Like every other command's -h flag. To achieve this, pass the command's
usage function to the cmdflag package, since that package is used by
multiple commands and cannot directly access *base.Command.

This also lets us get rid of testFlag1 and testFlag2, and instead have
contiguous raw strings for the test and testflag help docs.

Fixes #26999.

Change-Id: I2ebd66835ee61fa83270816a01fa312425224bb3
Reviewed-on: https://go-review.googlesource.com/c/144558
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Alan Donovan <adonovan@google.com>

This change fixes a TestFormat failure in fmt_test by adding a
recently introduced new known format (%q for syntax.Error).

Fixes #28621

Change-Id: I026ec88c334549a957a692c1652a860c57e23dae
Reviewed-on: https://go-review.googlesource.com/c/147837Reviewed-by: Ian Lance Taylor <iant@golang.org>

In ARM64 ABI, R18 is the "platform register", the use of which is
OS specific. The OS could choose to reserve this register. In
practice, it seems fine to use R18 on Linux but not on darwin (iOS).

Rename R18 to R18_PLATFORM to prevent accidental use. There is no
R18 usage within the standard library (besides tests, which are
updated).

Fixes #26110

Change-Id: Icef7b9549e2049db1df307a0180a3c90a12d7a84
Reviewed-on: https://go-review.googlesource.com/c/147218
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Dead-code eliminating labels is tricky because there might
be gotos that can still reach them.

Bug probably introduced with CL 91056

Fixes #28616

Change-Id: I6680465134e3486dcb658896f5172606cc51b104
Reviewed-on: https://go-review.googlesource.com/c/147817
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Iskander Sharipov <iskander.sharipov@intel.com>

This change adds the vet-lite command (the future cmd/vet) and all its
dependencies from x/tools, but not its tests and their dependencies.
It was created with these commands:

  $ (cd $GOPATH/src/golang.org/x/tools && git checkout c76e1ad)
  $ cd GOROOT/src/cmd
  $ govendor add $(go list -deps golang.org/x/tools/go/analysis/cmd/vet-lite | grep golang.org/x/tools)
  $ rm -fr $(find vendor/golang.org/x/tools/ -name testdata)
  $ rm $(find vendor/golang.org/x/tools/ -name \*_test.go)

I feel sure I am holding govendor wrong. Please advise.

A followup CL will make cmd/vet behave like vet-lite, initially just
for users that opt in, and soon after for all users, at which point
cmd/vet will be replaced in its entirety by a copy of vet-lite's small
main.go.

In the meantime, anyone can try the new tool using these commands:

 $ go build cmd/vendor/golang.org/x/tools/go/analysis/cmd/vet-lite
 $ export GOVETTOOL=$(which vet-lite)
 $ go vet your/project/...

Change-Id: Iea168111a32ce62f82f9fb706385ca0f368bc869
Reviewed-on: https://go-review.googlesource.com/c/147444Reviewed-by: Russ Cox <rsc@golang.org>

This improves the asm implementations for memmove and memclr on
ppc64x through use of vsx loads and stores when size is >= 32 bytes.
For memclr, dcbz is used when the size is >= 512 and aligned to 128.

Memclr/64       13.3ns ± 0%     10.7ns ± 0%   -19.55%  (p=0.000 n=8+7)
Memclr/96       14.9ns ± 0%     11.4ns ± 0%   -23.49%  (p=0.000 n=8+8)
Memclr/128      16.3ns ± 0%     12.3ns ± 0%   -24.54%  (p=0.000 n=8+8)
Memclr/160      17.3ns ± 0%     13.0ns ± 0%   -24.86%  (p=0.000 n=8+8)
Memclr/256      20.0ns ± 0%     15.3ns ± 0%   -23.62%  (p=0.000 n=8+8)
Memclr/512      34.2ns ± 0%     10.2ns ± 0%   -70.20%  (p=0.000 n=8+8)
Memclr/4096      178ns ± 0%       23ns ± 0%   -87.13%  (p=0.000 n=8+8)
Memclr/65536    2.67µs ± 0%     0.30µs ± 0%   -88.89%  (p=0.000 n=7+8)
Memclr/1M       43.2µs ± 0%     10.0µs ± 0%   -76.85%  (p=0.000 n=8+8)
Memclr/4M        173µs ± 0%       40µs ± 0%   -76.88%  (p=0.000 n=8+8)
Memclr/8M        349µs ± 0%       82µs ± 0%   -76.58%  (p=0.000 n=8+8)
Memclr/16M       701µs ± 7%      672µs ± 0%    -4.05%  (p=0.040 n=8+7)
Memclr/64M      2.70ms ± 0%     2.67ms ± 0%    -0.96%  (p=0.000 n=8+7)

