doc/go1.21: document macOS and Windows requirements

For #58645.
Fixes #57125.
Fixes #57003.
Fixes #57004.

Change-Id: I365929ca36aeb32e9ecd19f870e70869933ba6a0
Reviewed-on: https://go-review.googlesource.com/c/go/+/500239
Reviewed-by: Eli Bendersky <eliben@google.com>
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Bypass: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Carlos Amedee <carlos@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>

doc/go1.21: document that os.TempDir now uses GetTempPath2W

Updates #56899

Change-Id: Ibde69cd55c81ac0bb757b28b28d69463778dd117
Reviewed-on: https://go-review.googlesource.com/c/go/+/500255
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Quim Muntal <quimmuntal@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>

net: filter disabled interfaces in Windows DNS client

The Go DNS resolver on Windows should filter disabled
interfaces. Otherwise disabled TUN devices, VPNs will be also
considered as valid nameservers and finally timedout.

Fixes #56160

(Originally from Zhiyuan Zheng <zhzy0077@hotmail.com>
 in https://go.dev/cl/442375)

Co-authored-by: Zhiyuan Zheng <zhzy0077@hotmail.com>
GitHub-Last-Rev: db158625bb46692cbbc09e3fe4dfa023c733f4e6
GitHub-Pull-Request: golang/go#56161
Change-Id: I7becebc55c8ac612c670c533855f7e6ca397a496
Reviewed-on: https://go-review.googlesource.com/c/go/+/500375
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
Auto-Submit: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>

go/types, types2: remove version check for more lenient constant handling in inference

For #58671.
Fixes #60566.

Change-Id: I746f99cdfd44b204dc90350fcfb3867e9b8b1da8
Reviewed-on: https://go-review.googlesource.com/c/go/+/499997
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

go/types, types2: report inference instead of assignment failure when types can't match

Interface inference must only be used if we don't require exact
unification, otherwise we may infer types (that are reasonable)
but then fail with an assignment error.

Only checking if exact is set for defined (named) types is not
sufficient, we must also check outside. Oversight.

Fixes #60562.

Change-Id: I208a74bf7ed80bcb976ba9cc172715c83f9e3d0a
Reviewed-on: https://go-review.googlesource.com/c/go/+/499996
Auto-Submit: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>

cmd/api: make check even pickier about api/next/*.txt

As spotted in CL 499981, 3 out of 51 of the api/next/*.txt files
ended up with a blank line at the end. It's possible it would've
been more if human reviewers didn't catch them.

Since there's no formatter for these files, the only way to help
catch things is to make the check pickier (as done in CL 431335).
It can be loosened to let in useful blank lines if needed in the
future.

Change-Id: Iae7ee8e782b32707c576150914539ac4cc0faec4
Reviewed-on: https://go-review.googlesource.com/c/go/+/500115
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
Run-TryBot: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>

math/big: rename Int.ToFloat64 to Float64

The "To" prefix was a relic of the first draft
that I failed to make consistent with the unprefixed
name used in the proposal. Fortunately iant spotted
it during the API audit.

Updates #56984
Updates #60560

Change-Id: Ifa6eeddf6dd5f0637c0568e383f9a4bef88b10f9
Reviewed-on: https://go-review.googlesource.com/c/go/+/500116
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Alan Donovan <adonovan@google.com>

cmd/go: return more errors from ReadModFile, loadModFile

Return more errors instead of base.Fatalf, so we can handle them
in the callers.

For #57001.

Change-Id: If3e63d3f64188148f5d750991f9cb1175790d89d
Reviewed-on: https://go-review.googlesource.com/c/go/+/499983
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>

doc/go1.21: crypto release notes

Updates #58645

Change-Id: Ib7e2baba41bb327d8fc466afb1e117fe2f22e1c9
Reviewed-on: https://go-review.googlesource.com/c/go/+/499637
Reviewed-by: Damien Neil <dneil@google.com>
Auto-Submit: Roland Shoemaker <roland@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Roland Shoemaker <roland@golang.org>

go/types, types2: use exact matching when unifying constraint method signatures

Fixes #60556.

Change-Id: I203a0bc79eff607654c3e8350d259e694cb035b2
Reviewed-on: https://go-review.googlesource.com/c/go/+/499995
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>

cmd/go: add gover.Max and gover.MaxToolchain

I've inlined these by hand enough times now.

For #57001.

Change-Id: Ia274bd444f12b07ce14eeb2b5a66546880db8f77
Reviewed-on: https://go-review.googlesource.com/c/go/+/499982
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>

cmd/go: add base.Error and base.Fatal

Many many places in the go command use

base.Errorf("go: %v", err)

or

base.Fatalf("go: %v", err)

Introduce Error(error) and Fatal(error) to do this
and update all call sites (global search and replace).

The new Error gives us the opportunity to unwrap
a multierror and add the go prefix to each line,
which is the motivation for this change.
(We want to start returning a multierror from LoadModFile
and LoadModGraph.)

For #57001.

Change-Id: I9613653b94808224146077c30d22f814d4e19eed
Reviewed-on: https://go-review.googlesource.com/c/go/+/499980
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>

cmd/go: move CreateWorkFile from modload to workcmd

In workcmd it can deal with automatic version switching.

For #57001.

Change-Id: I5027690cf744d6d73f87e837c76ea7083ed56aba
Reviewed-on: https://go-review.googlesource.com/c/go/+/499979
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>

cmd/go: move version constants from modload to gover

For #57001.

Change-Id: Ia76478b8eaa934b7e1dc1e9cd7fe8a2428fc291a
Reviewed-on: https://go-review.googlesource.com/c/go/+/499978
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>

api: promote next to go1.21

Change-Id: I30cbd14bd2b3903d7960c5d860d76ce7f3d4730a
Reviewed-on: https://go-review.googlesource.com/c/go/+/499981
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Dmitri Shuralyov <dmitshur@golang.org>

runtime: update skips for TestGdbBacktrace

One issue simply has a reworded message, probably from a new version of
GDB. Another is a new issue.

Fixes #60553.
Fixes #58698.
Updates #39204.

Change-Id: I8389aa981fab5421f57ee761bfb5e1dd237709ed
Reviewed-on: https://go-review.googlesource.com/c/go/+/499975
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>

cmd/internal/obj/arm: handle HAUTO etc. in addpool

HAUTO should be handled the same way as other stack offsets for
adding to constant pool. Add the missing cases.

Fixes #57955.

Change-Id: If7fc82cafb2bbf0a6121e73e353b8825cb36b5bc
Reviewed-on: https://go-review.googlesource.com/c/go/+/463138
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>

cmd/internal/testdir: pass if GOEXPERIMENT=cgocheck2 is set

Some testdir tests fail if GOEXPERIMENT=cgocheck2 is set. Fix this by
skipping these tests.

