,
OADDSTR,
OADDR,
OANDAND,
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
OBYTES2STR,
ORUNES2STR,
OSTR2BYTES,
OSTR2RUNES,
OCAP,
OCOMPLIT,
OMAPLIT,
OSTRUCTLIT,
OARRAYLIT,
OSLICELIT,
OPTRLIT,
OCONV,
OCONVIFACE,
OCONVNOP,
ODOT,
OEQ,
ONE,
OLT,
OLE,
OGT,
OGE,
OKEY,
cmd/compile: add OSTRUCTKEY for keyed struct literals Previously, we used OKEY nodes to represent keyed struct literal elements. The field names were represented by an ONAME node, but this is clumsy because it's the only remaining case where ONAME was used to represent a bare identifier and not a variable. This CL introduces a new OSTRUCTKEY node op for use in struct literals. These ops instead store the field name in the node's own Sym field. This is similar in spirit to golang.org/cl/20890. Significant reduction in allocations for struct literal heavy code like package unicode: name old time/op new time/op delta Template 345ms ± 6% 341ms ± 6% ~ (p=0.141 n=29+28) Unicode 200ms ± 9% 184ms ± 7% -7.77% (p=0.000 n=29+30) GoTypes 1.04s ± 3% 1.05s ± 3% ~ (p=0.096 n=30+30) Compiler 4.47s ± 9% 4.49s ± 6% ~ (p=0.890 n=29+29) name old user-ns/op new user-ns/op delta Template 523M ±13% 516M ±17% ~ (p=0.400 n=29+30) Unicode 334M ±27% 314M ±30% ~ (p=0.093 n=30+30) GoTypes 1.53G ±10% 1.52G ±10% ~ (p=0.572 n=30+30) Compiler 6.28G ± 7% 6.34G ±11% ~ (p=0.300 n=30+30) name old alloc/op new alloc/op delta Template 44.5MB ± 0% 44.4MB ± 0% -0.35% (p=0.000 n=27+30) Unicode 39.2MB ± 0% 34.5MB ± 0% -11.79% (p=0.000 n=26+30) GoTypes 125MB ± 0% 125MB ± 0% -0.12% (p=0.000 n=29+30) Compiler 515MB ± 0% 515MB ± 0% -0.10% (p=0.000 n=29+30) name old allocs/op new allocs/op delta Template 426k ± 0% 424k ± 0% -0.39% (p=0.000 n=29+30) Unicode 374k ± 0% 323k ± 0% -13.67% (p=0.000 n=29+30) GoTypes 1.21M ± 0% 1.21M ± 0% -0.14% (p=0.000 n=29+29) Compiler 4.40M ± 0% 4.39M ± 0% -0.13% (p=0.000 n=29+30) Passes toolstash/buildall. Change-Id: Iba4ee765dd1748f67e52fcade1cd75c9f6e13fa9 Reviewed-on: https://go-review.googlesource.com/30974 Run-TryBot: Matthew Dempsky <mdempsky@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
2016-10-12 15:48:18 -07:00
OSTRUCTKEY,
OLEN,
OMUL,
OLSH,
ORSH,
OAND,
OANDNOT,
ONEW,
ONOT,
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
OBITNOT,
OPLUS,
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
ONEG,
OOROR,
OPAREN,
ORUNESTR,
OREAL,
OIMAG,
OCOMPLEX:
break
// Discardable as long as we know it's not division by zero.
case ODIV, OMOD:
if Isconst(n.Right, CTINT) && n.Right.Val().U.(*Mpint).CmpInt64(0) != 0 {
break
}
if Isconst(n.Right, CTFLT) && n.Right.Val().U.(*Mpflt).CmpFloat64(0) != 0 {
break
}
return false
// Discardable as long as we know it won't fail because of a bad size.
case OMAKECHAN, OMAKEMAP:
if Isconst(n.Left, CTINT) && n.Left.Val().U.(*Mpint).CmpInt64(0) == 0 {
break
}
return false
// Difficult to tell what sizes are okay.