Memmove/32      6.59ns ± 0%    5.84ns ± 0%  -11.34%  (p=0.029 n=4+4)
Memmove/64      7.91ns ± 0%    6.97ns ± 0%  -11.92%  (p=0.029 n=4+4)
Memmove/128     10.5ns ± 0%     8.8ns ± 0%  -16.24%  (p=0.029 n=4+4)
Memmove/256     21.0ns ± 0%    12.9ns ± 0%  -38.57%  (p=0.029 n=4+4)
Memmove/512     28.4ns ± 0%    26.2ns ± 0%   -7.75%  (p=0.029 n=4+4)
Memmove/1024    48.2ns ± 1%    39.4ns ± 0%  -18.26%  (p=0.029 n=4+4)
Memmove/2048    85.4ns ± 0%    69.0ns ± 0%  -19.20%  (p=0.029 n=4+4)
Memmove/4096     159ns ± 0%     128ns ± 0%  -19.50%  (p=0.029 n=4+4)

Change-Id: I8c1adf88790845bf31444a15249456006eb5bf8b
Reviewed-on: https://go-review.googlesource.com/c/141217
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Michael Munday <mike.munday@ibm.com>

On AIX, cmd/link needs two information in order to generate a dynamic
import, the library and its object needed. Currently, cmd/link isn't
able to retrieve this object only with the name of the library.
Therefore, the library pattern in cgo_import_dynamic must be
"lib.a/obj.o".

Change-Id: Ib8b8aaa9807c9fa6af46ece4e312d58073ed6ec1
Reviewed-on: https://go-review.googlesource.com/c/146957
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

Previously we panicked if the number of methods present for an embedded
field was >= 32. This change removes that limit and now StructOf
dynamically calls itself to create space for the number of methods.

Fixes #25402

Change-Id: I3b1deb119796d25f7e6eee1cdb126327b49a0b5e
GitHub-Last-Rev: 16da71ad6b23563f3ed26f1914adf41e3d42de69
GitHub-Pull-Request: golang/go#26865
Reviewed-on: https://go-review.googlesource.com/c/128479
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

Updates socks_bundle.go to git rev 26e67e7 for:
- 26e67e7 internal/socks: fix socket descriptor leakage in Dialer.Dial

Change-Id: I9ab27a85504d77f1ca2e97cb005f5e37fd3c3ff4
Reviewed-on: https://go-review.googlesource.com/c/147717
Run-TryBot: Mikio Hara <mikioh.public.networking@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

We collapse OpOffPtrs during generic rewrites.
However, we also use disjoint at the same time.
Instead of waiting for all OpOffPtrs to be collapsed
before the disjointness rules can kick in,
burrow through all OpOffPtrs immediately.

Change-Id: I60d0a70a9b4605b1817db7c4aab0c0d789651c90
Reviewed-on: https://go-review.googlesource.com/c/145206
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Michael Munday <mike.munday@ibm.com>
Reviewed-by: Keith Randall <khr@golang.org>

performance differences are in measurement noise as per benchcmp:

benchmark                            old ns/op     new ns/op     delta
BenchmarkUDP6LinkLocalUnicast-12     5012          5009          -0.06%

Fixes #28535

Change-Id: Id022e2ed089ce8388a2398e755848ec94e77e653
Reviewed-on: https://go-review.googlesource.com/c/146941
Run-TryBot: Mikio Hara <mikioh.public.networking@gmail.com>
Reviewed-by: Mikio Hara <mikioh.public.networking@gmail.com>

convT2E16 and other specialized type-to-interface routines
accept a type/itab argument and return a complete interface value.
However, we know enough in the routine to do without the type.
And the caller can construct the interface value using the type.

Doing so shrinks the call sites of ten of the specialized convT2x routines.
It also lets us unify the empty and non-empty interface routines.