Change-Id: I58d4ef0cceb86bcf93220b4a44de9b9dc4879b16
Reviewed-on: https://go-review.googlesource.com/c/go/+/499675
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>

cmd/cgo/internal/testerrors: pass if GOEXPERIMENT=cgocheck2 is set

The tests in this package are meant to check cgocheck and cgocheck2
mode, so they're of course sensitive to whether they're set.
Currently, the test will set GOEXPERIMENT=cgocheck2 for tests of
cgocheck2 mode, but won't *unset* cgocheck2 mode if it's already in
the environment for tests that expect it to be off. This means

  GOEXPERIMENT=cgocheck2 go test cmd/cgo/internal/testerrors

fails.

Fix this by removing cgocheck2 from GOEXPERIMENT if it's set and the
test case expects it to be unset.

Change-Id: If663e41b791fb89df9940bc5356a566e2a54a77a
Reviewed-on: https://go-review.googlesource.com/c/go/+/499557
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>

cmd/link/internal/ld: pass correct offset to codesign.Sign

Previously, codesign.Sign was called with Segtext.Fileoff and
Segtext.Filelen. However, both variables do not contain the
complete __TEXT segment, as it excludes padding and header.
Therefore, we now store a reference to the complete segment
in mstext when it is created and pass its offset (which should
always be 0) and filesize to codesign.Sign.

Fixes #59555

Change-Id: Iad88f142705949dcc0b192b811337df9b4be08cf
GitHub-Last-Rev: 37a048d58e01a3befa4db574e5fb48ab8d6013a1
GitHub-Pull-Request: golang/go#59581
Reviewed-on: https://go-review.googlesource.com/c/go/+/484015
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>

doc/go1.21: correct links in go/build note

Change-Id: Ie84cad994c1ffb2c91ce45b2dfb9257e6a7f309b
Reviewed-on: https://go-review.googlesource.com/c/go/+/499957
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
TryBot-Bypass: Ian Lance Taylor <iant@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>

doc/go1.21: fix io/fs.ReadDirFS link formatting

Change-Id: I348ac0c0e098f027b8d608a260ee9437d2b77539
Reviewed-on: https://go-review.googlesource.com/c/go/+/499835
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>

cmd/go: propagate gover.TooNewError from modload.LoadModGraph

For #57001.

Change-Id: I639190b5f035139ba42a93ca03dd8a4c747556ea
Reviewed-on: https://go-review.googlesource.com/c/go/+/499678
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>

doc: remove relnote TODO by merging with another section

This change was already described in the linker section.

Change-Id: I3e36dbd66b3b53c5f0063b0d67d8773383f1f77f
Reviewed-on: https://go-review.googlesource.com/c/go/+/499817
Auto-Submit: Eli Bendersky <eliben@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>
TryBot-Bypass: Eli Bendersky <eliben@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>

doc/go1.21: reformat flag init change note

It reads better as a single paragraph. Also added links.

Change-Id: Id6ba54111b63fbd536423272f674e0eb64b087a7
Reviewed-on: https://go-review.googlesource.com/c/go/+/499956
TryBot-Bypass: Ian Lance Taylor <iant@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>

doc/go1.21: remove some references to reverted CLs

Change-Id: Id9a8871af9cc3ac26be539877f2eea3462db6f3d
Reviewed-on: https://go-review.googlesource.com/c/go/+/499955
TryBot-Bypass: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>

doc/go1.21: fix type in closing tag

Change-Id: Ibcdedb8222a80c0478f783f4a1800e472dbe062c
Reviewed-on: https://go-review.googlesource.com/c/go/+/499935
Reviewed-by: Eli Bendersky <eliben@google.com>
TryBot-Bypass: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>

doc: add release note for go:wasmimport

Change-Id: I9c6fc4c7e24d0e78086d95dccd61e80bd5ddd856
Reviewed-on: https://go-review.googlesource.com/c/go/+/499815
Auto-Submit: Eli Bendersky‎ <eliben@golang.org>
Reviewed-by: Eli Bendersky <eliben@google.com>
Reviewed-by: Johan Brandhorst-Satzkorn <johan.brandhorst@gmail.com>
TryBot-Bypass: Eli Bendersky‎ <eliben@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>

cmd/go/internal/toolchain: avoid importing modcmd

modcmd is a high-level command, but toolchain is a low-level building
block. A dependency from toolchain on modcmd makes it very difficult
to call from other lower-level packages without creating an import
cycle.

Instead, use modfetch.Download in place of modcmd.DownloadModule.

For #57001.

Change-Id: I9694706d7225b269f26dc68814894613a3329abb
Reviewed-on: https://go-review.googlesource.com/c/go/+/499316
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>

cmd/go: fix TestScript/gotoolchain_net when $PATH contains golang.org/dl/go1.20.1

For #57001.

Change-Id: Idcc291c872864bcce7e6d411cdd208db3b331282
Reviewed-on: https://go-review.googlesource.com/c/go/+/499855
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Bypass: Bryan Mills <bcmills@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>

doc: add release note for cgo error on methods on C types

Change-Id: Ibc5f23f37e5bd45034771a0205eadac3072d6870
Reviewed-on: https://go-review.googlesource.com/c/go/+/499816
TryBot-Bypass: Eli Bendersky <eliben@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
Auto-Submit: Eli Bendersky <eliben@google.com>

runtime: only increment extraMInUse when actually in use

Currently lockextra always increments extraMInUse, even if the M won't
be used (or doesn't even exist), such as in addExtraM. addExtraM fails
to decrement extraMInUse, so it stays elevated forever.

Fix this bug and simplify the model by moving extraMInUse out of
lockextra to getExtraM, where we know the M will actually be used.

While we're here, remove the nilokay argument from getExtraM, which is
always false.

Fixes #60540.

Change-Id: I7a5d97456b3bc6ea1baeb06b5b2975e3b8dd96a0
Reviewed-on: https://go-review.googlesource.com/c/go/+/499677
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

doc/go1.21: document PGO GA and devirtualization

For #59959.
For #58645.

Change-Id: I574153ef2fd61a5e90ec281fca065c42fce22cc1
Reviewed-on: https://go-review.googlesource.com/c/go/+/498263
Reviewed-by: Eli Bendersky <eliben@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

doc/go1.21: document that os.DirFS implements fs.Read{File,Dir}FS

Also add a missing </a> in the preceding section.

For #53761

Change-Id: I8e64b86b5b32067f954d58cf9adf86cb4d2eeb2b
Reviewed-on: https://go-review.googlesource.com/c/go/+/499735
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Auto-Submit: Eli Bendersky <eliben@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>

cmd/go: delete cmd/go/internal/slices in favor of slices

Some files already use "slices", others use "cmd/go/internal/slices".
(Some files are using more than slices.Clip and must use "slices".)
Use "slices" consistently and delete cmd/go/internal/slices.

