misc/cgo/testcarchive: permit SIGQUIT for TestSignalForwardingExternal

Occasionally the signal will be sent to a Go thread, which will cause
the program to exit with SIGQUIT rather than SIGSEGV.

Add TestSignalForwardingGo to test the case where the signal is
expected to be delivered to a Go thread.

Fixes #53907

Change-Id: Iaefb964c2be4a815c11c507fa89648f8a7740ba9
Reviewed-on: https://go-review.googlesource.com/c/go/+/419014
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>

net: allow single dot in resolv.conf search statement

A single dot in a search statement doesn't affect DNS lookup, so just ignore it and a syntax error.

Fixes #54124

Change-Id: Idd43bd34c5c16af50cba51f0b6e24f992eec6e57
Reviewed-on: https://go-review.googlesource.com/c/go/+/423875
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>

sync: use atomic.Uint64 for WaitGroup state

So it's guaranteed to have 64-bit alignment, simplify the code without
losing any performance:

name                     old time/op    new time/op    delta
WaitGroupUncontended-8     3.84ns ± 2%    3.82ns ± 1%   ~     (p=0.159 n=10+10)
WaitGroupAddDone-8         33.2ns ± 3%    33.0ns ± 3%   ~     (p=0.564 n=9+10)
WaitGroupAddDoneWork-8     39.3ns ± 1%    39.3ns ± 1%   ~     (p=1.000 n=8+9)
WaitGroupWait-8            0.70ns ± 3%    0.70ns ± 2%   ~     (p=0.720 n=9+10)
WaitGroupWaitWork-8        7.93ns ± 1%    7.99ns ± 3%   ~     (p=0.271 n=10+10)
WaitGroupActuallyWait-8     135ns ± 2%     135ns ± 1%   ~     (p=0.897 n=10+10)

Change-Id: I446b53fa92873419aadd592f45e51398f8ad8652
Reviewed-on: https://go-review.googlesource.com/c/go/+/424835
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>

runtime: fast clock_gettime on FreeBSD, split getHPETTimecounter

Call only initHPETTimecounter on the system stack.
Use O_CLOEXEC flag when opening the HPET device.

FreeBSD 12.3-RELEASE-p2, AMD FX-8300
paulzhol@relic:~/go/src/time % ~/gocode/bin/benchcmp old_hpet.txt new_hpet.txt
benchcmp is deprecated in favor of benchstat: https://pkg.go.dev/golang.org/x/perf/cmd/benchstat
benchmark                   old ns/op     new ns/op     delta
BenchmarkNow-8              1420          1397          -1.62%
BenchmarkNowUnixNano-8      1421          1404          -1.20%
BenchmarkNowUnixMilli-8     1423          1405          -1.26%
BenchmarkNowUnixMicro-8     1423          1404          -1.34%

Update #50947

Change-Id: I553b5427fb0b86d7e070af4516b36326bc0aaf00
Reviewed-on: https://go-review.googlesource.com/c/go/+/391856
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>

runtime: add and use runtime/internal/sys.NotInHeap

Updates #46731

Change-Id: Ic2208c8bb639aa1e390be0d62e2bd799ecf20654
Reviewed-on: https://go-review.googlesource.com/c/go/+/421878
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>

internal/buildcfg: enable unified IR by default

Unified IR is now at feature parity with the non-unified frontend.

Change-Id: I2a8cd41267a66f5eb1f425b89d18c6d501676e07
Reviewed-on: https://go-review.googlesource.com/c/go/+/422235
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>

cmd/compile/internal/types2: use go/types/testdata/spec tests

Since the spec tests are now identical between the two type checkers,
remove the local copy of the spec tests and (for now) use the tests
in go/types/testdata/spec instead. Eventually we may decide to move
all tests out of the type checker directories and place them in a
shared space (e.g. internal/types/testdata).

For #54511.

Change-Id: Ibc62230553a5ff215160dcc97696b333ae9cf685
Reviewed-on: https://go-review.googlesource.com/c/go/+/424676
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>

go/types, types2: consolidate testdata/spec test files

Besides applying gofmt (which doesn't damage the tests in this case),
this removes a TODO for a set of nil-related tests.

The test files are now identical.

For #54511.

Change-Id: I0a286ad607f317f43972c1f5ee741b4f5bc9576d
Reviewed-on: https://go-review.googlesource.com/c/go/+/424675
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>

go/types, types2: consolidate testdata/examples test files

Besides applying gofmt (which doesn't damage the tests in this case),
the ERROR comments in the types2 files now match the go/types files.
But because types2 still reports some errors are different positions,
the checking code now allows for some position discrepancy (similar
to what we do for other tests).

Except for an outstanding TODO in go/types/testdata/examples/types.go
the test files are now identical.

For #54511.

Change-Id: I5748e0f678d11c5c0bdf4fdf28bd04f0b11b5b23
Reviewed-on: https://go-review.googlesource.com/c/go/+/424674
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>

cmd/compile/internal/syntax: handle missing index like in go/parser

Instead of simply reporting an error but otherwise dropping the
index expression from the parse tree when an index is missing
(as in: x[]), create an index expression with a "bad expression"
as index. This matches the behavior of go/parser and permits the
use of the same test case for both parsers.

(It would be simpler to adjust the go/parser to match the syntax
parser's behavior, but that would break backward-compatibility
of the go/parser.)

Adjust the affected test files.

For #54511.

Change-Id: If7668973794604593e869a24b560da92e100b812
Reviewed-on: https://go-review.googlesource.com/c/go/+/424654
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>

crypto/internal/boring: update to newer boringcrypto, add arm64

Update the version of BoringCrypto to boringssl tag
fips-20210429, for which FIPS approval is "in process".

Add GOARCH=arm64 BoringCrypto support.

Shuffle build to run as distinct steps in Docker so that
we can rerun later parts like build-goboring.sh without
rerunning all the setup.

Strip unnecessary parts of the syso using --strip-unneeded,
which cuts the amd64 syso from 10MB to 2.5MB.
The new arm64 syso is 2MB, so even though we added a new
architecture, we have half as much syso to distribute.

Change-Id: I0f7327389a3a485b82577abea9153d006418298f
Reviewed-on: https://go-review.googlesource.com/c/go/+/423362
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>

go/types: match types2 error msg formatting for conversion errors

This resolves a TODO. Also, it allows go/types and types2 to use
identical test files in testdata/spec, a prerequisition for sharing
them eventually.

For #54511.

Change-Id: Ia9d11b15ada237c18b6ce8f72732f70134a1a06f
Reviewed-on: https://go-review.googlesource.com/c/go/+/424537
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

internal/buildcfg: support GOPPC64=power10

This does not enable any new functionality. It should
behave identically to GOPPC64=power9.

Updates #44549

Change-Id: I9a860544527fcfe97cbaf89686459d40dcf9593e
Reviewed-on: https://go-review.googlesource.com/c/go/+/352791
Run-TryBot: Paul Murphy <murp@ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>

runtime: remove p padding field for atomic fields alignment

CL 424396 and CL 424397 changed timer0When/timerModifiedEarliest to
atomic.Uint64, just they're guaranted to have 64-bit alignment.

Change-Id: Idaff1059da2aac84520b9b0e34f9721a74dbba5a
Reviewed-on: https://go-review.googlesource.com/c/go/+/424794
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>

cmd/compile: remove OFORUNTIL

Not used any more.