Cuts 12k off cmd/go.

name                         old time/op  new time/op  delta
ConvT2ESmall-8               2.96ns ± 2%  2.34ns ± 4%  -21.01%  (p=0.000 n=175+189)
ConvT2EUintptr-8             3.00ns ± 3%  2.34ns ± 4%  -22.02%  (p=0.000 n=189+187)
ConvT2ELarge-8               21.3ns ± 7%  21.5ns ± 5%   +1.02%  (p=0.000 n=200+197)
ConvT2ISmall-8               2.99ns ± 4%  2.33ns ± 3%  -21.95%  (p=0.000 n=193+184)
ConvT2IUintptr-8             3.02ns ± 3%  2.33ns ± 3%  -22.82%  (p=0.000 n=198+190)
ConvT2ILarge-8               21.7ns ± 5%  22.2ns ± 4%   +2.31%  (p=0.000 n=199+198)
ConvT2Ezero/zero/16-8        2.96ns ± 2%  2.33ns ± 3%  -21.11%  (p=0.000 n=174+187)
ConvT2Ezero/zero/32-8        2.96ns ± 1%  2.35ns ± 4%  -20.62%  (p=0.000 n=163+193)
ConvT2Ezero/zero/64-8        2.99ns ± 2%  2.34ns ± 4%  -21.78%  (p=0.000 n=183+188)
ConvT2Ezero/zero/str-8       3.27ns ± 3%  2.54ns ± 3%  -22.32%  (p=0.000 n=195+192)
ConvT2Ezero/zero/slice-8     3.46ns ± 4%  2.81ns ± 3%  -18.96%  (p=0.000 n=197+164)
ConvT2Ezero/zero/big-8       88.4ns ±20%  90.0ns ±20%   +1.84%  (p=0.000 n=196+198)
ConvT2Ezero/nonzero/16-8     12.6ns ± 3%  12.3ns ± 3%   -2.34%  (p=0.000 n=167+196)
ConvT2Ezero/nonzero/32-8     12.3ns ± 4%  11.9ns ± 3%   -2.95%  (p=0.000 n=187+193)
ConvT2Ezero/nonzero/64-8     14.2ns ± 6%  13.8ns ± 5%   -2.94%  (p=0.000 n=198+199)
ConvT2Ezero/nonzero/str-8    27.2ns ± 5%  26.8ns ± 5%   -1.33%  (p=0.000 n=200+198)
ConvT2Ezero/nonzero/slice-8  33.3ns ± 8%  33.1ns ± 6%   -0.82%  (p=0.000 n=199+200)
ConvT2Ezero/nonzero/big-8    88.8ns ±22%  90.2ns ±18%   +1.58%  (p=0.000 n=200+199)


Neligible toolspeed impact.

name        old alloc/op      new alloc/op      delta
Template         35.4MB ± 0%       35.3MB ± 0%  -0.06%  (p=0.008 n=5+5)
Unicode          29.1MB ± 0%       29.1MB ± 0%    ~     (p=0.310 n=5+5)
GoTypes           122MB ± 0%        122MB ± 0%  -0.08%  (p=0.008 n=5+5)
Compiler          514MB ± 0%        513MB ± 0%  -0.02%  (p=0.008 n=5+5)
SSA              1.94GB ± 0%       1.94GB ± 0%  -0.01%  (p=0.008 n=5+5)
Flate            24.2MB ± 0%       24.2MB ± 0%    ~     (p=0.548 n=5+5)
GoParser         28.5MB ± 0%       28.5MB ± 0%  -0.05%  (p=0.016 n=5+5)
Reflect          86.3MB ± 0%       86.2MB ± 0%  -0.02%  (p=0.008 n=5+5)
Tar              34.9MB ± 0%       34.9MB ± 0%    ~     (p=0.095 n=5+5)
XML              47.1MB ± 0%       47.1MB ± 0%  -0.05%  (p=0.008 n=5+5)
[Geo mean]       81.0MB            81.0MB       -0.03%

