When running GOOS=android go test <pkg>, the go tool will build the test(s)
for Android and run it through the exec wrapper. The exec wrappers takes care
of pushing the binary to the device and running it. That's how the normal
builder with an Android device attached works: setting GOOS=android.
Some x/mobile tests uses the Android SDK and then use an Android specific way
to run integration tests with Java. Those tests don't run on GOOS=android.
To solve that dilemma, define a special builder without GOOS=android set
and use it (only) for x/mobile.
Then, disable the GOOS=android builder for x/mobile to avoid wasting
device time better spent on the main repository.
Similar to CL 164740.
While here, correct two GOHOSTOS values and rename the existing
android builders to match the GOOS-GOARCH-suffix form.
Updates golang/go#30482
Change-Id: I3712584c62c3765d5605cccb642aa6f119826ae8
Reviewed-on: https://go-review.googlesource.com/c/build/+/164660
Run-TryBot: Elias Naur <mail@eliasnaur.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
By adding gradle and Android build tools to PATH, setting ANDROID_HOME,
and adding the OpenAL headers to CGO_CFLAGS.
Change-Id: Ifeaa1b82f4108afc3787eee4449b3317185bbaec
Reviewed-on: https://go-review.googlesource.com/c/build/+/164658
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Now that we hace Android emulator builders on GCE, drop the old
ones running on hardware.
Also update the configuration of the remaining builders now that
androidtest.bash is no longer necessary.
Change-Id: I09cab08b28952f967205d6c35f36eb6e05a17950
Reviewed-on: https://go-review.googlesource.com/c/164397
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Now that we can do nested virtualization on GCE, that means we can run
the Android emulator (which requires KVM) on GCE and at least get
fast trybots and such for android-amd64.
Updates golang/go#23824
Change-Id: I0da38c7fa0f15492230a31291d2921ba72f2151d
Reviewed-on: https://go-review.googlesource.com/c/163738
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
This change adds an extra linux-ppc64le builder running on POWER9.
Fixesgolang/go#30138
Change-Id: I23bdfc116bb89c6d96be330c2c87827c47705b6b
Reviewed-on: https://go-review.googlesource.com/c/163757
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
This adds scripts to create a new builder host image that acts like
Container-Optimized OS (has docker, runs konlet on startup) but with a
Debian 9 kernel + userspace that permits KVM for nested
virtualization.
Updates golang/go#15581 (solaris)
Updates golang/go#23060 (dragonfly)
Updates golang/go#30262 (riscv)
Updates golang/go#30267 (fuchsia)
Updates golang/go#23824 (android)
Change-Id: Ib1d3a250556703856083c222be2a70c4e8d91884
Reviewed-on: https://go-review.googlesource.com/c/163301
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
This adds a linux-amd64 COS builder that should be just like our
existing linux-amd64 COS builder except that it's using a forked image
that has the VMX license bit enabled for nested virtualization. (GCE
appears to be using the license mechanism as some sort of opt-in
mechanism for features that aren't yet GA; might go away?)
Once this is in, it won't do any new builds as regular+trybot builders
are disabled. But it means I can then use gomote + debugnewvm to work
on preparing the other four image types.
Updates golang/go#15581 (solaris)
Updates golang/go#23060 (dragonfly)
Updates golang/go#30262 (riscv)
Updates golang/go#30267 (fuchsia)
Updates golang/go#23824 (android)
Change-Id: Ic55f17eea17908dba7f58618d8cd162a2ed9b015
Reviewed-on: https://go-review.googlesource.com/c/162959
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
The new BuildConfig.MinimumGoVersion field specifies the minimum
Go version the builder is allowed to use. It's useful when some
of the builders are too new, and do not support all of the supported
Go releases (e.g., openbsd-amd64-64 and freebsd-amd64-12_0 currently
require Go 1.11 and don't work on Go 1.10).
It only needs to be set when a builder doesn't support all supported
Go releases, since we don't typically test unsupported Go releases.
To allow cmd/coordinator to use this field and filter out work it
receives from maintner/maintnerd's GoFindTryWork RPC call,
we add a GoVersion slice to apipb.GerritTryWorkItem,
and populate it in maintapi.apiService.GoFindTryWork method.