Fixes #53860

Change-Id: Id0b1c3ed30b576d6c5f08f064d1262de337262b3
Reviewed-on: https://go-review.googlesource.com/c/go/+/418374
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

cmd/compile: rip out support for OpVarKill from ssa backend

Change-Id: I008479a7516d8379186ce630748e503d94d3b1e8
Reviewed-on: https://go-review.googlesource.com/c/go/+/419235
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>

cmd/compile: stop using VARKILL

With the introduction of stack objects, VARKILL information is
no longer needed.

With stack objects, an object is dead when there are no more static
references to it, and the stack scanner can't find any live pointers
to it. VARKILL information isn't used to establish live ranges for
address-taken variables any more. In effect, the last static reference
*is* the VARKILL, and there's an additional dynamic liveness check
during stack scanning.

Next CL will actually rip out the VARKILL opcodes.

Change-Id: I030a2ab867445cf4e0e69397911f8a2e2f0ed07b
Reviewed-on: https://go-review.googlesource.com/c/go/+/419234
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Keith Randall <khr@golang.org>

cmd/compile: don't use OFORUNTIL when implementing range loops

We don't need this special loop construct anymore now that we do
conservative GC scanning of the top of stack. Rewrite instead to a simple
pointer increment on every iteration. This leads to having a potential
past-the-end pointer at the end of the last iteration, but that value
immediately goes dead after the loop condition fails, and the past-the-end
pointer is never live across any call.

This simplifies and speeds up loops.

R=go1.20

TODO: actually delete all support for OFORUNTIL. It is now never generated,
but code to handle it (e.g. in ssagen) is still around.

TODO: in "for _, x := range" loops, we could get rid of the index
altogether and use a "pointer to the last element" reference to determine
when the loop is complete.

Fixes #53409

Change-Id: Ifc141600ff898a8bc6a75f793e575f8862679ba1
Reviewed-on: https://go-review.googlesource.com/c/go/+/414876
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

cmd/compile: enable more inlining for unified IR

The non-unified frontend had repeated issues with inlining and
generics (#49309, #51909, #52907), which led us to substantially
restrict inlining when shape types were present.

However, these issues are evidently not present in unified IR's
inliner, and the safety restrictions added for the non-unified
frontend can simply be disabled in unified mode.

Fixes #54497.

Change-Id: I8e6ac9f3393c588bfaf14c6452891b9640a9d1bd
Reviewed-on: https://go-review.googlesource.com/c/go/+/424775
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>

cmd/compile: fix devirtualization bug with unified IR

As a consistency check in devirtualization, when we determine `i` (of
interface type `I`) always has dynamic type `T`, we insert a type
assertion `i.(T)`. This emits an itab check for `go:itab.T,I`, but
it's always true (and so SSA optimizes it away).

However, if `I` is instead the generic interface type `I[T]`, then
`go:itab.T,I[int]` and `go:itab.T,I[go.shape.int]` are equivalent but
distinct itabs. And notably, we'll have originally created the
interface value using the former; but the (non-dynamic) TypeAssertExpr
created by devirtualization would ultimately emit a comparison against
the latter. This comparison would then evaluate false, leading to a
spurious type assertion panic at runtime.

The comparison is just meant as an extra safety check, so it should be
safe to just disable. But for now, it's simpler/safer to just punt on
devirtualization in this case. (The non-unified frontend doesn't
devirtualize this either.)

Change-Id: I6a8809bcfebc9571f32e289fa4bc6a8b0d21ca46
Reviewed-on: https://go-review.googlesource.com/c/go/+/424774
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>

cmd/compile/internal/noder: set ir.Name.DictIndex for unified IR

For local variables of derived type, Delve relies on ir.Name.DictIndex
being set to the type's rtype index within the function's dictionary.
This CL implements that functionality within unified IR.

Manually double checked that Delve behaves correctly, at least as far
as I can tell from casual use. Specifically, I confirmed that running
the test program from TestDictIndex, stepping into testfn, and then
running `print mapvar` prints `map[int]main.CustomInt []`, which
matches the behavior under GOEXPERIMENT=nounified. (Also compare that
when ir.Name.DictIndex is *not* set by unified IR, `print mapvar`
instead prints `map[int]go.shape.int []`.)

Fixes #54514.

Change-Id: I90d443945895abfba04dc018f15e00217930091c
Reviewed-on: https://go-review.googlesource.com/c/go/+/424735
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

cmd/compile/internal/noder: pointer shaping for unified IR

This CL implements pointer shaping in unified IR, corresponding to the
existing pointer shaping implemented in the non-unified frontend.

For example, if `func F[T any]` is instantiated as both `F[*int]` and
`F[*string]`, we'll now generate a single `F[go.shape.*uint8]` shaped
function that can be used by both.

Fixes #54513.

Change-Id: I2cef5ae411919e6dc5bcb3cac912abecb4cd5218
Reviewed-on: https://go-review.googlesource.com/c/go/+/424734
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>

runtime: remove dead code and unnecessary checks for amd64

Use amd64 assembly header to remove unnecessary cpu flags checks
and dead code that is guaranteed to not be executed when compiling
for specific microarchitectures.

name                  old time/op  new time/op  delta
BytesCompare/1-12     3.88ns ± 1%  3.18ns ± 1%  -18.15%  (p=0.008 n=5+5)
BytesCompare/2-12     3.89ns ± 1%  3.21ns ± 2%  -17.66%  (p=0.008 n=5+5)
BytesCompare/4-12     3.89ns ± 0%  3.17ns ± 0%  -18.62%  (p=0.008 n=5+5)
BytesCompare/8-12     3.44ns ± 2%  3.39ns ± 1%   -1.36%  (p=0.008 n=5+5)
BytesCompare/16-12    3.40ns ± 1%  3.14ns ± 0%   -7.77%  (p=0.008 n=5+5)
BytesCompare/32-12    3.90ns ± 1%  3.65ns ± 0%   -6.19%  (p=0.008 n=5+5)
BytesCompare/64-12    4.96ns ± 1%  4.71ns ± 2%   -4.98%  (p=0.008 n=5+5)
BytesCompare/128-12   6.42ns ± 0%  5.99ns ± 4%   -6.75%  (p=0.008 n=5+5)
BytesCompare/256-12   9.36ns ± 0%  7.40ns ± 0%  -20.97%  (p=0.008 n=5+5)
BytesCompare/512-12   15.9ns ± 1%  11.4ns ± 1%  -28.36%  (p=0.008 n=5+5)
BytesCompare/1024-12  27.0ns ± 0%  19.3ns ± 0%  -28.36%  (p=0.008 n=5+5)
BytesCompare/2048-12  50.2ns ± 0%  43.3ns ± 0%  -13.71%  (p=0.008 n=5+5)
[Geo mean]            7.13ns       6.07ns       -14.86%

