CL 658035 added TestCgoCallbackPprof, which is consistently failing on
solaris. runtime/pprof maintains a list of platforms where CPU profiling
does not work properly. Since this test requires CPU profiling, skip the
this test on those platforms.
For #72870.
For #72876.
For #72871.
Change-Id: I6a6a636cbf6b16abcbba8771178fe1d001be9d9b
Reviewed-on: https://go-review.googlesource.com/c/go/+/658415
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/658435
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Dmitri Shuralyov <dmitshur@google.com>
TryBot-Bypass: Dmitri Shuralyov <dmitshur@google.com>
mp.isExtraInC is intended to indicate that this M has no Go frames at
all; it is entirely executing in C.
If there was a cgocallback to Go and then a cgocall to C, such that the
leaf frames are C, that is fine. e.g., traceback can handle this fine
with SetCgoTraceback (or by simply skipping the C frames).
However, we currently mismanage isExtraInC, unconditionally setting it
on return from cgocallback. This means that if there are two levels of
cgocallback, we end up running Go code with isExtraInC set.
1. C-created thread calls into Go function 1 (via cgocallback).
2. Go function 1 calls into C function 1 (via cgocall).
3. C function 1 calls into Go function 2 (via cgocallback).
4. Go function 2 returns back to C function 1 (returning via the remainder of cgocallback).
5. C function 1 returns back to Go function 1 (returning via the remainder of cgocall).
6. Go function 1 is now running with mp.isExtraInC == true.
The fix is simple; only set isExtraInC on return from cgocallback if
there are no more Go frames. There can't be more Go frames unless there
is an active cgocall out of the Go frames.
For #72870.
Fixes#72871.
Cq-Include-Trybots: luci.golang.try:go1.23-linux-amd64-longtest
Change-Id: I6a6a636c4e7ba75a29639d7036c5af3738033467
Reviewed-on: https://go-review.googlesource.com/c/go/+/658035
Reviewed-by: Cherry Mui <cherryyz@google.com>
Commit-Queue: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit 577bb3d0ce576b2ca311e58dd942f189838b80fc)
Reviewed-on: https://go-review.googlesource.com/c/go/+/658055
Auto-Submit: Dmitri Shuralyov <dmitshur@google.com>
https://go.dev/doc/godebug#go-123 documents changes to winsymlink and
winreadlinkvolume in Go 1.23.
This fixes the registered "changed" minor version to Go 1.23,
so that defaults when building a Go 1.22 module are correct.
Fixes#72937
Change-Id: I5d5bf31ca04f9e95208fb0fdaad2232f9db653ff
Reviewed-on: https://go-review.googlesource.com/c/go/+/659035
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
(cherry picked from commit 2e749a645a6d03c7ac11bb172c4591564061b29e)
Reviewed-on: https://go-review.googlesource.com/c/go/+/658976
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Auto-Submit: Dmitri Shuralyov <dmitshur@google.com>
Unlike request headers, where we are allowed to leniently accept
a bare LF in place of a CRLF, chunked bodies must always use CRLF
line terminators. We were already enforcing this for chunk-data lines;
do so for chunk-size lines as well. Also reject bare CRs anywhere
other than as part of the CRLF terminator.
Fixes CVE-2025-22871
Fixes#72010
For #71988
Change-Id: Ib0e21af5a8ba28c2a1ca52b72af8e2265ec79e4a
Reviewed-on: https://go-review.googlesource.com/c/go/+/652998
Reviewed-by: Jonathan Amsterdam <jba@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit d31c805535f3fde95646ee4d87636aaaea66847b)
Reviewed-on: https://go-review.googlesource.com/c/go/+/657216
This is to update module version to the fixed x/net.
For #71984
Change-Id: I8f4357f14a7d44a782c131eb856b50a103be2f2d
Reviewed-on: https://go-review.googlesource.com/c/go/+/654796
Reviewed-by: Junyang Shao <shaojunyang@google.com>
Auto-Submit: Junyang Shao <shaojunyang@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
0b6d719 [internal-branch.go1.23-vendor] proxy, http/httpproxy: do not mismatch IPv6 zone ids against hosts
Fixes CVE-2025-22870
For #71985
Change-Id: Ib72c96bd0ab44d9ed2ac1428e0a9fc245464b3fc
Reviewed-on: https://go-internal-review.googlesource.com/c/go/+/2141
Commit-Queue: Damien Neil <dneil@google.com>
Reviewed-by: Roland Shoemaker <bracewell@google.com>
Reviewed-by: Neal Patel <nealpatel@google.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/654695
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Junyang Shao <shaojunyang@google.com>
CL 652181 accidentally missed this iPhone only code.