case OMAKESLICE:
return false
cmd/compile: optimize make+copy pattern to avoid memclr match: m = make([]T, x); copy(m, s) for pointer free T and x==len(s) rewrite to: m = mallocgc(x*elemsize(T), nil, false); memmove(&m, &s, x*elemsize(T)) otherwise rewrite to: m = makeslicecopy([]T, x, s) This avoids memclear and shading of pointers in the newly created slice before the copy. With this CL "s" is only be allowed to bev a variable and not a more complex expression. This restriction could be lifted in future versions of this optimization when it can be proven that "s" is not referencing "m". Triggers 450 times during make.bash.. Reduces go binary size by ~8 kbyte. name old time/op new time/op delta MakeSliceCopy/mallocmove/Byte 71.1ns ± 1% 65.8ns ± 0% -7.49% (p=0.000 n=10+9) MakeSliceCopy/mallocmove/Int 71.2ns ± 1% 66.0ns ± 0% -7.27% (p=0.000 n=10+8) MakeSliceCopy/mallocmove/Ptr 104ns ± 4% 99ns ± 1% -5.13% (p=0.000 n=10+10) MakeSliceCopy/makecopy/Byte 70.3ns ± 0% 68.0ns ± 0% -3.22% (p=0.000 n=10+9) MakeSliceCopy/makecopy/Int 70.3ns ± 0% 68.5ns ± 1% -2.59% (p=0.000 n=9+10) MakeSliceCopy/makecopy/Ptr 102ns ± 0% 99ns ± 1% -2.97% (p=0.000 n=9+9) MakeSliceCopy/nilappend/Byte 75.4ns ± 0% 74.9ns ± 2% -0.63% (p=0.015 n=9+9) MakeSliceCopy/nilappend/Int 75.6ns ± 0% 76.4ns ± 3% ~ (p=0.245 n=9+10) MakeSliceCopy/nilappend/Ptr 107ns ± 0% 108ns ± 1% +0.93% (p=0.005 n=9+10) Fixes #26252 Change-Id: Iec553dd1fef6ded16197216a472351c8799a8e71 Reviewed-on: https://go-review.googlesource.com/c/go/+/146719 Reviewed-by: Keith Randall <khr@golang.org> Run-TryBot: Martin Möhrmann <moehrmann@google.com> TryBot-Result: Gobot Gobot <gobot@golang.org>
2018-10-23 13:50:07 +02:00
case OMAKESLICECOPY:
return false
}
if !candiscard(n.Left) || !candiscard(n.Right) || !candiscardlist(n.Ninit) || !candiscardlist(n.Nbody) || !candiscardlist(n.List) || !candiscardlist(n.Rlist) {
return false
}
return true
}
// Rewrite
// go builtin(x, y, z)
// into
// go func(a1, a2, a3) {
// builtin(a1, a2, a3)
// }(x, y, z)
// for print, println, and delete.
//
// Rewrite
// go f(x, y, uintptr(unsafe.Pointer(z)))
// into
// go func(a1, a2, a3) {
// builtin(a1, a2, uintptr(a3))
// }(x, y, unsafe.Pointer(z))
// for function contains unsafe-uintptr arguments.
var wrapCall_prgen int
// The result of wrapCall MUST be assigned back to n, e.g.
// n.Left = wrapCall(n.Left, init)
func wrapCall(n *Node, init *Nodes) *Node {
if n.Ninit.Len() != 0 {
walkstmtlist(n.Ninit.Slice())
init.AppendNodes(&n.Ninit)
}
isBuiltinCall := n.Op != OCALLFUNC && n.Op != OCALLMETH && n.Op != OCALLINTER
// origArgs keeps track of what argument is uintptr-unsafe/unsafe-uintptr conversion.