name                old speed      new speed      delta
Count/10-12          723MB/s ± 0%   704MB/s ± 1%  -2.73%  (p=0.008 n=5+5)
Count/32-12         2.21GB/s ± 0%  2.12GB/s ± 2%  -4.21%  (p=0.008 n=5+5)
Count/4K-12         1.03GB/s ± 0%  1.03GB/s ± 1%    ~     (p=1.000 n=5+5)
Count/4M-12         1.04GB/s ± 0%  1.02GB/s ± 2%    ~     (p=0.310 n=5+5)
Count/64M-12        1.02GB/s ± 0%  1.01GB/s ± 1%  -1.00%  (p=0.016 n=5+5)
CountEasy/10-12      779MB/s ± 0%   768MB/s ± 1%  -1.48%  (p=0.008 n=5+5)
CountEasy/32-12     2.15GB/s ± 0%  2.09GB/s ± 1%  -2.71%  (p=0.008 n=5+5)
CountEasy/4K-12     45.1GB/s ± 1%  45.2GB/s ± 1%    ~     (p=0.421 n=5+5)
CountEasy/4M-12     36.4GB/s ± 1%  36.5GB/s ± 1%    ~     (p=0.690 n=5+5)
CountEasy/64M-12    16.1GB/s ± 2%  16.4GB/s ± 1%    ~     (p=0.056 n=5+5)
CountSingle/10-12   2.15GB/s ± 2%  2.22GB/s ± 1%  +3.37%  (p=0.008 n=5+5)
CountSingle/32-12   5.86GB/s ± 1%  5.76GB/s ± 1%  -1.55%  (p=0.008 n=5+5)
CountSingle/4K-12   54.6GB/s ± 1%  55.0GB/s ± 1%    ~     (p=0.548 n=5+5)
CountSingle/4M-12   45.9GB/s ± 4%  46.4GB/s ± 2%    ~     (p=0.548 n=5+5)
CountSingle/64M-12  17.3GB/s ± 1%  17.2GB/s ± 2%    ~     (p=1.000 n=5+5)
[Geo mean]          5.11GB/s       5.08GB/s       -0.53%

name          old speed      new speed      delta
Equal/1-12     200MB/s ± 0%   188MB/s ± 1%   -6.11%  (p=0.008 n=5+5)
Equal/6-12    1.20GB/s ± 0%  1.13GB/s ± 1%   -6.38%  (p=0.008 n=5+5)
Equal/9-12    1.67GB/s ± 3%  1.74GB/s ± 1%   +3.83%  (p=0.008 n=5+5)
Equal/15-12   2.82GB/s ± 1%  2.89GB/s ± 1%   +2.63%  (p=0.008 n=5+5)
Equal/16-12   2.96GB/s ± 1%  3.08GB/s ± 1%   +3.95%  (p=0.008 n=5+5)
Equal/20-12   3.33GB/s ± 1%  3.54GB/s ± 1%   +6.36%  (p=0.008 n=5+5)
Equal/32-12   4.57GB/s ± 0%  5.26GB/s ± 1%  +15.09%  (p=0.008 n=5+5)
Equal/4K-12   62.0GB/s ± 1%  65.9GB/s ± 2%   +6.29%  (p=0.008 n=5+5)
Equal/4M-12   23.6GB/s ± 2%  24.8GB/s ± 4%   +5.43%  (p=0.008 n=5+5)
Equal/64M-12  11.1GB/s ± 2%  11.3GB/s ± 1%   +1.69%  (p=0.008 n=5+5)
[Geo mean]    3.91GB/s       4.03GB/s        +3.11%

name                              old speed      new speed      delta
IndexByte/10-12                   2.64GB/s ± 0%  2.69GB/s ± 0%   +1.67%  (p=0.008 n=5+5)
IndexByte/32-12                   6.79GB/s ± 0%  6.27GB/s ± 0%   -7.57%  (p=0.008 n=5+5)
IndexByte/4K-12                   56.2GB/s ± 0%  56.9GB/s ± 0%   +1.27%  (p=0.008 n=5+5)
IndexByte/4M-12                   40.1GB/s ± 1%  41.7GB/s ± 1%   +4.05%  (p=0.008 n=5+5)
IndexByte/64M-12                  17.5GB/s ± 0%  17.7GB/s ± 1%     ~     (p=0.095 n=5+5)
IndexBytePortable/10-12           2.06GB/s ± 1%  2.16GB/s ± 1%   +5.08%  (p=0.008 n=5+5)
IndexBytePortable/32-12           1.40GB/s ± 1%  1.54GB/s ± 1%  +10.05%  (p=0.008 n=5+5)
IndexBytePortable/4K-12           3.99GB/s ± 0%  4.08GB/s ± 0%   +2.16%  (p=0.008 n=5+5)
IndexBytePortable/4M-12           4.05GB/s ± 1%  4.08GB/s ± 2%     ~     (p=0.095 n=5+5)
IndexBytePortable/64M-12          3.80GB/s ± 1%  3.81GB/s ± 0%     ~     (p=0.421 n=5+5)
IndexRune/10-12                    746MB/s ± 1%   752MB/s ± 0%   +0.85%  (p=0.008 n=5+5)
IndexRune/32-12                   2.33GB/s ± 0%  2.42GB/s ± 0%   +3.66%  (p=0.008 n=5+5)
IndexRune/4K-12                   44.4GB/s ± 0%  44.2GB/s ± 0%     ~     (p=0.095 n=5+5)
IndexRune/4M-12                   36.2GB/s ± 1%  36.3GB/s ± 2%     ~     (p=0.841 n=5+5)
IndexRune/64M-12                  16.2GB/s ± 2%  16.3GB/s ± 2%     ~     (p=0.548 n=5+5)
IndexRuneASCII/10-12              2.57GB/s ± 0%  2.58GB/s ± 0%   +0.63%  (p=0.008 n=5+5)
IndexRuneASCII/32-12              6.00GB/s ± 0%  6.30GB/s ± 1%   +4.98%  (p=0.008 n=5+5)
IndexRuneASCII/4K-12              56.7GB/s ± 0%  56.8GB/s ± 1%     ~     (p=0.151 n=5+5)
IndexRuneASCII/4M-12              41.6GB/s ± 1%  41.7GB/s ± 2%     ~     (p=0.151 n=5+5)
IndexRuneASCII/64M-12             17.7GB/s ± 1%  17.6GB/s ± 1%     ~     (p=0.222 n=5+5)
Index/10-12                       1.06GB/s ± 1%  1.06GB/s ± 0%     ~     (p=0.310 n=5+5)
Index/32-12                       3.57GB/s ± 0%  3.56GB/s ± 1%     ~     (p=0.056 n=5+5)
Index/4K-12                       1.02GB/s ± 2%  1.03GB/s ± 0%     ~     (p=0.690 n=5+5)
Index/4M-12                       1.04GB/s ± 0%  1.03GB/s ± 1%     ~     (p=1.000 n=4+5)
Index/64M-12                      1.02GB/s ± 0%  1.02GB/s ± 0%     ~     (p=0.905 n=5+4)
IndexEasy/10-12                   1.12GB/s ± 2%  1.15GB/s ± 1%   +3.10%  (p=0.008 n=5+5)
IndexEasy/32-12                   3.14GB/s ± 2%  3.13GB/s ± 1%     ~     (p=0.310 n=5+5)
IndexEasy/4K-12                   47.6GB/s ± 1%  47.7GB/s ± 2%     ~     (p=0.310 n=5+5)
IndexEasy/4M-12                   36.4GB/s ± 1%  36.3GB/s ± 2%     ~     (p=0.690 n=5+5)
IndexEasy/64M-12                  16.1GB/s ± 1%  16.4GB/s ± 5%     ~     (p=0.151 n=5+5)
[Geo mean]                        6.39GB/s       6.46GB/s        +1.11%

Change-Id: Ic1ca62f5cc719d87e2c4aeff25ad73507facff82
Reviewed-on: https://go-review.googlesource.com/c/go/+/397576
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>

runtime: convert timer0When/timerModifiedEarliest to atomic.Int64

So they match with when/nextwhen fields of timer struct.