For trybots on the Go repo, the GoVersion field is determined from
the branch name. For "release-branch.goX.Y" branches, it parses out
the major-minor Go version from the branch name. For master and
other branches, it assumes the latest Go release.
For trybots on subrepos, we already have the Go version information
available, so use it directly.
Afterwards, all that's left is to modify newTrySet in cmd/coordinator
to make use of BuildConfig.MinimumGoVersion and work.GoVersion to skip
builders that are too new for the Go version that needs to be tested.
Fixesgolang/go#29265
Change-Id: I50b01830647e33e37e9eb8b89e0f2518812fa44f
Reviewed-on: https://go-review.googlesource.com/c/155463
Run-TryBot: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
I built and tested the v7 image from CL 152037 per the instructions in
env/plan9-386/README.
Change-Id: Ia4dcb1ac8429ab3e54e8ed21719d3ad894529927
Reviewed-on: https://go-review.googlesource.com/c/152238
Run-TryBot: Bryan C. Mills <bcmills@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Only the "oauth2" and "build" repos for now, as a test. We'll lock
down policy more later and decide when to do this automatically.
Also, this currently only runs buildlets which run in our GCP project,
because we're not yet proxying the a localhost:3000 port from the
reverse buildlets to an authenticated TLS connection back to our
module proxy service on GKE.
Updates golang/go#14594Fixesgolang/go#29637
Change-Id: I6f05da2186b38dc8056081252563a82c50f0ce05
Reviewed-on: https://go-review.googlesource.com/c/157438
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Bryan C. Mills <bcmills@google.com>
Reviewed-by: Andrew Bonventre <andybons@golang.org>
Install GDB from the ports collection, the in-tree was removed from the base system
in FreeBSD 12.0, and the older releases have an unsupported (6.1.1 or older version).
Updates golang/go#25289
Change-Id: I8899a2a70c0b9ee9171621206431ca24a7e97704
Reviewed-on: https://go-review.googlesource.com/c/155977
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Writing the snapshot to GCS is timing out for the Windows/ARM builder.
It tars about 500MB of data, resulting in about 175MB of data, which
it then sends to GCS. It takes about 4:30 to tar the workdir/go directory
with the built-in tar command, leaving only 30s to transfer 175MB
across the web, so it's not surprising that it times out.
Updates golang/go#26148
Change-Id: I7e70aeaf1ac6334c17fdca64ae9ab946c3d87d9a
Reviewed-on: https://go-review.googlesource.com/c/153518
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
We care most about testing the dev.boringcrypto branches on linux/amd64.
Stop testing them on all builders, as it's not worth the resource use.
Continue to test on one additional non-linux/amd64 port,
namely windows/386, to ensure that fallbacks work.
Fixesgolang/go#26791
Change-Id: I4556886f55175173de52e35cf57833b22843d586
Reviewed-on: https://go-review.googlesource.com/c/149739
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
This change takes care of a TODO comment. It makes the builder
configuration more centralized and contained in the dashboard
package.
Invert it for consistency with BuildConfig.BuildRepo method.
Updates golang/go#26791
Change-Id: I46368adadb85f2ec730da4fc0abe5fd6a112a7c7
Reviewed-on: https://go-review.googlesource.com/c/149738
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
By now, OpenBSD 6.4 has proven reliable, so we want it to be used
for building subrepos as well.
Updates golang/go#26833
Change-Id: I026207865514a026f86936140a16dd707f9594ea
Reviewed-on: https://go-review.googlesource.com/c/148888
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
OpenBSD 6.4 is the current OpenBSD release, while 6.2 has just stopped
being supported. The new 6.4 builder has proven to be stable and
reliable, so start using it for trybots as well.
Looking at the non-trybot build times for the Go repository with
OpenBSD 6.2 and 6.4 builders to get a sense of the build speed,
it seems OpenBSD 6.4 is about 15% slower (1200~ seconds on 6.2 vs
1380~ seconds on 6.4). This might be due to security-oriented changes
in OpenBSD 6.3 and 6.4 that were done to mitigate Spectre/Meltdown
vulnerabilities.