name        old allocs/op     new allocs/op     delta
Template           349k ± 0%         349k ± 0%  -0.08%  (p=0.008 n=5+5)
Unicode            340k ± 0%         340k ± 0%    ~     (p=0.111 n=5+5)
GoTypes           1.28M ± 0%        1.28M ± 0%  -0.09%  (p=0.008 n=5+5)
Compiler          4.92M ± 0%        4.92M ± 0%  -0.08%  (p=0.008 n=5+5)
SSA               15.3M ± 0%        15.3M ± 0%  -0.03%  (p=0.008 n=5+5)
Flate              233k ± 0%         233k ± 0%    ~     (p=0.500 n=5+5)
GoParser           292k ± 0%         292k ± 0%  -0.06%  (p=0.008 n=5+5)
Reflect           1.05M ± 0%        1.05M ± 0%  -0.02%  (p=0.008 n=5+5)
Tar                344k ± 0%         343k ± 0%  -0.06%  (p=0.008 n=5+5)
XML                430k ± 0%         429k ± 0%  -0.08%  (p=0.008 n=5+5)
[Geo mean]         809k              809k       -0.05%

name        old object-bytes  new object-bytes  delta
Template          507kB ± 0%        507kB ± 0%  -0.04%  (p=0.008 n=5+5)
Unicode           225kB ± 0%        225kB ± 0%    ~     (all equal)
GoTypes          1.85MB ± 0%       1.85MB ± 0%  -0.08%  (p=0.008 n=5+5)
Compiler         6.75MB ± 0%       6.75MB ± 0%  +0.01%  (p=0.008 n=5+5)
SSA              21.4MB ± 0%       21.4MB ± 0%  -0.02%  (p=0.008 n=5+5)
Flate             328kB ± 0%        328kB ± 0%  -0.03%  (p=0.008 n=5+5)
GoParser          403kB ± 0%        402kB ± 0%  -0.06%  (p=0.008 n=5+5)
Reflect          1.41MB ± 0%       1.41MB ± 0%  -0.03%  (p=0.008 n=5+5)
Tar               457kB ± 0%        457kB ± 0%  -0.05%  (p=0.008 n=5+5)
XML               601kB ± 0%        600kB ± 0%  -0.16%  (p=0.008 n=5+5)
[Geo mean]       1.05MB            1.04MB       -0.05%


Change-Id: I677a4108c0ecd32617549294036aa84f9214c4fe
Reviewed-on: https://go-review.googlesource.com/c/147360
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Martin Möhrmann <moehrmann@google.com>

Consider these functions:

func f(a any) int
func g(a any) int

Prior to this change, since f and g have identical signatures,
they would share a single generated func type.

types.SubstAny makes a shallow type copy, even after instantiation,
f and g share a single generated Result type.
So if you instantiate f with any=T, call dowidth,
instantiate g with any=U, and call dowidth,
and if sizeof(T) != sizeof(U),
then the Offset of the result for f is now wrong.

I don't believe this happens at all right now, but it bit me hard when
experimenting with some other compiler changes.
And it's hard to debug. It results in rare stack corruption, causing
problems far from the actual source of the problem.

To fix this, change SubstAny to make deep copies of TSTRUCTs.

name        old alloc/op      new alloc/op      delta
Template         35.3MB ± 0%       35.4MB ± 0%  +0.23%  (p=0.008 n=5+5)
Unicode          29.1MB ± 0%       29.1MB ± 0%  +0.16%  (p=0.008 n=5+5)
GoTypes           122MB ± 0%        122MB ± 0%  +0.16%  (p=0.008 n=5+5)
Compiler          513MB ± 0%        514MB ± 0%  +0.19%  (p=0.008 n=5+5)
SSA              1.94GB ± 0%       1.94GB ± 0%  +0.01%  (p=0.008 n=5+5)
Flate            24.2MB ± 0%       24.2MB ± 0%  +0.08%  (p=0.008 n=5+5)
GoParser         28.5MB ± 0%       28.5MB ± 0%  +0.24%  (p=0.008 n=5+5)
Reflect          86.2MB ± 0%       86.3MB ± 0%  +0.09%  (p=0.008 n=5+5)
Tar              34.9MB ± 0%       34.9MB ± 0%  +0.13%  (p=0.008 n=5+5)
XML              47.0MB ± 0%       47.1MB ± 0%  +0.18%  (p=0.008 n=5+5)
[Geo mean]       80.9MB            81.0MB       +0.15%