Updates #53821

Change-Id: Iad0cceb129796745774facfbbfe5756df3a320b4
Reviewed-on: https://go-review.googlesource.com/c/go/+/423117
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>

cmd/dist: enable testsanitizer test on ppc64le

Even though the -race option works for ppc64le, some of the
testsanitizer test have failed in the past on our builders. These
same failures can't be reproduced on other systems.

This is an experiment to re-enable this test on ppc64le to see if
it still fails on the builders.

Updates #45040

Change-Id: I0729bec5864e6c4cc752968485e89179df027063
Reviewed-on: https://go-review.googlesource.com/c/go/+/424534
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Archana Ravindar <aravind5@in.ibm.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

go/printer: avoid allocating for every position printed

printer.print is an overloaded method for multiple purposes.
When fed a position, it updates the current position.
When fed a string, it prints the string.
When fed a token, it prints the token. And so on.

However, this overloading comes at a significant cost.
Because the parameters are a list of the `any` interface type,
any type which is not of pointer or interface kind will allocate when
passed as an argument, as interfaces can only contain pointers.

A large portion of the arguments passed to the print method are of type
token.Pos, whose underlying type is int - so it allocates.
Removing those allocations has a significant benefit,
at the cost of some verbosity in the code:

name      old time/op    new time/op    delta
Print-16    6.10ms ± 2%    5.39ms ± 2%  -11.72%  (p=0.000 n=10+10)

name      old speed      new speed      delta
Print-16  8.50MB/s ± 2%  9.63MB/s ± 2%  +13.28%  (p=0.000 n=10+10)

name      old alloc/op   new alloc/op   delta
Print-16     443kB ± 0%     332kB ± 0%  -25.10%  (p=0.000 n=10+9)

name      old allocs/op  new allocs/op  delta
Print-16     17.3k ± 0%      3.5k ± 0%  -80.10%  (p=0.000 n=10+10)

There should be more significant speed-ups left, particularly for the
token.Token, string, and whiteSpace types fed to the same method.
They are left for a future CL, in case this kind of optimization is not
a path we want to take.

Change-Id: I3ff8387242c5a935bb003e60e0813b7b9c65402e
Reviewed-on: https://go-review.googlesource.com/c/go/+/412557
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: hopehook <hopehook@qq.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

hash/crc32: remove single-use {castagnoli,ieee}ArchImpl bools

These are only used once right after being set in castagnoliInit and
ieeeInit, respectively.

Change-Id: Ifc5df1fe3040b97b1d252cc6d6d28d8fbc132fa5
Reviewed-on: https://go-review.googlesource.com/c/go/+/423854
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>

cmd/internal/osinfo: use unix.ByteSliceToString

The golang.org/x/sys/unix package is already imported for Utsname and
Uname. Use ByteSliceToString from that package as well to replace the
locally defined utsString helper which serves the same purpose and
matches ByteSliceToString's implementation.

Change-Id: I5d9de186a5aeb1feed1387beedefbcd260fe22ff
Reviewed-on: https://go-review.googlesource.com/c/go/+/415654
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>

debug/pe: use saferio to set symbol slice capacity

No test case because the problem can only happen for invalid data.
Let the fuzzer find cases like this.

For #47653
Fixes #53530

Change-Id: If1cebbbcabb188fec8be30ef043c8c4c935a9564
Reviewed-on: https://go-review.googlesource.com/c/go/+/413995
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Alex Brainman <alex.brainman@gmail.com>
Reviewed-by: Bryan Mills <bcmills@google.com>

time: fix LoadLocation documentation formatting

The documentation for LoadLocation contains an enumerated list,
but does not render as such because it's missing leading spaces.

Output verified with the go doc command and godoc server.

Change-Id: I88b61d34048b7d01ee5cd77c32849af266e2f4c5
Reviewed-on: https://go-review.googlesource.com/c/go/+/423297
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>

internal/poll: remove fallback path in accept

Support for operating system versions requiring the fallback to
CloseOnExec/SetNonblock was dropped from recent Go versions. The minimum
Linux kernel version is 2.6.32 as of Go 1.18. FreeBSD 10 is no longer
supported as of Go 1.13.

Follows a similar change for net.sysSocket in CL 403634 and
syscall.Socket in CL 422374.

For #45964

Change-Id: I60848415742a1d8204e1fda585462ff35ad6722f
Reviewed-on: https://go-review.googlesource.com/c/go/+/422375
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>

syscall: implement EpollCreate using EpollCreate1 on all linux platforms

Most newer architectures (e.g. arm64, riscv64, loong64) don't provide the
epoll_create syscall. Some systems (e.g. Android) block it even if it
were available. In the kernel, the epoll_create syscall is implemented
[1] the same way EpollCreate is implemented in this package for
platforms without the epoll_create syscall. The epoll_create1 syscall is
available since Linux kernel 2.6.27 and the minimum required kernel
version is 2.6.32 since Go 1.18 (see #45964). Thus, avoid the separate
wrapper and consistently implement EpollCreate using EpollCreate1.

[1] https://elixir.bootlin.com/linux/v5.15-rc1/source/fs/eventpoll.c#L2006

The same change was already done in CL 349809 for golang.org/x/sys/unix.

For #45964

Change-Id: I5463b208aa7ae236fa2c175d6d3ec6568f1840b0
Reviewed-on: https://go-review.googlesource.com/c/go/+/411594
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>

cmd/go: avoid passing testing.T to isCaseSensitive and goVersion

The previous implementation of isCaseSensitive called t.Fatalf in the
wrong place, causing tests after the first to proceed past an error
determining case-sensitivity. That could lead to confusing errors.

(Moreover, I would like to try to disentangle the script engine from
testing.T so that I can also use it to generate serving contents in
the replacement for vcs-test.golang.org.)

The implementation of goVersion called ts.fatalf, which is probably
fine but prevents the script environment from being computed outside
of a test, as we might want to do for debugging and other scripting.

For #27494.

Change-Id: Ibfee0704523fdcd6174b544ff84267216435025b
Reviewed-on: https://go-review.googlesource.com/c/go/+/419874
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>

cmd/go: run tests when cmd/go is cross-compiled

When the GOOS or GOARCH of the cmd/go test binary does not match the
GOOS or GOARCH of the installed 'go' binary itself, the test currently
attempts to trick 'go test' into thinking that there were no test
functions to run.

That makes it very difficult to discover how to actually run the
tests, which in turn makes it difficult to diagnose and fix
regressions in, say, the linux-386-longtest builders. (We have had a
few of those lately, and they shouldn't be as much of an ordeal to fix
as they currently are.)

There are three underlying problems:

1. cmd/go uses its own GOOS and GOARCH to figure out which variant of
   other tools to use, and the cache keys for all installed tools and
   libraries include the IDs of the tools used to build them. So when
   cmd/go's GOARCH changes, all installed tools and binaries appear
   stale *even if* they were just installed by invoking the native
   cmd/go with the appropriate GOARCH value set.

2. The "go/build" library used by cmd/go toggles its default
   CGO_ENABLED behavior depending on whether the GOOS and GOARCH being
   imported match runtime.GOOS and runtime.GOARCH.

3. A handful of cmd/go tests explicitly use gccgo, but the user's
   installed gccgo binary cannot necessarily cross-compile to the same
   platforms as cmd/go.

