This adds a debug check to throw immediately if any pointers are added
to the gcWork buffer after the mark completion barrier. The intent is
to catch the source of the cached GC work that occasionally produces
"P has cached GC work at end of mark termination" failures.
The result should be that we get "throwOnGCWork" throws instead of "P
has cached GC work at end of mark termination" throws, but with useful
stack traces.
This should be reverted before the release. I've been unable to
reproduce this issue locally, but this issue appears fairly regularly
on the builders, so the intent is to catch it on the builders.
This probably slows down the GC slightly.
For #27993.
Change-Id: I5035e14058ad313bfbd3d68c41ec05179147a85c
Reviewed-on: https://go-review.googlesource.com/c/149969
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
This CL adds CFGs to ssa.html.
It execs dot to generate SVG,
which then gets inlined into the html.
Some standard naming and javascript hacks
enable integration with the rest of ssa.html.
Clicking on blocks highlights the relevant
part of the CFG, and vice versa.
Sample output and screenshots can be seen in #28177.
CFGs can be turned on with the suffix mask:
:* - dump CFG for every phase
:lower - just the lower phase
:lower-layout - lower through layout
:w,x-y - phases w and x through y
Calling dot after every pass is noticeably slow,
instead use the range of phases.
Dead blocks are not displayed on CFG.
User can zoom and pan individual CFG
when the automatic adjustment has failed.
Dot-related errors are reported
without bringing down the process.
Fixes#28177
Change-Id: Id52c42d86c4559ca737288aa10561b67a119c63d
Reviewed-on: https://go-review.googlesource.com/c/142517
Run-TryBot: Yury Smolsky <yury@smolsky.by>
Reviewed-by: David Chase <drchase@google.com>
In race mode, these functions are defined and declared in
different packages, which therefore don't have implicit arg maps.
When they are defer'd, and the stack needs to move, the runtime
fails with missing stack maps. This CL adds arg maps (FUNCDATA)
to them.
Updates #28848
Change-Id: I0271563b7e78e7797ce2990c303dced957efaa86
Reviewed-on: https://go-review.googlesource.com/c/150457
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
The playground is not using GOOS=js, so it is not able to use the
package syscall/js. Examples that depend on syscall/js should not
show a "Run" button.
Fixes#28526.
Change-Id: I8b2fcdd0c0ee517a5c3864bf459f813129542389
Reviewed-on: https://go-review.googlesource.com/c/148918
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Return a PathError instead of an unwrapped syscall.EINVAL if the path
ends with dots.
As suggested by Roger Peppe in CL 150158.
Change-Id: I4d82a6ff64a979b67a843a1cc4fea58ed9326aed
Reviewed-on: https://go-review.googlesource.com/c/150160
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Prohibiting RemoveAll with paths that end in ".." was added with
CL 137442 in this release cycle, but it worked before and it should
continue to work.
Also run TestRemoveAllDot on all systems; the test is not specific to
the use of unlinkat and friends.
Change-Id: I277784c8915cd748fec318d2936062440d5d1fde
Reviewed-on: https://go-review.googlesource.com/c/150497
Reviewed-by: Tobias Klauser <tobias.klauser@gmail.com>
CL 150417 was submitted before I could recommend this change to remove
an unnecessary allocation.
Updates #28849
Change-Id: I4cd655f62bb3d00eda6c997f074785385bceee0c
Reviewed-on: https://go-review.googlesource.com/c/150498
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
The js/wasm architecture does not support signals at all, but there are
already some signal constants defined because of stdlib dependencies.
This change adds a dummy constant for syscall.SIGTERM as well, to make
js/wasm compatible with more existing Go code.
There is the Go proverb "Syscall must always be guarded with build
tags.", so code should not expect syscall.SIGTERM to exist. Still,
adding SIGTERM should do more good than harm.
Fixes#28719.
Change-Id: I3554b484f96a21427491c04eb1dd57e7af5bd62f
Reviewed-on: https://go-review.googlesource.com/c/150477
Run-TryBot: Richard Musiol <neelance@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
We should be able to build docker after this get applied.