Change-Id: I69ec680106ad2924276f7473e6547a3a907efc96
Reviewed-on: https://go-review.googlesource.com/c/go/+/499715
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>

cmd/go: make toolchain less special in MVS

We were using the omission of toolchain from the MVS graph
as a signal that toolchain was not mentioned on the go get line,
but not including it in the graph causes various problems,
and it may be reintroduced to the graph during operations like
pruning conversion, after which its presence is not a good signal
about whether it was mentioned on the go get command line.
Fix all this irregularity by explicitly telling WriteGoMod whether
the command line mentioned toolchain instead.

For #57001.

Change-Id: I74084637c177c30918fdb114a0d9030cdee7324e
Reviewed-on: https://go-review.googlesource.com/c/go/+/499575
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>

go/types, types2: better error message for some inference failures

For a unification failure involving a constraint, rather than just
reporting (for instance)

        S does not match []E

now report the inferred type for the type parameter, use spec
terminology when referring to the constraint, and print the
constraint in full:

        S (type func()) does not satisfy ~[]E

There's more we can do, but this is better than what we had.

For #60542.

Change-Id: I033369fa0dfc475f0ec0da0582e8cbefb109f3cf
Reviewed-on: https://go-review.googlesource.com/c/go/+/499639
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>

doc/go1.21: remove math.Compare TODOs

Change-Id: I975b437240df03da84f2ce7c47f7a0097a597f25
Reviewed-on: https://go-review.googlesource.com/c/go/+/499638
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
TryBot-Bypass: Ian Lance Taylor <iant@google.com>

go1.21.html: describe testing/slogtest package

Change-Id: I734aae67d410073056388852cd94776c7483b4b5
Reviewed-on: https://go-review.googlesource.com/c/go/+/498273
Run-TryBot: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>
TryBot-Bypass: Jonathan Amsterdam <jba@google.com>

doc/go1.21: document type inference changes

For #39661.
For #41176.
For #51593.
For #52397.
For #57192.
For #58645.
For #58650.
For #58671.
For #59338.
For #59750.
For #60353.

Change-Id: Ib731c9f2879beb541f44cb10e40c36a8677d3ad4
Reviewed-on: https://go-review.googlesource.com/c/go/+/499282
TryBot-Bypass: Robert Griesemer <gri@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>

doc/go1.21: mention improvement to C-to-Go calls

For #51676.
For #58645.

Change-Id: I9045051b5a25c6dfc833eef13e6c105a0d8ae763
Reviewed-on: https://go-review.googlesource.com/c/go/+/499716
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

go1.21.html: remove log/slog package "changes"

Since log/slog doesn't appear in earlier Go versions, we don't need
to document these proposals as changes.

Change-Id: Iae6ed9ca6bd22d8c46c21deece9ddedf7d739032
Reviewed-on: https://go-review.googlesource.com/c/go/+/498272
Run-TryBot: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>
TryBot-Bypass: Jonathan Amsterdam <jba@google.com>

Revert "math: add Compare and Compare32"

This reverts CL 467515. Now that we have cmp.Compare,
we don't need math.Compare or math.Compare32 after all.

For #56491
Fixes #60519

Change-Id: I8ed33464adfc6d69bd6b328edb26aa2ee3d234d9
Reviewed-on: https://go-review.googlesource.com/c/go/+/499416
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Eli Bendersky <eliben@google.com>

cmd/link: support R_PPC64_REL24_P9NOTOC

This can be treated identically to R_PPC64_REL24_NOTOC as stubs
are generated based on GOPPC64 and -buildmode.

Change-Id: I3c42a9bf3c08d107fb656e6bf59b2307783bd9af
Reviewed-on: https://go-review.googlesource.com/c/go/+/497915
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Paul Murphy <murp@ibm.com>

doc/go1.21: fix link in WTF-8 section

Change-Id: I8fbba7ca5a148ff6523be412991af0c089e7f7ea
Reviewed-on: https://go-review.googlesource.com/c/go/+/499516
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

runtime: preserve R29 in the write barrier flush path on ppc64

Surprisingly, it usually survived the call to flush a write
barrier.  Usually.

Fixes #60368

Change-Id: I4792a57738e5829c79baebae4d13b62abe9526b6
Reviewed-on: https://go-review.googlesource.com/c/go/+/499679
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>

runtime: fix alignment code in memmove_riscv64.s

The riscv64 implementation of memmove has two optimizations that are
applied when both source and destination pointers share the same alignment
but that alignment is not 8 bytes. Both optimizations attempt to align the
source and destination pointers to 8 byte boundaries before performing 8 byte
aligned loads and stores.  Both optimizations are incorrect.  The first
optimization is applied when the destination pointer is smaller than the source
pointer.  In this case the code increments both pointers by (pointer & 3) bytes
rather than (8 - (pointer & 7)) bytes.  The second optimization is applied
when the destination pointer is larger than the source pointer.  In this
case the existing code decrements the pointers by (pointer & 3) bytes instead
of (pointer & 7).

This commit fixes both optimizations avoiding unaligned 8 byte accesses.
As this particular optimization is not covered by any of the existing
benchmarks a new benchmark, BenchmarkMemmoveUnalignedSrcDst,
is provided that exercises both optimizations. Results of the new
benchmark, which were run on a SiFive HiFive Unmatched A00 with 16GB of RAM
running Ubuntu 23.04 are presented below.