To address the cache-invalidation problem, we modify the test variant
of cmd/go to use the host's native toolchain (as indicated by the new
TESTGO_GOHOSTOS and TESTGO_GOHOSTARCH environment variables) instead
of the toolchain matching the test binary itself. That allows a test
cmd/go binary compiled with GOARCH=386 to use libraries and tools
cross-compiled by the native toolchain, so that

$ GOARCH=386 go install std cmd

suffices to make the packages in std and cmd non-stale in the
tests.

To address the CGO_ENABLED mismatch, we set CGO_ENABLED explicitly in
the test's environment whenever it may differ from the default. Since
script tests that use cgo are already expected to use a [cgo]
condition, setting the environment to match that condition fixes the
cgo-specific tests.

To address the gccgo-specific cross-compilation failures, we add a new
script condition, [cross], which evaluates to true whenever the
platform of the test binary differs from that of the native toolchain.
We can then use that condition to explicitly skip the handful of gccgo
tests that fail under cross-compilation.

Fixes #53936.

Change-Id: I8633944f674eb5941ccc95df928991660e7e8137
Reviewed-on: https://go-review.googlesource.com/c/go/+/356611
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>

all: remove duplicate "the" words in comments

Following CL 424454, using command

   rg --multiline " the\s{1,}the " *
   rg --multiline " the\s{1,}//\s{1,}the " *

all the words "the" that are repeated in comments are found.

Change-Id: I60b769b98f04c927b4c228e10f37faf190964069
Reviewed-on: https://go-review.googlesource.com/c/go/+/423836
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>

cmd/compile/internal/noder: shape-based stenciling for unified IR

This CL switches unified IR to use shape-based stenciling with runtime
dictionaries, like the existing non-unified frontend. Specifically,
when instantiating generic functions and types `X[T]`, we now also
instantiated shaped variants `X[shapify(T)]` that can be shared by
`T`'s with common underlying types.

For example, for generic function `F`, `F[int](args...)` will be
rewritten to `F[go.shape.int](&.dict.F[int], args...)`.

For generic type `T` with method `M` and value `t` of type `T[int]`,
`t.M(args...)` will be rewritten to `T[go.shape.int].M(t,
&.dict.T[int], args...)`.

Two notable distinctions from the non-unified frontend:

1. For simplicity, currently shaping is limited to simply converting
type arguments to their underlying type. Subsequent CLs will implement
more aggressive shaping.

2. For generic types, a single dictionary is generated to be shared by
all methods, rather than separate dictionaries for each method. I
originally went with this design because I have an idea of changing
interface calls to pass the itab pointer via the closure
register (which should have zero overhead), and then the interface
wrappers for generic methods could use the *runtime.itab to find the
runtime dictionary that corresponds to the dynamic type. This would
allow emitting fewer method wrappers.

However, this choice does have the consequence that currently even if
a method is unused and its code is pruned by the linker, it may have
produced runtime dictionary entries that need to be kept alive anyway.

I'm open to changing this to generate per-method dictionaries, though
this would require changing the unified IR export data format; so it
would be best to make this decision before Go 1.20.

The other option is making the linker smarter about pruning unneeded
dictionary entries, like how it already prunes itab entries. For
example, the runtime dictionary for `T[int]` could have a `R_DICTTYPE`
meta-relocation against symbol `.dicttype.T[go.shape.int]` that
declares it's a dictionary associated with that type; and then each
method on `T[go.shape.T]` could have `R_DICTUSE` meta-relocations
against `.dicttype.T[go.shape.T]+offset` indicating which fields
within dictionaries of that type need to be preserved.

Change-Id: I369580b1d93d19640a4b5ecada4f6231adcce3fd
Reviewed-on: https://go-review.googlesource.com/c/go/+/421821
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

internal/poll: remove detection of buggy splice on old Linux versions

The splice syscall is buggy prior to Linux 2.6.29. CL 113999 added a
workaround to detect buggy versions and disable use of splice for these.
As of Go 1.18 the minumum Linux version is 2.6.32. Thus, a non-buggy
implementation of the splice syscall can be assumed.

For #45964
Fixes #54505

Change-Id: Ied3a3334da7a3f7fa1280b7c5b1dfb9030219336
Reviewed-on: https://go-review.googlesource.com/c/go/+/422979
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>

os: only add file descriptors which are set to non-blocking mode to the netpoller

Either ones where kind == kindNonBlock or those we've successfully called syscall.SetNonblock() on.
Restore blocking behavior if we detect an error registering with the netpoller and our flow was
successful in setting the inital syscall.SetNonblock().

Update #54100

Change-Id: I08934e4107c7fb36c15a7ca23ac880490b4df235
Reviewed-on: https://go-review.googlesource.com/c/go/+/420334
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Goutnik <dgoutnik@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Yuval Pavel Zholkover <paulzhol@gmail.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Auto-Submit: Ian Lance Taylor <iant@golang.org>

debug/buildinfo: implement for Plan 9 a.out

Plan 9 a.out was not implemented for debug/buildinfo, which
was causing test failures on Plan 9. This adds an implementation,
and causes the tests to pass.

Fixes #53949

Change-Id: I90a307ef9babf8cf381f8746d731cac2206b234a
Reviewed-on: https://go-review.googlesource.com/c/go/+/418014
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Benny Siegert <bsiegert@gmail.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>

runtime: convert profbuf.eof to atomic type

Updates #53821

Change-Id: I271faaedbf8b8efca5fc765496eaf45c94927edf
Reviewed-on: https://go-review.googlesource.com/c/go/+/423891
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>

runtime: convert profbuf.overflowTime to atomic type

Updates #53821

Change-Id: I916549d831f84d4f1439433aea6a61ff5301d80c
Reviewed-on: https://go-review.googlesource.com/c/go/+/423890
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>

runtime: convert profbuf.overflow to atomic type

Updates #53821

Change-Id: Ib9c3be3d25c836636a59fdfacc63ba75edd9e016
Reviewed-on: https://go-review.googlesource.com/c/go/+/423889
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>

runtime: convert m.cgoCallersUse to atomic type

Updates #53821

Change-Id: I99b01f8e91b798e73275635c8a63fcdc4a8df9f1
Reviewed-on: https://go-review.googlesource.com/c/go/+/423888
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>

runtime: delete repeated word

Change-Id: I0d7005481fc61b4c7f3286674c022334c8c0cae0
Reviewed-on: https://go-review.googlesource.com/c/go/+/424494
Reviewed-by: hopehook <hopehook@qq.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: hopehook <hopehook@qq.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>

runtime: convert p.timerModifiedEarliest to atomic type

Updates #53821

Change-Id: Iac0d7a3871d9e3ee0ba38ee7ab989faca9c89666
Reviewed-on: https://go-review.googlesource.com/c/go/+/424397
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>

crypto/internal/boring/bcache: make Cache type-safe using generics

Generics lets us write Cache[K, V] instead of using unsafe.Pointer,
which lets us remove all the uses of package unsafe around the
uses of the cache.

I tried to do Cache[*K, *V] instead of Cache[K, V] but that was not possible.

Change-Id: If3b54cf4c8d2a44879a5f343fd91ecff096537e9
Reviewed-on: https://go-review.googlesource.com/c/go/+/423357
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>

runtime: gofmt -w -s

Change-Id: I1226ff66fd0c64984939793eb8ef96c08d030fa1
Reviewed-on: https://go-review.googlesource.com/c/go/+/424399
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
Run-TryBot: hopehook <hopehook@qq.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

cmd/compile: gofmt -w -s

Change-Id: I4dad103d23121a21b04800ec157487fdf79f89a6
Reviewed-on: https://go-review.googlesource.com/c/go/+/424398
Run-TryBot: hopehook <hopehook@qq.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

run: set GOENV=off when running 'go tool dist env'

'go tool' sets environment variables, including the GOAMD64 value
from the user's go.env file.