Updates #13192
Change-Id: I5378d3518fac52d6bd4c97828884c1b382b7ace5
GitHub-Last-Rev: 210b7bc2e1
GitHub-Pull-Request: golang/go#28546
Reviewed-on: https://go-review.googlesource.com/c/146898
Reviewed-by: Jiang Ma <ma.jiang@zte.com.cn>
Reviewed-by: Clément Chigot <clement.chigot@atos.net>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
FormatMediaType used rune&0x80==0 to check if parameter values consisted
of valid ascii charaters. Comparing strings using their runes instead of
their bytes leads to some non-ascii strings to pass as valid.
E.g. the rune for 'Ą' is 0x104, 0x104 & 0x80 => 0. Its byte
representation is 0xc4 0x84, both of which result in non zero values
when masked with 0x80
Fixes#28849
Change-Id: Ib9fb4968bcbbec0197d81136f380d40a2a56c14b
Reviewed-on: https://go-review.googlesource.com/c/150417
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
ValueOf() panics if x is not one of the expected types.
Change-Id: I1105e46bd09a5ab13c162b77c1c50cc45bce27a2
GitHub-Last-Rev: 34a88ce820
GitHub-Pull-Request: golang/go#28846
Reviewed-on: https://go-review.googlesource.com/c/150138
Reviewed-by: Richard Musiol <neelance@gmail.com>
This commit performs replace double quote to backquote,
so now all examples looks consistent.
Change-Id: I8cf760ce1bdeff9619a88e531161b9516385241b
GitHub-Last-Rev: e3e636cebb
GitHub-Pull-Request: golang/go#28879
Reviewed-on: https://go-review.googlesource.com/c/150397
Reviewed-by: Rob Pike <r@golang.org>
Run-TryBot: Rob Pike <r@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Change-Id: I7a1046f5e0aedbbdd1106a616de410fe4e0cb7d8
Reviewed-on: https://go-review.googlesource.com/c/92295
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
This commit adds examples for Match, Find,
FindAllSubmatch, FindSubmatch and Match functions.
Change-Id: I2bdf8c3cee6e89d618109397378c1fc91aaf1dfb
GitHub-Last-Rev: 33f34b7adc
GitHub-Pull-Request: golang/go#28837
Reviewed-on: https://go-review.googlesource.com/c/150020
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Change-Id: I70afd2f7b6783926174c4e66565b711cffeb97c5
Reviewed-on: https://go-review.googlesource.com/c/150141
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
cmd/vet, now simplified to a single function call is now authoritative,
not a copy of vet-lite.
The update-xtools.sh script now uses the imports of cmd/vet as the
roots for vendoring.
Change-Id: I4faef3fcf3db10b3a3930726e8d0720a3c8395da
Reviewed-on: https://go-review.googlesource.com/c/150297
Run-TryBot: Alan Donovan <adonovan@google.com>
Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
- restore and rework cmd/vet/doc.go, which was clobbered during the vet-lite switch.
- document go vet -vettool=prog flag and how to run an alternative checker.
- make 'go vet -help' show how to list vet tool's flags. Example:
$ go vet -help
usage: go vet [-n] [-x] [-vettool prog] [build flags] [vet flags] [packages]
Run 'go help vet' for details.
Run 'go tool vet help' for the vet tool's flags.
$ go vet -vettool=~/bin/myvet -help
usage: go vet [-n] [-x] [-vettool prog] [build flags] [vet flags] [packages]
Run 'go help vet' for details.
Run '~/bin/myvet help' for the vet tool's flags.
Updates #28840
Change-Id: Ieb79dfe29e1df074f865bc9a9d47b44199675d7d
Reviewed-on: https://go-review.googlesource.com/c/147018
Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Mainly to pull the fix for the regression in #28792.
Change-Id: If71ae783fd9a9e3935186b49fdf501ba098235a2
Reviewed-on: https://go-review.googlesource.com/c/150161
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Alan Donovan <adonovan@google.com>
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>
CL 146020 changed the behavior of RemoveAll("") on unix systems using
the *at functions to return syscall.EINVAL instead of nil. Adjust the
*at implementation to retain this behavior as is the case on the *noat
systems.