MemmoveUnalignedSrcDst/f_16_0-4            39.48n ±  5%   43.47n ± 2%  +10.13% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_16_0-4            45.39n ±  5%   41.55n ± 4%   -8.47% (p=0.000 n=10)
MemmoveUnalignedSrcDst/f_16_1-4          1230.50n ±  1%   83.44n ± 5%  -93.22% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_16_1-4            69.34n ±  4%   67.83n ± 8%        ~ (p=0.436 n=10)
MemmoveUnalignedSrcDst/f_16_4-4          2349.00n ±  1%   72.09n ± 4%  -96.93% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_16_4-4          2357.00n ±  0%   77.61n ± 4%  -96.71% (p=0.000 n=10)
MemmoveUnalignedSrcDst/f_16_7-4          1235.00n ±  0%   62.02n ± 2%  -94.98% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_16_7-4          1246.00n ±  0%   84.05n ± 6%  -93.25% (p=0.000 n=10)
MemmoveUnalignedSrcDst/f_64_0-4            49.96n ±  2%   50.01n ± 2%        ~ (p=0.755 n=10)
MemmoveUnalignedSrcDst/b_64_0-4            52.06n ±  3%   51.65n ± 3%        ~ (p=0.631 n=10)
MemmoveUnalignedSrcDst/f_64_1-4          8105.50n ±  0%   97.63n ± 1%  -98.80% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_64_1-4            84.07n ±  4%   84.90n ± 5%        ~ (p=0.315 n=10)
MemmoveUnalignedSrcDst/f_64_4-4          9192.00n ±  0%   86.16n ± 3%  -99.06% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_64_4-4          9195.50n ±  1%   91.88n ± 5%  -99.00% (p=0.000 n=10)
MemmoveUnalignedSrcDst/f_64_7-4          8106.50n ±  0%   78.44n ± 9%  -99.03% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_64_7-4          8107.00n ±  0%   99.19n ± 1%  -98.78% (p=0.000 n=10)
MemmoveUnalignedSrcDst/f_256_0-4           90.95n ±  1%   92.16n ± 8%        ~ (p=0.123 n=10)
MemmoveUnalignedSrcDst/b_256_0-4           96.09n ± 12%   94.90n ± 2%        ~ (p=0.143 n=10)
MemmoveUnalignedSrcDst/f_256_1-4         35492.5n ±  0%   133.5n ± 0%  -99.62% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_256_1-4           128.7n ±  1%   130.1n ± 1%   +1.13% (p=0.005 n=10)
MemmoveUnalignedSrcDst/f_256_4-4         36599.0n ±  0%   123.0n ± 1%  -99.66% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_256_4-4         36675.5n ±  0%   130.7n ± 1%  -99.64% (p=0.000 n=10)
MemmoveUnalignedSrcDst/f_256_7-4         35555.5n ±  0%   121.6n ± 2%  -99.66% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_256_7-4         35584.0n ±  0%   139.1n ± 1%  -99.61% (p=0.000 n=10)
MemmoveUnalignedSrcDst/f_4096_0-4          956.3n ±  2%   960.8n ± 1%        ~ (p=0.306 n=10)
MemmoveUnalignedSrcDst/b_4096_0-4          1.015µ ±  2%   1.012µ ± 2%        ~ (p=0.076 n=10)
MemmoveUnalignedSrcDst/f_4096_1-4        584.406µ ±  0%   1.002µ ± 1%  -99.83% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_4096_1-4          1.044µ ±  1%   1.040µ ± 2%        ~ (p=0.090 n=10)
MemmoveUnalignedSrcDst/f_4096_4-4       585113.5n ±  0%   988.6n ± 2%  -99.83% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_4096_4-4        586.521µ ±  0%   1.044µ ± 1%  -99.82% (p=0.000 n=10)
MemmoveUnalignedSrcDst/f_4096_7-4       585374.5n ±  0%   986.2n ± 0%  -99.83% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_4096_7-4        584.595µ ±  1%   1.055µ ± 0%  -99.82% (p=0.000 n=10)
MemmoveUnalignedSrcDst/f_65536_0-4         54.83µ ±  0%   55.00µ ± 0%   +0.31% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_65536_0-4         56.54µ ±  0%   56.64µ ± 0%   +0.19% (p=0.011 n=10)
MemmoveUnalignedSrcDst/f_65536_1-4       9450.51µ ±  0%   58.25µ ± 0%  -99.38% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_65536_1-4         56.65µ ±  0%   56.68µ ± 0%        ~ (p=0.353 n=10)
MemmoveUnalignedSrcDst/f_65536_4-4       9449.48µ ±  0%   58.24µ ± 0%  -99.38% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_65536_4-4       9462.91µ ±  0%   56.69µ ± 0%  -99.40% (p=0.000 n=10)
MemmoveUnalignedSrcDst/f_65536_7-4       9477.37µ ±  0%   58.26µ ± 0%  -99.39% (p=0.000 n=10)
MemmoveUnalignedSrcDst/b_65536_7-4       9467.96µ ±  0%   56.68µ ± 0%  -99.40% (p=0.000 n=10)
geomean                                    11.16µ         509.8n       -95.43%

Change-Id: Idfa1873b81fece3b2b1a0aed398fa5663cc73b83
Reviewed-on: https://go-review.googlesource.com/c/go/+/498377
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>

doc/go1.21: mention upgrade to Unicode 15.0.0

For #48621
For #55079

Change-Id: I279b4fbf07d2fa9de9577559a4e17a57f5692aa3
Reviewed-on: https://go-review.googlesource.com/c/go/+/499618
Reviewed-by: Eli Bendersky <eliben@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
TryBot-Bypass: Ian Lance Taylor <iant@google.com>

doc/go1.21: document new bytes.Buffer methods

For #53685

Change-Id: I237297d19afeb36ad738074d0c61caa7012f65ac
Reviewed-on: https://go-review.googlesource.com/c/go/+/499617
Reviewed-by: Eli Bendersky <eliben@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Bypass: Ian Lance Taylor <iant@google.com>

doc/go1.21: mention debug/elf changes

For #55107
For #56887
For #56892

Change-Id: Ibcca34c931ed3595d877c4573bba221dec66bb5a
Reviewed-on: https://go-review.googlesource.com/c/go/+/499616
Reviewed-by: Eli Bendersky <eliben@google.com>
TryBot-Bypass: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>

doc/go1.21: reading from debug/pe uninitialized section fails

For #47653

Change-Id: Id44c9dba58966f43f188030a53343d890a6ffde7
Reviewed-on: https://go-review.googlesource.com/c/go/+/499419
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Bypass: Ian Lance Taylor <iant@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>

doc/go1.21: mention that embedded files implement ReadAt

For #57803

Change-Id: I8e33f4dd3fc3071bfbf4d2848faefbe8488f5742
Reviewed-on: https://go-review.googlesource.com/c/go/+/499418
TryBot-Bypass: Ian Lance Taylor <iant@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>

doc/go1.21: mention flag.BoolFunc

For #53747

Change-Id: Ia5e2f89c1184f2dfd6d672b838b0dbb579e6c954
Reviewed-on: https://go-review.googlesource.com/c/go/+/499417
Reviewed-by: Eli Bendersky <eliben@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Bypass: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>

doc/go1.21: document io/fs formatting functions

Also document the new String methods that call them.

For #54451

Change-Id: I5cd7e0fc6c84097bba6d29c4d6012ed3c8bb1e0d
Reviewed-on: https://go-review.googlesource.com/c/go/+/499177
Reviewed-by: Eli Bendersky <eliben@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Bypass: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>

doc/go1.21: mention directive handling in go/{ast,build}

For #56986
For #59033

Change-Id: I7d03fe34d418aff97a551b236b5d43506e402871
Reviewed-on: https://go-review.googlesource.com/c/go/+/499415
TryBot-Bypass: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>

runtime: fix race in BenchmarkSetType* benchmarks

Currently the BenchmarkSetType* benchmarks are racy: they call
heapBitsSetType on an allocation that might be in a span in-use for
allocation on another P. Because heap bits are bits but are written
byte-wise non-atomically (because a P assumes it has total ownership of
a span's bits), two threads can race writing the same heap bitmap byte
creating incorrect metadata.