'go tool dist test' then rebuilds and reinstalls the toolchain and
standard library based on those variables. It should not; instead, it
should test exactly the configuration installed by the make scripts.

Fixes #54084.

Change-Id: I7cc8a21cc1d8331e06d7b7c55b14d170f8e2faab
Reviewed-on: https://go-review.googlesource.com/c/go/+/420055
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>

runtime: refactor ARM VDSO call setup to helper

We have a very complex process to make VDSO calls on ARM. Create a
wrapper helper function which reduces duplication and allows for
additional calls from other packages.

vdsoCall has a few differences from the original code in
walltime/nanotime:

* It does not use R0-R3, as they are passed through as arguments to fn.
* It does not save g if g.m.gsignal.stack.lo is zero. This may occur if
it called at startup on g0 between assigning g0.m.gsignal and setting
its stack.

For #49182

Change-Id: I51aca514b4835b71142011341d2f09125334d30f
Reviewed-on: https://go-review.googlesource.com/c/go/+/362795
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

cmd/go: skip test cases that depend on gopkg.in

Updates #54503.

Change-Id: Ie13d028b09260d2d316c343b3ea812bb9cce4e6d
Reviewed-on: https://go-review.googlesource.com/c/go/+/424594
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>

crypto/subtle: add XORBytes

Export cipher.xorBytes as subtle.XORBytes, for proposal #53021,
to provide fast XOR to cryptography libraries outside crypto/cipher.

Along with the move, implement the alignment check TODO
in xor_generic.go, so that systems with neither unaligned
accesses nor custom assembly can still XOR a word at a time
in word-based algorithms like GCM. This removes the need
for the separate cipher.xorWords.

Fixes #53021.

Change-Id: I58f80a922f1cff671b5ebc6168eb046e702b5a4c
Reviewed-on: https://go-review.googlesource.com/c/go/+/421435
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Alan Donovan <adonovan@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>

crypto/internal/subtle: rename to crypto/internal/alias

This avoids an import conflict with crypto/subtle.
CL 424175 does the same for x/crypto.

Change-Id: Id4a319b3283b8affaaf769062388325b31fe1715
Reviewed-on: https://go-review.googlesource.com/c/go/+/424194
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>

net: revert "reenable TestLookupLongTXT"

Test is still flaky.

For #22857

Change-Id: Ic0d979778eb4c2d3779b18a983e7077789ae08a4
Reviewed-on: https://go-review.googlesource.com/c/go/+/424535
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>

runtime: convert p.timer0When to atomic type

Updates #53821

Change-Id: I523ec61116d290ecf7b7e3eb96e468695766cb4d
Reviewed-on: https://go-review.googlesource.com/c/go/+/424396
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

runtime: convert m.preemptGen to atomic type

Updates #53821

Change-Id: I134dac3b1eb35f2da00e5ef8f4c264f08d4f65b6
Reviewed-on: https://go-review.googlesource.com/c/go/+/423887
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>

runtime: convert freezing to atomic type

Updates #53821

Change-Id: I77fcdb972b8920e1fb42248ce5bd2c3d2d0bd27e
Reviewed-on: https://go-review.googlesource.com/c/go/+/423885
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>

runtime/trace: convert tracing.enabled to atomic type

Updates #53821

Change-Id: I8a063ae94568cd2ea65c2e891618069a96139891
Reviewed-on: https://go-review.googlesource.com/c/go/+/423884
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>

runtime: convert netpoll netpollInited to atomic type

Updates #53821

Change-Id: Ifa2e5f5d4047117b1887c1e56851355547bb4f33
Reviewed-on: https://go-review.googlesource.com/c/go/+/423881
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>

runtime: convert kqueue netpollWakeSig to atomic type

Updates #53821

Change-Id: I85d7444be36967e1e7e0ff2ce2f19b73581ecdde
Reviewed-on: https://go-review.googlesource.com/c/go/+/423880
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>

runtime: convert windows netpollWakeSig to atomic type

Updates #53821

Change-Id: Ifc5bc551a4bcf82bf10d2e670ad2644eb9c6b334
Reviewed-on: https://go-review.googlesource.com/c/go/+/423879
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>

runtime: convert solaris netpollWakeSig to atomic type

Updates #53821

Change-Id: Ic2799c125267dc5b13b265db41fbe8bf7c08b8a5
Reviewed-on: https://go-review.googlesource.com/c/go/+/423878
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>

runtime: convert aix netpollWakeSig to atomic type

Updates #53821

Change-Id: Ic073871ed2638ca22e6cb057dd8297f27582e78f
Reviewed-on: https://go-review.googlesource.com/c/go/+/423877
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>

runtime: convert linux netpollWakeSig to atomic type

Updates #53821

Change-Id: If4090393a127c2f468c8ae5ba478a9f59d73b945
Reviewed-on: https://go-review.googlesource.com/c/go/+/423876
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

net/http: return ErrNoCookie from Request.Cookie when name is ""

Request.Cookie(name string) will return the first cookie
when cookie name is "". Since readCookies in
file net/http/cookie.go at line 247 return all cookies
when second parameter is a empty string.

To fix it, Return ErrNoCookie from Request.Cookie(""),
instead of the first cookie in the request.

Fixes #53181

Change-Id: Ie623ca4c53da64ef7623a7863292a2d771f76832
GitHub-Last-Rev: 01098cd5dbcc8d65e9c0893e0586601584e5f8b9
GitHub-Pull-Request: golang/go#53183
Reviewed-on: https://go-review.googlesource.com/c/go/+/409754
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Damien Neil <dneil@google.com>

net: reenable TestLookupLongTXT

This test was disabled for flakiness.
The underlying code has been rewritten.
Reenabling the test to see if it has been fixed.

Fixes #22857.

Change-Id: If488ee5f4692cdf4ed534b8f08b4a54b80a60943
Reviewed-on: https://go-review.googlesource.com/c/go/+/423454
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Damien Neil <dneil@google.com>

runtime: convert m.signalPending to atomic type

Updates #53821

Change-Id: Ib455be9ca7120ded7c77d34556eff977aa61faa3
Reviewed-on: https://go-review.googlesource.com/c/go/+/423886
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

runtime: convert mSpanStateBox.s to atomic type

Updates #53821

Change-Id: I02f31a7a8295deb3e840565412abf10ff776c2c3
Reviewed-on: https://go-review.googlesource.com/c/go/+/424395
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Keith Randall <khr@google.com>

runtime: convert g.parkingOnChan to atomic type

Updates #53821

Change-Id: I54de39b984984fb3c160aba5afacb90131fd47c4
Reviewed-on: https://go-review.googlesource.com/c/go/+/424394
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>

cmd/go: propagate match errors in 'go run'

Fixes #51604.

Change-Id: I3bc86652c62d2b329d9c2db5ea443d56cf17f8d1
Reviewed-on: https://go-review.googlesource.com/c/go/+/418094
Reviewed-by: Nooras Saba‎ <saba@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>

internal/testenv: add a test for the GoTool function

GoTool was added in CL 20967, and revised in CL 21292, for #14901.

