Minor cleanup: remove the symbol attribute AttrSeenGlobal, since it is
redundant with the existing attribute AttrOnList (no need to have what
amounts to a separate flag for checking the same property).
Change-Id: Ia269b64de37c2bb4a2314bbecf3d2091c6d57424
Reviewed-on: https://go-review.googlesource.com/c/go/+/239477
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
For content-addressable symbols with relocations, we build a
content hash based on its content and relocations. Depending on
the category of the referenced symbol, we choose different hash
algorithms such that the hash is globally consistent.
For now, we only support content-addressable symbols with
relocations when the current package's import path is known, so
that the symbol names are fully expanded. Otherwise, if the
referenced symbol is a named symbol whose name is not fully
expanded, the hash won't be globally consistent, and can cause
erroneous collisions. This is fine for now, as the deduplication
is just an optimization, not a requirement for correctness (until
we get to type descriptors).
Change-Id: I639e4e03dd749b5d71f0a55c2525926575b1ac30
Reviewed-on: https://go-review.googlesource.com/c/go/+/243142
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Jeremy Faller <jeremy@golang.org>
This CL introduces content-addressable symbols (a.k.a. hashed
symbols) to object files. Content-addressable symbols are
identified and referenced by their content hashes, instead of by
names.
In the object file, a new pseudo-package index PkgIdxHashed is
introduced, for content-addressable symbols, and a new block is
added to store their hashes. The hashes are used by the linker to
identify and deduplicate the symbols.
For now, we only support content-addressable symbols that are
always locally defined (i.e. no cross-package references).
As a proof of concept, make string constant symbols content-
addressable.
Change-Id: Iaf53efd74c0ffb54fa95f784628cc84e95844536
Reviewed-on: https://go-review.googlesource.com/c/go/+/242079
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Jeremy Faller <jeremy@golang.org>
Mark compiler-generated ".stmp_%d" and "<fn>.stkobj" symbols as
AttrStatic, so as to tell the linker that they do not need to be
inserted into its name lookup tables.
Change-Id: I59ffd11659b2c54c2d0ad41275d05c3f919e3b88
Reviewed-on: https://go-review.googlesource.com/c/go/+/240497
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
All callers to gdata knew the kind of node they were working with,
so all calls to gdata have been replaced with more specific calls.
Some OADDR nodes were constructed solely for the purpose of
passing them to gdata for unwrapping. In those cases, we can now
cut to the chase.
Passes toolstash-check.
Change-Id: Iacc1abefd7f748cb269661a03768d3367319b0b0
Reviewed-on: https://go-review.googlesource.com/c/go/+/228888
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
The previous change moved code around to create slicesym.
This change simplifies slicesym and its callsites
by accepting an int64 for lencap instead of a node,
and by removing all the calls to gdata.
It also stops modifying n,
which avoids the need to make a copy of it.
Passes toolstash-check.
Change-Id: I4d25454d11b4bb8941000244443e3c99eef4bdd0
Reviewed-on: https://go-review.googlesource.com/c/go/+/227550
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
This change mostly moves code around to unify it.
A subsequent change will simplify and improve slicesym.
Passes toolstash-check.
Change-Id: I84a877ea747febb2b571d4089ba6d905b51b27ec
Reviewed-on: https://go-review.googlesource.com/c/go/+/227549
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
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>
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 cmd 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: I51a389860b9676cfa1b84722f5fb84d3c4ee9e28
Reviewed-on: https://go-review.googlesource.com/c/go/+/190098
Reviewed-by: Austin Clements <austin@google.com>
It is convenient to have a seekable writer. A later CL will make
use of Seek.
Change-Id: Iba0107ce2975d9a451d97f16aa91a318dd4c90e2
Reviewed-on: https://go-review.googlesource.com/c/go/+/196028
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Jeremy Faller <jeremy@golang.org>
Currently, at the end of compilation, the compiler writes out the
export data, the linker object file header, then does more
code/data generation, then writes the main content of the linker
object file. This CL refactors it to finish all the code/data
generation before writing any output file.
A later CL will inject some code that operates on all defined
symbols before writing the output. This ensures all the symbols
are available at that point.
Change-Id: I97d946553fd0ffd298234c520219540d29783576
Reviewed-on: https://go-review.googlesource.com/c/go/+/196027
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Stack object symbol is named as <FunctionName>.stkobj. If the
function itself is not DUPOK, its stack object symbol should only
be defined in the package where the function is defined,
therefore no duplicates.
If in the future we change the stack object symbol to
content-hash naming, as other gcdata symbols, we can mark it as
DUPOK.
