Darwin doesn't save/restore the AVX-512 mask registers when
processing a signal. Go generates lots of signals, which cause
the mask registers to randomly get clobbered. Better to pretend
that AVX-512 isn't supported.
Update golang/go#49233
Change-Id: I0dda77a6f205e10eefccd82e6f4f0f184dae42d9
Reviewed-on: https://go-review.googlesource.com/c/sys/+/361255
Trust: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Martin Möhrmann <martin@golang.org>
The raw syscall returned the two pipes whereas the libc call
takes a pointer to a location to write the two pipes.
When we switched over from raw syscalls to libc calls, this
change in behavior was missed.
Fixesgolang/go#43498
Change-Id: Icee2204dcb8be8fc94be0df106e1ff061cafa446
Reviewed-on: https://go-review.googlesource.com/c/sys/+/281432
Trust: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
That CL worked around a bug present in the
OS X Yosemite Public Beta versions 1 and 2.
Beta 3 (released today) has fixed the bug.
LGTM=bradfitz
R=golang-codereviews, bradfitz
CC=golang-codereviews
https://golang.org/cl/144010043
These are being deleted from go.sys because in general they can
only be implemented in close coordination with the runtime.
LGTM=0intro
R=golang-codereviews, 0intro
CC=golang-codereviews
https://golang.org/cl/132810043
These are being deleted from go.sys because in general they can
only be implemented in close coordination with the runtime.
LGTM=dave
R=rsc, dave, josharian
CC=golang-codereviews
https://golang.org/cl/129500043
These are being deleted from go.sys because in general they can
only be implemented in close coordination with the runtime.
LGTM=alex.brainman
R=alex.brainman
CC=golang-codereviews
https://golang.org/cl/132010043
If we use a local type, it won't compare properly with errors from
the rest of the standard library. Errors are the one type from syscall
that propagates through the system, so it's important to have only
one type for them.
Ditto for syscall.Signal.
LGTM=dave
R=rsc, dave
CC=golang-codereviews
https://golang.org/cl/123490043
If we use a local type, it won't compare properly with errors from
the rest of the standard library. Errors are the one type from syscall
that propagates through the system, so it's important to have only
one type for them.
mkerrors_windows.sh is gone, so:
- rename zerrors_windows.go and delete its DO NOT EDIT mark
- delete the contentless zerrors_windows_*.go files
LGTM=rsc
R=rsc
CC=alex.brainman, golang-codereviews
https://golang.org/cl/122600043
If we use a local type, it won't compare properly with errors from
the rest of the standard library. Errors are the one type from syscall
that propagates through the system, so it's important to have only
one type for them.
Will apply the corresponding change to the other packages once
this is approved.
Also delete some deprecated code. This is a new package; no need
to carry the past.
LGTM=rsc
R=rsc
CC=golang-codereviews
https://golang.org/cl/126250043
For this particular system call, which uses the net package,
the expected error is syscall.EPERM not unix.EPERM.
With this change, go test passes on my linux.
LGTM=iant, rsc
R=rsc, iant
CC=golang-codereviews
https://golang.org/cl/125270043
Also provide stubs to call loadlibrary and getprocaddress, which are implemented
for package syscall by runtime/syscall_windows.goc.
Should get the windows go.sys build closer to green.
LGTM=rsc
R=golang-codereviews, rsc
CC=golang-codereviews
https://golang.org/cl/126170043
They were missed yesterday when adding the tags for all the other unix variants.
LGTM=aram
R=rsc, aram
CC=golang-codereviews
https://golang.org/cl/126150043
The environment is global state that is owned by the standard syscall package.
With this change, go test passes on darwin in the unix directory.
LGTM=rsc
R=rsc
CC=golang-codereviews
https://golang.org/cl/129900043
It's a peculiar environment that probably doesn't belong here.
We can bring it back easily if we need it.
LGTM=dave, rsc
R=rsc, dave
CC=golang-codereviews
https://golang.org/cl/128110043
Darwin/amd64 now builds. No others tested yet.
mkall.sh also runs correctly. The updates, which are
substantial, are not yet applied to avoid unnecessary churn.
Restore mkerrors.sh, which is needed in the unix directory.
LGTM=rsc
R=rsc
CC=golang-codereviews
https://golang.org/cl/122410044
Semi-automatic migration from package syscall to package {plan9,windows,unix}.
No builds attempted yet, but this gets a lot of noise behind us so subsequent
CLs will be more concise and easier to follow.
Subsequent CLs will have semantic content.
LGTM=rsc
R=golang-codereviews, rsc
CC=golang-codereviews
https://golang.org/cl/121520043
This CL copies to each package of go.sys the files from syscall it will need.
Different directories have different files, but these:
mkall.sh
str.go
syscall.go
mksyscall.pl
race.go
race0.go
syscall_test.go
are copied to all three.
No changes yet, these are just copies. They are not ready to use yet:
package names are wrong, for starters. But this clean copy will make
it easier to follow the changes as the packages are enabled.
LGTM=rsc
R=golang-codereviews, rsc
CC=golang-codereviews
https://golang.org/cl/126960043
This subrepo is the ultimate home for some of the experimental components
being developed in go.exp now, and also existing pieces such as go vet.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/9481043
Keith Randall