For #71961
For #71962
Change-Id: I567f8bb38958907442e69494da330d5199d11f54
Reviewed-on: https://go-review.googlesource.com/c/go/+/653136
Reviewed-by: Michael Pratt <mpratt@google.com>
Commit-Queue: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
It's used by the SWIG CI build, at least, and it's an easy fix.
[The change in x_cgo_sys_thread_create is gone, as that function was
already compliant in 1.23.]
Fixes#71962
For #71961
Change-Id: Id21071a5aef216b35ecf0e9cd3e05d08972d92fe
Reviewed-on: https://go-review.googlesource.com/c/go/+/652181
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
(cherry picked from commit 76c70282538bf4cccd6f98b5b26df7f5a7f2cebd)
Reviewed-on: https://go-review.googlesource.com/c/go/+/652935
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Fixes#71915
For #71905
Change-Id: I50a418f8552e071c6e5011af5b9accc7d41548d0
Reviewed-on: https://go-review.googlesource.com/c/go/+/651855
Reviewed-by: Ian Lance Taylor <iant@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
(cherry picked from commit 194696f1d1f6e5609f96d0fb0192595e7e0f5b90)
Reviewed-on: https://go-review.googlesource.com/c/go/+/652875
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
Currently method values aren't correctly handled in Seq because we call
canRangeFunc on the reciever type, not the method value type, when we're
handling a method value. reflect.Value.Type has the logic to obtain the
method value type from the Value.
This change slightly refactors reflect.Value.Type into a separate
function so we can obtain the correct type as an abi.Type and pass it
off to canRangeFunc (and canRangeFunc2).
For #71874.
Fixes#71875.
Change-Id: Ie62dfca2a84b8f2f816bb87ff1ed1a58a7bb8122
Reviewed-on: https://go-review.googlesource.com/c/go/+/651416
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
(cherry picked from commit d93f6df0cc4f33127ef76fa994edd54d7726d0a9)
Reviewed-on: https://go-review.googlesource.com/c/go/+/651498
Reviewed-by: Firuze Sayan <sayanfiruze@gmail.com>
As of CL 650835, the pidfd test child no longer sends SIGCHLD on exit.
Per clone(2), "If [the child termination] signal is specified as
anything other than SIGCHLD, then the parent process must specify the
__WALL or __WCLONE options when waiting for the child with wait(2)."
Align with this requirement.
For #71848.
For #71828.
Change-Id: I6a6a636c739e4a59abe1533fe429a433e8588939
Reviewed-on: https://go-review.googlesource.com/c/go/+/651415
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Michael Pratt <mpratt@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit e1e65ae3ee5f977c31f3651233cc7ff2a0b579ca)
Reviewed-on: https://go-review.googlesource.com/c/go/+/651495
Avoid a spurious SIGCHLD the first time we start a process.
For #71828Fixes#71848
Change-Id: I744100d21bf6aaaaafc99bc5eec9f9f807a50682
Reviewed-on: https://go-review.googlesource.com/c/go/+/651055
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
The existing code for recover from deferrangefunc was broken in
several ways.
1. the code following a deferrangefunc call did not check the return
value for an out-of-band value indicating "return now" (i.e., recover
was called)
2. the returned value was delivered using a bespoke ABI that happened
to match on register-ABI platforms, but not on older stack-based
ABI.
3. the returned value was the wrong width (1 word versus 2) and
type/value(integer 1, not a pointer to anything) for deferrangefunc's
any-typed return value (in practice, the OOB value check could catch
this, but still, it's sketchy).
This -- using the deferreturn lookup method already in place for
open-coded defers -- turned out to be a much-less-ugly way of
obtaining the desired transfer of control for recover().
TODO: we also could do this for regular defer, and delete some code.
Fixes#71839
Change-Id: If7d7ea789ad4320821aab3b443759a7d71647ff0
Reviewed-on: https://go-review.googlesource.com/c/go/+/650476
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/651496
The timespec argument takes the remainder in nanoseconds, not
microseconds. Convert the remaining time to nsec.