CL 145022 has helped recover some lost performance by re-enabling SMT.
OpenBSD trybots seem to take around 6 minutes now, so there's
some room to make them slower before they're the bottleneck.
Try to update for now, and if it becomes a bottleneck, then
we can prioritize investigating ways to improve OpenBSD 6.4
builder performance further if possible.
References:
• https://www.openbsd.org/64.html
• https://www.openbsd.org/63.html
• https://man.openbsd.org/sysctl.2#HW_SMT_2Fixesgolang/go#26833
Change-Id: I42c136fab800df46866a95017a8f70047b37f847
Reviewed-on: https://go-review.googlesource.com/c/144777
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
BuildConfig grew way too large and has too many slices & such to be a
value type. Make it a pointer type, which matches HostConfig.
Change-Id: Ie625bece9d6d8c1ec6cff26e77416bd9b1f256d8
Reviewed-on: https://go-review.googlesource.com/c/145077
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Disable Spectre/Meltdown mitigations and enable simultaneous
multithreading (SMT) on the OpenBSD 6.4 builder image.
This was done by setting hw.smt=1 in /etc/sysctl.conf when building
the image.
I considered an alternative approach of doing this at runtime via
a cmd/buildlet special case for OpenBSD. It would've likely been
faster to implement if I had the idea at the beginning. However,
by the time I saw it, I had already started to build a new OpenBSD 6.4
image with the /etc/sysctl.conf file, then tested it via debugnewvm.
At this point, there's no longer a time saving advantage, so I decided
to prefer the v2 image because it keeps OpenBSD-specific configuration
more contained in env/openbsd-* directory, rather than spreading it out
between there and the cmd/buildlet runtime code. It seems to be a
slightly better option.
The times to do a full build, as measured via cmd/debugnewvm, were:
6.2 = 19m25.218s
6.4 = 28m49.565s
6.4 with hw.smt=1 = 22m55.909s
That should translate to faster trybot runs, which is important
for open source project health.
The 386 environment doesn't need to be updated because it doesn't
support hw.smt:
sysctl: hw.smt: value is not available
Attempting to set it anyway should be harmless and okay to do,
in case it happens to get supported in the future.
We're only using amd64 environment for trybots, so it's okay
for the purposes of golang/go#28403.
Reference: https://man.openbsd.org/sysctl.2#HW_SMT_2Fixesgolang/go#28403
Change-Id: I7bc4cbf83ccbdb3aa9dd19eeabd88feb1c425811
Reviewed-on: https://go-review.googlesource.com/c/145022
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Previously, they were running a 6.4 snapshot release from October 4, 2018,
added in CL 139857.
The final release of OpenBSD 6.4 has been released on October 18, 2018,
and can be used now. See https://www.openbsd.org/64.html.
Fixup unbound variable in env/openbsd-amd64/make.bash:
$ ./make.bash
./make.bash: line 57: PKG_ADD_OPTIONS: unbound variable
The amd64 architecture image was tested with cmd/debugnewvm
and passed all tests successfully.
Updates golang/go#26833.
Change-Id: I0e3b1ed5aa9b2e624d9afbe46a8a344f88ddcc56
Reviewed-on: https://go-review.googlesource.com/c/143458
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
This change adds amd64 and 386 OpenBSD 6.4 snapshot builders.
The snapshots are from October 4, 2018:
/pub/OpenBSD/snapshots/amd64/BUILDINFO:
Build date: 1538669162 - Thu Oct 4 16:06:02 UTC 2018
/pub/OpenBSD/snapshots/i386/BUILDINFO:
Build date: 1538681766 - Thu Oct 4 19:36:06 UTC 2018
These are meant to be replaced with the final version of OpenBSD 6.4
when it is released.
Modify make.bash to support fetching snapshot versions.
The amd64 architecture image was tested with cmd/debugnewvm
and passed all tests successfully.
Updates golang/go#26833.