Change-Id: I5aee96578940e2f76e7115d96cd2716021672c03
Reviewed-on: https://go-review.googlesource.com/c/go/+/191437
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Renaming the method makes clear, both to readers and to vet,
that this method is not the implementation of io.Seeker:
it cannot fail.
Working toward making the tree vet-safe instead of having
so many exceptions in cmd/vet/all/whitelist.
For #31916.
Change-Id: I3e6ad7264cb0121b4b76935450cccb71d533e96b
Reviewed-on: https://go-review.googlesource.com/c/go/+/176108
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Currently, when the compiler emits a symbol name in the object
file, it uses "". for the package path of the package being
compiled. This is then expanded in the linker to the actual
package path.
With CL 173938, it does not need an allocation if the symbol name
does not need expansion. In many cases, the compiler actually
knows the package path (through the -p flag), so we could just
write it out in compile time, without fixing it up in the linker.
This reduces allocations in the linker.
In case that the package path is not known (compiler's -p flag is
missing, or the object file is generated by the assembler), the
linker still does the expansion.
This reduces ~100MB allocations (~10% inuse_space) in linking
k8s.io/kubernetes/cmd/kube-apiserver on Linux/AMD64.
Also makes the linker a little faster: linking cmd/go on
Linux/AMD64:
Real 1.13 ± 1% 1.11 ± 1% -2.13% (p=0.000 n=10+10)
User 1.17 ± 3% 1.14 ± 5% -3.14% (p=0.003 n=10+10)
Sys 0.34 ±15% 0.34 ±15% ~ (p=0.986 n=10+10)
The caveat is that the object files get slightly bigger. On
Linux/AMD64, runtime.a gets 2.1% bigger, cmd/compile/internal/ssa
(which has a longer import path) gets 2.8% bigger.
This reveals that when building an unnamed plugin (e.g.
go build -buildmode=plugin x.go), the go command passes different
package paths to the compiler and to the linker. Before this CL
there seems nothing obviously broken, but given that the compiler
already emits the package's import path in various places (e.g.
debug info), I guess it is possible that this leads to some
unexpected behavior. Now that the compiler writes the package
path in more places, this disagreement actually leads to
unresolved symbols. Adjust the go command to use the same package
path for both compiling and linking.
Change-Id: I19f08981f51db577871c906e08d9e0fd588a2dd8
Reviewed-on: https://go-review.googlesource.com/c/go/+/174657
Reviewed-by: Austin Clements <austin@google.com>
We compile package sort as part of the compiler bootstrap,
to make sure the compiler uses a consistent sort algorithm
no matter what version of Go it is compiled against.
(This matters for elements that compare "equal" but are distinguishable.)
Package sort was compiled in such a way as to disallow
sort.Slice entirely during bootstrap (at least with some compilers),
while cmd/internal/obj was compiled in such a way as to
make obj.SortSlice available to all compilers, precisely because
sort.Slice was not. This is all highly confusing.
Simplify by making sort.Slice available all the time.
Followup to CL 169137 and #30440
(and also CL 40114 and CL 73951).
Change-Id: I127f4e02d6c71392805d256c3a90ef7c51f9ba0c
Reviewed-on: https://go-review.googlesource.com/c/go/+/174525
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Stack object generation code was always using the local package name
for its symbol. Normally that doesn't matter, as we usually only
compile functions in the local package. But for wrappers, the compiler
generates functions which live in other packages. When there are two
other packages with identical functions to wrap, the same name appears
twice, and the compiler goes boom.
Fixes#31252
Change-Id: I7026eebabe562cb159b8b6046cf656afd336ba25
Reviewed-on: https://go-review.googlesource.com/c/go/+/171464
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Currently, WriteObjFile deduplicates symbols by name. This is a
strange and unexpected place to do this. But, worse, there's no
checking that it's reasonable to deduplicate two symbols, so this
makes it incredibly easy to mask errors involving duplicate symbols.
Dealing with duplicate symbols is better left to the linker. We're
also about to introduce multiple symbols with the same name but
different ABIs/versions, which would make this deduplication more
complicated. We just removed the only part of the compiler that
actually depended on this behavior.
This CL removes symbol deduplication from WriteObjFile, since it is no
longer needed.
For #27539.
Change-Id: I650c550e46e83f95c67cb6c6646f9b2f7f10df30
Reviewed-on: https://go-review.googlesource.com/c/146558
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
Currently, liveness produces a distinct obj.LSym for each GC bitmap
for each function. These are then named by content hash and only
ultimately deduplicated by WriteObjFile.