Fixes#71727
Change-Id: I36cbbe3a088830c5e3afcc9516ef42e96ee21268
Reviewed-on: https://go-review.googlesource.com/c/go/+/648915
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Mauri de Souza Meneguzzo <mauri870@gmail.com>
Reviewed-by: Axel Busch <axel.busch@ibm.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Vishwanatha HD <vishwanatha.hd@ibm.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/649395
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Remove the branching instruction from p256NegCond which made it variable
time. The technique used matches that used in p256MovCond.
Fixes#71383Fixes#71423
Fixes CVE-2025-22866
Change-Id: Ibc2a46814d856cbbdaf6cc0c5a415ed5d42ca793
Reviewed-on: https://go-review.googlesource.com/c/go/+/643735
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Paul Murphy <murp@ibm.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit 6fc23a3cff5e38ff72923fee50f51254dcdc6e93)
Reviewed-on: https://go-review.googlesource.com/c/go/+/645515
Reviewed-by: Carlos Amedee <carlos@golang.org>
Newer git versions (at least git 2.47.1) do not send all the matching tags
for a shallow fetch of a specific hash anymore. The go command assumes
that git servers do this. Since that assumption is broken, use the local
copy of the remote refs list to augment the tags sent by the server.
This makes the cmd/go/internal/modfetch tests pass again with newer git.
For #71261Fixes#71263
Change-Id: I9fd4f3fd7beeb68a522938599f8f3acd887d0b26
Reviewed-on: https://go-review.googlesource.com/c/go/+/642437
Reviewed-by: Michael Matloob <matloob@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Russ Cox <rsc@golang.org>
(cherry picked from commit bd80d8956f3062d2b2bff2d7da6b879dfa909f12)
Reviewed-on: https://go-review.googlesource.com/c/go/+/642696
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
We can't coalesce a non-WB store with a subsequent Move, as the
result of the store might be the source of the move.
There's a simple codegen test. Not sure how we might do a real test,
as all the repro's I've come up with are very expensive and unreliable.
Fixes#71230
Change-Id: If18bf181a266b9b90964e2591cd2e61a7168371c
Reviewed-on: https://go-review.googlesource.com/c/go/+/642197
Reviewed-by: Keith Randall <khr@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/642498
When an HTTP redirect changes the host of a request, we drop
sensitive headers such as Authorization from the redirected request.
Fix a bug where a chain of redirects could result in sensitive
headers being sent to the wrong host:
1. request to a.tld with Authorization header
2. a.tld redirects to b.tld
3. request to b.tld with no Authorization header
4. b.tld redirects to b.tld
3. request to b.tld with Authorization header restored
Thanks to Kyle Seely for reporting this issue.
For #70530
Fixes ##71211
Fixes CVE-2024-45336
Reviewed-on: https://go-internal-review.googlesource.com/c/go/+/1641
Reviewed-by: Roland Shoemaker <bracewell@google.com>
Reviewed-by: Tatiana Bradley <tatianabradley@google.com>
Commit-Queue: Roland Shoemaker <bracewell@google.com>
Change-Id: I326544358de71ff892d9e9fe338252a5dd04001f
Reviewed-on: https://go-internal-review.googlesource.com/c/go/+/1764
Reviewed-on: https://go-review.googlesource.com/c/go/+/643104
Auto-Submit: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
When checking URI constraints, use netip.ParseAddr, which understands
zones, unlike net.ParseIP which chokes on them. This prevents zone IDs
from mistakenly satisfying URI constraints.
Thanks to Juho Forsén of Mattermost for reporting this issue.
For #71156Fixes#71208
Fixes CVE-2024-45341
Change-Id: Iecac2529f3605382d257996e0fb6d6983547e400
Reviewed-on: https://go-internal-review.googlesource.com/c/go/+/1700
Reviewed-by: Tatiana Bradley <tatianabradley@google.com>
Reviewed-by: Damien Neil <dneil@google.com>
(cherry picked from commit 22ca55d396ba801e6ae9b2bd67a059fcb30562fd)
Reviewed-on: https://go-internal-review.googlesource.com/c/go/+/1762
Reviewed-by: Roland Shoemaker <bracewell@google.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/643103
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Currently injectglist emits all the trace events before actually calling
casgstatus on each goroutine. This is a problem, since tracing can
observe an inconsistent state (gstatus does not match tracer's 'emitted
an event' state).