name        old allocs/op     new allocs/op     delta
Template           348k ± 0%         349k ± 0%  +0.38%  (p=0.008 n=5+5)
Unicode            340k ± 0%         340k ± 0%  +0.21%  (p=0.008 n=5+5)
GoTypes           1.27M ± 0%        1.28M ± 0%  +0.27%  (p=0.008 n=5+5)
Compiler          4.90M ± 0%        4.92M ± 0%  +0.36%  (p=0.008 n=5+5)
SSA               15.3M ± 0%        15.3M ± 0%  +0.03%  (p=0.008 n=5+5)
Flate              232k ± 0%         233k ± 0%  +0.14%  (p=0.008 n=5+5)
GoParser           291k ± 0%         292k ± 0%  +0.42%  (p=0.008 n=5+5)
Reflect           1.05M ± 0%        1.05M ± 0%  +0.14%  (p=0.008 n=5+5)
Tar                343k ± 0%         344k ± 0%  +0.22%  (p=0.008 n=5+5)
XML                428k ± 0%         430k ± 0%  +0.36%  (p=0.008 n=5+5)
[Geo mean]         807k              809k       +0.25%

Change-Id: I62134db642206cded01920dc1d8a7da61f7ca0ac
Reviewed-on: https://go-review.googlesource.com/c/147038
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>

This change cleans up references to MSpan, MCache, and MCentral in the
docs via a bunch of sed invocations to better reflect the Go names for
the equivalent structures (i.e. mspan, mcache, mcentral) and their
methods (i.e. MSpan_Sweep -> mspan.sweep).

Change-Id: Ie911ac975a24bd25200a273086dd835ab78b1711
Reviewed-on: https://go-review.googlesource.com/c/147557Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

TestTracebackAncestors has a ~0.1% chance of failing with more
goroutines in the traceback than expected. This happens because
there's a window between each goroutine starting its child and that
goroutine actually exiting. The test captures its own stack trace
after everything is "done", but if this happens during that window, it
will include the goroutine that's in the process of being torn down.
Here's an example of such a failure:
https://build.golang.org/log/fad10d0625295eb79fa879f53b8b32b9d0596af8

This CL fixes this by recording the goroutines that are expected to
exit and removing them from the stack trace. With this fix, this test
passed 15,000 times with no failures.

Change-Id: I71e7c6282987a15e8b74188b9c585aa2ca97cbcd
Reviewed-on: https://go-review.googlesource.com/c/147517
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

This documentation example was broken:
https://golang.org/pkg/image/png/#example_Decode.
It did not have the "io" package imported,
The package was referenced in the result type of the function.

The "playExample" function did not inspect
the result types of declared functions.

This CL adds inspecting of parameters and result types of functions.

Fixes #28492
Updates #9679

Change-Id: I6d8b11bad2db8ea8ba69039cfaa914093bdd5132
Reviewed-on: https://go-review.googlesource.com/c/146118
Run-TryBot: Yury Smolsky <yury@smolsky.by>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>

Updates #28601

Change-Id: I734fc5ded153126d384f0df912ecd4d208005e49
Reviewed-on: https://go-review.googlesource.com/c/147537
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>

This change re-introduces (temporarily) a work-around for recursive
alias type declarations, originally in https://golang.org/cl/35831/
(intended as fix for #18640). The work-around was removed later
for a more comprehensive cycle detection check. That check
contained a subtle error which made the code appear to work,
while in fact creating incorrect types internally. See #25838
for details.

By re-introducing the original work-around, we eliminate problems
with many simple recursive type declarations involving aliases;
specifically cases such as #27232 and #27267. However, the more
general problem remains.

This CL also fixes the subtle error (incorrect variable use when
analyzing a type cycle) mentioned above and now issues a fatal
error with a reference to the relevant issue (rather than crashing
later during the compilation). While not great, this is better
than the current status. The long-term solution will need to
address these cycles (see #25838).

As a consequence, several old test cases are not accepted anymore
by the compiler since they happened to work accidentally only.
This CL disables parts or all code of those test cases. The issues
are: #18640, #23823, and #24939.

One of the new test cases (fixedbugs/issue27232.go) exposed a
go/types issue. The test case is excluded from the go/types test
suite and an issue was filed (#28576).

Updates #18640.
Updates #23823.
Updates #24939.
Updates #25838.
Updates #28576.

Fixes #27232.
Fixes #27267.

