Stowage/go
2
0
Fork 0
mirror of https://github.com/golang/go.git synced 2025-12-08 06:10:04 +00:00
Code Issues Projects Releases Packages Wiki Activity

cmd/compile/internal/ssa: combine load + op on AMD64

Browse source 
On AMD64 Most operation can have one operand in memory.
Combine load and dependand operation into one new operation,
where possible. I've seen no significant performance changes on go1,
but this allows to remove ~1.8kb code from go tool. And in math package
I see e. g.:

Remainder-6            70.0ns ± 0%   64.6ns ± 0%   -7.76%  (p=0.000 n=9+1
Change-Id: I88b8602b1d55da8ba548a34eb7da4b25d59a297e
Reviewed-on: https://go-review.googlesource.com/36793
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
This commit is contained in:
Ilya Tocar 2017-02-10 13:17:20 -06:00
parent a9292b833b
commit 21c71d7788
5 changed files with 1160 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

20
src/cmd/compile/internal/ssa/gen/AMD64Ops.go
View file

@ -127,6 +127,7 @@ func init() {
flagsgpax = regInfo{inputs: nil, clobbers: ax, outputs: []regMask{gp &^ ax}}
gpload = regInfo{inputs: []regMask{gpspsb, 0}, outputs: gponly}
gp21load = regInfo{inputs: []regMask{gp, gpspsb, 0}, outputs: gponly}
gploadidx = regInfo{inputs: []regMask{gpspsb, gpsp, 0}, outputs: gponly}
gpstore = regInfo{inputs: []regMask{gpspsb, gpsp, 0}}
@ -138,6 +139,7 @@ func init() {
fp01 = regInfo{inputs: nil, outputs: fponly}
fp21 = regInfo{inputs: []regMask{fp, fp}, outputs: fponly}
fp21load = regInfo{inputs: []regMask{fp, gpspsb, 0}, outputs: fponly}
fpgp = regInfo{inputs: fponly, outputs: gponly}
gpfp = regInfo{inputs: gponly, outputs: fponly}
fp11 = regInfo{inputs: fponly, outputs: fponly}
@ -177,6 +179,13 @@ func init() {
{name: "MOVSDstoreidx1", argLength: 4, reg: fpstoreidx, asm: "MOVSD", aux: "SymOff"}, // fp64 indexed by i store
{name: "MOVSDstoreidx8", argLength: 4, reg: fpstoreidx, asm: "MOVSD", aux: "SymOff"}, // fp64 indexed by 8i store
{name: "ADDSDmem", argLength: 3, reg: fp21load, asm: "ADDSD", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true}, // fp32 arg0 + tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "ADDSSmem", argLength: 3, reg: fp21load, asm: "ADDSS", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true}, // fp32 arg0 + tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "SUBSSmem", argLength: 3, reg: fp21load, asm: "SUBSS", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true}, // fp32 arg0 - tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "SUBSDmem", argLength: 3, reg: fp21load, asm: "SUBSD", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true}, // fp64 arg0 - tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "MULSSmem", argLength: 3, reg: fp21load, asm: "MULSS", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true}, // fp32 arg0 * tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "MULSDmem", argLength: 3, reg: fp21load, asm: "MULSD", aux: "SymOff", resultInArg0: true, faultOnNilArg1: true}, // fp64 arg0 * tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
// binary ops
{name: "ADDQ", argLength: 2, reg: gp21sp, asm: "ADDQ", commutative: true, clobberFlags: true}, // arg0 + arg1
{name: "ADDL", argLength: 2, reg: gp21sp, asm: "ADDL", commutative: true, clobberFlags: true}, // arg0 + arg1
@ -279,6 +288,17 @@ func init() {
{name: "ROLWconst", argLength: 1, reg: gp11, asm: "ROLW", aux: "Int16", resultInArg0: true, clobberFlags: true}, // arg0 rotate left auxint, rotate amount 0-15
{name: "ROLBconst", argLength: 1, reg: gp11, asm: "ROLB", aux: "Int8", resultInArg0: true, clobberFlags: true}, // arg0 rotate left auxint, rotate amount 0-7
{name: "ADDLmem", argLength: 3, reg: gp21load, asm: "ADDL", typ: "UInt32", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true}, // arg0 + tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "ADDQmem", argLength: 3, reg: gp21load, asm: "ADDQ", typ: "UInt64", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true}, // arg0 + tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "SUBQmem", argLength: 3, reg: gp21load, asm: "SUBQ", typ: "UInt64", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true}, // arg0 - tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "SUBLmem", argLength: 3, reg: gp21load, asm: "SUBL", typ: "UInt32", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true}, // arg0 - tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "ANDLmem", argLength: 3, reg: gp21load, asm: "ANDL", typ: "UInt32", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true}, // arg0 & tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "ANDQmem", argLength: 3, reg: gp21load, asm: "ANDQ", typ: "UInt64", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true}, // arg0 & tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "ORQmem", argLength: 3, reg: gp21load, asm: "ORQ", typ: "UInt64", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true}, // arg0 | tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "ORLmem", argLength: 3, reg: gp21load, asm: "ORL", typ: "UInt32", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true}, // arg0 | tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "XORQmem", argLength: 3, reg: gp21load, asm: "XORQ", typ: "UInt64", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true}, // arg0 ^ tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
{name: "XORLmem", argLength: 3, reg: gp21load, asm: "XORL", typ: "UInt32", aux: "SymOff", resultInArg0: true, clobberFlags: true, faultOnNilArg1: true}, // arg0 ^ tmp, tmp loaded from arg1+auxint+aux, arg2 = mem
// unary ops
{name: "NEGQ", argLength: 1, reg: gp11, asm: "NEGQ", resultInArg0: true, clobberFlags: true}, // -arg0
{name: "NEGL", argLength: 1, reg: gp11, asm: "NEGL", resultInArg0: true, clobberFlags: true}, // -arg0