This change fixes the problem by having injectglist do what every other
scheduler function does, and that's wrap each call to casgstatus in
traceAcquire/traceRelease.
For #70883.
Fixes#71147.
Change-Id: I857e96cec01688013597e8efc0c4c3d0b72d3a70
Reviewed-on: https://go-review.googlesource.com/c/go/+/638558
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit f025d19e7b3f0c66242760c213cc2b54cb100f69)
Reviewed-on: https://go-review.googlesource.com/c/go/+/641378
Auto-Submit: Michael Pratt <mpratt@google.com>
Updates #71077Fixes#71104
Change-Id: I6a6a465685f3bd50a5bb35a160f87b59b74fa6af
Reviewed-on: https://go-review.googlesource.com/c/go/+/639655
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Damien Neil <dneil@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Auto-Submit: Damien Neil <dneil@google.com>
Reviewed-by: Joel Sing <joel@sing.id.au>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/640315
Reviewed-by: Filippo Valsorda <filippo@golang.org>
This is not a cherry pick, because the code has changed on tip.
For #70763Fixes#70789
Change-Id: If9fcfee17e86a746cf8c72293dc34f80b430f6e6
Reviewed-on: https://go-review.googlesource.com/c/go/+/635397
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Damien Neil <dneil@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Currently, at a cgo callback where there is already a Go frame on
the stack (i.e. C->Go->C->Go), we require that at the inner Go
callback the SP is within the g0's stack bounds set by a previous
callback. This is to prevent that the C code switches stack while
having a Go frame on the stack, which we don't really support. But
this could also happen when we cannot get accurate stack bounds,
e.g. when pthread_getattr_np is not available. Since the stack
bounds are just estimates based on the current SP, if there are
multiple C->Go callbacks with various stack depth, it is possible
that the SP of a later callback falls out of a previous call's
estimate. This leads to runtime throw in a seemingly reasonable
program.
This CL changes it to save the old g0 stack bounds at cgocallback,
update the bounds, and restore the old bounds at return. So each
callback will get its own stack bounds based on the current SP,
and when it returns, the outer callback has the its old stack
bounds restored.
Also, at a cgo callback when there is no Go frame on the stack,
we currently always get new stack bounds. We do this because if
we can only get estimated bounds based on the SP, and the stack
depth varies a lot between two C->Go calls, the previous
estimates may be off and we fall out or nearly fall out of the
previous bounds. But this causes a performance problem: the
pthread API to get accurate stack bounds (pthread_getattr_np) is
very slow when called on the main thread. Getting the stack bounds
every time significantly slows down repeated C->Go calls on the
main thread.
This CL fixes it by "caching" the stack bounds if they are
accurate. I.e. at the second time Go calls into C, if the previous
stack bounds are accurate, and the current SP is in bounds, we can
be sure it is the same stack and we don't need to update the bounds.
This avoids the repeated calls to pthread_getattr_np. If we cannot
get the accurate bounds, we continue to update the stack bounds
based on the SP, and that operation is very cheap.
On a Linux/AMD64 machine with glibc:
name old time/op new time/op delta
CgoCallbackMainThread-8 96.4µs ± 3% 0.1µs ± 2% -99.92% (p=0.000 n=10+9)
Updates #68285.
Updates #68587.
Fixes#69988.
Change-Id: I3422badd5ad8ff63e1a733152d05fb7a44d5d435
Reviewed-on: https://go-review.googlesource.com/c/go/+/600296
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
(cherry picked from commit 76a8409eb81eda553363783dcdd9d6224368ae0e)
Reviewed-on: https://go-review.googlesource.com/c/go/+/635775
Flips the pkgReader.enableAlias flag to true when reading unified IR.
This was disabled while resolving #66873. This resolves the TODO to
flip it back to true.
Fixes#70394Fixes#70517
Updates #66873
Change-Id: Ifd52b0f9510d6bcf151de1c9a18d71ab548c14e4
Reviewed-on: https://go-review.googlesource.com/c/go/+/604099
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
(cherry picked from commit 209ed1a9c75d17046285c416b74a14bb89799757)
Reviewed-on: https://go-review.googlesource.com/c/go/+/631855
Commit-Queue: Tim King <taking@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Fix a regression that appeared in 1.23 when it comes to the stack traces
shown in the trace viewer. In 1.22 and earlier, the viewer was always
showing end stack traces. In 1.23 and later the viewer started to
exclusively show start stack traces.
