]> Cypherpunks repositories - gostls13.git/log
gostls13.git
3 years agotest: add test of select inside generic function
Keith Randall [Fri, 3 Dec 2021 21:57:25 +0000 (13:57 -0800)]
test: add test of select inside generic function

Make sure that we can import/export selects for generics.

Change-Id: Ibf36e98fc574ce9275820aa426b3e6703b0aae6d
Reviewed-on: https://go-review.googlesource.com/c/go/+/369101
Trust: Keith Randall <khr@golang.org>
Trust: Dan Scales <danscales@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dan Scales <danscales@google.com>
3 years agodoc/go1.18: change github.com/golang/go/issues to golang.org/issue
Hossein Zolfi [Fri, 3 Dec 2021 19:21:11 +0000 (22:51 +0330)]
doc/go1.18: change github.com/golang/go/issues to golang.org/issue

Updates #47694

Change-Id: I84547ff995e2a302cf229fefefd28a7220a17028
Reviewed-on: https://go-review.googlesource.com/c/go/+/369016
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Trust: Austin Clements <austin@google.com>

3 years agocmd/compile/abi-internal: mention SSE restriction on Plan 9
Austin Clements [Fri, 3 Dec 2021 18:40:10 +0000 (13:40 -0500)]
cmd/compile/abi-internal: mention SSE restriction on Plan 9

Change-Id: I2be08b88b5147cf37ac55b7472d63503739c9f05
Reviewed-on: https://go-review.googlesource.com/c/go/+/369156
Trust: Austin Clements <austin@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
3 years agodoc/go1.18: catch up with runtime/compiler commits and API changes
Austin Clements [Fri, 3 Dec 2021 18:39:42 +0000 (13:39 -0500)]
doc/go1.18: catch up with runtime/compiler commits and API changes

For #47694.

Change-Id: I21b1af1807d4da2fb2f4d9b961d44a21d715d7d2
Reviewed-on: https://go-review.googlesource.com/c/go/+/369155
Trust: Austin Clements <austin@google.com>
Trust: Jeremy Faller <jeremy@golang.org>
Reviewed-by: Jeremy Faller <jeremy@golang.org>
3 years ago.github: use multiple issue templates
Sean Liao [Tue, 23 Nov 2021 19:52:07 +0000 (20:52 +0100)]
.github: use multiple issue templates

There are currently multiple issue templates floating around for
different projects, these can sometimes be hard to find.

Fixes #29839

Change-Id: I6600b6f78842736d81d35e6a64247d00706d9e0e
Reviewed-on: https://go-review.googlesource.com/c/go/+/366736
Trust: Russ Cox <rsc@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Run-TryBot: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

3 years agodoc/go1.18: mention fuzzing in the release notes
Katie Hockman [Fri, 3 Dec 2021 17:23:44 +0000 (12:23 -0500)]
doc/go1.18: mention fuzzing in the release notes

Also make a few small formatting fixes.

Change-Id: Iad99d030312393af3b6533f2cd00f09aea0f2a7d
Reviewed-on: https://go-review.googlesource.com/c/go/+/369074
Trust: Katie Hockman <katie@golang.org>
Run-TryBot: Katie Hockman <katie@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>
3 years agodoc/go1.18: reorganize runtime/compiler a little
Austin Clements [Fri, 3 Dec 2021 14:04:26 +0000 (09:04 -0500)]
doc/go1.18: reorganize runtime/compiler a little

For #47694.

Change-Id: Iedf85f522f7c79ae0a61b4fc1f2f092cf7b613df
Reviewed-on: https://go-review.googlesource.com/c/go/+/368696
Trust: Austin Clements <austin@google.com>
Reviewed-by: Jeremy Faller <jeremy@golang.org>
3 years agotest: re-enable a bunch of tests with types2
Dan Scales [Sat, 27 Nov 2021 18:33:59 +0000 (10:33 -0800)]
test: re-enable a bunch of tests with types2

Enable a bunch of types2-related error tests to run successfully, so
they no longer have to be disabled in run.go.

 - directive.go: split it into directive.go and directive2.go, since the
   possible errors are now split across the parser and noder2, so they
   can't all be reported in one file.

 - linkname2.go: similarly, split it into linkname2.go and linkname3.go
   for the same reason.

 - issue16428.go, issue17645.go, issue47201.dir/bo.go: handle slightly
   different wording by types2

 - issue5609.go: handle slight different error (array length must be
   integer vs. array bound too large).

 - float_lit3.go: handle slightly different wording (overflows
   float vs cannot convert to float)

I purposely didn't try to fix tests yet where there are extra or missing
errors on different lines, since that is not easy to make work for both
-G=3 and -G=0. In a later change, will flip to make the types2 version
match correctly, vs. the -G=0 version.

Change-Id: I6079ff258e3b90146335b9995764e3b1b56cda59
Reviewed-on: https://go-review.googlesource.com/c/go/+/368455
Trust: Dan Scales <danscales@google.com>
Run-TryBot: Dan Scales <danscales@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
3 years agoruntime: fix missing pprof labels
Felix Geisendörfer [Mon, 20 Sep 2021 14:09:47 +0000 (16:09 +0200)]
runtime: fix missing pprof labels

Use gp.m.curg instead of the gp when recording cpu profiler stack
traces. This ensures profiler labels are captured when systemstack or similar
is executing on behalf of the current goroutine.

After this there are still rare cases of samples containing the labelHog
function, so more work might be needed. This patch should fix ~99% of the
problem.

Also change testCPUProfile interface a little to allow the new test to
re-run with a longer duration if it fails during a -short run.

Fixes #48577.

Change-Id: I3dbc9fd5af3c513544e822acaa43055b2e00dfa9
Reviewed-on: https://go-review.googlesource.com/c/go/+/367200
Trust: Michael Pratt <mpratt@google.com>
Trust: Bryan Mills <bcmills@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
3 years agoos: test that LookupEnv reports all keys found in Environ
Bryan C. Mills [Tue, 30 Nov 2021 21:33:51 +0000 (16:33 -0500)]
os: test that LookupEnv reports all keys found in Environ

For #49886

Change-Id: Ie3a7f12a0d30ec719caf375e7be30cc4a5796c3f
Reviewed-on: https://go-review.googlesource.com/c/go/+/367850
Trust: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

3 years agobuiltin: document "any" and "comparable"
Ian Lance Taylor [Thu, 2 Dec 2021 22:52:31 +0000 (14:52 -0800)]
builtin: document "any" and "comparable"

Fixes #49927

Change-Id: I8b34cf13b3bc6338309f005648ca3ee6852927f0
Reviewed-on: https://go-review.googlesource.com/c/go/+/368954
Trust: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

3 years agocmd/compile: set PPC64's MAXWIDTH as other architectures
Cherry Mui [Fri, 3 Dec 2021 00:05:27 +0000 (19:05 -0500)]
cmd/compile: set PPC64's MAXWIDTH as other architectures

PPC64's MAXWIDTH is set as 1<<60 whereas on other 64-bit
architetures it is set as 1<<50. Set to 1<<50 for consistency. The
toolchain cannot handle such large program anyway.

May fix PPC64 build.

Change-Id: Ic3972a089b2f14a96e4ded57ef218d763c924a6a
Reviewed-on: https://go-review.googlesource.com/c/go/+/368955
Trust: Cherry Mui <cherryyz@google.com>
Trust: Dan Scales <danscales@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dan Scales <danscales@google.com>
3 years agoruntime/race: rebuild darwin syso to work around macOS 12 malloc reserved address
Cherry Mui [Tue, 30 Nov 2021 23:58:53 +0000 (18:58 -0500)]
runtime/race: rebuild darwin syso to work around macOS 12 malloc reserved address

On macOS 12 a new malloc implementation (nano) is used by default,
and apparently it reserves address range
0x600000000000-0x600020000000, which conflicts with the address
range that TSAN uses for Go. Work around the issue by changing the
address range slightly.

The actual change is made on LLVM at https://reviews.llvm.org/D114825 .
This CL includes syso's built with the patch applied.

Fixes #49138.