I don't fully understand what problem the GoTool function was added to
solve: the discussion on that issue was pretty sparse, but it seems
like when we run tests of GOROOT packages they always know their own
location relative to GOROOT (and thus always know where to find the
'go' tool).

Lacking that understanding, I don't want to change its behavior, but I
do at least want to verify that it resolves to the real 'go' tool in
the common case (running 'go test' on a package in GOROOT/src).

For #50892
For #50893
Updates #14901

Change-Id: I06d831e6765be631dfc4854d7fddc3d27fc1de34
Reviewed-on: https://go-review.googlesource.com/c/go/+/381834
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

test/codegen: updated multiple tests to verify on ppc64,ppc64le

Updated multiple tests in test/codegen: math.go, mathbits.go, shift.go
and slices.go to verify on ppc64/ppc64le as well

Change-Id: Id88dd41569b7097819fb4d451b615f69cf7f7a94
Reviewed-on: https://go-review.googlesource.com/c/go/+/412115
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Archana Ravindar <aravind5@in.ibm.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Paul Murphy <murp@ibm.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>

time: optimize GoString

Optimize Time.GoString by avoiding multiple calls to absDate.

name        old time/op    new time/op    delta
GoString-8     313ns ± 2%     197ns ± 1%  -37.08%  (p=0.008 n=5+5)

name        old alloc/op   new alloc/op   delta
GoString-8     80.0B ± 0%     80.0B ± 0%     ~     (all equal)

name        old allocs/op  new allocs/op  delta
GoString-8      1.00 ± 0%      1.00 ± 0%     ~     (all equal)

Fixes #54436

Change-Id: I8e6f8e7bbb9857b4bc0cdf6ed29a6b2415775db7
Reviewed-on: https://go-review.googlesource.com/c/go/+/423634
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Joseph Tsai <joetsai@digital-static.net>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>

all: use "noopt" build tag for checking optimization disabled

Fixes #49390

Change-Id: Ie5a5e097635c9fdcf4509455007283009a7d3021
Reviewed-on: https://go-review.googlesource.com/c/go/+/423256
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

runtime: add symbolic constants for freebsd/amd64 syscalls

Brings freebsd/amd64 on par with other OSes/archs.

Updates #53479

Change-Id: I05164c887d776dd4a0ae51eafd4014042b782999
Reviewed-on: https://go-review.googlesource.com/c/go/+/413366
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>

cmd/go: add go generate -skip flag

Following proposal discussion in #38687, add go generate -skip
to allow easier skipping of specific //go:generate directives.

Fixes #38687.

Change-Id: Ied5b4042965dd6a2b93c1c517045fccae2d95c3f
Reviewed-on: https://go-review.googlesource.com/c/go/+/421440
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Rob Pike <r@golang.org>

runtime: add symbolic constants for freebsd/386 syscalls

Brings freebsd/386 on par with other OSes/archs.

Updates #53479

Change-Id: I16a7d051233e06e589b19919e68f07b6bf92f73c
Reviewed-on: https://go-review.googlesource.com/c/go/+/413365
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>

encoding/base64: optimize decodemap memory set

The existing implementation has an execution time higher in the benchmark than this one.
This is an optimized implementation using the copy() function and a constant 256 bytes string with the values to be copied.

```
name           old time/op    new time/op    delta
NewEncoding-4     329ns ± 1%     231ns ± 1%  -29.72%  (p=0.008 n=5+5)

name           old speed      new speed      delta
NewEncoding-4   778MB/s ± 1%  1108MB/s ± 1%  +42.29%  (p=0.008 n=5+5)
```

Fixes #53211

Change-Id: I80fe62aa40623125ef81ae9164a8405eed30b71b
GitHub-Last-Rev: 55dce6f636ad29a70daa935169710c22f44ab31f
GitHub-Pull-Request: golang/go#53212
Reviewed-on: https://go-review.googlesource.com/c/go/+/410194
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Auto-Submit: Keith Randall <khr@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>

internal/xcoff: use saferio to read string table

No test case because the problem can only happen for invalid data.
Let the fuzzer find cases like this.

For #47653
Fixes #52526

Change-Id: Id90a5e391a4f684f45f8d7e32608eb03b8081076
Reviewed-on: https://go-review.googlesource.com/c/go/+/413875
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>

debug/macho, internal/saferio: limit slice allocation

Don't allocate slices that are too large; choose a smaller capacity
and build the slice using append. Use this in debug/macho to avoid
over-allocating if a fat header is incorrect.

No debug/macho test case because the problem can only happen for
invalid data. Let the fuzzer find cases like this.

For #47653
Fixes #52523

Change-Id: I372c9cdbdda8626a3225e79d713650beb350ebc7
Reviewed-on: https://go-review.googlesource.com/c/go/+/413874
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Tobias Klauser <tobias.klauser@gmail.com>

archive/tar: delete unreachable continue code

Change-Id: Id492ee4e614a38880a6a5830371dcd9a8b37129a
Reviewed-on: https://go-review.googlesource.com/c/go/+/422214
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Joseph Tsai <joetsai@digital-static.net>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Run-TryBot: hopehook <hopehook@qq.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: hopehook <hopehook@qq.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>

debug/pe, internal/saferio: use saferio to read PE section data

For #47653
Fixes #53189

Change-Id: If35b968fc53e4c96b18964cfb020cdc003b881bf
Reviewed-on: https://go-review.googlesource.com/c/go/+/412014
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Alex Brainman <alex.brainman@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>

all: clean up unreachable Continue after Fatal

As CL 422214 did, this CL intends to clean up the rest
unreachable "Continue" after Fatal.

Change-Id: I3b7e1b59bdfccb185e20525ce113e241d277dad3
Reviewed-on: https://go-review.googlesource.com/c/go/+/422514
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: hopehook <hopehook@qq.com>

runtime: process ptr bitmaps one word at a time

[This is a retry of CL 407036 + its revert CL 422394. The only
content change is the 1-line change in cmd/internal/obj/objfile.go.]

Read the bitmaps one uintptr at a time instead of one byte at a time.

Performance so far:
 Allocation heavy, no retention: ~30% faster in heapBitsSetType
 Scan heavy, ~no allocation: ~even in scanobject

Change-Id: I04d899e1dbd23e989e9f552cdc1880318779c14c
Reviewed-on: https://go-review.googlesource.com/c/go/+/422635
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>

runtime: redo heap bitmap

[this is a retry of CL 407035 + its revert CL 422395. The content is unchanged]

Use just 1 bit per word to record the ptr/nonptr bitmap.
Use word-sized operations to manipulate the bitmap, so we can operate
on up to 64 ptr/nonptr bits at a time.

Use a separate bitmap, one bit per word of the ptr/nonptr bitmap,
to encode a no-more-pointers signal. Since we can check 64 ptr/nonptr
bits at once, knowing the exact last pointer location is not necessary.

As a followon CL, we should make the gcdata bitmap an array of
uintptr instead of an array of byte, so we can load 64 bits of it at once.
Similarly for the processing of gc programs.

Change-Id: Ica5eb622f5b87e647be64f471d67b02732ef8be6
Reviewed-on: https://go-review.googlesource.com/c/go/+/422634
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@golang.org>

net/http/httputil: add ReverseProxy.Rewrite

Add a new Rewrite hook to ReverseProxy, superseding the Director hook.