For various reasons (see next commit), we want to remove this
deduplication behavior from WriteObjFile. Furthermore, it's
inefficient to produce these duplicate symbols in the first place.
GC bitmaps are the only source of duplicate symbols in the compiler.
This commit eliminates these duplicate symbols by declaring them in
the Ctxt symbol hash just like every other obj.LSym. As a result, all
GC bitmaps with the same content now refer to the same obj.LSym.
The next commit will remove deduplication from WriteObjFile.
For #27539.
Change-Id: I4f15e3d99530122cdf473b7a838c69ef5f79db59
Reviewed-on: https://go-review.googlesource.com/c/146557
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
I don't know how this file wasn't gofmted.
Change-Id: I9b3765ae63970b7bc4dc87107f546e64a78e2830
Reviewed-on: https://go-review.googlesource.com/c/146497
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Nowadays there are better ways to safely run untrusted Go programs, like
NaCl and gVisor.
Change-Id: I20c45f13a50dbcf35c343438b720eb93e7b4e13a
Reviewed-on: https://go-review.googlesource.com/c/142717
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
This used to be used by cmd/vet and some assembly generation tests, but
those were removed in CL 37691 and CL 107336. No point in keeping an
unneeded flag around.
Fixes#28220.
Change-Id: I59f8546954ab36ea61ceba81c10d6e16d74b966a
Reviewed-on: https://go-review.googlesource.com/c/142677
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Rework how the compiler+runtime handles stack-allocated variables
whose address is taken.
Direct references to such variables work as before. References through
pointers, however, use a new mechanism. The new mechanism is more
precise than the old "ambiguously live" mechanism. It computes liveness
at runtime based on the actual references among objects on the stack.
Each function records all of its address-taken objects in a FUNCDATA.
These are called "stack objects". The runtime then uses that
information while scanning a stack to find all of the stack objects on
a stack. It then does a mark phase on the stack objects, using all the
pointers found on the stack (and ancillary structures, like defer
records) as the root set. Only stack objects which are found to be
live during this mark phase will be scanned and thus retain any heap
objects they point to.
A subsequent CL will remove all the "ambiguously live" logic from
the compiler, so that the stack object tracing will be required.
For this CL, the stack tracing is all redundant with the current
ambiguously live logic.
Update #22350
Change-Id: Ide19f1f71a5b6ec8c4d54f8f66f0e9a98344772f
Reviewed-on: https://go-review.googlesource.com/c/134155
Reviewed-by: Austin Clements <austin@google.com>
I was wrong. There was a need to loop here.
Fixes#24761
Change-Id: If13b3ab72febde930bdaebdddd1c05e0d0446020
Reviewed-on: https://go-review.googlesource.com/105615
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
The standard library has plenty of polished encoder/decoder
implementations. No need for another ad-hoc one.
I considered using encoding/gob instead, but these strings go into the
package data part of the object file, so it's important they don't
contain "\n$$\n". Package json escapes newlines in strings, so it's
safe to use here.
Change-Id: I998655524ccee7365c2c8e9a843e6975e95a3e62
Reviewed-on: https://go-review.googlesource.com/106463
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
CL 40693 added concurrent backend compilation support,
and used it for user-provided functions.
Autogenerated functions were still compiled serially.
This CL brings them into the fold.
As of this CL, when requested,
no functions are compiled serially.
There generally aren't many autogenerated functions.
When there are, this CL can help a lot,
because autogenerated functions are usually short.
Many short functions is the best case scenario
for concurrent compilation; see CL 41192.
One example of such a package comes from Dave Cheney's benchjuju:
github.com/juju/govmomi/vim25/types.
It has thousands of autogenerated functions.
This CL improves performance on the entire benchmark
by around a second on my machine at c=8, or about ~5%.
Updates #15756
Change-Id: Ia21e302b2469a9ed743df02244ec7ebde55b32f3
Reviewed-on: https://go-review.googlesource.com/41503
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>
By always writing out pack files, the object file format can be
simplified somewhat. In particular, the export data format will no
longer require escaping, because the pack file provides appropriate
framing.
This CL does not affect build systems that use -pack, which includes
all major Go build systems (cmd/go, gb, bazel).
Also, existing package import logic already distinguishes pack/object
files based on file contents rather than file extension.
The only exception is cmd/pack, which specially handled object files
created by cmd/compile when used with the 'c' mode. This mode is
extended to now recognize the pack files produced by cmd/compile and
handle them as before.
Passes toolstash-check.
Updates #21705.
Updates #24512.
Change-Id: Idf131013bfebd73a5cde7e087eb19964503a9422
Reviewed-on: https://go-review.googlesource.com/102236
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
As reported by unparam.