Change-Id: I7b367d6e042b0db39a691c71601c98e4f8728a70
Reviewed-on: https://go-review.googlesource.com/c/go/+/367916
Trust: Cherry Mui <cherryyz@google.com>
Reviewed-by: Austin Clements <austin@google.com>
3 years agodoc/go1.18: mention new -asan option
Ian Lance Taylor [Thu, 2 Dec 2021 19:39:38 +0000 (11:39 -0800)]
doc/go1.18: mention new -asan option

For #44853
For #47694

Change-Id: Ia76246218b1361d8bdf510bbfc5178c83cdd3eec
Reviewed-on: https://go-review.googlesource.com/c/go/+/368834
Trust: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Jeremy Faller <jeremy@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
3 years agocrypto/x509: fix comments on certDirectories
Shang Jian Ding [Wed, 6 Oct 2021 00:52:20 +0000 (19:52 -0500)]
crypto/x509: fix comments on certDirectories

CL 205237 allowed SSL_CERT_DIR to be a colon delimited list of
directories. In the case that SSL_CERT_DIR is unset, the change
also made certDirectories to all be loaded rather than stopping
after successfully reading at least one file from a directory.
This update fixes code comments on the certDirectories package
level variables to reflect current behavior.

Fixes #48808

Change-Id: Id92f875545272fc6205d9955d03ea7bf844f15eb
Reviewed-on: https://go-review.googlesource.com/c/go/+/354140
Reviewed-by: Emmanuel Odeke <emmanuel@orijtech.com>
Reviewed-by: Katie Hockman <katie@golang.org>
Run-TryBot: Emmanuel Odeke <emmanuel@orijtech.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Trust: Katie Hockman <katie@golang.org>

3 years agodoc/go1.18: remove some TODOs for changes we aren't mentioning
Jeremy Faller [Thu, 2 Dec 2021 18:18:33 +0000 (13:18 -0500)]
doc/go1.18: remove some TODOs for changes we aren't mentioning

Updates #47694

Change-Id: I8fdbded4aaf974425ee317a1cdd940a5ac88c0b3
Reviewed-on: https://go-review.googlesource.com/c/go/+/368794
Trust: Jeremy Faller <jeremy@golang.org>
Run-TryBot: Jeremy Faller <jeremy@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agodoc/go1.18: add docs on FieldByIndexErr
Jeremy Faller [Wed, 1 Dec 2021 22:03:15 +0000 (17:03 -0500)]
doc/go1.18: add docs on FieldByIndexErr

From:
https://go-review.googlesource.com/c/go/+/357962/

Updates #47694

Change-Id: I2d43dcbd28c03457b55eada26e87f7710a113b0c
Reviewed-on: https://go-review.googlesource.com/c/go/+/368364
Trust: Jeremy Faller <jeremy@golang.org>
Run-TryBot: Jeremy Faller <jeremy@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agodoc/go1.18: add docs on new reflect.Value methods
Jeremy Faller [Wed, 1 Dec 2021 21:55:37 +0000 (16:55 -0500)]
doc/go1.18: add docs on new reflect.Value methods

From:

https://go-review.googlesource.com/c/go/+/352131/

Updates #47694

Change-Id: I7c7811c49900049f5ef7fc906fe25d2ffd73c7b9
Reviewed-on: https://go-review.googlesource.com/c/go/+/368363
Trust: Jeremy Faller <jeremy@golang.org>
Run-TryBot: Jeremy Faller <jeremy@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Fabio Falzoi <fabio.falzoi84@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

3 years agodoc/go1.18: add docs on MapIter.Reset
Jeremy Faller [Wed, 1 Dec 2021 21:43:47 +0000 (16:43 -0500)]
doc/go1.18: add docs on MapIter.Reset

From:

https://go-review.googlesource.com/c/go/+/321891

Updates #47694

Change-Id: I2cdd5d9f3da62a5690cd2ef921ed48957c602d64
Reviewed-on: https://go-review.googlesource.com/c/go/+/368360
Trust: Jeremy Faller <jeremy@golang.org>
Run-TryBot: Jeremy Faller <jeremy@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agodoc/go1.18: add docs on User.GroupIds
Jeremy Faller [Wed, 1 Dec 2021 21:33:33 +0000 (16:33 -0500)]
doc/go1.18: add docs on User.GroupIds

Cobbled up some release notes from:

https://go-review.googlesource.com/c/go/+/330753/

Updates #47694

Change-Id: I249e4b96b7d51185b07c11a734505677e9e36315
Reviewed-on: https://go-review.googlesource.com/c/go/+/368358
Trust: Jeremy Faller <jeremy@golang.org>
Run-TryBot: Jeremy Faller <jeremy@golang.org>
Reviewed-by: Kirill Kolyshkin <kolyshkin@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

3 years agodoc/go1.18: add docs on Cookie.Valid
Jeremy Faller [Wed, 1 Dec 2021 21:27:55 +0000 (16:27 -0500)]
doc/go1.18: add docs on Cookie.Valid

Updates #47694

Change-Id: Ibf3e25fbccf82e4abd7bd76b2ea6ceb4b05d4664
Reviewed-on: https://go-review.googlesource.com/c/go/+/368357
Trust: Jeremy Faller <jeremy@golang.org>
Run-TryBot: Jeremy Faller <jeremy@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agodoc/go1.18: add deprecation note for net.Error.Temporary
Jeremy Faller [Wed, 1 Dec 2021 18:56:53 +0000 (13:56 -0500)]
doc/go1.18: add deprecation note for net.Error.Temporary

Updates #47694

Change-Id: Ia25ad49f688efa3d60d83290095648711704b478
Reviewed-on: https://go-review.googlesource.com/c/go/+/368316
Trust: Jeremy Faller <jeremy@golang.org>
Run-TryBot: Jeremy Faller <jeremy@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agodoc/go1.18: add Conn.NetConn documentation to crypto/tls section
Jeremy Faller [Wed, 1 Dec 2021 18:48:22 +0000 (13:48 -0500)]
doc/go1.18: add Conn.NetConn documentation to crypto/tls section

Updates #47694

Change-Id: Iead44baa8b2a06ecf7b92d343ed9117f0fc0793e
Reviewed-on: https://go-review.googlesource.com/c/go/+/368315
Trust: Jeremy Faller <jeremy@golang.org>
Trust: Katie Hockman <katie@golang.org>
Run-TryBot: Jeremy Faller <jeremy@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Katie Hockman <katie@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Agniva De Sarker <agniva.quicksilver@gmail.com>
3 years agodoc/go1.18: add constraints package documentation
Jeremy Faller [Wed, 1 Dec 2021 18:41:10 +0000 (13:41 -0500)]
doc/go1.18: add constraints package documentation

Updates #47694

Change-Id: I3239023dad194c317e271e6093eff2cfbed1a4b9
Reviewed-on: https://go-review.googlesource.com/c/go/+/368314
Trust: Jeremy Faller <jeremy@golang.org>
Run-TryBot: Jeremy Faller <jeremy@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agodoc/go1.18: add information on new pacer.
Jeremy Faller [Wed, 1 Dec 2021 17:44:04 +0000 (12:44 -0500)]
doc/go1.18: add information on new pacer.

This is based off Michael's notes.

Updates #47694

Change-Id: I6e7944f85b776e8481829a2fafd177a49557c6ca
Reviewed-on: https://go-review.googlesource.com/c/go/+/368156
Trust: Jeremy Faller <jeremy@golang.org>
Run-TryBot: Jeremy Faller <jeremy@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
3 years agocmd/compile: report channel size errors correctly for -G=3
Dan Scales [Mon, 29 Nov 2021 19:45:17 +0000 (11:45 -0800)]
cmd/compile: report channel size errors correctly for -G=3

First, we need to set base.Pos in varDecl() and typeDecl(), so it will
be correct if we need to report type size errors while converting types.
Changed error calls in types/sizes.go to use Errorf, not ErrorfAt, since
we want to use base.Pos (which will set from t.Pos(), if that is
available).

Second, we need to add an extra call CalcSize(t1.Elem()) in the
TCHANARGS case of CalcSize(). We can use CalcSize() rather than
CheckSize(), since we know the top-level recursive type will have been
calculated by the time we process the fake TCHANARGS type. In -G=0 mode,
the size of the channel element has often been calculated because of
some other processing (but not in the case of #49767). But in -G=3 mode,
we just calculate sizes during the single noder2 pass, so we are more
likely to have not gotten to calculating the size of the element yet,
depending on the order of processing of the deferredTypeStack.

Fixes the tests fixedbugs/issue{42058a,42058b}.go that were
disabled for -G=3 mode.

Had to add exceptions in stdlib_test.go for go/types and types2, because
the types2 typechecker does not know about type size limits.

Fixes #49814
Fixes #49771
Updates #49767

Change-Id: I77d058e8ceff68a58c4c386a8cf46799c54b04c3
Reviewed-on: https://go-review.googlesource.com/c/go/+/367955
Trust: Dan Scales <danscales@google.com>
Run-TryBot: Dan Scales <danscales@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
3 years agodoc/go1.18: add short-circuit evaluation for text/template
Jeremy Faller [Fri, 19 Nov 2021 21:51:47 +0000 (16:51 -0500)]
doc/go1.18: add short-circuit evaluation for text/template

This description is based on https://golang.org/cl/321490.