Change-Id: I6c2d10da98bfc6f4f445c755fcaab17fc7b214c5
Reviewed-on: https://go-review.googlesource.com/c/147286Reviewed-by: Matthew Dempsky <mdempsky@google.com>

Change-Id: I5f7d2155c2c3a47dabdf16fe46b122ede81de4fc
Reviewed-on: https://go-review.googlesource.com/c/147284Reviewed-by: Matthew Dempsky <mdempsky@google.com>

As of 07e738ec all spans are allocated out of a treap, and not just
large spans or spans for large objects. Also, now we have a separate
treap for spans that have been scavenged.

Change-Id: I9c2cb7b6798fc536bbd34835da2e888224fd7ed4
Reviewed-on: https://go-review.googlesource.com/c/142958Reviewed-by: Austin Clements <austin@google.com>

Change-Id: Ib9dcdc76095f6718f1cdc83349503f52567c76d4
Reviewed-on: https://go-review.googlesource.com/c/60801
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>

This uses the mutator utilization distribution to compute the p99.9,
p99, and p95 mutator utilization topograph lines and display them
along with the MMU.

Change-Id: I8c7e0ec326aa4bc00619ec7562854253f01cc802
Reviewed-on: https://go-review.googlesource.com/c/60800
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>

This adds support for computing the quantiles of a mutator utilization
distribution.

Change-Id: Ia8b3ed14bf415c234e2f567360fd1b361d28bd40
Reviewed-on: https://go-review.googlesource.com/c/60799
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>

Change-Id: I672240487172380c9eef61837b41698021aaf834
Reviewed-on: https://go-review.googlesource.com/c/60798
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>

The current MMU analysis considers all Ps together, so if, for
example, one of four Ps is blocked, mutator utilization is 75%.
However, this is less useful for understanding the impact on
individual goroutines because that one blocked goroutine could be
blocked for a very long time, but we still appear to have good
utilization.

Hence, this introduces a new flag that does a "per-P" analysis where
the utilization of each P is considered independently. The MMU is then
the combination of the MMU for each P's utilization function.

Change-Id: Id67b980d4d82b511d28300cdf92ccbb5ae8f0c78
Reviewed-on: https://go-review.googlesource.com/c/60797
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>

Change-Id: I4ba963b003cb25b39d7575d423f17930d84f3f69
Reviewed-on: https://go-review.googlesource.com/c/60796
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>

This adds the ability to click a point on the MMU graph to show a list
of the worst 10 mutator utilization windows of the selected size. This
list in turn links to the trace viewer to drill down on specifically
what happened in each specific window.

Change-Id: Ic1b72d8b37fbf2212211c513cf36b34788b30133
Reviewed-on: https://go-review.googlesource.com/c/60795
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>

This will let the trace viewer show specifically when poor utilization
happened and link to specific instances in the trace.

Change-Id: I1f03a0f9d9a7570009bb15762e7b8b6f215e9423
Reviewed-on: https://go-review.googlesource.com/c/60793
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>

This further optimizes MMU construction by first computing a
low-resolution summary of the utilization curve. This "band" summary
lets us compute the worst-possible window starting in each of these
low-resolution bands (even without knowing where in the band the
window falls). This in turn lets us compute precise minimum mutator
utilization only in the worst low-resolution bands until we can show
that any remaining bands can't possibly contain a worse window.

This slows down MMU construction for small traces, but these are
reasonably fast to compute either way. For large traces (e.g.,
150,000+ utilization changes) it's significantly faster.

Change-Id: Ie66454e71f3fb06be3f6173b6d91ad75c61bda48
Reviewed-on: https://go-review.googlesource.com/c/60792
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>

This commit speeds up MMU construction by ~10X (and reduces the number
of windows considered by ~20X) by using an observation about the
maximum slope of the windowed mutator utilization function to advance
the window time in jumps if the window's current mean mutator
utilization is much larger than the current minimum.

Change-Id: If3cba5da0c4adc37b568740f940793e491e96a51
Reviewed-on: https://go-review.googlesource.com/c/60791
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>

This adds an endpoint to the trace tool that plots the minimum mutator
utilization curve using information on mark assists and GC pauses from
the trace.

This commit implements a fairly straightforward O(nm) algorithm for
computing the MMU (and tests against an even more direct but slower
algorithm). Future commits will extend and optimize this algorithm.