Passes toolstash -cmp on std cmd.
Change-Id: I55473e1eed096ed1c3e431aed2cbf0b6b5444b91
Reviewed-on: https://go-review.googlesource.com/97895
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
The linker will refuse to work on objects larger than
2e9 bytes (see issue #9862 for why).
With this change, the compiler gives a useful error
message explaining this, instead of leaving it to the
linker to give a cryptic message later.
Fixes#1700.
Change-Id: I3933ce08ef846721ece7405bdba81dff644cb004
Reviewed-on: https://go-review.googlesource.com/74330
Reviewed-by: Robert Griesemer <gri@golang.org>
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
* replace a copy of IsMethod with a call of it.
* a few more switches where they simplify the code.
* prefer composite literals over "n := new(...); n.x = y; ...".
* use defers to get rid of three goto labels.
* rewrite updateHasCall into two funcs to remove gotos.
Passes toolstash-check on std cmd.
Change-Id: Icb5442a89a87319ef4b640bbc5faebf41b193ef1
Reviewed-on: https://go-review.googlesource.com/72070
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
The core dump reader would like a bunch of ideal int
constants to be available in dwarf.
Makes the go binary 0.9% bigger.
Update #14517
Change-Id: I00cdfc7f53bcdc56fccba576c1d33010f03bdd95
Reviewed-on: https://go-review.googlesource.com/69270
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Prior to this CL, the compiler and assembler
were sloppy about the LSym.Type for LSyms
containing static data.
The linker then fixed this up, converting
Sxxx and SBSS to SDATA, and SNOPTRBSS to SNOPTRDATA
if it noticed that the symbol had associated data.
It is preferable to just get this right in cmd/compile
and cmd/asm, because it removes an unnecessary traversal
of the symbol table from the linker (see #14624).
Do this by touching up the LSym.Type fixes in
LSym.prepwrite and Link.Globl.
I have confirmed by instrumenting the linker
that the now-eliminated code paths were unreached.
And an additional check in the object file writing code
will help preserve that invariant.
There was a case in the Windows linker,
with internal linking and cgo,
where we were generating SNOPTRBSS symbols with data.
For now, convert those at the site at which they occur
into SNOPTRDATA, just like they were.
Does not pass toolstash-check,
but does generate identical linked binaries.
No compiler performance changes.
Change-Id: I77b071ab103685ff8e042cee9abb864385488872
Reviewed-on: https://go-review.googlesource.com/40864
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Alex Brainman <alex.brainman@gmail.com>
Reviewed-by: Michael Hudson-Doyle <michael.hudson@canonical.com>
This CL adds initial support for concurrent backend compilation.
BACKGROUND
The compiler currently consists (very roughly) of the following phases:
1. Initialization.
2. Lexing and parsing into the cmd/compile/internal/syntax AST.
3. Translation into the cmd/compile/internal/gc AST.
4. Some gc AST passes: typechecking, escape analysis, inlining,
closure handling, expression evaluation ordering (order.go),
and some lowering and optimization (walk.go).
5. Translation into the cmd/compile/internal/ssa SSA form.
6. Optimization and lowering of SSA form.
7. Translation from SSA form to assembler instructions.
8. Translation from assembler instructions to machine code.
9. Writing lots of output: machine code, DWARF symbols,
type and reflection info, export data.
Phase 2 was already concurrent as of Go 1.8.
Phase 3 is planned for eventual removal;
we hope to go straight from syntax AST to SSA.
Phases 5–8 are per-function; this CL adds support for
processing multiple functions concurrently.
The slowest phases in the compiler are 5 and 6,
so this offers the opportunity for some good speed-ups.
Unfortunately, it's not quite that straightforward.
In the current compiler, the latter parts of phase 4
(order, walk) are done function-at-a-time as needed.
Making order and walk concurrency-safe proved hard,
and they're not particularly slow, so there wasn't much reward.
To enable phases 5–8 to be done concurrently,
when concurrent backend compilation is requested,
we complete phase 4 for all functions
before starting later phases for any functions.
Also, in reality, we automatically generate new
functions in phase 9, such as method wrappers
and equality and has routines.
Those new functions then go through phases 4–8.
This CL disables concurrent backend compilation
after the first, big, user-provided batch of
functions has been compiled.
This is done to keep things simple,
and because the autogenerated functions
tend to be small, few, simple, and fast to compile.
USAGE
Concurrent backend compilation still defaults to off.
To set the number of functions that may be backend-compiled
concurrently, use the compiler flag -c.
In future work, cmd/go will automatically set -c.