Change-Id: Iea456605ace08f25d4eba37f9a962e6b9af95369
Reviewed-on: https://go-review.googlesource.com/c/139857
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Not yet the defaults or trybots until we see they're working regularly
for a few days and compare timings, etc.
Updates golang/go#25289
Change-Id: I4baed066b5f6d9f01e5454b716cacda4986620d0
Reviewed-on: https://go-review.googlesource.com/138258
Reviewed-by: Yuval Pavel Zholkover <paulzhol@gmail.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
This is a starting point for adding an aix-ppc64 reverse builder.
There can be further adjustments made to it as needed.
Updates golang/go#27160.
Change-Id: I8a61adbe21f1a08265df9a206d94bb04e880a508
Reviewed-on: https://go-review.googlesource.com/131335
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
plan9-386-v6 is a new plan9-386 image built with the change
done in CL 129098.
I tested it with cmd/debugnewvm on latest Go master commit,
and all stdlib tests passed, including TestVariousDeadlines.
(cmd/... tests didn't finish in time before cmd/debugnewvm
timed out.) Running it 3 more times, there was still one
TestVariousDeadlines failure and 2 successes, so we'll have
to keep watching it.
Updates golang/go#26945.
Change-Id: Ibd1e990e60c3a69a5fddede18d7a121fa319e79d
Reviewed-on: https://go-review.googlesource.com/130137
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
We don't need to run all the tests in a trybot. Run just enough to get
good coverage, without going over 5 minutes.
Any regressions elsewhere will be caught by the build.golang.org
(slower) runs.
Change-Id: I32f1fc17681bfb509844d2bd35b05c950806d283
Reviewed-on: https://go-review.googlesource.com/121938
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Windows Defender scans all new files written to disk which slows down
build performance.
- Uninstall Windows Defender as part of image building
- Refactor/Cleanup registry section for consistency
- Promote new image version to dashboard (built in staging/prod)
fixesgolang/go#26055
Change-Id: I3e16b9a9581824c20abed5e8ffae1efd46c6dd09
Reviewed-on: https://go-review.googlesource.com/121937
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Jakob Borg kindly supplied a new iPhone 6 to replace the ailing
iPhone 5 (arm) and iPhone 6 (arm64) that currenctly runs the iOS
builders.
The new phone runs iOS 10 that can run both arm and arm64, so
instead of replacing the device ids specified with GOIOS_DEVICE_ID,
simply remove them altogether.
Change-Id: I36cf574a902b953a3d2bf2a5de50741dff1b940d
Reviewed-on: https://go-review.googlesource.com/119415
Run-TryBot: Elias Naur <elias.naur@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
s390x notably was missing. Also add the mobile ones.
Change-Id: Id6be52b73fac897980d4e94ac5c5b07027f5b4e7
Reviewed-on: https://go-review.googlesource.com/114315
Reviewed-by: Bryan C. Mills <bcmills@google.com>
This change downloads/extracts the Visual Studio Build Tools to C:\godep\vs
Verified by running C:\godep\vs\VC\Tools\MSVC\14.14.26428\bin\Hostx64\x86\cl.exe
Related: golang/go#20982
Change-Id: I9dcb4afc0971317ff39381f089867a08c402974a
Reviewed-on: https://go-review.googlesource.com/112036
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Once containers run on COS instead of Kubernetes, one name (Kube*) is
wrong and the other (GCE) is ambiguous. So rename them now to be more
specific.
No behavior changes. Just renaming in this step, to reduce size of
next CL.
Updates golang/go#25108
Change-Id: Ib09eb682ef74acbbf6ed50b46074f834ef5e0c0b
Reviewed-on: https://go-review.googlesource.com/111639
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Rename the 8branch builders from “8branch” to “8_0” (in keeping with
the version suffixes for FreeBSD and Darwin), and fix an apparent typo
in the 386 SSHUsername field.
Updates golang/go#25206.
Updates golang/go#25109.
Change-Id: I6053c8de17dcb3c7ff32da05c51902ac4bcc45fe
Reviewed-on: https://go-review.googlesource.com/110626
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
cmd/go takes ~188 seconds on windows-amd64, which is over the
default 180 second limit. Bump up the timout rate.