Updates #47694

Change-Id: I48656cd487d2fccf0b0d3390f350f1bc6f2b0080
Reviewed-on: https://go-review.googlesource.com/c/go/+/365738
Trust: Jeremy Faller <jeremy@golang.org>
Trust: Dmitri Shuralyov <dmitshur@golang.org>
Run-TryBot: Jeremy Faller <jeremy@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
3 years agoall: update vendored golang.org/x/tools
Robert Findley [Thu, 2 Dec 2021 17:36:28 +0000 (12:36 -0500)]
all: update vendored golang.org/x/tools

Update the vendored x/tools to pick up CL 364678, which updates vet
analyzers following a change to the underlying of type parameters.

This also pulls in significant changes to the typeutil package to
support new constructs in typeutil.Map, but this is not used by vet.

The following commands were used:

  go get -d golang.org/x/tools@e212aff8fd146c44ddb0167c1dfbd5531d6c9213
  go mod tidy
  go mod vendor

Fixes #49855

Change-Id: I3ffc59f3693710c83b81d390999aeabc8043723b
Reviewed-on: https://go-review.googlesource.com/c/go/+/368774
Trust: Robert Findley <rfindley@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>
Run-TryBot: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

3 years agoruntime: break out new minimum heap size into a goexperiment
Michael Anthony Knyszek [Wed, 1 Dec 2021 15:26:31 +0000 (15:26 +0000)]
runtime: break out new minimum heap size into a goexperiment

The new minimum heap of 512 KiB has been the cause of some build
slowdown (~1%) and microbenchmark slowdown (usually ~0%, up to ~50%)
because of two reasons:
1. Applications with lots of small short-lived processes execute many
   more GC cycles.
2. Applications with heaps <4 MiB GC up to 8x more often.

In many ways these consequences are inevitable given how GOGC works,
however we need to investigate more as to whether the apparent slowdowns
are indeed unavoidable or if the GC has issues scaling down, which it's
too late for for this release.

Given that this release is already huge, it's OK to push this back.
We'll take a closer look at it next cycle, so place block it behind a
new goexperiment to allow users and ourselves to easily experiment with
it.

Fixes #49744.
Updates #44167.

Change-Id: Ibad51f7873de7517490c89802f3c593834e77ff0
Reviewed-on: https://go-review.googlesource.com/c/go/+/368137
Trust: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: David Chase <drchase@google.com>
3 years agomisc/ios: bump min version
Changkun Ou [Thu, 2 Dec 2021 15:01:09 +0000 (16:01 +0100)]
misc/ios: bump min version

For #48076

Change-Id: I5edaa43af82c3478555c309a001a3d1b16de3d64
Reviewed-on: https://go-review.googlesource.com/c/go/+/368615
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

3 years agoruntime: print errno on clock_gettime failure on OpenBSD
Austin Clements [Wed, 1 Dec 2021 18:41:45 +0000 (13:41 -0500)]
runtime: print errno on clock_gettime failure on OpenBSD

For #49532.

Change-Id: I5afc64c987f0519903128550a7dac3a0f5e592cf
Reviewed-on: https://go-review.googlesource.com/c/go/+/368334
Trust: Austin Clements <austin@google.com>
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Joel Sing <joel@sing.id.au>
Reviewed-by: Cherry Mui <cherryyz@google.com>
3 years agocmd/compile: fix case where g.curDecl should be saved/restored
Dan Scales [Wed, 1 Dec 2021 18:53:17 +0000 (10:53 -0800)]
cmd/compile: fix case where g.curDecl should be saved/restored

When we set g.curDecl for the type params created during fillinMethods
for an instantiated type, we need to save/restore its value, because
fillinMethods() may be called while processing a typeDecl. We want the
value of g.curDecl to continue to be correct for type params created in
the typeDecl. Because of ordering issues, not restoring g.curDecl
happens to cause problems (which don't always show up visibly) exactly
when a type param is not actually used in a type declaration.

Cleared g.curDecl to "" at the later points in typeDecl() and
funcDecl(). This allows adding asserts that g.curDecl is always empty
("") when we set it in typeDecl() and funcDecl(), and always non-empty
when we use it in typ0().

Fixes #49893

Change-Id: Ic2fb1df791585bd257f2b86ffaae0453c31705c5
Reviewed-on: https://go-review.googlesource.com/c/go/+/368454
Trust: Dan Scales <danscales@google.com>
Run-TryBot: Dan Scales <danscales@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
3 years agonet: remove arbitrary deadlines in PacketConn tests
Bryan C. Mills [Wed, 1 Dec 2021 16:04:39 +0000 (11:04 -0500)]
net: remove arbitrary deadlines in PacketConn tests

When TestPacketConn was added (in CL 6501057) it included arbitrary
100ms deadlines. Those deadlines were arbitrarily increased to 500ms
in CL 4922.

If the test is actually provoking a deadlock, allowing it to deadlock
will give us a more useful goroutine dump. Otherwise, the deadlines
don't seem all that useful — they appear to increase code coverage,
but have no effect on the test in the typical case, and can only
cause flakes on particularly-slow machines.

For #43627

Change-Id: I83de5217c54c743b83adddf51d4f6f2bd5b91732
Reviewed-on: https://go-review.googlesource.com/c/go/+/368215
Trust: Bryan C. Mills <bcmills@google.com>
Run-TryBot: Bryan C. Mills <bcmills@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agoos/exec: in TestContextCancel, dump goroutines on failure
Bryan C. Mills [Wed, 1 Dec 2021 19:15:08 +0000 (14:15 -0500)]
os/exec: in TestContextCancel, dump goroutines on failure

If this test fails, we want to know exactly what the os/exec
goroutines are doing. Panicking gives us a goroutine dump,
whereas t.Fatal does not.

While we're here, use exponential backoff instead of a hard-coded 1ms
sleep. We want to give the OS enough time to actually terminate the
subprocess.

For #42061

Change-Id: I3d50a71ac314853c68a935218e7f97ce18b08b5b
Reviewed-on: https://go-review.googlesource.com/c/go/+/368317
Trust: Bryan C. Mills <bcmills@google.com>
Run-TryBot: Bryan C. Mills <bcmills@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agodoc/go1.18: mention stack trace change
Cherry Mui [Wed, 1 Dec 2021 16:39:51 +0000 (11:39 -0500)]
doc/go1.18: mention stack trace change

For #47694.

Change-Id: I06cac88d5328c8c0e38212ad801d70bc36cdfc6f
Reviewed-on: https://go-review.googlesource.com/c/go/+/368234
Trust: Cherry Mui <cherryyz@google.com>
Reviewed-by: Austin Clements <austin@google.com>
3 years agogcimporters: allow reusing empty interfaces on the RHS of type decls
Robert Findley [Tue, 30 Nov 2021 22:48:51 +0000 (17:48 -0500)]
gcimporters: allow reusing empty interfaces on the RHS of type decls

We guard against caching or reusing interfaces on the RHS of a type
declaration, because for such interfaces the base type is used as the
interface method receiver type. However, we don't need to do this for
empty interfaces. By refining our guard, we can allow importing the
predeclared 'any' type on the RHS of a type declaration.

Update tests to add more coverage for importing generic export data.
Some accomodation had to be made for the unified builder, which does not
yet fully support generics in export data.

Fixes #49888

Change-Id: I51f329de464fc7309f95991b839ab55868c2924f
Reviewed-on: https://go-review.googlesource.com/c/go/+/367851
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>
3 years agocmd/go/internal/modfetch: remove legacy code
Baokun Lee [Wed, 1 Dec 2021 09:48:55 +0000 (17:48 +0800)]
cmd/go/internal/modfetch: remove legacy code

ReadFileRevs function is no longer used.

Change-Id: Ibac6319dca4cf8010195e7c2fb502655494fb728
Reviewed-on: https://go-review.googlesource.com/c/go/+/367756
Run-TryBot: Baokun Lee <bk@golangcn.org>
Reviewed-by: Bryan C. Mills <bcmills@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Trust: Baokun Lee <bk@golangcn.org>

3 years agoruntime: fix racy allgs access on weak memory architectures
Austin Clements [Wed, 1 Dec 2021 13:56:19 +0000 (08:56 -0500)]
runtime: fix racy allgs access on weak memory architectures

Currently, markroot is very clever about accessing the allgs slice to
find stack roots. Unfortunately, on weak memory architectures, it's a
little too clever and can sometimes read a nil g, causing a fatal
panic.