Fix this by forcing every value we're writing heap bits for into a large
object. Large object spans will never be written to concurrently unless
they're freed first.

Also, while we're here, refactor the benchmarks a bit. Use generics to
eliminate the reflect nastiness in gc_test.go, and pass b.ResetTimer
down into the test to get slightly more accurate results.

Fixes #60050.

Change-Id: Ib7d6249b321963367c8c8ca88385386c8ae9af1c
Reviewed-on: https://go-review.googlesource.com/c/go/+/497215
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

cmd/go: introduce WriteOpts argument for WriteGoMod

This CL is a no-op, just adding the new options and plumbing it through.
'go get' will use this option to let commitRequirements know whether
toolchain was mentioned explicitly on the command line.

For #57001.

Change-Id: Iee7145f3335e899704df3e98fb840f1aa4063b0c
Reviewed-on: https://go-review.googlesource.com/c/go/+/499555
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

cmd/go: clean up handling of go mod tidy -go=v during modload

Slight cleanup of the special case for "mod tidy" during modload
of a module with no go version. Now it's clearer what is special.

For #57001.

Change-Id: I4a3ad0a948ccb5c62365f2a7f7f7de5582cc57c9
Reviewed-on: https://go-review.googlesource.com/c/go/+/499535
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>

cmd/go: add detection of toolchain switch loops

This happens mainly during testing because the virtual
toolchain switch is not terribly robust, and if you accidentally
try to exec "1.23" instead of "go1.23" it will let you, but it
won't work right.

Of course, although we feel pretty good about the non-test
implementation, perhaps it has a toolchain switch loop lurking too,
or perhaps one will be introduced in the future.

To handle the test bug, and just in case we have a real bug later,
add detection of toolchain switch loops with clear messages.

Also fixes a bug in setting the -lang flag properly when invoking
the Go compiler: this is the first test using 'go 1.21.x' lines
during a build.

For #57001.

Change-Id: I0ece3dd718596689a23b677cf08ddf32ea97bc57
Reviewed-on: https://go-review.googlesource.com/c/go/+/498436
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

runtime: make TestCPUMetricsSleep even more lenient

This test was introduced as a regression test for #60276. However, it
was quite flaky on a number of different platforms because there are
myriad ways the runtime can eat into time one might expect is completely
idle.

This change re-enables the test, but makes it much more resilient.
Because the issue we're testing for is persistent, we now require 10
consecutive failures to count. Any single success counts as a test
success. This change also makes the test's idle time bound more lenient,
allowing for a little bit of time to be eaten up. The regression we're
testing for results in nearly zero idle time being accounted for.

If this is still not good enough to eliminate flakes, this test should
just be deleted.

For #60276.
Fixes #60376.

Change-Id: Icd81f0c9970821b7f386f6d27c8a566fee4d0ff7
Reviewed-on: https://go-review.googlesource.com/c/go/+/498274
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>

cmd/go: scan $PATH to find version list in NewerToolchain

NewerToolchain needs a list of candidate toolchains.
Currently it always consults the module version list, using the network.
When GOTOOLCHAIN=path, it should probably not do this,
both because =path implies we don't want to use the network
and because not every released version will be in $PATH.
Instead, scan $PATH to find the available versions.

For #57001.

Change-Id: I478612c88d1504704a3f53fcfc73d8d4eedae493
Reviewed-on: https://go-review.googlesource.com/c/go/+/499296
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Bypass: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>

doc: document WER and SEH changes in Go 1.21

While here, I've removed the CL 472195 TODO, which I marked as
RELNOTE=yes by mistake.

For #57441
For #57302

Change-Id: I7563140bf307f8d732f0154d02b8fa0735527323
Reviewed-on: https://go-review.googlesource.com/c/go/+/499515
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Quim Muntal <quimmuntal@gmail.com>

internal/coverage/pods: sort counter files first by origin, then name

This patch fixes a problem with the way pods (clumps of related
coverage meta+counter data files) are collected, which was causing
problems for "go tool covdata subtract".

A subtract operation such as "go tool covdata subtract -i=dir1,dir2
-o=out" works by loading in all the counter data files from "dir1"
before any of the data files from "dir2" are loaded. The sorting
function in the pods code was sorting counter files for a given pod
based purely on name, which meant that differences in process ID
assignment could result in some files from "dir2" being presented
before "dir1". The fix is to change the sorting compare function to
prefer origin directory over filename.

Fixes #60526.

Change-Id: I2226ea675fc99666a9a28e6550d823bcdf2d6977
Reviewed-on: https://go-review.googlesource.com/c/go/+/499317
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>

doc: document new panic behavior of flag definition after Set

For #57411

Change-Id: I56c112bb03dde24c2e2643c9b72ce06158a8e717
Reviewed-on: https://go-review.googlesource.com/c/go/+/499278
TryBot-Bypass: Keith Randall <khr@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Keith Randall <khr@google.com>

doc: document change of cgocheck=2 from debug to experiment

Change-Id: I98bdfe8ab338a5b92c466a5aacfa95266171e5c3
Reviewed-on: https://go-review.googlesource.com/c/go/+/499279
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Bypass: Keith Randall <khr@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>

doc: mention the new purego implementation of hash/maphash

Change-Id: Ieedddf52ae09cf5a8e9347f0f30aff21e8369787
Reviewed-on: https://go-review.googlesource.com/c/go/+/499281
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Bypass: Keith Randall <khr@golang.org>

doc: mention improvements in checking R15 use in assembly when dynamic linking

Change-Id: I223649818763c0620875c034a77a8865252e3e69
Reviewed-on: https://go-review.googlesource.com/c/go/+/499280
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Bypass: Keith Randall <khr@golang.org>

go/types, types2: use exact unification for component types

This change defines two unification modes used to control unification:

- assign  set when unifying types involved in an assignment
- exact   if set, types unify if they can be made identical

Currently, unification is inexact: when a defined type is compared
against a type literal, the underlying type of the defined type is
considered. When channel types are compared, the channel direction
is ignored. And when defined types are compared where one (or both)
are interfaces, interface unification is used.

By contrast, exact unification requires types to match exactly:
if they can be unified, the types must be identical (with suitable
type arguments).

Exact unification is required when comparing component types.
For instance, when unifying func(x P) with func(x Q), the two
signatures unify only if P is identical to Q per Go's assignment
rules.

Until now we have ignored exact unification and made due with inexact
unification everywhere, even for component types. In some cases this
led to infinite recursions in the unifier, which we guarded against
with a depth limit (and unification failure).

Go's assignmemt rules allow inexact matching at the top-level but
require exact matching for element types.

This change passes 'assign' to the unifier when unifying parameter
against argument types because those follow assignment rules.
When comparing constraints, inexact unification is used as before.