Change-Id: Ib0e5289ab6c4d6127d18375e3a72453818b984a8
Reviewed-on: https://go-review.googlesource.com/110415
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Go 1.11 will not support Windows XP, as pre-announced (per policy) at
https://golang.org/doc/go1.10#ports which says:
Go 1.10 is the last release that will run on Windows XP or Windows
Vista. Go 1.11 will require Windows 7 or later.
It was fun while it lasted.
Fixesgolang/go#10267
Change-Id: I1929088be00c938621cc106acdb276ffcb93c8cc
Reviewed-on: https://go-review.googlesource.com/95576
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Add an android/386 and an android/amd64 builder, both running on an
emulator. Use the new GOANDROID_ADB_FLAGS to direct the build to
either a device (android/arm, android/arm64) or an emulator
(android/386, android/amd64).
Change-Id: I8e25aa5efcbbb186c71c4e108a37ae11b0246c98
Reviewed-on: https://go-review.googlesource.com/88775
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
The slow test directory was previously skipped in 7b0a998e (CL
82395) for the slower + VM builders, but I forgot Windows XP.
The test directory is slow and doesn't have port-specific code. It's
mostly low-level compiler tests. We can safely skip it on most
builders, especially the slow & scarce ones.
Change-Id: I5c830509628e7eb1ab9649d31a47970afe135c33
Reviewed-on: https://go-review.googlesource.com/84335
Reviewed-by: Andrew Bonventre <andybons@golang.org>
This adds a darwin-amd64-race builder. Of the four platforms that
support the race detector, this was the only one missing a builder.
To make up for the extra work, skip some redundant tests that the Mac
(and other) builds do. The test directory is really slow, and doesn't
have platform-specific code. We have amd64 & 386 test coverage for
that directory via the Linux builders, which is fine. No need to do it
on some of the heavier VM builders, or for trybots.
Fixesgolang/go#17674
Change-Id: I888119b3fdd28884b87a59d78746ab1fef6960c7
Reviewed-on: https://go-review.googlesource.com/82395
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Also update the bootstrap toolchain (verified in VMware VM locally).
Updates golang/go#20852
Change-Id: Iac875419ed8dc51fb9db09d8948067f24c601e17
Reviewed-on: https://go-review.googlesource.com/82018
Reviewed-by: Andrew Bonventre <andybons@golang.org>
Until issues are fixed. No point in wasting resources on column of
red.
Updates golang/go#22968
Change-Id: I2de5a4d8aca6fcbdab2f43d87818c7d8f333f8bc
Reviewed-on: https://go-review.googlesource.com/81637
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Also clean up some redundant comments, move trybots from OpenBSD
6.0 to 6.2, and add some missing MaxAtOnce caps.
Updates golang/go#20852
Change-Id: I3d862ca7a7172952ec2254b583e4bf6102b026e6
Reviewed-on: https://go-review.googlesource.com/80876
Reviewed-by: Andrew Bonventre <andybons@golang.org>
freebsd.org says that the current supported releases are:
> Production: 11.1, 11.0, 10.4, 10.3
Previously we had builders for 9.3, 10.1, and 11.0.
We'll keep FreeBSD 9.3 for now (as it's still supported by the last
two released Go versions), but we'll stop testing FreeBSD 9.3 at tip
shortly here (that will be a separate change).
This CL updates our builders from 11.0 to 11.1 and 10.1 to 10.3,
spanning the range from the old to newest supported release. We'll
stop testing 10.4 and 11.0, as that's a bit overkill.
Another reason we're doing 10.3 instead of 10.4 is because the Go 1.9
release notes (https://golang.org/doc/go1.9#freebsd) said that future
Go releases (Go 1.10) will require FreeBSD 10.3+, so we need to test
FreeBSD 10.3, being the oldest. (Go has historically "supported"
FreeBSD releases a bit longer than they're officially supported by
FreeBSD, partly due to our out-of-sync release cycles.)