Specifically, gcMarkRootPrepare snapshots the length of allgs during
STW and then markroot accesses allgs up to this length during
concurrent marking. During concurrent marking, allgadd can append to
allgs *without synchronizing with markroot*, but the argument is that
the markroot access should be safe because allgs only grows
monotonically and existing entries in allgs never change.

This reasoning is insufficient on weak memory architectures. Suppose
thread 1 calls allgadd during concurrent marking and that allgs is
already at capacity. On thread 1, append will allocate a new slice
that initially consists of all nils, then copy the old backing store
to the new slice (write A), then allgadd will publish the new slice to
the allgs global (write B). Meanwhile, on thread 2, markroot reads the
allgs slice base pointer (read A), computes an offset from that base
pointer, and reads the value at that offset (read B). On a weak memory
machine, thread 2 can observe write B *before* write A. If the order
of events from thread 2's perspective is write B, read A, read B,
write A, then markroot on thread 2 will read a nil g and then panic.

Fix this by taking a snapshot of the allgs slice header in
gcMarkRootPrepare while the world is stopped and using that snapshot
as the list of stack roots in markroot. This eliminates all read/write
concurrency around the access in markroot.

Alternatively, we could make markroot use the atomicAllGs API to
atomically access the allgs list, but in my opinion it's much less
subtle to just eliminate all of the interesting concurrency around the
allgs access.

Fixes #49686.
Fixes #48845.
Fixes #43824.
(These are all just different paths to the same ultimate issue.)

Change-Id: I472b4934a637bbe88c8a080a280aa30212acf984
Reviewed-on: https://go-review.googlesource.com/c/go/+/368134
Trust: Austin Clements <austin@google.com>
Trust: Bryan C. Mills <bcmills@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
3 years agoruntime: give the subprocess more time in TestSpuriousWakeupsNeverHangSemasleep
Bryan C. Mills [Tue, 30 Nov 2021 21:10:47 +0000 (16:10 -0500)]
runtime: give the subprocess more time in TestSpuriousWakeupsNeverHangSemasleep

Issue #27250 reproduced readily enough to keep the subprocess hung
indefinitely when it occurred, so the timeout does not need to be
short to maintain test fidelity. On the other hand, on heavily loaded
systems it might take a while for the kernel to actually start the
subprocess, and it might also take a while for control flow to return
to the test after the subprocess exits.

We can reduce noise from this test in two ways:

1. Measure the timeout from closer to when the subprocess actually
starts sleeping, instead of when we started creating the subprocess.

2. Use a longer timeout, since it doesn't actually need to be short.

Fixes #38921
Updates #27250

Change-Id: I01c11ae82d0cdc6e7def2da6544b4d07201b35e8
Reviewed-on: https://go-review.googlesource.com/c/go/+/367849
Trust: Bryan C. Mills <bcmills@google.com>
Run-TryBot: Bryan C. Mills <bcmills@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
3 years agoruntime: fix riscv64 sigaction mask field offset
Rhys Hiltner [Tue, 30 Nov 2021 00:48:34 +0000 (16:48 -0800)]
runtime: fix riscv64 sigaction mask field offset

The Linux kernel for riscv64 does not include an sa_restorer field on
its sigaction struct, and expects sa_mask to come immediately after the
sa_flags field. Arrange the fields of the sigaction struct that are
known to the kernel so they appear at the correct byte offsets, and so
they agree with the output of "go tool cgo -godefs".

Follow the example set by the mips/mipsle port to leave the sa_restorer
field in place, but at an offset where it won't hurt anything.

Fixes #49709

Change-Id: I9bb0d7dbd7439d07e3a204461c7d790f33fd4977
Reviewed-on: https://go-review.googlesource.com/c/go/+/367635
Run-TryBot: Rhys Hiltner <rhys@justin.tv>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Joel Sing <joel@sing.id.au>
Trust: Ian Lance Taylor <iant@golang.org>

3 years agonet: do not use sendfile on iOS
Cherry Mui [Tue, 30 Nov 2021 22:30:57 +0000 (17:30 -0500)]
net: do not use sendfile on iOS

Apparently, on the iOS builder sendfile causes a SIGSYS signal
(instead of returning ENOSYS). Disabling it for now so we can
make progress on iOS. We can revisit if sendfile is actually
broken on iOS and whether it is beneficial.

Updates #49616.

Change-Id: I3883fad0ce35e3f0aa352301eb499a1afa0225a1
Reviewed-on: https://go-review.googlesource.com/c/go/+/368054
Trust: Cherry Mui <cherryyz@google.com>
Reviewed-by: Emmanuel Odeke <emmanuel@orijtech.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Changkun Ou <mail@changkun.de>
Reviewed-by: Roland Shoemaker <roland@golang.org>
3 years agocmd/go/internal/modload: fix up main-module checks from CL 334932
Bryan C. Mills [Mon, 29 Nov 2021 19:26:44 +0000 (14:26 -0500)]
cmd/go/internal/modload: fix up main-module checks from CL 334932

Some critical Version == "" checks were missing in mvs.go, causing
mvs.Req to fail to retain requirements provided by older versions of
main modules.

A few checks also ought to be rotated to put the less expensive
string-equality checks before the more expensive map lookups.

Fixes #48511

Change-Id: Ib8de9d49a6413660792c003866bfcf9ab7f82ee2
Reviewed-on: https://go-review.googlesource.com/c/go/+/368136
Trust: Bryan C. Mills <bcmills@google.com>
Run-TryBot: Bryan C. Mills <bcmills@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
3 years agocmd/compile: fix identical to recognize any and interface{}
Cuong Manh Le [Tue, 30 Nov 2021 09:58:36 +0000 (16:58 +0700)]
cmd/compile: fix identical to recognize any and interface{}

Currently, identical handles any and interface{} by checking against
Types[TINTER]. This is not always true, since when two generated
interface{} types may not use the same *Type instance.

Instead, we must check whether Type is empty interface or not.

Fixes #49875

Change-Id: I28fe4fc0100041a01bb03da795cfe8232b515fc4
Reviewed-on: https://go-review.googlesource.com/c/go/+/367754
Trust: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
3 years agoruntime: skip TestSignalIgnoreSIGTRAP on all OpenBSD builders
Tobias Klauser [Tue, 30 Nov 2021 22:21:39 +0000 (23:21 +0100)]
runtime: skip TestSignalIgnoreSIGTRAP on all OpenBSD builders

TestSignalIgnoreSIGTRAP is flaky on OpenBSD and the cause is suspected
to be a kernel bug. This test is currently only skipped on the previous
OpenBSD 6.2 and 6.4 builders for #17496. In the meantime the OpenBSD
builders were upgraded to more recent OpenBSD versions (currently 6.8
and 7.0). It seems the issue is still present in these OpenBSD versions
and there is no obvious workaround in Go. Thus, skip the flaky test on
OpenBSD in general.

Updates #17496
Updates #49725

Change-Id: I3577d287dcfaad7a81679db2e71540854fce065a
Reviewed-on: https://go-review.googlesource.com/c/go/+/367115
Trust: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Emmanuel Odeke <emmanuel@orijtech.com>
Reviewed-by: Bryan C. Mills <bcmills@google.com>
3 years agomisc/cgo/test: further reduce likeliness of hang in Test9400
Paul E. Murphy [Fri, 19 Nov 2021 22:33:42 +0000 (16:33 -0600)]
misc/cgo/test: further reduce likeliness of hang in Test9400

As suggested by #49680, a GC could be in-progress when we
disable GC. Force a GC after we pause to ensure we don't
hang in this case.

For #49695

Change-Id: I4fc4c06ef2ac174217c3dcf7d58c7669226e2d24
Reviewed-on: https://go-review.googlesource.com/c/go/+/367874
Run-TryBot: Paul Murphy <murp@ibm.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Trust: Paul Murphy <murp@ibm.com>

3 years agoruntime: add invalidptr=0 for TestGcZombieReporting
Chaoqun Han [Thu, 25 Nov 2021 13:16:28 +0000 (21:16 +0800)]
runtime: add invalidptr=0 for TestGcZombieReporting

pointers in zombies slice may cross-span, add invalidptr=0 for avoiding the badPointer check

Fixes #49613

Change-Id: Ifb1931922170e87e799e2e6081dc85dab3890205
Reviewed-on: https://go-review.googlesource.com/c/go/+/367044
Trust: Emmanuel Odeke <emmanuel@orijtech.com>
Run-TryBot: Emmanuel Odeke <emmanuel@orijtech.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Bryan C. Mills <bcmills@google.com>
Reviewed-by: Emmanuel Odeke <emmanuel@orijtech.com>
3 years agocmd/go: adjust BuildInfo.Settings
Russ Cox [Mon, 25 Oct 2021 20:02:07 +0000 (16:02 -0400)]
cmd/go: adjust BuildInfo.Settings

Make Settings more closely align with command-line flags
and environment variables.