Showing only the start stack traces made it impossible to see the last
stack trace produced by a goroutine. It also made it hard to understand
why a goroutine went off-cpu, as one had to hunt down the next running
slice of the same goroutine.
Emit end stack traces in addition to start stack traces to fix the
issue.
Fixes#70592
Change-Id: Ib22ea61388c1d94cdbc99fae2d207c4dce011a59
Reviewed-on: https://go-review.googlesource.com/c/go/+/631895
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Felix Geisendörfer <felix.geisendoerfer@datadoghq.com>
Reviewed-by: Nick Ripley <nick.ripley@datadoghq.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
(cherry picked from commit 6405e60ca6be798c1f8c1d0365fd63b89b524df5)
Reviewed-on: https://go-review.googlesource.com/c/go/+/632075
Reviewed-by: Veronica Silina <veronicasilina@google.com>
Auto-Submit: Veronica Silina <veronicasilina@google.com>
The failures in #70288 are consistent with and strongly imply
stack corruption during fault handling, and debug prints show
that the Go code run during fault handling is running about
300 bytes above the bottom of the goroutine stack.
That should be okay, but that implies the DLL code that called
Go's handler was running near the bottom of the stack too,
and maybe it called other deeper things before or after the
Go handler and smashed the stack that way.
stackSystem is already 4096 bytes on amd64;
making it match that on 386 makes the flaky failures go away.
It's a little unsatisfying not to be able to say exactly what is
overflowing the stack, but the circumstantial evidence is
very strong that it's Windows.
For #70288.
Fixes#70475.
Change-Id: Ife89385873d5e5062a71629dbfee40825edefa49
Reviewed-on: https://go-review.googlesource.com/c/go/+/627375
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit 7eeb0a188eb644486da9f77bae0375d91433d0bf)
Reviewed-on: https://go-review.googlesource.com/c/go/+/632196
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Veronica Silina <veronicasilina@google.com>
getOrAddWeakHandle is very careful about keeping its input alive across
the operation, but not very careful about keeping the heap-allocated
handle it creates alive. In fact, there's a window in this function
where it is *only* visible via the special. Specifically, the window of
time between when the handle is stored in the special and when the
special actually becomes visible to the GC.
(If we fail to add the special because it already exists, that case is
fine. We don't even use the same handle value, but the one we obtain
from the attached GC-visible special, *and* we return that value, so it
remains live.)
For #70455.
Fixes#70469.
Change-Id: Iadaff0cfb93bcaf61ba2b05be7fa0519c481de82
Reviewed-on: https://go-review.googlesource.com/c/go/+/630316
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Currently it's possible for weak->strong conversions to create more GC
work during mark termination. When a weak->strong conversion happens
during the mark phase, we need to mark the newly-strong pointer, since
it may now be the only pointer to that object. In other words, the
object could be white.
But queueing new white objects creates GC work, and if this happens
during mark termination, we could end up violating mark termination
invariants. In the parlance of the mark termination algorithm, the
weak->strong conversion is a non-monotonic source of GC work, unlike the
write barriers (which will eventually only see black objects).
This change fixes the problem by forcing weak->strong conversions to
block during mark termination. We can do this efficiently by setting a
global flag before the ragged barrier that is checked at each
weak->strong conversion. If the flag is set, then the conversions block.
The ragged barrier ensures that all Ps have observed the flag and that
any weak->strong conversions which completed before the ragged barrier
have their newly-minted strong pointers visible in GC work queues if
necessary. We later unset the flag and wake all the blocked goroutines
during the mark termination STW.
There are a few subtleties that we need to account for. For one, it's
possible that a goroutine which blocked in a weak->strong conversion
wakes up only to find it's mark termination time again, so we need to
recheck the global flag on wake. We should also stay non-preemptible
while performing the check, so that if the check *does* appear as true,
it cannot switch back to false while we're actively trying to block. If
it switches to false while we try to block, then we'll be stuck in the
queue until the following GC.
All-in-all, this CL is more complicated than I would have liked, but
it's the only idea so far that is clearly correct to me at a high level.
This change adds a test which is somewhat invasive as it manipulates
mark termination, but hopefully that infrastructure will be useful for
debugging, fixing, and regression testing mark termination whenever we
do fix it.