Additionally, also make sure RemoveAll("") on systems not using the "at
functions (e.g. nacl and js/wasm) follow the same behavior (which wasn't
the case previously).
Fixes#28830
Change-Id: I8383c1423fefe871d18ff49134a1d23077ec6867
Reviewed-on: https://go-review.googlesource.com/c/150158
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: roger peppe <rogpeppe@gmail.com>
Procedure names should reflect what they do; function names
should reflect what they return. Functions are used in
expressions, often in things like if's, so they need
to read appropriately.
if CheckHMAC(a, b, key)
is unhelpful because we can't deduce whether CheckHMAC
returns true on error or non-error; instead
if ValidHMAC(a, b, key)
makes the point clear and makes a future mistake
in using the routine less likely.
https://www.lysator.liu.se/c/pikestyle.html
Change-Id: I7c4b1981c90c8d7475ddd8ec18dee3db2e0f42df
GitHub-Last-Rev: 32199a418b
GitHub-Pull-Request: golang/go#28823
Reviewed-on: https://go-review.googlesource.com/c/149857
Reviewed-by: Filippo Valsorda <filippo@golang.org>
CL 138675 added a call to runtime.save_g which uses thread local
storage to store g. On iOS however, that storage was not initialized
yet. Move the call to below _cgo_init where it is set up.
Change-Id: I14538d3e7d56ff35a6fa02c47bca306d24c38010
Reviewed-on: https://go-review.googlesource.com/c/150157
Run-TryBot: Elias Naur <elias.naur@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
Currently, both asm and compile have a -symabis flag, but in asm it's
a boolean flag that means to generate a symbol ABIs file and in the
compiler its a string flag giving the path of the symbol ABIs file to
consume. I'm worried about this false symmetry biting us in the
future, so rename asm's flag to -gensymabis.
Updates #27539.
Change-Id: I8b9c18a852d2838099718f8989813f19d82e7434
Reviewed-on: https://go-review.googlesource.com/c/149818
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Now that the build system has been updated to install x/tools in
$GOPATH (CL 149658), depend on it being there and don't ignore
failures to build the tool.
Update to CL 149097.
Change-Id: I72fde347217533697068b6a6773696cc2b83e9ed
Reviewed-on: https://go-review.googlesource.com/c/150017
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Alan Donovan <adonovan@google.com>
- Adds 'go vet -json' and 'go vet -c=1' flags
- Removes the pkgfact analyzer, included by mistake.
Change-Id: Id3f1879af479efc567ea0508a1de7a37db5bee89
Reviewed-on: https://go-review.googlesource.com/c/149961
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Go get in mod-enabled packages lets you do go get "pkg@<hash>" or "pkg@<branch>".
Go internally will switch the hash or branch into a pseudo version.
Go mod download should do the same. The bug lay in the fact that the disk cache
was not being written when Go converted the hash/branch into a pseudo version.
Fixes#27947
Change-Id: I94c29a5c95f69ab18a9cd7a2ecade128047c5e36
GitHub-Last-Rev: 668634b3e7
GitHub-Pull-Request: golang/go#28042
Reviewed-on: https://go-review.googlesource.com/c/140257
Reviewed-by: Bryan C. Mills <bcmills@google.com>
Run-TryBot: Bryan C. Mills <bcmills@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
This field, which matches the IDs used by go list, will enable all vet
drivers to produce JSON output in a consistent format (a map from
package ID to analysis name to result).
Change-Id: Icac703b944de55df42c996dc2f672005014ad57a
Reviewed-on: https://go-review.googlesource.com/c/149960
Reviewed-by: Russ Cox <rsc@golang.org>
The standard build assumes the variable is unset.
Make it so, like we do for GOFLAGS, GOBIN, and so on.