In 'assign' mode, when comparing element types, the unifyier is
called recursively, this time with the 'exact' mode set, causing
element types to be compared exactly. If unification succeeds for
element types, they are identical (with suitable type arguments).

This change fixes #60460. It also fixes a bug in the test for
issue #60377. We also don't need to rely anymore on the recursion
depth limit (a temporary fix) for #59740. Finally, because we use
exact unification when comparing element types which are channels,
errors caused by assignment failures (due to inexact inference which
succeeded when it shouldn't have) now produce the correct inference
error.

Fixes #60460.
For #60377.
For #59740.

Change-Id: Icb6a9b4dbd34294f99328a06d52135cb499cab85
Reviewed-on: https://go-review.googlesource.com/c/go/+/498895
Reviewed-by: Robert Findley <rfindley@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

go/types, types2: add unifyMode to unifier, pass it through

Pass a mode parameter through all unifier calls but make no use of it.
When unifying type elements (components of composite types), use emode,
which currently is set to mode.

Preparatory step to fix #60460. Factoring out this mechanical change
will make the actual fix smaller and easier to review and understand.
Because this change doesn't affect the behavior of the unifier, it is
safe.

For #60460.

Change-Id: I5b67499d93025be2128c14cc00bcc3b8cc9f44b2
Reviewed-on: https://go-review.googlesource.com/c/go/+/498955
Auto-Submit: Robert Griesemer <gri@google.com>
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>

cmd/cgo: rewrite pointer passing rules in terms of pinning

With the introduction of runtime.Pinner, we need to update the cgo
pointer passing rules to accomodate the new functionality. These rule
changes are easier to describe if the rest of the pointer passing rules
are described in terms of pinning as well (Go memory is implicitly
pinned when a pointer to it is passed to a C function, and implicitly
unpinned when that function returns).

For #46787.

Change-Id: I263f03412bc9165f19c9ada72fb005ed0483a8ee
Reviewed-on: https://go-review.googlesource.com/c/go/+/498116
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Michael Knyszek <mknyszek@google.com>

doc/go1.21: fix HTML tag

Fix a broken HTML tag.

Change-Id: I9e7c65c2cc12c7b14ca205f78b78ac1e9c92a3ad
Reviewed-on: https://go-review.googlesource.com/c/go/+/499355
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Bypass: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>

cmd/go: lock in new toolchain semantics

The design doc says 'toolchain' lines apply even if the default
toolchain is older than the one specified in the toolchain line.
However, that leads to various confusing behavior and security issues.
Instead, treat toolchain as a min go version that only applies
in the current module (not in dependencies).

As an example of confusing behavior / security issue, if I install
Go 1.30 and then run 'go build' in a module I've checked out,
I expect to use Go 1.30 or newer, not to silently use an older toolchain
that may have security problems fixed in Go 1.30.
Making toolchain a min establishes that guarantee.

Also clean up the tests quite a bit.

Finally drop + from the acceptable version suffixes; we use + for +auto and +path.

For #57001.

Change-Id: Ia92c66be75d6d0e31cb4e2c0aa936fa4ec5c0a8c
Reviewed-on: https://go-review.googlesource.com/c/go/+/498260
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>

cmd/go: switch to newer toolchain in go get as needed

If we run 'go get go@1.40' or 'go get m@v' where m has a go.mod
that says 'go 1.40', we need to write a new go.mod that says 'go 1.40'.
But we can't be sure we know how to write a Go 1.40-compatible go.mod.
Instead, download the latest point release of Go 1.40 and invoke it to
finish the get command.

For #57001.

Change-Id: I4133fc3c2ecf91226a6c09a3086275ecc517e223
Reviewed-on: https://go-review.googlesource.com/c/go/+/498118
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>

cmd/compile: remove special treatment for cover vars in initorder

This patch reverts a portion of the changes in CL 443715, specifically
the code in initorder that treats coverage counter variables as special
with respect to init order. The special casing is no longer needed
now after a change to the way coverage instrumention is done (the go and
cover cmds now make sure that coverage variables appear first in
the compilation order).

Updates #56293.

Change-Id: Idf803ff4c1a095e88d455a6adcd63991687eb288
Reviewed-on: https://go-review.googlesource.com/c/go/+/499216
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>

cmd/{cover,go}: revise fix for pkg init order change with -cover

This patch contains a revised fix for issue #56293, switching to a
scheme in which coverage counter variables and meta-data variables are
written to a separate output file as opposed to being tacked onto the
end of an existing rewritten source file.

The advantage of writing counter vars to a separate file is that the
Go command can then present that file as the first source file to the
compiler when the package is built; this will ensure that counter
variable are treated as lexically "before" any other variable that
might call an instrumented function as part of its initializer.

Updates #56293.

Change-Id: Iccb8a6532b976d36ccbd5a2a339882d1f5d19477
Reviewed-on: https://go-review.googlesource.com/c/go/+/499215
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>

runtime: clarify Pinner doc

Rewrite in idiomatic form, fix a typo, break into paragraph, rephrase
slightly.

Change-Id: If146db86aa84214e61128a3bc3e9905396d44710
Reviewed-on: https://go-review.googlesource.com/c/go/+/498716
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Jonathan Amsterdam <jba@google.com>
Reviewed-by: Ansiwen <ansiwen@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>

doc/go1.21: mention errors.ErrUnsupported

Also mention errors that implement it.

For #41198

Change-Id: I4f01b112f53b19e2494b701bb012cb2cb52f8962
Reviewed-on: https://go-review.googlesource.com/c/go/+/498775
Reviewed-by: Eli Bendersky <eliben@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>

crypto/tls: fix cipher suite check when doing 0-RTT resumption

Change-Id: Ia50898308b80149f862457f9cd9f1123da4e6b6f
Reviewed-on: https://go-review.googlesource.com/c/go/+/498215
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Macrombi Lux <macrentals502@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Filippo Valsorda <filippo@golang.org>

doc/go1.21: mention maps package

For #57436

Change-Id: I99e8b0819c76f1ccf12154a894c9c4c9d68124d4
Reviewed-on: https://go-review.googlesource.com/c/go/+/498602
Reviewed-by: Eli Bendersky <eliben@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>

doc/go1.21: mention multipath TCP support

For #56539
For #59166

Change-Id: Ief392464916a1a74a8fcc6c3c7bdb213e8c6ef98
Reviewed-on: https://go-review.googlesource.com/c/go/+/498601
Run-TryBot: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Matthieu Baerts <matthieu.baerts@tessares.net>

cmd/go/internal/modload: break more cycles in readModGraph

Before CL 471595, modload.readModGraph in module with graph pruning
enabled only ever chased down transitive dependencies of unpruned
roots, so pruned dependencies couldn't cause cycles and we didn't
need to dedup them in the loading queue.

However, in 'go get' we are now passing in a set of upgraded modules
to unprune, and those upgraded modules can potentially contain cycles,
leading to an infinite loop during loading.