For #69803.
Fixes#70323.
Change-Id: Ie314e6fd357c9e2a07a9be21f217f75f7aba8c4a
Reviewed-on: https://go-review.googlesource.com/c/go/+/623615
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
(cherry picked from commit 80d306da50aef6334bcb65fb02f5728cb9513691)
Reviewed-on: https://go-review.googlesource.com/c/go/+/627615
TryBot-Bypass: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Dmitri Shuralyov <dmitshur@google.com>
syscall.SyscallN is implemented by runtime.syscall_syscalln, which makes
sure that the variadic argument doesn't escape.
There is no need to worry about the lifetime of the elements of the
variadic argument, as the compiler will keep them live until the
function returns.
For #70197Fixes#70202
Change-Id: I12991f0be12062eea68f2b103fa0a794c1b527eb
Reviewed-on: https://go-review.googlesource.com/c/go/+/625297
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Alex Brainman <alex.brainman@gmail.com>
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit 7fff741016c8157e107cce8013ee3ca621725384)
Reviewed-on: https://go-review.googlesource.com/c/go/+/630196
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Quim Muntal <quimmuntal@gmail.com>
This stops the test from failing with a known failure mode, and
creates time to look into what the next steps should be, if any.
For #69840Fixes#70239
Change-Id: I060903d256ed65c5dfcd70ae76eb361cab63186f
Reviewed-on: https://go-review.googlesource.com/c/go/+/625197
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Eric Grosse <grosse@gmail.com>
(cherry picked from commit bea9b91f0f4be730c880edbe496ab25c9b742cad)
Reviewed-on: https://go-review.googlesource.com/c/go/+/627575
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Bypass: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
The old test relied on naming conventions. The new test
uses an explicit parent pointer chain initialized when the
closures are created (in the same place that the names
used in the older fragile test were assigned).
Fixes#70198.
Change-Id: Ie834103c7096e4505faaff3bed1fc6e918a21211
Reviewed-on: https://go-review.googlesource.com/c/go/+/622656
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/625535
Auto-Submit: Dmitri Shuralyov <dmitshur@google.com>
syscall.SyscallX consumes a lot of stack space, which is a problem
because they are nosplit functions. They used to use less stack space,
but CL 563315, that landed in Go 1.23, increased the stack usage by a
lot.
This CL reduces the stack usage back to the previous level.
Fixes#69848
Updates #69813
Change-Id: Iddedd28b693c66a258da687389768055c493fc2e
Reviewed-on: https://go-review.googlesource.com/c/go/+/618497
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit fa7343aca326aad061ab877c1a4cebb96c4355c1)
Reviewed-on: https://go-review.googlesource.com/c/go/+/623516
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Fix a regression introduced in CL 598515 causing runtime.MutexProfile
stack traces to omit their root frames.
In most cases this was merely causing the `runtime.goexit` frame to go
missing. But in the case of runtime._LostContendedRuntimeLock, an empty
stack trace was being produced.
Add a test that catches this regression by checking for a stack trace
with the `runtime.goexit` frame.
Also fix a separate problem in expandFrame that could cause
out-of-bounds panics when profstackdepth is set to a value below 32.
There is no test for this fix because profstackdepth can't be changed at
runtime right now.
Fixes#69865
Change-Id: I1600fe62548ea84981df0916d25072c3ddf1ea1a
Reviewed-on: https://go-review.googlesource.com/c/go/+/611615
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Nick Ripley <nick.ripley@datadoghq.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit c64ca8c6ef13723b9f25f4b5e1c7b6986b958d2e)
Reviewed-on: https://go-review.googlesource.com/c/go/+/621276
Reviewed-by: Cherry Mui <cherryyz@google.com>
CL 622235 would fix#70000 while resulting in one extra sendfile(2) system
call when sendfile(2) returns (>0, EAGAIN).
That's also why I left sendfile_bsd.go behind, and didn't make it line up
with other two implementations: sendfile_linux.go and sendfile_solaris.go.
Unlike sendfile(2)'s on Linux and Solaris that always return (0, EAGAIN),
sendfile(2)'s on *BSD and macOS may return (>0, EAGAIN) when using a socket
marked for non-blocking I/O. In that case, the current code will try to re-call
sendfile(2) immediately, which will most likely get us a (0, EAGAIN).