- Change command-line flags to begin with -

- Change syntax of build lines to use Key=Value instead of Key<tab>Value.

- Change CGO_ENABLED to 0/1, matching environment variable,
  instead of false/true.

- Add GOOS and GOARCH.
  These are technically redundant, in that they can be extracted
  from the binary in other ways most of the time, but not always:
  GOOS=ios and GOOS=darwin may produce binaries that are
  difficult to tell apart. In any case, it's a lot easier to have them
  directly in the settings list than derive them from other parts
  of the binary.

- Add GOEXPERIMENT.
  These could be inferred from the tags list, but the experiments
  are being removed from the tags list.

- Change the tags list to match the -tags command-line argument.

- Add msan and race, echoing the -msan and -race arguments
  (always 'true' when present, omitted when false).

- Add GO$GOARCH when set.

Change-Id: Icb59ef4faa5c22407eadd94147b7e53cf4344ce6
Reviewed-on: https://go-review.googlesource.com/c/go/+/358539
Trust: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Bryan C. Mills <bcmills@google.com>
3 years agomisc/cgo/test: reduce likeliness of hang in Test9400
Paul E. Murphy [Fri, 19 Nov 2021 22:33:42 +0000 (16:33 -0600)]
misc/cgo/test: reduce likeliness of hang in Test9400

If a GC triggers while spinning in RewindAndSetgid, it may result in
this test hanging. Avoid it by disabling the collector before entering
the uninterruptable ASM conditional wait.

Fixes #49695

Change-Id: Ie0a03653481fb746f862469361b7840f4bfa8b67
Reviewed-on: https://go-review.googlesource.com/c/go/+/365836
Run-TryBot: Paul Murphy <murp@ibm.com>
TryBot-Result: Go Bot <gobot@golang.org>
Trust: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agonet/http: eliminate arbitrary timeout in TestClientWriteShutdown
Bryan C. Mills [Mon, 29 Nov 2021 20:17:36 +0000 (15:17 -0500)]
net/http: eliminate arbitrary timeout in TestClientWriteShutdown

This test occasionally hangs on the darwin-arm64-11_0-toothrot
builder. When it does, it fails with the unhelpful error message
"timeout" instead of a useful goroutine dump.

This change eliminates the use of an arbitrary timeout channel, so
that if (and probably when) the test hangs again we will get more
useful logs to diagnose the root cause.

For #49860

Change-Id: I23f6f1c81209f0b2dbe565e1dfb26b1b2eff0187
Reviewed-on: https://go-review.googlesource.com/c/go/+/367615
Trust: Bryan C. Mills <bcmills@google.com>
Run-TryBot: Bryan C. Mills <bcmills@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
3 years agoruntime: keep //go:cgo_unsafe_args arguments alive to prevent GC
Jason A. Donenfeld [Mon, 29 Nov 2021 20:07:04 +0000 (15:07 -0500)]
runtime: keep //go:cgo_unsafe_args arguments alive to prevent GC

When syscall's DLL.FindProc calls into syscall_getprocaddress with a
byte slice pointer, we need to keep those bytes alive. Otherwise the GC
will collect the allocation, and we wind up calling `GetProcAddress` on
garbage, which showed up as various flakes in the builders. It turns out
that this problem extends to many uses of //go:cgo_unsafe_args
throughout, on all platforms. So this patch fixes the issue by keeping
non-integer pointer arguments alive through their invocation in
//go:cgo_unsafe_args functions.

Fixes #49731.

Change-Id: I93e4fbc2e8e210cb3fc53149708758bb33f2f9c7
Reviewed-on: https://go-review.googlesource.com/c/go/+/367654
Trust: Jason A. Donenfeld <Jason@zx2c4.com>
Run-TryBot: Jason A. Donenfeld <Jason@zx2c4.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Patrik Nyblom <pnyb@google.com>
TryBot-Result: Go Bot <gobot@golang.org>

3 years agocmd/compile: change iexportVersionGeneric to 2
Dan Scales [Mon, 29 Nov 2021 20:04:42 +0000 (12:04 -0800)]
cmd/compile: change iexportVersionGeneric to 2

Don't expect/allow generics-related info in iexportVersion 1, now that
we increased the export version to 2.

Fixes #49853

Change-Id: I9bacee7f8e7cb9bb3b02a00084fad77edd220121
Reviewed-on: https://go-review.googlesource.com/c/go/+/367634
Trust: Dan Scales <danscales@google.com>
Run-TryBot: Dan Scales <danscales@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
3 years agogo/types, types2: handle case of no specific target types in conversion
Robert Findley [Mon, 29 Nov 2021 22:39:19 +0000 (17:39 -0500)]
go/types, types2: handle case of no specific target types in conversion

Avoid a panic by handling the case of no specific target type in a type
parameter to type parameter conversions.

Fixes #49864

Change-Id: I117dd80cc9d47c8c1e168f1caf0f281726270c84
Reviewed-on: https://go-review.googlesource.com/c/go/+/367616
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>

3 years agogo/types: ensure that constructed type parameters are immutable
Robert Findley [Mon, 29 Nov 2021 19:52:54 +0000 (14:52 -0500)]
go/types: ensure that constructed type parameters are immutable

TypeParam.iface may mutate TypeParam.bound in the event that the type
parameter bound is not an interface.

Ensure that iface() is called before the type-checking pass returns, and
before NewTypeParam or TypeParam.SetConstraint exits.

Fixes #49788

Change-Id: I72279acf5f0223161671c04887bc2c3df4158927
Reviewed-on: https://go-review.googlesource.com/c/go/+/367614
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>
3 years agogo/types: consider structural restrictions in Implements
Robert Findley [Mon, 29 Nov 2021 17:21:46 +0000 (12:21 -0500)]
go/types: consider structural restrictions in Implements

Fixes #49786

Change-Id: I4559d013399deda48bcb97aef3427ecf87a3ef26
Reviewed-on: https://go-review.googlesource.com/c/go/+/367515
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>
3 years agogo/types: better error message for missing ~ in constraint
Robert Griesemer [Fri, 26 Nov 2021 21:27:12 +0000 (13:27 -0800)]
go/types: better error message for missing ~ in constraint

This is a port of CL 366758 from types2 to go/types.

For #49179.

Change-Id: I7e1c6ffb392d5c535cf901004b7acbe8c3be9b0f
Reviewed-on: https://go-review.googlesource.com/c/go/+/367199
Trust: Robert Griesemer <gri@golang.org>
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
3 years agogo/types: better error position for instantiation failure
Robert Griesemer [Fri, 26 Nov 2021 21:14:01 +0000 (13:14 -0800)]
go/types: better error position for instantiation failure

This is a port of CL 366757 from types2 to go/types,
adjusted for the different handling of index expressions
in go/types.

For #49179.

Change-Id: Ic859eb09683134d055e28c8e0cb1f3814a87dc5c
Reviewed-on: https://go-review.googlesource.com/c/go/+/367198
Trust: Robert Griesemer <gri@golang.org>
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
3 years agogo/types: produce empty type set for invalid ~T
Robert Griesemer [Fri, 26 Nov 2021 20:43:06 +0000 (12:43 -0800)]
go/types: produce empty type set for invalid ~T

This is a clean port of CL 366278 from types2 to go/types.

For #49739.

Change-Id: I2e2cb739c02fcc07e012499c7b65b13b057875ea
Reviewed-on: https://go-review.googlesource.com/c/go/+/367197
Trust: Robert Griesemer <gri@golang.org>
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
3 years agogo/types: report types for mismatched call and return statements
Robert Griesemer [Thu, 25 Nov 2021 00:48:00 +0000 (16:48 -0800)]
go/types: report types for mismatched call and return statements

This is a port of CL 364874 from types2 to go/types with various
adjustments:

- the error position for "not enough arguments" in calls is the
  closing ) rather than the position of the last provided argument

- the ERROR comments in tests are positioned accordingly

- the reg. expression for matching error strings accepts newlines
  for the . pattern (added s flag)

For #48834.
For #48835.