Change-Id: I4ad5695f8021b08bd1a35dd99112970a813d247c
Reviewed-on: https://go-review.googlesource.com/c/149959
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
The reason the 386 trybot was happy but 'GOARCH=386 go test cmd/vet'
was not is that CgoEnabled defaults to false in a cross build;
I have no idea why. Now we ask the go command for the effective
value so that the test works in both cases.
Also, remove stale comment.
Fixes#28829
Change-Id: I1210af34da6986f47924059de5c1f08b2824ace9
Reviewed-on: https://go-review.googlesource.com/c/149958
Run-TryBot: Alan Donovan <adonovan@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
In general we don't assume that the go command knows the
specific version of the compiler being used, including which
experiments the compiler was built with. Let the compiler tell us,
instead of importing cmd/internal/objabi from cmd/go.
Replacement for CL 128735.
Change-Id: Iaa07f46e19764d0fb14a1c89979bea7bb7139b9c
Reviewed-on: https://go-review.googlesource.com/c/149338
Reviewed-by: Bryan C. Mills <bcmills@google.com>
A bug in the old code was indirectly causing a confusing print,
but CL 131635 fixed the print instead of the surrounding code.
Fix the surrounding code, restore the old print, and test that the
error is actually reported (it was being ignored in a direct go get
but displaying in go build).
Change-Id: I03c21380fce481060c443b0cc820f3617497fdd9
Reviewed-on: https://go-review.googlesource.com/c/149317
Reviewed-by: Bryan C. Mills <bcmills@google.com>
The compiler's Format test verifies that the correct format
strings for the given arguments are used in the compiler
sources. The format strings are fairly specialized which is
why we cannot use go vet; and the mapping is based on a
hard-wired map.
In the past, if that map got out of sync with the compiler
sources, it was necessary to manually update the map. This
change introduces an update mechanism which simply requires
the test to be run with the -u flag.
(Formerly, the -u flag was used to automatically rewrite
format strings; now we use -r for that.)
Change-Id: I9259566a6120a13cf34b143875975ada62697890
Reviewed-on: https://go-review.googlesource.com/c/149460
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Show two larger application examples. One example that
could be used in a CLI, the other in a long running
service. These demonstarates different strategies for
handling DB.Ping errors in context.
Fixes#23738
Change-Id: Id01213caf1f47917239a7506b01d30e37db74d31
Reviewed-on: https://go-review.googlesource.com/c/101216
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
The package doc for the testing package doesn't have a simple
example demonstrating how to write a test with an expectation. The doc
has simple examples for benchmarks, examples, and skipping, and it would be
useful for people new to writing tests in Go.
Also moved the skip example further down because it references tests and
benchmarks but benchmarks haven't been discussed in detail until the
next section. Skip is also a less used feature and it seems misplaced to
sit so high up in the package documentation. As an example, Skip is used
570 times the Go code repository which is significantly less than Error
and Fatal that are used 23,303 times.
Also changed 'sample' to 'simple' in other places in the package documentation
to keep the language used consistent when describing the small examples.
Fixes#27839
Change-Id: Ie01a3751986ee61adf2a2f2eda59cc182342baa7
GitHub-Last-Rev: 7357bfdcd2
GitHub-Pull-Request: golang/go#27840
Reviewed-on: https://go-review.googlesource.com/c/137175
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Emmanuel Odeke <emm.odeke@gmail.com>
Select linux/arm64 for the asm test.
Disable the cgo test for now.
Will fix properly in a follow-up.
Filed Issue 28829 to track it.
Updates #28829
Change-Id: Ic05f619700b06e91c43f8c150b089b8e77d92c85
Reviewed-on: https://go-review.googlesource.com/c/149937
Run-TryBot: Alan Donovan <adonovan@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
On reflect documentation page only this function
doesn't have description, this commit add simple description.
Change-Id: Idcda89ddd1f6fdd1938c4030e89ebdc186255ce6
GitHub-Last-Rev: 1553b834bb
GitHub-Pull-Request: golang/go#28818
Reviewed-on: https://go-review.googlesource.com/c/149721
Reviewed-by: Ian Lance Taylor <iant@golang.org>
CL 129682 removed go mod fix but unfortunately
we hadn't updated the source code hence running
go mod -fix
would suggest
go mod fix
which is a nonexistent command.