Furthermore, this CL has been intentionally written
so that the c=1 path has no backend concurrency whatsoever,
not even spawning any goroutines.
This helps ensure that, should problems arise
late in the development cycle,
we can simply have cmd/go set c=1 always,
and revert to the original compiler behavior.
MUTEXES
Most of the work required to make concurrent backend
compilation safe has occurred over the past month.
This CL adds a handful of mutexes to get the rest of the way there;
they are the mutexes that I didn't see a clean way to avoid.
Some of them may still be eliminable in future work.
In no particular order:
* gc.funcsymsmu. The global funcsyms slice is populated
lazily when we need function symbols for closures.
This occurs during gc AST to SSA translation.
The function funcsym also does a package lookup,
which is a source of races on types.Pkg.Syms;
funcsymsmu also covers that package lookup.
This mutex is low priority: it adds a single global,
it is in an infrequently used code path, and it is low contention.
Since funcsyms may now be added in any order,
we must sort them to preserve reproducible builds.
* gc.largeStackFramesMu. We don't discover until after SSA compilation
that a function's stack frame is gigantic.
Recording that error happens basically never,
but it does happen concurrently.
Fix with a low priority mutex and sorting.
* obj.Link.hashmu. ctxt.hash stores the mapping from
types.Syms (compiler symbols) to obj.LSyms (linker symbols).
It is accessed fairly heavily through all the phases.
This is the only heavily contended mutex.
* gc.signatlistmu. The global signatlist map is
populated with types through several of the concurrent phases,
including notably via ngotype during DWARF generation.
It is low priority for removal.
* gc.typepkgmu. Looking up symbols in the types package
happens a fair amount during backend compilation
and DWARF generation, particularly via ngotype.
This mutex helps us to avoid a broader mutex on types.Pkg.Syms.
It has low-to-moderate contention.
* types.internedStringsmu. gc AST to SSA conversion and
some SSA work introduce new autotmps.
Those autotmps have their names interned to reduce allocations.
That interning requires protecting types.internedStrings.
The autotmp names are heavily re-used, and the mutex
overhead and contention here are low, so it is probably
a worthwhile performance optimization to keep this mutex.
TESTING
I have been testing this code locally by running
'go install -race cmd/compile'
and then doing
'go build -a -gcflags=-c=128 std cmd'
for all architectures and a variety of compiler flags.
This obviously needs to be made part of the builders,
but it is too expensive to make part of all.bash.
I have filed #19962 for this.
REPRODUCIBLE BUILDS
This version of the compiler generates reproducible builds.
Testing reproducible builds also needs automation, however,
and is also too expensive for all.bash.
This is #19961.
Also of note is that some of the compiler flags used by 'toolstash -cmp'
are currently incompatible with concurrent backend compilation.
They still work fine with c=1.
Time will tell whether this is a problem.
NEXT STEPS
* Continue to find and fix races and bugs,
using a combination of code inspection, fuzzing,
and hopefully some community experimentation.
I do not know of any outstanding races,
but there probably are some.
* Improve testing.
* Improve performance, for many values of c.
* Integrate with cmd/go and fine tune.
* Support concurrent compilation with the -race flag.
It is a sad irony that it does not yet work.
* Minor code cleanup that has been deferred during
the last month due to uncertainty about the
ultimate shape of this CL.
PERFORMANCE
Here's the buried lede, at last. :)
All benchmarks are from my 8 core 2.9 GHz Intel Core i7 darwin/amd64 laptop.
First, going from tip to this CL with c=1 has almost no impact.