origArgs := make([]*Node, n.List.Len())
t := nod(OTFUNC, nil, nil)
for i, arg := range n.List.Slice() {
s := lookupN("a", i)
if !isBuiltinCall && arg.Op == OCONVNOP && arg.Type.Etype == TUINTPTR && arg.Left.Type.IsUnsafePtr() {
origArgs[i] = arg
arg = arg.Left
n.List.SetIndex(i, arg)
}
t.List.Append(symfield(s, arg.Type))
}
wrapCall_prgen++
sym := lookupN("wrap·", wrapCall_prgen)
fn := dclfunc(sym, t)
args := paramNnames(t.Type)
for i, origArg := range origArgs {
if origArg == nil {
continue
}
arg := nod(origArg.Op, args[i], nil)
arg.Type = origArg.Type
args[i] = arg
}
call := nod(n.Op, nil, nil)
if !isBuiltinCall {
call.Op = OCALL
call.Left = n.Left
}
call.List.Set(args)
fn.Nbody.Set1(call)
funcbody()
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
fn = typecheck(fn, ctxStmt)
typecheckslice(fn.Nbody.Slice(), ctxStmt)
xtop = append(xtop, fn)
call = nod(OCALL, nil, nil)
call.Left = fn.Func.Nname
call.List.Set(n.List.Slice())
call = typecheck(call, ctxStmt)
call = walkexpr(call, init)
return call
}
// substArgTypes substitutes the given list of types for
// successive occurrences of the "any" placeholder in the
// type syntax expression n.Type.
// The result of substArgTypes MUST be assigned back to old, e.g.
// n.Left = substArgTypes(n.Left, t1, t2)
func substArgTypes(old *Node, types_ ...*types.Type) *Node {
n := old.copy()
for _, t := range types_ {
dowidth(t)
}
n.Type = types.SubstAny(n.Type, &types_)
if len(types_) > 0 {
Fatalf("substArgTypes: too many argument types")
}
return n
}
// canMergeLoads reports whether the backend optimization passes for
// the current architecture can combine adjacent loads into a single
// larger, possibly unaligned, load. Note that currently the
// optimizations must be able to handle little endian byte order.
func canMergeLoads() bool {
switch thearch.LinkArch.Family {
case sys.ARM64, sys.AMD64, sys.I386, sys.S390X:
return true
case sys.PPC64:
// Load combining only supported on ppc64le.
return thearch.LinkArch.ByteOrder == binary.LittleEndian
}
return false
}
// isRuneCount reports whether n is of the form len([]rune(string)).
// These are optimized into a call to runtime.countrunes.
func isRuneCount(n *Node) bool {
cmd/compile: bulk rename This change does a bulk rename of several identifiers in the compiler. See #27167 and https://docs.google.com/document/d/19_ExiylD9MRfeAjKIfEsMU1_RGhuxB9sA0b5Zv7byVI/ for context and for discussion of these particular renames. Commands run to generate this change: gorename -from '"cmd/compile/internal/gc".OPROC' -to OGO gorename -from '"cmd/compile/internal/gc".OCOM' -to OBITNOT gorename -from '"cmd/compile/internal/gc".OMINUS' -to ONEG gorename -from '"cmd/compile/internal/gc".OIND' -to ODEREF gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTR' -to OBYTES2STR gorename -from '"cmd/compile/internal/gc".OARRAYBYTESTRTMP' -to OBYTES2STRTMP gorename -from '"cmd/compile/internal/gc".OARRAYRUNESTR' -to ORUNES2STR gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTE' -to OSTR2BYTES gorename -from '"cmd/compile/internal/gc".OSTRARRAYBYTETMP' -to OSTR2BYTESTMP gorename -from '"cmd/compile/internal/gc".OSTRARRAYRUNE' -to OSTR2RUNES gorename -from '"cmd/compile/internal/gc".Etop' -to ctxStmt gorename -from '"cmd/compile/internal/gc".Erv' -to ctxExpr gorename -from '"cmd/compile/internal/gc".Ecall' -to ctxCallee gorename -from '"cmd/compile/internal/gc".Efnstruct' -to ctxMultiOK gorename -from '"cmd/compile/internal/gc".Easgn' -to ctxAssign gorename -from '"cmd/compile/internal/gc".Ecomplit' -to ctxCompLit Not altered: parameters and local variables (mostly in typecheck.go) named top, which should probably now be called ctx (and which should probably have a named type). Also not altered: Field called Top in gc.Func. gorename -from '"cmd/compile/internal/gc".Node.