Change-Id: I64362ecf605bcf9d89b8dc121432e0131bd5da1b
Reviewed-on: https://go-review.googlesource.com/c/go/+/367196
Trust: Robert Griesemer <gri@golang.org>
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
3 years agogo/types: restore original assignment error messages
Robert Griesemer [Thu, 25 Nov 2021 00:18:57 +0000 (16:18 -0800)]
go/types: restore original assignment error messages

This is the missing portion of the port of CL 351669
from types2 to go/types, now that we have a local flag
to control for compiler error messages.

Mostly a clean port but for adjustments to error reporting
which requires error codes in go/types.

Prerequisite for port of CL 364874.

Change-Id: I5fc8c83003e4396351f42e9adb08f4ebc8a05653
Reviewed-on: https://go-review.googlesource.com/c/go/+/367195
Trust: Robert Griesemer <gri@golang.org>
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
3 years agoruntime: fix preemption sensitivity in TestTinyAllocIssue37262
Austin Clements [Mon, 22 Nov 2021 20:33:01 +0000 (15:33 -0500)]
runtime: fix preemption sensitivity in TestTinyAllocIssue37262

TestTinyAllocIssue37262 assumes that all of its allocations will come
from the same tiny allocator (that is, the same P), and that nothing
else will allocate from that tiny allocator while it's running. It can
fail incorrectly if these assumptions aren't met.

Fix this potential test flakiness by disabling preemption during this
test.

As far as I know, this has never happened on the builders. It was
found by mayMoreStackPreempt.

Change-Id: I59f993e0bdbf46a9add842d0e278415422c3f804
Reviewed-on: https://go-review.googlesource.com/c/go/+/366994
Trust: Austin Clements <austin@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
3 years agospec: adjust section on package unsafe for type parameters
Robert Griesemer [Sun, 28 Nov 2021 23:25:09 +0000 (15:25 -0800)]
spec: adjust section on package unsafe for type parameters

Change-Id: I562d4648756e710020ee491f3801896563a89baa
Reviewed-on: https://go-review.googlesource.com/c/go/+/367395
Trust: Robert Griesemer <gri@golang.org>
Trust: Dan Scales <danscales@google.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agospec: rules for index expressions, len, cap, with type parameter types
Robert Griesemer [Wed, 24 Nov 2021 04:24:40 +0000 (20:24 -0800)]
spec: rules for index expressions, len, cap, with type parameter types

We want to support some special cases for index expressions, len, and
cap on operands of type parameters (such as indexing a value constrained
by byte slices and strings), hence the extra rules.

Change-Id: I4a07dc7e64bb47361b021d606c52eae1784d5430
Reviewed-on: https://go-review.googlesource.com/c/go/+/366814
Trust: Robert Griesemer <gri@golang.org>
Trust: Dan Scales <danscales@google.com>
Reviewed-by: Dan Scales <danscales@google.com>
3 years agospec: corrections to various sections
Robert Griesemer [Sun, 28 Nov 2021 22:17:11 +0000 (14:17 -0800)]
spec: corrections to various sections

- fix definition of "specific types" and add more examples
- state that a parameterized function must be instantiated
  when used as a function value
- remove duplicate word ("can can" -> "can")

Thanks to @danscales for finding these.

Change-Id: Ideb41efc35a3e67694d3bc97e462454feae37c44
Reviewed-on: https://go-review.googlesource.com/c/go/+/367394
Trust: Robert Griesemer <gri@golang.org>
Trust: Dan Scales <danscales@google.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Dan Scales <danscales@google.com>
3 years agovendor: update golang.org/x/net to tip
Ian Lance Taylor [Tue, 23 Nov 2021 20:34:04 +0000 (12:34 -0800)]
vendor: update golang.org/x/net to tip

This brings in a fix for OpenBSD that lets it correctly gather network
interface information.

For #42064

Change-Id: Ib88fd2f494bb2ee86fd2725d8375b2df1404c4ca
Reviewed-on: https://go-review.googlesource.com/c/go/+/366756
Trust: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: Bryan C. Mills <bcmills@google.com>
TryBot-Result: Go Bot <gobot@golang.org>

3 years agodoc/go1.18: use <code> instead of <pre> for target specification
Jason A. Donenfeld [Mon, 29 Nov 2021 16:05:41 +0000 (11:05 -0500)]
doc/go1.18: use <code> instead of <pre> for target specification

For #47694.

Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Change-Id: I38c2fd9b57fbbacf220a2bc679f67e2dfdcc7cb1
Reviewed-on: https://go-review.googlesource.com/c/go/+/367514
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
3 years agocmd/link: merge note sections into one segment
Meng Zhuo [Tue, 30 Mar 2021 11:58:35 +0000 (19:58 +0800)]
cmd/link: merge note sections into one segment

The ld from binutils merges note sections into one PT_NOTE
segment.
We should do that for consistency with binutils.

Change-Id: I45703525c720972d49c36c4f10ac47d1628b5698
Reviewed-on: https://go-review.googlesource.com/c/go/+/265957
Trust: Meng Zhuo <mzh@golangcn.org>
Trust: Emmanuel Odeke <emmanuel@orijtech.com>
Run-TryBot: Meng Zhuo <mzh@golangcn.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
3 years agoA+C: add Bharath Kumar Uppala (individual CLA)
Bharath Kumar Uppala [Thu, 25 Nov 2021 04:39:47 +0000 (10:09 +0530)]
A+C: add Bharath Kumar Uppala (individual CLA)

Change-Id: I4943e943892bd29bca2afafddb62f3060bc153e9
Reviewed-on: https://go-review.googlesource.com/c/go/+/367074
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Meng Zhuo <mzh@golangcn.org>
3 years agospec: fix link for instantiations
syumai [Sat, 27 Nov 2021 11:16:57 +0000 (11:16 +0000)]
spec: fix link for instantiations

This change corrects the link `Instantiantions` to `Instantiations` in the spec.

Change-Id: Ib0ed03420ae401d20af1ea723c5487018b2f462d
GitHub-Last-Rev: b84316c818b4aba022362fd09fac0d2b85da1a81
GitHub-Pull-Request: golang/go#49816
Reviewed-on: https://go-review.googlesource.com/c/go/+/367274
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Emmanuel Odeke <emmanuel@orijtech.com>
3 years agocmd/go: fix a typo in mod_lazy_new_import.txt
tangxi666 [Sat, 27 Nov 2021 05:20:05 +0000 (05:20 +0000)]
cmd/go: fix a typo in mod_lazy_new_import.txt

x/y -> a/y

Change-Id: If24970623731098bb72345b8f4c8518b563bbec8
GitHub-Last-Rev: 8a35dddbc155d97f03c08c06d945596584bc2e6d
GitHub-Pull-Request: golang/go#49813
Reviewed-on: https://go-review.googlesource.com/c/go/+/367202
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Emmanuel Odeke <emmanuel@orijtech.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>

3 years agotesting: simplify fuzzResult.String to avoid unnecessarily using fmt.Sprintf
sivchari [Sat, 27 Nov 2021 23:02:16 +0000 (23:02 +0000)]
testing: simplify fuzzResult.String to avoid unnecessarily using fmt.Sprintf

Change-Id: I16b6bfb6b0f02672c894b20845aa14d8dd1979b4
GitHub-Last-Rev: 75ab90123a8a5ad42e96795b756d3a9e898aaa06
GitHub-Pull-Request: golang/go#49819
Reviewed-on: https://go-review.googlesource.com/c/go/+/367314
Reviewed-by: Emmanuel Odeke <emmanuel@orijtech.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Emmanuel Odeke <emmanuel@orijtech.com>
TryBot-Result: Go Bot <gobot@golang.org>

3 years agodoc: go1.18 release notes for CLs 323318/332771
Carlo Alberto Ferraris [Mon, 22 Nov 2021 01:26:48 +0000 (01:26 +0000)]
doc: go1.18 release notes for CLs 323318/332771

Updates #47694

Change-Id: Ib76737996a701906117e096eb6d05b388576a874
GitHub-Last-Rev: 05b96499651bb4e90617460a72f1df9f2c8615f3
GitHub-Pull-Request: golang/go#49588
Reviewed-on: https://go-review.googlesource.com/c/go/+/363840
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Trust: Daniel Martí <mvdan@mvdan.cc>

3 years agocmd/go/internal: gofmt after CL 355010
Tobias Klauser [Thu, 25 Nov 2021 11:22:36 +0000 (12:22 +0100)]
cmd/go/internal: gofmt after CL 355010

Change-Id: I25902cc4e7a3d2a78b467825b723cd72b310e2a3
Reviewed-on: https://go-review.googlesource.com/c/go/+/367094
Trust: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Emmanuel Odeke <emmanuel@orijtech.com>
3 years agobufio: mention that panic at slicing means underlying reader is broken
Keith Randall [Fri, 26 Nov 2021 15:36:09 +0000 (07:36 -0800)]
bufio: mention that panic at slicing means underlying reader is broken

Fixes #49795

Change-Id: I2b4fd14f0ed36b643522559bebf5ce52b1d7b304
Reviewed-on: https://go-review.googlesource.com/c/go/+/367214
Trust: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agocmd/doc: fix "builtin" package parsing
Roi Martin [Thu, 25 Nov 2021 12:47:39 +0000 (13:47 +0100)]
cmd/doc: fix "builtin" package parsing

As stated in the code, "The builtin package needs special treatment: its
symbols are lower case but we want to see them, always". Thus, cmd/doc
forces the -u flag if the package being queried is called "builtin".
However, this happens after having already parsed the package. This
CL forces the -u flag just after parsing the command arguments and
before parsing any package.