This should be useful for debugging and understanding mutator
utilization issues like #14951, #14812, #18155. #18534, #21107,
particularly once follow-up CLs add trace cross-referencing.

Change-Id: Ic2866869e7da1e6c56ba3e809abbcb2eb9c4923a
Reviewed-on: https://go-review.googlesource.com/c/60790
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>

Change-Id: I19ff9d231c4cc779b0737802c3c40ee2e00934dd
Reviewed-on: https://go-review.googlesource.com/c/147477Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

zgoos_aix.go is missing GoosJs, the order of GoosAndroid and GoosAix is
mixed up in all files and GoosHurd was added after CL 146023 introduced
GOOS=hurd.

Change-Id: I7e2f5a15645272e9020cfca86e44c364fc072a2b
Reviewed-on: https://go-review.googlesource.com/c/147397
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

The kernel on some Samsung S9+ models reports support for arm64 8.1
atomics, but in reality only some of the cores support them. Go
programs scheduled to cores without support will crash with SIGILL.

This change unconditionally disables the optimization on Android.
A better fix is to precisely detect the offending chipset.

Fixes #28431

Change-Id: I35a1273e5660603824d30ebef2ce7e429241bf1f
Reviewed-on: https://go-review.googlesource.com/c/147377
Run-TryBot: Elias Naur <elias.naur@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>

At each comparison, we're making a copy of the whole string.
Instead, use unsafe to share the string backing store with a []byte.

It reduces the test time from ~4sec to ~1sec on my machine
(darwin/amd64).  Some builders were having much more trouble with this
test (>3min), it may help more there.

Fixes #26174
Fixes #28573
Fixes #26155
Update #26473

Change-Id: Id5856fd26faf6ff46e763a088f039230556a4116
Reviewed-on: https://go-review.googlesource.com/c/147358Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

Was only ever filled with one Etype (TFORW) and only used
in one place. Easier to just check for TFORW.

Change-Id: Icc96da3a22b0af1d7e60bc5841c744916c53341e
Reviewed-on: https://go-review.googlesource.com/c/147285Reviewed-by: Martin Möhrmann <moehrmann@google.com>

The Fatalf mechanism already prints "compiler internal error:"
when reporting an error. There's no need to have "internal error"
in the error message passed to Fatalf calls. Removed them.

Fixes #28575.

Change-Id: I12b1bea37bc839780f257c27ef9e2005bf334925
Reviewed-on: https://go-review.googlesource.com/c/147287
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Martin Möhrmann <moehrmann@google.com>

Follow-up on https://golang.org/cl/146782.

TBR=bradfitz

Change-Id: Idaf5488fedfc05d6ff71706fa0bcd70bf98ab25a
Reviewed-on: https://go-review.googlesource.com/c/147283Reviewed-by: Robert Griesemer <gri@golang.org>

Currently, WriteObjFile deduplicates symbols by name. This is a
strange and unexpected place to do this. But, worse, there's no
checking that it's reasonable to deduplicate two symbols, so this
makes it incredibly easy to mask errors involving duplicate symbols.
Dealing with duplicate symbols is better left to the linker. We're
also about to introduce multiple symbols with the same name but
different ABIs/versions, which would make this deduplication more
complicated. We just removed the only part of the compiler that
actually depended on this behavior.

This CL removes symbol deduplication from WriteObjFile, since it is no
longer needed.

For #27539.

Change-Id: I650c550e46e83f95c67cb6c6646f9b2f7f10df30
Reviewed-on: https://go-review.googlesource.com/c/146558
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>

Currently, liveness produces a distinct obj.LSym for each GC bitmap
for each function. These are then named by content hash and only
ultimately deduplicated by WriteObjFile.

For various reasons (see next commit), we want to remove this
deduplication behavior from WriteObjFile. Furthermore, it's
inefficient to produce these duplicate symbols in the first place.

GC bitmaps are the only source of duplicate symbols in the compiler.
This commit eliminates these duplicate symbols by declaring them in
the Ctxt symbol hash just like every other obj.LSym. As a result, all
GC bitmaps with the same content now refer to the same obj.LSym.

The next commit will remove deduplication from WriteObjFile.

For #27539.

Change-Id: I4f15e3d99530122cdf473b7a838c69ef5f79db59
Reviewed-on: https://go-review.googlesource.com/c/146557
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>