Fixesgolang/go#19303Fixesgolang/go#22854
Change-Id: I69589000b1c87aecf054857f514cc48ccedf752f
Reviewed-on: https://go-review.googlesource.com/80297
Reviewed-by: Andrew Bonventre <andybons@golang.org>
Its image has been rebuilt with latest upstream NetBSD changes.
(See CLs 64750, 77475)
Updates golang/go#20852
Change-Id: I8d68313326ec29a97731d2a79198dfdae5b789a6
Reviewed-on: https://go-review.googlesource.com/79436
Reviewed-by: Benny Siegert <bsiegert@gmail.com>
Reviewed-by: Andrew Bonventre <andybons@golang.org>
It's doing no good being red all the time, especially as nobody is
working on it.
Updates golang/go#22689
Updates golang/go#19938
Change-Id: Icb947878d85e920f24ea458eb0f319844ca5bd60
Reviewed-on: https://go-review.googlesource.com/78575
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Enable public key auth for windows buildlets. Previously the
username/password was hardcoded.
This change enables the `/connect-ssh` http endpoint on the buildlet to setup
authorized_keys and uses the coordinator to proxy ssh traffic to the
buildlet.
Validation:
- run local coordinator (with patches to dev env to act like GCE) with
changes
- build windows buildlet, post to own bucket
- execute:
gomote -localdev create windows-amd64-2012
ssh -p 2222 user-jrjohnson-windows-amd64-2012-0@127.0.0.1
Fixesgolang/go#21247
Change-Id: If7adc7e6296a071d50412d9fa9d46ecd9e6cc3f3
Reviewed-on: https://go-review.googlesource.com/52350
Reviewed-by: Alex Brainman <alex.brainman@gmail.com>
The newly restored iPhone 6 device now runs iOS 11 which can't run
arm binaries. Add another device (iPhone 5) for darwin/arm builds
and use GOIOS_DEVICE_ID to distinguish between them.
While we're here, update the model numbers to reflect reality.
Change-Id: Iaef33469d19b99354d0dd72577d31289d179c82c
Reviewed-on: https://go-review.googlesource.com/67090
Reviewed-by: David Crawshaw <crawshaw@golang.org>
Running multiple iOS devices on a single host was too ambitious for
the brittle test harness. This CL reverts most of the split from CL
57297 except the device model names are updated to reflect reality.
I'm sorry.
Change-Id: I4e9b1d677eb7c57178f112dedb62580fc74a461d
Reviewed-on: https://go-review.googlesource.com/59510
Reviewed-by: Sarah Adams <shadams@google.com>
The iOS builder is horribly slow. Speed it up by splitting the
single builder in two, one for arm and one for arm64.
Change-Id: Iabe5f305433c5c613b71712609a339dc2d6abc89
Reviewed-on: https://go-review.googlesource.com/57297
Reviewed-by: Sarah Adams <shadams@google.com>
I didn't realize that the freebsd-arm builder was actually called
freebsd-arm-paulzhol before and I mistakenly created the replacement
buildet-based builder as "freebsd-arm".
Name it how it was.
Change-Id: Ib7d73947ce599c82c628ab323001213e1b4ee20b
Reviewed-on: https://go-review.googlesource.com/53358
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
No point snapshotting for non-sharded builders.
And snapshotting is too slow and hits the timeout on freebsd-arm.
Updates golang/go#21191
Change-Id: I225aa6af493b92aa45589491c69853e13d513dd1
Reviewed-on: https://go-review.googlesource.com/53493
Reviewed-by: Yuval Pavel Zholkover <paulzhol@gmail.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
The GCE bug is fixed and these VMs boot again.
(Previously: golang.org/cl/50011)
Change-Id: I3f330b8d0c6a6ae6708b2299bba13945a06fa36f
Reviewed-on: https://go-review.googlesource.com/53270
Reviewed-by: Herbie Ong <herbie@google.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Brad Fitzpatrick
Elias Naur
Carlos Eduardo Seo
Dmitri Shuralyov
Bryan C. Mills
Yuval Pavel Zholkover
Jordan Rhee
Tobias Klauser
Quentin Perez
Jeff Johnson
Elias Naur
Alberto Donizetti
Andrew Bonventre