Fixes #49796.

Change-Id: If690a900c7cfd1700feecb9529bd4344c3c249d1
Reviewed-on: https://go-review.googlesource.com/c/go/+/367134
Reviewed-by: Rob Pike <r@golang.org>
Run-TryBot: Rob Pike <r@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Trust: Emmanuel Odeke <emmanuel@orijtech.com>

3 years agodoc/go1.18: document non-cooperative preemption on windows/arm{,64}
Jason A. Donenfeld [Thu, 25 Nov 2021 12:05:58 +0000 (13:05 +0100)]
doc/go1.18: document non-cooperative preemption on windows/arm{,64}

For #47694.
Updates #49759.

Change-Id: I7accd81b8ea6c31e4a2b5e155cf93fe9c447813b
Reviewed-on: https://go-review.googlesource.com/c/go/+/367095
Trust: Jason A. Donenfeld <Jason@zx2c4.com>
Reviewed-by: Emmanuel Odeke <emmanuel@orijtech.com>
3 years agotest/fixedbugs: fix go directive of issue16008.go
sivchari [Fri, 26 Nov 2021 07:45:44 +0000 (07:45 +0000)]
test/fixedbugs: fix go directive of issue16008.go

This change modifies issue16008.go
I fixed // go:noinline to //go:noinline

Change-Id: Ic133eec51f0a7c4acf8cb22d25473ca08f1e916c
GitHub-Last-Rev: dd1868f2ca1f9ca7e2d6d1bfc15c601649896fdd
GitHub-Pull-Request: golang/go#49801
Reviewed-on: https://go-review.googlesource.com/c/go/+/367174
Reviewed-by: Alberto Donizetti <alb.donizetti@gmail.com>
Reviewed-by: Emmanuel Odeke <emmanuel@orijtech.com>
Trust: Alberto Donizetti <alb.donizetti@gmail.com>
Run-TryBot: Alberto Donizetti <alb.donizetti@gmail.com>
TryBot-Result: Go Bot <gobot@golang.org>

3 years agoA+C: add Chaoqun Han (individual CLA)
Chaoqun Han [Thu, 25 Nov 2021 01:41:06 +0000 (09:41 +0800)]
A+C: add Chaoqun Han (individual CLA)

Change-Id: Id4aa067ef84510a31992d7d32cc697dd8b8342f1
Reviewed-on: https://go-review.googlesource.com/c/go/+/367035
Reviewed-by: Meng Zhuo <mzh@golangcn.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Trust: Meng Zhuo <mzh@golangcn.org>
Run-TryBot: Meng Zhuo <mzh@golangcn.org>
TryBot-Result: Go Bot <gobot@golang.org>

3 years agoruntime: support non-cooperative preemption on windows/arm64
Jason A. Donenfeld [Tue, 23 Nov 2021 19:57:24 +0000 (20:57 +0100)]
runtime: support non-cooperative preemption on windows/arm64

This adds support for injecting asynchronous preemption calls on
windows/arm64. This code exactly follows sigctxt.pushCall for POSIX OSes
on arm64.

Fixes #49759.

Change-Id: Id35ff6bc105c1db9d7ed2918d3ecab0e4e9a9431
Reviewed-on: https://go-review.googlesource.com/c/go/+/366735
Trust: Jason A. Donenfeld <Jason@zx2c4.com>
Run-TryBot: Jason A. Donenfeld <Jason@zx2c4.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Patrik Nyblom <pnyb@google.com>
3 years agoruntime: support non-cooperative preemption on windows/arm
Jason A. Donenfeld [Tue, 23 Nov 2021 19:04:24 +0000 (20:04 +0100)]
runtime: support non-cooperative preemption on windows/arm

This adds support for injecting asynchronous preemption calls on
windows/arm. This code follows sigctxt.pushCall for POSIX OSes
on arm, except we subtract 1 from IP, just as in CL 273727.

Updates #10958.
Updates #24543.
Updates #49759.

Change-Id: Id0c2aed28662f50631b8c8cede3b4e6f088dafea
Reviewed-on: https://go-review.googlesource.com/c/go/+/366734
Trust: Jason A. Donenfeld <Jason@zx2c4.com>
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Patrik Nyblom <pnyb@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
3 years agocmd/dist: add buildtag parsing test
Russ Cox [Wed, 24 Nov 2021 15:35:21 +0000 (10:35 -0500)]
cmd/dist: add buildtag parsing test

Forgot to 'git add' this test written as part of CL 359314.

For #41184.

Change-Id: I2ebd48fd62a2053c8b16e5a8c48c1e11d1b86d5b
Reviewed-on: https://go-review.googlesource.com/c/go/+/366894
Trust: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agocmd/compile/internal/types2: better error message for missing ~ in constraint
Robert Griesemer [Tue, 23 Nov 2021 23:55:11 +0000 (15:55 -0800)]
cmd/compile/internal/types2: better error message for missing ~ in constraint

If a constraint could be satisfied if one of its type elements
had a ~, provide this information in the error message.

Fixes #49179.

Change-Id: I59f1a855a0646ad7254a978420b0334f1f52ec22
Reviewed-on: https://go-review.googlesource.com/c/go/+/366758
Trust: Robert Griesemer <gri@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
3 years agocmd/compile/internal/types2: better error position for instantiation failure
Robert Griesemer [Tue, 23 Nov 2021 21:54:02 +0000 (13:54 -0800)]
cmd/compile/internal/types2: better error position for instantiation failure

- Thread type argument expressions (rather than posLists) through various
type-checker functions so we can provide a better error position.

- Adjust signatures that expect a syntax.Pos to accept a poser instead
to avoid gratuituous conversions from expressions to positions.

- Rename targsx to xlist so we use xlist consistently for expression
lists.

First step in providing a better error message for the issue below.

For #49179.

Change-Id: I8fc685a2ee4f5640f4abd35568ba32bcb34e9e84
Reviewed-on: https://go-review.googlesource.com/c/go/+/366757
Trust: Robert Griesemer <gri@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
3 years agoruntime: skip TestTimePprof on illumos
Tobias Klauser [Wed, 24 Nov 2021 14:37:21 +0000 (15:37 +0100)]
runtime: skip TestTimePprof on illumos

On illumos nanotime calls libc, like on the other systems for which
TestTimePprof is skipped.

For #43118

Change-Id: I370d3f098a261185920cb1e3e3402d16200e301a
Reviewed-on: https://go-review.googlesource.com/c/go/+/366737
Trust: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agodoc/go1.18: add Clone doc
jiahua wang [Tue, 16 Nov 2021 07:36:37 +0000 (15:36 +0800)]
doc/go1.18: add Clone doc

For #47694

Change-Id: I3b135f6ff199d7a9746726b131fbe7dd97a8e95e
Reviewed-on: https://go-review.googlesource.com/c/go/+/364254
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Trust: Heschi Kreinick <heschi@google.com>

3 years agocmd/compile/internal/types2: produce empty type set for invalid ~T
Robert Griesemer [Tue, 23 Nov 2021 03:02:56 +0000 (19:02 -0800)]
cmd/compile/internal/types2: produce empty type set for invalid ~T

If ~T is not permitted because the underlying type of T is not the
same as T, there is no type that satisfies ~T. Besides reporting an
error, also ensure that the corresponding type set is empty.

For #49739.

Change-Id: I127f75f170902e7989f7fe7b352dabda9f72e2a5
Reviewed-on: https://go-review.googlesource.com/c/go/+/366278
Trust: Robert Griesemer <gri@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
3 years agocmd/compile/internal/types2: report types for mismatched call and return statements
Robert Griesemer [Wed, 17 Nov 2021 23:23:12 +0000 (15:23 -0800)]
cmd/compile/internal/types2: report types for mismatched call and return statements

Thanks to emmanuel@orijtech.com who wrote the initial version of
this change (CL 354490).