name old time/op new time/op delta
Template 195ms ± 3% 194ms ± 5% ~ (p=0.370 n=30+29)
Unicode 86.6ms ± 3% 87.0ms ± 7% ~ (p=0.958 n=29+30)
GoTypes 548ms ± 3% 555ms ± 4% +1.35% (p=0.001 n=30+28)
Compiler 2.51s ± 2% 2.54s ± 2% +1.17% (p=0.000 n=28+30)
SSA 5.16s ± 3% 5.16s ± 2% ~ (p=0.910 n=30+29)
Flate 124ms ± 5% 124ms ± 4% ~ (p=0.947 n=30+30)
GoParser 146ms ± 3% 146ms ± 3% ~ (p=0.150 n=29+28)
Reflect 354ms ± 3% 352ms ± 4% ~ (p=0.096 n=29+29)
Tar 107ms ± 5% 106ms ± 3% ~ (p=0.370 n=30+29)
XML 200ms ± 4% 201ms ± 4% ~ (p=0.313 n=29+28)
[Geo mean] 332ms 333ms +0.10%
name old user-time/op new user-time/op delta
Template 227ms ± 5% 225ms ± 5% ~ (p=0.457 n=28+27)
Unicode 109ms ± 4% 109ms ± 5% ~ (p=0.758 n=29+29)
GoTypes 713ms ± 4% 721ms ± 5% ~ (p=0.051 n=30+29)
Compiler 3.36s ± 2% 3.38s ± 3% ~ (p=0.146 n=30+30)
SSA 7.46s ± 3% 7.47s ± 3% ~ (p=0.804 n=30+29)
Flate 146ms ± 7% 147ms ± 3% ~ (p=0.833 n=29+27)
GoParser 179ms ± 5% 179ms ± 5% ~ (p=0.866 n=30+30)
Reflect 431ms ± 4% 429ms ± 4% ~ (p=0.593 n=29+30)
Tar 124ms ± 5% 123ms ± 5% ~ (p=0.140 n=29+29)
XML 243ms ± 4% 242ms ± 7% ~ (p=0.404 n=29+29)
[Geo mean] 415ms 415ms +0.02%
name old obj-bytes new obj-bytes delta
Template 382k ± 0% 382k ± 0% ~ (all equal)
Unicode 203k ± 0% 203k ± 0% ~ (all equal)
GoTypes 1.18M ± 0% 1.18M ± 0% ~ (all equal)
Compiler 3.98M ± 0% 3.98M ± 0% ~ (all equal)
SSA 8.28M ± 0% 8.28M ± 0% ~ (all equal)
Flate 230k ± 0% 230k ± 0% ~ (all equal)
GoParser 287k ± 0% 287k ± 0% ~ (all equal)
Reflect 1.00M ± 0% 1.00M ± 0% ~ (all equal)
Tar 190k ± 0% 190k ± 0% ~ (all equal)
XML 416k ± 0% 416k ± 0% ~ (all equal)
[Geo mean] 660k 660k +0.00%
Comparing this CL to itself, from c=1 to c=2
improves real times 20-30%, costs 5-10% more CPU time,
and adds about 2% alloc.
The allocation increase comes from allocating more ssa.Caches.
name old time/op new time/op delta
Template 202ms ± 3% 149ms ± 3% -26.15% (p=0.000 n=49+49)
Unicode 87.4ms ± 4% 84.2ms ± 3% -3.68% (p=0.000 n=48+48)
GoTypes 560ms ± 2% 398ms ± 2% -28.96% (p=0.000 n=49+49)
Compiler 2.46s ± 3% 1.76s ± 2% -28.61% (p=0.000 n=48+46)
SSA 6.17s ± 2% 4.04s ± 1% -34.52% (p=0.000 n=49+49)
Flate 126ms ± 3% 92ms ± 2% -26.81% (p=0.000 n=49+48)
GoParser 148ms ± 4% 107ms ± 2% -27.78% (p=0.000 n=49+48)
Reflect 361ms ± 3% 281ms ± 3% -22.10% (p=0.000 n=49+49)
Tar 109ms ± 4% 86ms ± 3% -20.81% (p=0.000 n=49+47)
XML 204ms ± 3% 144ms ± 2% -29.53% (p=0.000 n=48+45)
name old user-time/op new user-time/op delta
Template 246ms ± 9% 246ms ± 4% ~ (p=0.401 n=50+48)
Unicode 109ms ± 4% 111ms ± 4% +1.47% (p=0.000 n=44+50)
GoTypes 728ms ± 3% 765ms ± 3% +5.04% (p=0.000 n=46+50)
Compiler 3.33s ± 3% 3.41s ± 2% +2.31% (p=0.000 n=49+48)
SSA 8.52s ± 2% 9.11s ± 2% +6.93% (p=0.000 n=49+47)
Flate 149ms ± 4% 161ms ± 3% +8.13% (p=0.000 n=50+47)
GoParser 181ms ± 5% 192ms ± 2% +6.40% (p=0.000 n=49+46)
Reflect 452ms ± 9% 474ms ± 2% +4.99% (p=0.000 n=50+48)
Tar 126ms ± 6% 136ms ± 4% +7.95% (p=0.000 n=50+49)
XML 247ms ± 5% 264ms ± 3% +6.94% (p=0.000 n=48+50)
name old alloc/op new alloc/op delta
Template 38.8MB ± 0% 39.3MB ± 0% +1.48% (p=0.008 n=5+5)
Unicode 29.8MB ± 0% 30.2MB ± 0% +1.19% (p=0.008 n=5+5)
GoTypes 113MB ± 0% 114MB ± 0% +0.69% (p=0.008 n=5+5)
Compiler 443MB ± 0% 447MB ± 0% +0.95% (p=0.008 n=5+5)