Director does not distinguish between the inbound and outbound request,
which makes it possible for headers added by Director to be inadvertently
removed before forwarding if they are listed in the inbound request's
Connection header. Rewrite accepts a value containing the inbound
and outbound requests, with hop-by-hop headers already removed from
the outbound request, avoiding this problem.

ReverseProxy's appends the client IP to the inbound X-Forwarded-For
header by default. Users must manually delete untrusted X-Forwarded-For
values. When used with a Rewrite hook, ReverseProxy now strips
X-Forwarded-* headers by default.

NewSingleHostReverseProxy creates a proxy that does not rewrite the
Host header of inbound requests. Changing this behavior is
cumbersome, as it requires wrapping the Director function created
by NewSingleHostReverseProxy. The Rewrite hook's ProxyRequest
parameter provides a SetURL method that provides equivalent
functionality to NewSingleHostReverseProxy, rewrites the Host
header by default, and can be more easily extended with additional
customizations.

Fixes #28168.
Fixes #50580.
Fixes #53002.

Change-Id: Ib84e2fdd1d52c610e3887af66f517d4a74e594d0
Reviewed-on: https://go-review.googlesource.com/c/go/+/407214
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>

strings, bytes: add CutPrefix and CutSuffix

Fixes #42537

Change-Id: Ie03c2614ffee30ebe707acad6b9f6c28fb134a45
Reviewed-on: https://go-review.googlesource.com/c/go/+/407176
Reviewed-by: Benny Siegert <bsiegert@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Changkun Ou <mail@changkun.de>
Reviewed-by: Ian Lance Taylor <iant@google.com>

all: fix some typos

Change-Id: I147622770587cff1551a0be189003666463f34e3
GitHub-Last-Rev: 71b5375983820445a331c46481a2a3a44fafb760
GitHub-Pull-Request: golang/go#54472
Reviewed-on: https://go-review.googlesource.com/c/go/+/424154
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
Auto-Submit: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>

cmd/go: stamp VCS information in test binaries when -buildvcs=true

(But still not when -buildvcs=auto, the default.)

Fixes #52648.

Change-Id: I87a87d4ea84e8bf9635a4f7c8674c9311c3e21be
Reviewed-on: https://go-review.googlesource.com/c/go/+/409177
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>

go/token: add (*FileSet).RemoveFile(*File) method

The design of FileSet encourages it to be used as a global variable.
Each call to AddFile consumes about 3KB, that is never returned,
even after an application no longer cares about the File.
This change adds a RemoveFile method that a long-running application
can use to release a File that is no longer needed, saving memory.

Fixes golang/go#53200

Change-Id: Ifd34d650fe0d18b1395f922a4cd02a535afbe560
Reviewed-on: https://go-review.googlesource.com/c/go/+/410114
Auto-Submit: Alan Donovan <adonovan@google.com>
Run-TryBot: Alan Donovan <adonovan@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>

cmd/go: restore TestLegacyMod as a script test

This test was apparently mistakenly removed without a replacement in
CL 213223, but its testdata was left in the tree. This change removes
the orphaned testdata subdirectory, and restores the test that
previously used that data as a self-contained script.

For #27494.

Change-Id: Ice81895a44c558aaab198b8ef7ec046d92f5d58f
Reviewed-on: https://go-review.googlesource.com/c/go/+/417658
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

cmd/go: convert TestAccidentalGitCheckout to a script test

As far as I can determine, this is the only existing non-script test
that relies on vcs-test.golang.org. I am attempting to eliminate the
vcs-test server, and to keep configuration straightforward I would
like to only set up its replacement for script tests.

For #27494.

Change-Id: Ib978228b9c3f403af5f1c445e08942cbe7915b62
Reviewed-on: https://go-review.googlesource.com/c/go/+/418103
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>

testing: do not print 'root' benchName during test of package testing

$ cd $GOROOT/src/testing
$ go test
root
root
root
root
PASS
$

The root prints have been happening since Go 1.14.
There is a test in sub_test.go that calls b.Run directly
with a benchmark named "root", which triggers the print.
Silence them.

Change-Id: I2f0c186f04c6139bc24fab0e91975fcf0a8e80fc
Reviewed-on: https://go-review.googlesource.com/c/go/+/421437
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

cmd/go/internal/mmap: use syscall.Mmap on solaris

Now that syscall.Mmap is defined on solaris (see CL 413374), use it in
mmapFile like on other Unix ports.

For #52875

Change-Id: Ic5c5a84da8613f0c6dc947a52b7fcca50af43d79
Reviewed-on: https://go-review.googlesource.com/c/go/+/413375
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>

test: add regression test for issue 43942

CL 356011 fixed this issue too.

Fixes #43942

Change-Id: I35ad397e78eeb80eff3a4217c4d40f15d40bdebb
Reviewed-on: https://go-review.googlesource.com/c/go/+/423814
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

go/types, types2: check integer constant literal overflow

Fixes #54280.

Change-Id: I44a31daaace50bc90c96cd36387bd1a009d6a287
Reviewed-on: https://go-review.googlesource.com/c/go/+/424055
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

go/types, types2: skip comparison for operands with invalid types

Fixes #54405.

Change-Id: Ia7b2709b83966fa080e41e3d4818527d1e8b49f9
Reviewed-on: https://go-review.googlesource.com/c/go/+/424054
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>

net/http: make Request.WithContext documentation less prescriptive

WithContext makes a shallow copy of a Request, and Clone makes a
deep copy. Both set the context of the new request. The distinction
between the two is clear, and it doesn't seem useful or necessary
to say that "it's rare to need WithContext".

Also update a couple locations that mention WithContext to mention
Clone as well.

Fixes #53413.

Change-Id: I89e6ddebd7d5ca6573e522fe48cd7f50cc645cdd
Reviewed-on: https://go-review.googlesource.com/c/go/+/412778
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>

go/types, types2: don't build unnecessary error strings in implements

When accessing (*Checker).implements from types.AssignableTo or
types.ComparableTo, we don't need to build error strings -- they won't
be used.

This string manipulation showed up as a hot spot in gopls completion,
which checks a lot of type predicates when searching for candidate
completions.

This CL yields the following results for gopls' completion benchmarks:

StructCompletion-8         24.7ms ±34%  26.0ms ±17%     ~     (p=0.447 n=10+9)
ImportCompletion-8         1.41ms ± 2%  1.45ms ± 4%   +2.42%  (p=0.027 n=8+9)
SliceCompletion-8          27.0ms ±18%  25.2ms ± 3%   -6.67%  (p=0.008 n=9+8)
FuncDeepCompletion-8       57.6ms ± 4%  22.4ms ± 4%  -61.18%  (p=0.000 n=8+9)
CompletionFollowingEdit-8   157ms ±13%   103ms ±15%  -34.70%  (p=0.000 n=10+10)

Notably, deep completion (which searches many candidates) is almost 3x
faster after this change.

Fixes #54172

Change-Id: If8303a411aed3a20bd91f7b61e346d703084166c
Reviewed-on: https://go-review.googlesource.com/c/go/+/423360
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>

bytes: add Clone function

The new Clone function returns a copy of b[:len(b)]
for the input byte slice b.
The result may have additional unused capacity.
Clone(nil) returns nil.

Fixes #45038

Change-Id: I0469a202d77a7b491f1341c08915d07ddd1f0300
Reviewed-on: https://go-review.googlesource.com/c/go/+/359675
Run-TryBot: Martin Möhrmann <martin@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Martin Möhrmann <moehrmann@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>