This change is following CL 354490 in idea but also contains various
simplifications, slightly improved printing of signature/type patterns,
adjustments for types2, and some fine-tuning of error positions.

Also adjusted several ERROR regexp patterns.

Fixes #48834.
Fixes #48835.

Change-Id: I31cf20c81753b1dc84836dbe83a39030ceb9db23
Reviewed-on: https://go-review.googlesource.com/c/go/+/364874
Trust: Robert Griesemer <gri@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Emmanuel Odeke <emmanuel@orijtech.com>
3 years agogo/types: print "nil" rather than "untyped nil"
Robert Griesemer [Tue, 23 Nov 2021 00:04:17 +0000 (16:04 -0800)]
go/types: print "nil" rather than "untyped nil"

This is a port of CL 366276 from types2 to go/types
with minor adjustments due to the slightly different
handling of nil in go/types.

It uses some more detailed error strings in stmt0.src;
the same changes are made to the corresponding types2
file.

For #48852.

Change-Id: I2cdf258799bcbe2d12bbadaf67b8b4504b356bd0
Reviewed-on: https://go-review.googlesource.com/c/go/+/366277
Trust: Robert Griesemer <gri@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
3 years agospec: type assertions and switches don't operate on type parameters
Robert Griesemer [Wed, 24 Nov 2021 04:56:33 +0000 (20:56 -0800)]
spec: type assertions and switches don't operate on type parameters

Change-Id: I11111b3617673be94508128489aed6488d518537
Reviewed-on: https://go-review.googlesource.com/c/go/+/366834
Trust: Robert Griesemer <gri@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agospec: adjust type identity rules for type parameters
Robert Griesemer [Wed, 24 Nov 2021 02:11:07 +0000 (18:11 -0800)]
spec: adjust type identity rules for type parameters

Change-Id: I5ffc7f26236487070447eaa0f6b14d1fab44c3c7
Reviewed-on: https://go-review.googlesource.com/c/go/+/366794
Trust: Robert Griesemer <gri@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agospec: add section on instantiation
Robert Griesemer [Sat, 20 Nov 2021 22:35:02 +0000 (14:35 -0800)]
spec: add section on instantiation

Change-Id: I2770da87b4c977b51dfa046f2f08283917675e1c
Reviewed-on: https://go-review.googlesource.com/c/go/+/365916
Trust: Robert Griesemer <gri@golang.org>
Trust: Dan Scales <danscales@google.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Dan Scales <danscales@google.com>
3 years agospec: adjust conversion rules for type parameters
Robert Griesemer [Sat, 20 Nov 2021 21:17:27 +0000 (13:17 -0800)]
spec: adjust conversion rules for type parameters

Change-Id: I7bfddf4be0d1d95419f312bb349ae2e16b74b795
Reviewed-on: https://go-review.googlesource.com/c/go/+/365915
Trust: Robert Griesemer <gri@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agospec: adjust assignability rules for type parameters
Robert Griesemer [Fri, 19 Nov 2021 00:11:10 +0000 (16:11 -0800)]
spec: adjust assignability rules for type parameters

Change-Id: I3c4d8bdb5e92ee7fdca9593fb043f94f467755e8
Reviewed-on: https://go-review.googlesource.com/c/go/+/365434
Trust: Robert Griesemer <gri@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agospec: adjust representability rules for type parameters
Robert Griesemer [Sat, 20 Nov 2021 21:29:02 +0000 (13:29 -0800)]
spec: adjust representability rules for type parameters

Change-Id: I4423a059527066c4418c195911f8184dfd3f5a15
Reviewed-on: https://go-review.googlesource.com/c/go/+/365914
Trust: Robert Griesemer <gri@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agospec: add definition of "specific types" of an interface
Robert Griesemer [Tue, 23 Nov 2021 20:02:40 +0000 (12:02 -0800)]
spec: add definition of "specific types" of an interface

The notion of specific types will be used to define rules for
assignability, convertability, etc. when type parameters are
involved.

Change-Id: Ic5c134261e2a9fe05cdf25efd342f052458ab5c8
Reviewed-on: https://go-review.googlesource.com/c/go/+/366754
Trust: Robert Griesemer <gri@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
3 years agocmd/go: fix bug in using the workfile flag with tests
Michael Matloob [Fri, 19 Nov 2021 21:09:52 +0000 (16:09 -0500)]
cmd/go: fix bug in using the workfile flag with tests

Tests do custom flag processing so we must process the workfile flag
after that happens.

Also fix an issue where errors weren't handled properly when the
workfile wasn't absolute (the go command should just exit), and where a
parse error was just dropped.

Fixes #48576

Change-Id: I3a94d8d3a515114b2c4cc0e73f63447df2fc6bc9
Reviewed-on: https://go-review.googlesource.com/c/go/+/366174
Trust: Michael Matloob <matloob@golang.org>
Run-TryBot: Michael Matloob <matloob@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Bryan C. Mills <bcmills@google.com>
3 years agodoc/go1.18: document that iOS 12 or newer is required
Dmitri Shuralyov [Wed, 24 Nov 2021 16:05:51 +0000 (11:05 -0500)]
doc/go1.18: document that iOS 12 or newer is required

For #47694.
Updates #49616.
Updates #48076.

Change-Id: I570564c3a54d3cd9cfc9b8267df9fbee3363b650
Reviewed-on: https://go-review.googlesource.com/c/go/+/366914
Trust: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
3 years agocmd/internal/objfile, debug/gosym: use the address of runtime.text as textStart
Cherry Mui [Tue, 23 Nov 2021 23:03:47 +0000 (18:03 -0500)]
cmd/internal/objfile, debug/gosym: use the address of runtime.text as textStart

Tools like objdump uses the pcln table to find the line number of
a given PC. For a PIE binary, at least in some cases such as on
macOS 12 with ld64-711, the table contains unrelocated address,
which does not match the address in the symbol table, causing the
lookup to fail.

In Go 1.18 the pcln table is essentually position independent,
except the start PC. Instead of reading the static content from
the table, use the PC of runtime.text from the symbol table.

While here, change the type of textStart to uint64. What matters
here is the word size of the target program, not the host, so it
shouldn't be uintptr.

Fixes #49700.

Change-Id: I517d79be7ba02dd4dd0275e75a11a136b08d76cd
Reviewed-on: https://go-review.googlesource.com/c/go/+/366695
Trust: Cherry Mui <cherryyz@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
3 years agocmd/internal/obj/x86: modify the threshold of assert loop for span6
zhouguangyuan [Mon, 22 Nov 2021 15:39:52 +0000 (23:39 +0800)]
cmd/internal/obj/x86: modify the threshold of assert loop for span6

Fixes: #49716
Change-Id: I7ed73f874c2ee1ee3f31c9c4428ed484167ca803
Reviewed-on: https://go-review.googlesource.com/c/go/+/366094
Reviewed-by: Cherry Mui <cherryyz@google.com>
Trust: Heschi Kreinick <heschi@google.com>

3 years agocmd/compile/internal/inline: revise closure inl position fix
Than McIntosh [Tue, 23 Nov 2021 13:19:45 +0000 (08:19 -0500)]
cmd/compile/internal/inline: revise closure inl position fix

This patch revises the fix for issue 46234, fixing a bug that was
accidentally introduced by CL 320913. When inlining a chunk of code
with a closure expression, we want to avoid updating the source
positions in the function being closed over, but we do want to update
the position for the ClosureExpr itself (since it is part of the
function we are inlining). CL 320913 unintentionally did away with the
closure expr source position update; here we restore it again.

Updates #46234.
Fixes #49171.

Change-Id: Iaa51bc498e374b9e5a46fa0acd7db520edbbbfca
Reviewed-on: https://go-review.googlesource.com/c/go/+/366494
Trust: Than McIntosh <thanm@google.com>
Trust: Dan Scales <danscales@google.com>
Reviewed-by: Dan Scales <danscales@google.com>
3 years agoruntime: skip TestTimePprof if nanotime calls libc
Ian Lance Taylor [Wed, 24 Nov 2021 00:30:17 +0000 (16:30 -0800)]
runtime: skip TestTimePprof if nanotime calls libc

Fixes #43118

Change-Id: I499bf335904e2b72a2a8876d0368fff5e69aa7fa
Reviewed-on: https://go-review.googlesource.com/c/go/+/366759
Trust: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Bryan C. Mills <bcmills@google.com>
TryBot-Result: Go Bot <gobot@golang.org>