After that, it goes to `dstFD.pd.waitWrite(dstFD.isFile)` below,
which should have been done in the first place.
Thus, the real problem that leads to #70000 is that the old code doesn't handle
the special case of sendfile(2) sending the exact number of bytes the caller requested.
Fixes#70000Fixes#70020
Change-Id: I6073d6b9feb58b3d7e114ec21e4e80d9727bca66
Reviewed-on: https://go-review.googlesource.com/c/go/+/622255
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Run-TryBot: Andy Pan <panjf2000@gmail.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/622697
The BSD implementation of poll.SendFile incorrectly halted
copying after succesfully writing one full chunk of data.
Adjust the copy loop to match the Linux and Solaris
implementations.
In testing, empirically macOS appears to sometimes return
EAGAIN from sendfile after successfully copying a full
chunk. Add a check to all implementations to return nil
after successfully copying all data if the last sendfile
call returns EAGAIN.
For #70000
For #70020
Change-Id: I57ba649491fc078c7330310b23e1cfd85135c8ff
Reviewed-on: https://go-review.googlesource.com/c/go/+/622235
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
(cherry picked from commit bd388c0216bcb33d7325b0ad9722a3be8155a289)
Reviewed-on: https://go-review.googlesource.com/c/go/+/622696
Currently, on Mach-O, the Go linker doesn't generate LC_UUID in
internal linking mode. This causes some macOS system tools unable
to track the binary, as well as in some cases the binary unable
to access local network on macOS 15.
This CL makes the linker start generate LC_UUID. Currently, the
UUID is generated if the -B flag is specified. And we'll make it
generate UUID by default in a later CL. The -B flag is currently
for generating GNU build ID on ELF, which is a similar concept to
Mach-O's UUID. Instead of introducing another flag, we just use
the same flag and the same setting. Specifically, "-B gobuildid"
will generate a UUID based on the Go build ID.
Updates #68678.
Fixes#69992.
Cq-Include-Trybots: luci.golang.try:go1.23-darwin-amd64_14,go1.23-darwin-arm64_13
Change-Id: I90089a78ba144110bf06c1c6836daf2d737ff10a
Reviewed-on: https://go-review.googlesource.com/c/go/+/618595
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Ingo Oeser <nightlyone@googlemail.com>
Reviewed-by: Than McIntosh <thanm@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit 20ed60311848ca40e51cb430fa602dd83a9c726f)
Reviewed-on: https://go-review.googlesource.com/c/go/+/622595
Reviewed-by: Michael Pratt <mpratt@google.com>
Goroutine profiles require checking in with the profiler before any
goroutine starts running. coroswitch is a place where a goroutine may
start running, but where we do not check in with the profiler, which
leads to crashes. Fix this by checking in with the profiler the same way
execute does.
For #69998.
Fixes#70001.
Change-Id: Idef6dd31b70a73dd1c967b56c307c7a46a26ba73
Reviewed-on: https://go-review.googlesource.com/c/go/+/622016
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit 2a98a1849f059ffa94ab23a1ab7d8fa0fd0b48dd)
Reviewed-on: https://go-review.googlesource.com/c/go/+/622375
Reviewed-by: Michael Pratt <mpratt@google.com>
This change switches isSending to be an atomic.Int32 instead of an
atomic.Uint8. The Int32 version is managed as a counter, which is
something that we couldn't do with Uint8 without adding a new intrinsic
which may not be available on all architectures.
That is, instead of only being able to support 8 concurrent timer
firings on the same timer because we only have 8 independent bits to set
for each concurrent timer firing, we can now have 2^31-1 concurrent
timer firings before running into any issues. Like the fact that each
bit-set was matched with a clear, here we match increments with
decrements to indicate that we're in the "sending on a channel" critical
section in the timer code, so we can report the correct result back on
Stop or Reset.
We choose an Int32 instead of a Uint32 because it's easier to check for
obviously bad values (negative values are always bad) and 2^31-1
concurrent timer firings should be enough for anyone.
Previously, we avoided anything bigger than a Uint8 because we could
pack it into some padding in the runtime.timer struct. But it turns out
that the type that actually matters, runtime.timeTimer, is exactly 96
bytes in size. This means its in the next size class up in the 112 byte
size class because of an allocation header. We thus have some free space
to work with. This change increases the size of this struct from 96
bytes to 104 bytes.