SSA 1.25GB ± 0% 1.26GB ± 0% +0.89% (p=0.008 n=5+5)
Flate 25.3MB ± 0% 25.9MB ± 1% +2.35% (p=0.008 n=5+5)
GoParser 31.7MB ± 0% 32.2MB ± 0% +1.59% (p=0.008 n=5+5)
Reflect 78.2MB ± 0% 78.9MB ± 0% +0.91% (p=0.008 n=5+5)
Tar 26.6MB ± 0% 27.0MB ± 0% +1.80% (p=0.008 n=5+5)
XML 42.4MB ± 0% 43.4MB ± 0% +2.35% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
Template 379k ± 0% 378k ± 0% ~ (p=0.421 n=5+5)
Unicode 322k ± 0% 321k ± 0% ~ (p=0.222 n=5+5)
GoTypes 1.14M ± 0% 1.14M ± 0% ~ (p=0.548 n=5+5)
Compiler 4.12M ± 0% 4.11M ± 0% -0.14% (p=0.032 n=5+5)
SSA 9.72M ± 0% 9.72M ± 0% ~ (p=0.421 n=5+5)
Flate 234k ± 1% 234k ± 0% ~ (p=0.421 n=5+5)
GoParser 316k ± 1% 315k ± 0% ~ (p=0.222 n=5+5)
Reflect 980k ± 0% 979k ± 0% ~ (p=0.095 n=5+5)
Tar 249k ± 1% 249k ± 1% ~ (p=0.841 n=5+5)
XML 392k ± 0% 391k ± 0% ~ (p=0.095 n=5+5)
From c=1 to c=4, real time is down ~40%, CPU usage up 10-20%, alloc up ~5%:
name old time/op new time/op delta
Template 203ms ± 3% 131ms ± 5% -35.45% (p=0.000 n=50+50)
Unicode 87.2ms ± 4% 84.1ms ± 2% -3.61% (p=0.000 n=48+47)
GoTypes 560ms ± 4% 310ms ± 2% -44.65% (p=0.000 n=50+49)
Compiler 2.47s ± 3% 1.41s ± 2% -43.10% (p=0.000 n=50+46)
SSA 6.17s ± 2% 3.20s ± 2% -48.06% (p=0.000 n=49+49)
Flate 126ms ± 4% 74ms ± 2% -41.06% (p=0.000 n=49+48)
GoParser 148ms ± 4% 89ms ± 3% -39.97% (p=0.000 n=49+50)
Reflect 360ms ± 3% 242ms ± 3% -32.81% (p=0.000 n=49+49)
Tar 108ms ± 4% 73ms ± 4% -32.48% (p=0.000 n=50+49)
XML 203ms ± 3% 119ms ± 3% -41.56% (p=0.000 n=49+48)
name old user-time/op new user-time/op delta
Template 246ms ± 9% 287ms ± 9% +16.98% (p=0.000 n=50+50)
Unicode 109ms ± 4% 118ms ± 5% +7.56% (p=0.000 n=46+50)
GoTypes 735ms ± 4% 806ms ± 2% +9.62% (p=0.000 n=50+50)
Compiler 3.34s ± 4% 3.56s ± 2% +6.78% (p=0.000 n=49+49)
SSA 8.54s ± 3% 10.04s ± 3% +17.55% (p=0.000 n=50+50)
Flate 149ms ± 6% 176ms ± 3% +17.82% (p=0.000 n=50+48)
GoParser 181ms ± 5% 213ms ± 3% +17.47% (p=0.000 n=50+50)
Reflect 453ms ± 6% 499ms ± 2% +10.11% (p=0.000 n=50+48)
Tar 126ms ± 5% 149ms ±11% +18.76% (p=0.000 n=50+50)
XML 246ms ± 5% 287ms ± 4% +16.53% (p=0.000 n=49+50)
name old alloc/op new alloc/op delta
Template 38.8MB ± 0% 40.4MB ± 0% +4.21% (p=0.008 n=5+5)
Unicode 29.8MB ± 0% 30.9MB ± 0% +3.68% (p=0.008 n=5+5)
GoTypes 113MB ± 0% 116MB ± 0% +2.71% (p=0.008 n=5+5)
Compiler 443MB ± 0% 455MB ± 0% +2.75% (p=0.008 n=5+5)
SSA 1.25GB ± 0% 1.27GB ± 0% +1.84% (p=0.008 n=5+5)
Flate 25.3MB ± 0% 26.9MB ± 1% +6.31% (p=0.008 n=5+5)
GoParser 31.7MB ± 0% 33.2MB ± 0% +4.61% (p=0.008 n=5+5)
Reflect 78.2MB ± 0% 80.2MB ± 0% +2.53% (p=0.008 n=5+5)
Tar 26.6MB ± 0% 27.9MB ± 0% +5.19% (p=0.008 n=5+5)
XML 42.4MB ± 0% 44.6MB ± 0% +5.20% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
Template 380k ± 0% 379k ± 0% -0.39% (p=0.032 n=5+5)
Unicode 321k ± 0% 321k ± 0% ~ (p=0.841 n=5+5)
GoTypes 1.14M ± 0% 1.14M ± 0% ~ (p=0.421 n=5+5)
Compiler 4.12M ± 0% 4.14M ± 0% +0.52% (p=0.008 n=5+5)
SSA 9.72M ± 0% 9.76M ± 0% +0.37% (p=0.008 n=5+5)
Flate 234k ± 1% 234k ± 1% ~ (p=0.690 n=5+5)
GoParser 316k ± 0% 317k ± 1% ~ (p=0.841 n=5+5)
Reflect 981k ± 0% 981k ± 0% ~ (p=1.000 n=5+5)
Tar 250k ± 0% 249k ± 1% ~ (p=0.151 n=5+5)
XML 393k ± 0% 392k ± 0% ~ (p=0.056 n=5+5)
Going beyond c=4 on my machine tends to increase CPU time and allocs