We have two options for a fix: we could either drop the 'unprune'
check in the enqueue operation (and instead expand the 'unprune'
requirements in a separate pass, as we do in workspace mode), or we
could check for cycles for all modules (not just the ones that are
naturally unpruned). The latter option makes it clearer that this
process must terminate, so we choose that.

(It may be possible to clean up and simplify the workspace-mode case
now that we are passing in the 'unprune' map, but for now we're
looking for a minimal fix for the Go 1.21 release.)

Fixes #60490.

Change-Id: I701f5d43a35e357f6c0c0c9d10b7aa088f917311
Reviewed-on: https://go-review.googlesource.com/c/go/+/499195
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>

internal/bytealg: fix alignment code in compare_riscv64.s

The riscv64 implementation of compare has an optimization that is
applied when both pointers share the same alignment but that alignment
is not 8 bytes.  In this case it tries to align both pointers to an 8 byte boundaries,
by individually comparing the first few bytes of each buffer.  Unfortunately,
the existing code is incorrect.  It adjusts the pointers by the wrong number
of bytes resulting, in most cases, in pointers that are not 8 byte aligned.

This commit fixes the issue by individually comparing the first
(8 - (pointer & 7)) bytes of each buffer rather than the first
(pointer & 7) bytes.

We also remove an unnecessary immediate MOV instruction.

This particular optimization is not covered by any of the existing
benchmarks so a new benchmark, benchmarkCompareBytesBigBothUnaligned,
is provided.  The benchmark tests the case where both pointers have
the same alignment but may not be 8 byte aligned.  Results of the
new benchmark along with some of the existing benchmarks generated on
a SiFive HiFive Unmatched A00 with 16GB of RAM running Ubuntu 23.04
are presented below.

CompareBytesEqual-4                           70.00n ±  6%   68.32n ±  0%   -2.40% (p=0.020 n=10)
CompareBytesToNil-4                           19.31n ±  0%   18.47n ±  0%   -4.35% (p=0.000 n=10)
CompareBytesEmpty-4                           16.79n ±  0%   15.95n ±  0%   -4.97% (p=0.000 n=10)
CompareBytesIdentical-4                       19.94n ± 15%   18.32n ± 13%   -8.15% (p=0.040 n=10)
CompareBytesSameLength-4                      37.93n ±  0%   42.44n ±  1%  +11.91% (p=0.000 n=10)
CompareBytesDifferentLength-4                 37.93n ±  0%   42.44n ±  0%  +11.89% (p=0.000 n=10)
CompareBytesBigUnaligned/offset=1-4           3.881m ± 14%   3.880m ± 15%        ~ (p=0.436 n=10)
CompareBytesBigUnaligned/offset=2-4           3.884m ±  0%   3.875m ±  0%        ~ (p=0.190 n=10)
CompareBytesBigUnaligned/offset=3-4           3.858m ±  1%   3.868m ±  1%        ~ (p=0.105 n=10)
CompareBytesBigUnaligned/offset=4-4           3.877m ±  1%   3.876m ±  0%        ~ (p=0.529 n=10)
CompareBytesBigUnaligned/offset=5-4           3.859m ±  0%   3.874m ±  0%   +0.39% (p=0.009 n=10)
CompareBytesBigUnaligned/offset=6-4           3.878m ±  1%   3.876m ±  0%        ~ (p=0.353 n=10)
CompareBytesBigUnaligned/offset=7-4           3.868m ±  1%   3.877m ±  0%        ~ (p=0.190 n=10)
CompareBytesBigBothUnaligned/offset=0-4       1.586m ±  0%   1.765m ±  0%  +11.30% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=1-4     153.132m ±  1%   1.765m ±  1%  -98.85% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=2-4     152.930m ±  1%   1.765m ±  1%  -98.85% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=3-4     152.093m ±  1%   1.769m ±  0%  -98.84% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=4-4       1.602m ±  0%   1.764m ±  0%  +10.11% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=5-4     152.314m ±  1%   1.768m ±  0%  -98.84% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=6-4     152.905m ±  1%   1.764m ±  1%  -98.85% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=7-4     152.951m ±  1%   1.804m ±  2%  -98.82% (p=0.000 n=10)
CompareBytesBig-4                             1.441m ± 21%   1.373m ± 55%        ~ (p=0.481 n=10)
CompareBytesBigIdentical-4                    19.94n ±  1%   19.10n ±  0%   -4.21% (p=0.001 n=10)
geomean                                       243.7µ         76.65µ        -68.54%

CompareBytesBigUnaligned/offset=1-4          257.7Mi ± 12%     257.7Mi ± 13%          ~ (p=0.424 n=10)
CompareBytesBigUnaligned/offset=2-4          257.5Mi ±  0%     258.1Mi ±  0%          ~ (p=0.190 n=10)
CompareBytesBigUnaligned/offset=3-4          259.2Mi ±  1%     258.5Mi ±  1%          ~ (p=0.105 n=10)
CompareBytesBigUnaligned/offset=4-4          257.9Mi ±  1%     258.0Mi ±  0%          ~ (p=0.529 n=10)
CompareBytesBigUnaligned/offset=5-4          259.1Mi ±  0%     258.1Mi ±  0%     -0.39% (p=0.008 n=10)
CompareBytesBigUnaligned/offset=6-4          257.9Mi ±  1%     258.0Mi ±  0%          ~ (p=0.353 n=10)
CompareBytesBigUnaligned/offset=7-4          258.5Mi ±  1%     257.9Mi ±  0%          ~ (p=0.190 n=10)
CompareBytesBigBothUnaligned/offset=0-4      630.6Mi ±  0%     566.6Mi ±  0%    -10.15% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=1-4      6.533Mi ±  1%   566.545Mi ±  1%  +8572.48% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=2-4      6.537Mi ±  1%   566.683Mi ±  1%  +8568.27% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=3-4      6.576Mi ±  1%   565.200Mi ±  0%  +8495.43% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=4-4      624.2Mi ±  0%     566.9Mi ±  0%     -9.18% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=5-4      6.566Mi ±  1%   565.758Mi ±  0%  +8516.41% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=6-4      6.542Mi ±  1%   567.036Mi ±  1%  +8567.35% (p=0.000 n=10)
CompareBytesBigBothUnaligned/offset=7-4      6.542Mi ±  1%   554.390Mi ±  2%  +8374.05% (p=0.000 n=10)
CompareBytesBig-4                            694.2Mi ± 18%     728.1Mi ± 35%          ~ (p=0.481 n=10)
CompareBytesBigIdentical-4                   47.83Ti ±  1%     49.92Ti ±  0%     +4.39% (p=0.002 n=10)
geomean                                      170.0Mi           813.8Mi         +378.66%

Change-Id: I0a2d0386d5ca1ffa249682a12ebd1533508e31e9
Reviewed-on: https://go-review.googlesource.com/c/go/+/497838
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Joel Sing <joel@sing.id.au>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: M Zhuo <mzh@golangcn.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: M Zhuo <mzh@golangcn.org>

go/types: don't use user string as format string

Fixes #60500.