(I'm not sure if runtime.timer is often allocated directly, but if it
is, we get lucky in the same way too. It's exactly 80 bytes in size,
which means its in the 96-byte size class, leaving us with some space to
work with.)
Fixes#69978
For #69969.
Related to #69880 and #69312 and #69882.
Change-Id: I9fd59cb6a69365c62971d1f225490a65c58f3e77
Cq-Include-Trybots: luci.golang.try:go1.23-linux-amd64-longtest
Reviewed-on: https://go-review.googlesource.com/c/go/+/621616
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit 6a49f81edc7aa8aa12e26a1a0ed8819a3e5c7b5e)
Reviewed-on: https://go-review.googlesource.com/c/go/+/621856
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
The Ticker Stop and Reset methods don't report a value,
so we don't need to track whether they are interrupting a send.
This includes a test that used to fail about 2% of the time on
my laptop when run under x/tools/cmd/stress.
For #69880Fixes#69882
Change-Id: Ic6d14b344594149dd3c24b37bbe4e42e83f9a9ad
Reviewed-on: https://go-review.googlesource.com/c/go/+/620136
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
(cherry picked from commit 48849e0866f64a40d04a9151e44e5a73acdfc17b)
Reviewed-on: https://go-review.googlesource.com/c/go/+/620137
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Fix a regression introduced in CL 572396 causing goroutine stacks not
getting null terminated.
This bug impacts callers that reuse the []StackRecord slice for multiple
calls to GoroutineProfile. See https://github.com/felixge/fgprof/issues/33
for an example of the problem.
Add a test case to prevent similar regressions in the future. Use null
padding instead of null termination to be consistent with other profile
types and because it's less code to implement. Also fix the
ThreadCreateProfile code path.
Fixes#69258
Change-Id: I0b9414f6c694c304bc03a5682586f619e9bf0588
Reviewed-on: https://go-review.googlesource.com/c/go/+/609815
Reviewed-by: Tim King <taking@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
(cherry picked from commit 49e542aa85b7c2d9f6cf50de00843b455bc1e635)
Reviewed-on: https://go-review.googlesource.com/c/go/+/621277
Reviewed-by: Cherry Mui <cherryyz@google.com>
Inside Google we have seen issues with QEMU user mode failing to wake a
parent waitid when this child exits with SYS_EXIT. This bug appears to
not affect SYS_EXIT_GROUP.
It is currently unclear if this is a general QEMU or specific to
Google's configuration, but SYS_EXIT and SYS_EXIT_GROUP are semantically
equivalent here, so we can use the latter here in case this is a general
QEMU bug.
For #68976.
For #69259.
Change-Id: I34e51088c9a6b7493a060e2a719a3cc4a3d54aa0
Reviewed-on: https://go-review.googlesource.com/c/go/+/617417
Reviewed-by: Ian Lance Taylor <iant@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
(cherry picked from commit 47a99359206f0dd41228deda0aa31f1e769cc156)
Reviewed-on: https://go-review.googlesource.com/c/go/+/617716
clone(CLONE_PIDFD) was added in Linux 5.2 and pidfd_open was added in
Linux 5.3. Thus our feature check for pidfd_open should be sufficient to
ensure that clone(CLONE_PIDFD) works.
Unfortuantely, some alternative Linux implementations may not follow
this strict ordering. For example, QEMU 7.2 (Dec 2022) added pidfd_open,
but clone(CLONE_PIDFD) was only added in QEMU 8.0 (Apr 2023).
Debian bookworm provides QEMU 7.2 by default.
For #68976.
Fixes#69259.
Change-Id: Ie3f3dc51f0cd76944871bf98690abf59f68fd7bf
Reviewed-on: https://go-review.googlesource.com/c/go/+/592078
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
(cherry picked from commit 7a5fc9b34deb8d9fe22c9d060a5839827344fcc2)
Reviewed-on: https://go-review.googlesource.com/c/go/+/612218
I've done some more testing of the new isSending field.
I'm not able to get more than 2 bits set. That said,
with this change it's significantly less likely to have even
2 bits set. The idea here is to clear the bit before possibly
locking the channel we are sending the value on, thus avoiding
some delay and some serialization.
For #69312
For #69333
Change-Id: I8b5f167f162bbcbcbf7ea47305967f349b62b0f4
Reviewed-on: https://go-review.googlesource.com/c/go/+/617596
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Commit-Queue: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