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/gc".Node.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/gc".nodeIsddd' -to nodeIsDDD gorename -from '"cmd/compile/internal/types".Field.Isddd' -to IsDDD gorename -from '"cmd/compile/internal/types".Field.SetIsddd' -to SetIsDDD gorename -from '"cmd/compile/internal/types".fieldIsddd' -to fieldIsDDD Not altered: function gc.hasddd, params and local variables called isddd Also not altered: fmt.go prints nodes using "isddd(%v)". cd cmd/compile/internal/gc; go generate I then manually found impacted comments using exact string match and fixed them up by hand. The comment changes were trivial. Passes toolstash-check. Fixes #27167. If this experiment is deemed a success, we will open a new tracking issue for renames to do at the end of the 1.13 cycles. Change-Id: I2dc541533d2ab0d06cb3d31d65df205ecfb151e8 Reviewed-on: https://go-review.googlesource.com/c/150140 Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
2018-11-18 08:34:38 -08:00
return Debug['N'] == 0 && !instrumenting && n.Op == OLEN && n.Left.Op == OSTR2RUNES
}
func walkCheckPtrAlignment(n *Node, init *Nodes, count *Node) *Node {
if !n.Type.IsPtr() {
Fatalf("expected pointer type: %v", n.Type)
}
elem := n.Type.Elem()
if count != nil {
if !elem.IsArray() {
Fatalf("expected array type: %v", elem)
}
elem = elem.Elem()
}
size := elem.Size()
if elem.Alignment() == 1 && (size == 0 || size == 1 && count == nil) {
return n
}
if count == nil {
count = nodintconst(1)
}
n.Left = cheapexpr(n.Left, init)
init.Append(mkcall("checkptrAlignment", nil, init, convnop(n.Left, types.Types[TUNSAFEPTR]), typename(elem), conv(count, types.Types[TUINTPTR])))
return n
}
var walkCheckPtrArithmeticMarker byte
func walkCheckPtrArithmetic(n *Node, init *Nodes) *Node {
// Calling cheapexpr(n, init) below leads to a recursive call
// to walkexpr, which leads us back here again. Use n.Opt to
// prevent infinite loops.
if opt := n.Opt(); opt == &walkCheckPtrArithmeticMarker {
return n
} else if opt != nil {
// We use n.Opt() here because today it's not used for OCONVNOP. If that changes,
// there's no guarantee that temporarily replacing it is safe, so just hard fail here.
Fatalf("unexpected Opt: %v", opt)
}
n.SetOpt(&walkCheckPtrArithmeticMarker)
defer n.SetOpt(nil)
// TODO(mdempsky): Make stricter. We only need to exempt
// reflect.Value.Pointer and reflect.Value.UnsafeAddr.
switch n.Left.Op {
case OCALLFUNC, OCALLMETH, OCALLINTER:
return n
}
if n.Left.Op == ODOTPTR && isReflectHeaderDataField(n.Left) {
return n
}
// Find original unsafe.Pointer operands involved in this
// arithmetic expression.
//
// "It is valid both to add and to subtract offsets from a
// pointer in this way. It is also valid to use &^ to round
// pointers, usually for alignment."
var originals []*Node
var walk func(n *Node)
walk = func(n *Node) {
switch n.Op {
case OADD:
walk(n.Left)
walk(n.Right)
case OSUB:
walk(n.Left)
case OAND:
if n.Implicit() { // was OANDNOT
walk(n.Left)
}
case OCONVNOP:
if n.Left.Type.IsUnsafePtr() {
n.Left = cheapexpr(n.Left, init)
originals = append(originals, convnop(n.Left, types.Types[TUNSAFEPTR]))
}
}
}
walk(n.Left)
n = cheapexpr(n, init)
slice := mkdotargslice(types.NewSlice(types.Types[TUNSAFEPTR]), originals)
slice.Esc = EscNone
init.Append(mkcall("checkptrArithmetic", nil, init, convnop(n, types.Types[TUNSAFEPTR]), slice))
// TODO(khr): Mark backing store of slice as dead. This will allow us to reuse
// the backing store for multiple calls to checkptrArithmetic.
return n
}
// checkPtr reports whether pointer checking should be enabled for
// function fn at a given level. See debugHelpFooter for defined
// levels.
func checkPtr(fn *Node, level int) bool {
return Debug_checkptr >= level && fn.Func.Pragma&NoCheckPtr == 0
}