This change fixes that to instead suggest
go mod tidy
Change-Id: Ie0d7c90805034e9fe6df24afaa15340c44d4f426
GitHub-Last-Rev: 5ae1340954
GitHub-Pull-Request: golang/go#28402
Reviewed-on: https://go-review.googlesource.com/c/144838
Reviewed-by: Emmanuel Odeke <emm.odeke@gmail.com>
Run-TryBot: Emmanuel Odeke <emm.odeke@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
This change deletes the legacy implementation of vet, replacing it
with a short main.go that merely selects the desired analyzers and
calls into the "unitchecker" implementation vendored from
golang.org/x/tools/go/analysis.
Unlike the full vet checker (x/tools/go/analysis/cmd/vet), the 'lite'
unitchecker cannot also be run standalone (as 'go tool vet' or
cmd/vet); it must be invoked by 'go vet'.
This design was chosen to avoid vendoring many
additional dependencies into GOROOT, in particular go/packages. If
go/packages should someday become part of the standard library, there
will be considerable opportunity for simplification.
This change also patches the vendored analysisflag package
(by adding patch.go) so that it fully supports the build
system's -V flag protocol.
Also:
- remove stale internal/unitchecker/ tree
(belonged in https://go-review.googlesource.com/c/149778).
- move vet legacy flags (-all, -v, -source, -tags) into analysisflags
as all drivers will need them, not just unitchecker.
I will upstream this change.
A sampling of tests from the cmd/vet testsuite have been preserved as
a smoke test, to ensure that each analyzer is being run, and for
convenience when evaluating changes. Comprehensive tests for each
analyzer live upstream in x/tools. The tests have been heavily reduced
and reorganized so that they conform to the structure required by 'go
vet'.
Change-Id: I84b38caeef733e65deb95234b3b87b5f61046def
Reviewed-on: https://go-review.googlesource.com/c/149609
Reviewed-by: Russ Cox <rsc@golang.org>
The old whole-page reclaimer was the only thing that used the busy
span lists. Remove them so nothing uses them any more.
Change-Id: I4007dd2be08b9ef41bfdb0c387215c73c392cc4c
Reviewed-on: https://go-review.googlesource.com/c/138960
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Rick Hudson <rlh@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
When we attempt to allocate an N page span (either for a large
allocation or when an mcentral runs dry), we first try to sweep spans
to release N pages. Currently, this can be extremely expensive:
sweeping a span to emptiness is the hardest thing to ask for and the
sweeper generally doesn't know where to even look for potentially
fruitful results. Since this is on the critical path of many
allocations, this is unfortunate.
This CL changes how we reclaim empty spans. Instead of trying lots of
spans and hoping for the best, it uses the newly introduced span marks
to efficiently find empty spans. The span marks (and in-use bits) are
in a dense bitmap, so these spans can be found with an efficient
sequential memory scan. This approach can scan for unmarked spans at
about 300 GB/ms and can free unmarked spans at about 32 MB/ms. We
could probably significantly improve the rate at which is can free
unmarked spans, but that's a separate issue.
Like the current reclaimer, this is still linear in the number of
spans that are swept, but the constant factor is now so vanishingly
small that it doesn't matter.
The benchmark in #18155 demonstrates both significant page reclaiming
delays, and object reclaiming delays. With "-retain-count=20000000
-preallocate=true -loop-count=3", the benchmark demonstrates several
page reclaiming delays on the order of 40ms. After this change, the
page reclaims are insignificant. The longest sweeps are still ~150ms,
but are object reclaiming delays. We'll address those in the next
several CLs.
Updates #18155.
Fixes#21378 by completely replacing the logic that had that bug.
Change-Id: Iad80eec11d7fc262d02c8f0761ac6998425c4064
Reviewed-on: https://go-review.googlesource.com/c/138959
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Rick Hudson <rlh@golang.org>