without impacting real time.
The CPU time numbers matter, because when there are many concurrent
compilation processes, that will impact the overall throughput.
The numbers above are in many ways the best case scenario;
we can take full advantage of all cores.
Fortunately, the most common compilation scenario is incremental
re-compilation of a single package during a build/test cycle.
Updates #15756
Change-Id: I6725558ca2069edec0ac5b0d1683105a9fff6bea
Reviewed-on: https://go-review.googlesource.com/40693
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Robert Griesemer <gri@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
dumptypestructs did several different jobs.
Split them into separate functions
and call them in turn.
Hand dumptypestructs a list of dcls,
rather than reading the global.
Rename dumpptabs for (marginal) clarity.
This is groundwork for compiling autogenerated
functions concurrently.
Passes toolstash-check.
Change-Id: I627a1dffc70a7e4b7b4436ab19af1406267f01dc
Reviewed-on: https://go-review.googlesource.com/41501
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
The only remaining uses of duintxx
are in the implementation of duintNN.
I hope to inline those once I figure out why
CL 40864 is broken.
Note that some uses of duintxx with width Widthint
were converted into duintptr.
I did that, since #19954 is officially going to move forward.
Passes toolstash-check.
Change-Id: Id25253b711ea589d0199b51be9a3c18ca1af59ce
Reviewed-on: https://go-review.googlesource.com/41398
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
This is an automated refactoring to eliminate
all dxxx calls in gc/obj.go that accept types.Sym
instead of obj.LSym parameters.
The refactoring was of the form:
gorename -from '"cmd/compile/internal/gc".duintxx' -to Duintxx
gorename -from '"cmd/compile/internal/gc".duintxxLSym' -to DuintxxLSym
eg -t t.go -w cmd/compile/internal/gc
gofmt -r 'DuintxxLSym -> duintxxLSym' -w cmd/compile/internal/gc
where t.go looked like:
func before(s *types.Sym, off int, v uint64, wid int) int {
return gc.Duintxx(s, off, v, wid)
}
func after(s *types.Sym, off int, v uint64, wid int) int {
return gc.DuintxxLSym(s.Linksym(), off, v, wid)
}
The rename/gofmt shenanigans were to work around
limitations and bugs in eg and gorename.
The resulting code in reflect.go looks temporarily ugly,
but it makes refactoring and cleanup opportunities
much clearer.
Next step is to rename all the dxxx methods to rename the -LSym suffix
and clean up reflect.go.
The renaming is left for a separate CL to make the changes in
this CL more obvious, and thus hopefully easier to review.
Passes toolstash-check.
Change-Id: Ib31a2b6fd146ed03a855d20ecb0433f0f74e2f10
Reviewed-on: https://go-review.googlesource.com/41396
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
There were only two versions, 0 and 1,
and the only user of version 1 was the assembler,
to indicate that a symbol was static.
Rename LSym.Version to Static,
and add it to LSym.Attributes.
Simplify call-sites.
Passes toolstash-check.
Change-Id: Iabd39918f5019cce78f381d13f0481ae09f3871f
Reviewed-on: https://go-review.googlesource.com/41201
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