Change-Id: I272c71071f239d5c9472e19ae25461001e7f6a8f
Reviewed-on: https://go-review.googlesource.com/c/go/+/499275
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>

syscall: fix ComputerName on Windows

GetComputerName expects n to be the size of the buffer, and
on output contains the number of characters copied to the buffer.

CL 493036 broke ComputerName by always setting n to 0.

Change-Id: I3f4b30d2f9825d321a6d28ec82bdc7b6294e04e4
Reviewed-on: https://go-review.googlesource.com/c/go/+/499035
Run-TryBot: Quim Muntal <quimmuntal@gmail.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Alex Brainman <alex.brainman@gmail.com>
Reviewed-by: Bryan Mills <bcmills@google.com>

cmd/go: update cover flag documentation

Cover flag documentation was not up to date, and the -covermode flag
documentation was missing.

Fixes #60438

Change-Id: I11e769fa2a32d074de7a748aa534afdfe2ca4e78
Reviewed-on: https://go-review.googlesource.com/c/go/+/498415
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>

doc/go1.21: document clear builtin and init order changes

Also move all the language changes to the same part of the release notes.

For #56351
For #57411

Change-Id: Id1c51b5eb8f7d85e61a2ae44ee7d73bb13036631
Reviewed-on: https://go-review.googlesource.com/c/go/+/498755
TryBot-Bypass: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Ian Lance Taylor <iant@golang.org>

cmd/go: always track visited packages in setPGOProfilePath

Currently we only track visited (copied) packages when a copy is
required. When a copy is not required, we will rewalk each package's
entire dependency graph every time we see it, which is terribly
inefficient.

Pull the visited package check up a level so that we visit packages only
once regardless of how many times they are visited.

Fixes #60455.
Fixes #60428.

Change-Id: I4e9b31eeeaa170db650c461a5de2ca984b9aba0f
Reviewed-on: https://go-review.googlesource.com/c/go/+/498735
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>

doc/go1.21: mention os package changes

Also mention WTF-8 support in the syscall package.

For #32558
For #58977
For #59971

Change-Id: Id1627889b5e498add498748d9bfc69fb58030b35
Reviewed-on: https://go-review.googlesource.com/c/go/+/498600
Reviewed-by: Eli Bendersky <eliben@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>

doc/go1.21: mention changes to the reflect package

Added Value.Clear, deprecated SliceHeader and StringHeader.

For #55002
For #56906

Change-Id: Ib7497aff830d56fad90c31ec28596e71a448e9ce
Reviewed-on: https://go-review.googlesource.com/c/go/+/498757
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

doc/go1.21: mention regexp.MarshalText and UnmarshalText

For #46159

Change-Id: Ia9cc0827a89d362532d1a662b791de8eebbfb2fe
Reviewed-on: https://go-review.googlesource.com/c/go/+/498756
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

all: update vendored dependencies

Generated with x/build/cmd/updatestd.

Updates #36905.

Change-Id: Iecd76e820c5f40a0f5e013684f7e7bef4c3fd482
Reviewed-on: https://go-review.googlesource.com/c/go/+/498598
Run-TryBot: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

doc/go1.21: remove TODOs for x/ repos

Remove x/ repos changes that aren't part of the 1.21 release.

Change-Id: Ia5cbf02ed6c0cc4dc0ff80d68403723b3f9ae0ba
Reviewed-on: https://go-review.googlesource.com/c/go/+/498596
Run-TryBot: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>

cmd/internal/testdir: stop manually adding GOROOT/bin to PATH

The go command already places $GOROOT/bin at the beginning of $PATH in
the test's environment as of Go 1.19¹, so there's no need for the test
to do it anymore. Start enjoying yet another benefit of using 'go test'.

¹ See go.dev/issue/57050.

For #56844.

Change-Id: If7732cd8b8979eabf185485d3c73858a4e546d69
Reviewed-on: https://go-review.googlesource.com/c/go/+/498271
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>

cmd/go: remove stray reference to mkalldocs.sh

Clean up a stray reference in the docs to 'mkalldocs.sh', which no
longer exists (was replaced by a Go program, in 676794f73e).

Change-Id: I058e86395b14756d7afc184f99f89dbc5ba0184e
Reviewed-on: https://go-review.googlesource.com/c/go/+/498715
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
Auto-Submit: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>

doc: release notes for my humble contributions to go1.21

Updates #58645

Change-Id: Ieb4e6062613e26826ba8373cb4c2fd9198d0c692
Reviewed-on: https://go-review.googlesource.com/c/go/+/496116
Reviewed-by: Eli Bendersky <eliben@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Alan Donovan <adonovan@google.com>
TryBot-Bypass: Eli Bendersky <eliben@google.com>

syscall: implement wasip1 Fcntl

CL 494915 broke non-blocking I/O on wasip1 for files
opened with os.NewFile. This is fixed by providing an
implementation of fcntl(F_GETFL) for wasip1.

Change-Id: I78979076b95495fd4b94814552e5f5b043270cd0
Reviewed-on: https://go-review.googlesource.com/c/go/+/498195
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Achille Roussel <achille.roussel@gmail.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>

misc/wasm: set PATH variable in exec

The PATH variable is required to run the testenv tests.
Set it for all the runtime invocations where we don't
already set it by inheriting from the environment.

For #59583
For #59907
For #60097

Change-Id: If582dd8f086e3f40bc58d555f6034dcffe6f8e5f
Reviewed-on: https://go-review.googlesource.com/c/go/+/498616
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Johan Brandhorst-Satzkorn <johan.brandhorst@gmail.com>

doc/go1.21: mention syscall package changes

One of the changes also affects the os package, so mention it
there too.

For #46259

Change-Id: I8041a5ce009725ab210118ee668fc94196d9ff82
Reviewed-on: https://go-review.googlesource.com/c/go/+/498376
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Eli Bendersky <eliben@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>

io: clarify that Read(nil) can return 0, EOF

Fixes #40385

Change-Id: I965b5db985fd4418a992e883073cbc8309b2cb88
Reviewed-on: https://go-review.googlesource.com/c/go/+/498355
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Rob Pike <r@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>

doc: add release note on dead global map variable deletion

Add a short release note blurb on linker dead-coding of global map
variables.

Updates #58645.

Change-Id: Ibfdd3e2b2b78517db4245b1d7eedd5e75a643b47
Reviewed-on: https://go-review.googlesource.com/c/go/+/498575
Reviewed-by: Eli Bendersky <eliben@google.com>
TryBot-Bypass: Than McIntosh <thanm@google.com>