The step_result json can currently be empty if the orchestrator build
is canceled while running bot_update. Trying to do
step_result.json.output in the finally block raises an "Uncaught
exception AttributeError".
Ex: https://ci.chromium.org/ui/p/chromium/builders/try/win10_chromium_x64_rel_ng/1121568/overview
This change checks for the json attr before trying to read json.output.
Bug: 1305332
Change-Id: I1dc277642ece1b96144be6381a7c4b8426cb2bc6
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3518372
Reviewed-by: Garrett Beaty <gbeaty@google.com>
Reviewed-by: Josip Sokcevic <sokcevic@google.com>
Commit-Queue: Stephanie Kim <kimstephanie@google.com>
This reverts commit 8b64ddec2a.
Reason for revert: it appears we need this, since .DEPS.git
specifies custom dependencies needed for tryjob.
Original change's description:
> Remove unused functionality in bot_update.py
>
> .DEPS.git has been removed in 2014 and no longer supported. It was
> deleted in https://codereview.chromium.org/497313003 (August 2014). Last
> entry in chromium/src is from tag 44.0.2403.47 in June 2015.
>
> Recipe-Nontrivial-Roll: build
> Recipe-Nontrivial-Roll: build_limited
> Recipe-Nontrivial-Roll: chrome_release
> Recipe-Nontrivial-Roll: chromiumos
> Recipe-Nontrivial-Roll: infra
> Change-Id: I3ffb8813af867df959098c02d399c8d745894aae
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3439573
> Reviewed-by: Aravind Vasudevan <aravindvasudev@google.com>
> Commit-Queue: Josip Sokcevic <sokcevic@google.com>
Change-Id: I36bcd352a984b4b762096197a384e86347418290
Recipe-Nontrivial-Roll: build
Recipe-Nontrivial-Roll: build_limited
Recipe-Nontrivial-Roll: chrome_release
Recipe-Nontrivial-Roll: chromiumos
Recipe-Nontrivial-Roll: infra
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3448707
Auto-Submit: Josip Sokcevic <sokcevic@google.com>
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
.DEPS.git has been removed in 2014 and no longer supported. It was
deleted in https://codereview.chromium.org/497313003 (August 2014). Last
entry in chromium/src is from tag 44.0.2403.47 in June 2015.
Recipe-Nontrivial-Roll: build
Recipe-Nontrivial-Roll: build_limited
Recipe-Nontrivial-Roll: chrome_release
Recipe-Nontrivial-Roll: chromiumos
Recipe-Nontrivial-Roll: infra
Change-Id: I3ffb8813af867df959098c02d399c8d745894aae
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3439573
Reviewed-by: Aravind Vasudevan <aravindvasudev@google.com>
Commit-Queue: Josip Sokcevic <sokcevic@google.com>
User could attach CC addresses to the gerrit change created.
BUG=1294382
TEST=train
Change-Id: I236658d80f41b27bf007bb344934ce835cd7b5de
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3440862
Reviewed-by: Josip Sokcevic <sokcevic@google.com>
Commit-Queue: Xinan Lin <linxinan@chromium.org>
The get_files_affected_by_patch method on the test API will return a
TestData object that determines the output of the "git diff" call. a
Test was added specifically for the get_files_affected_by_patch API
method to ensure that the test API operates as expected.
Change-Id: Ia1846700461757a72ed518894a5a38a8ca54fcc9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3422006
Auto-Submit: Garrett Beaty <gbeaty@google.com>
Reviewed-by: Gavin Mak <gavinmak@google.com>
Commit-Queue: Garrett Beaty <gbeaty@google.com>
gclient removed syntax validation flag in 2020:
https://crrev.com/c/2083589, so this can safely be removed.
R=aravindvasudev@google.com, gavinmak@google.com
Recipe-Nontrivial-Roll: chrome_release
Recipe-Nontrivial-Roll: build_limited
Recipe-Nontrivial-Roll: build
Change-Id: I2aced6cdb01d571ca361cc57e7c8715c9ca40552
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3419278
Reviewed-by: Gavin Mak <gavinmak@google.com>
Commit-Queue: Josip Sokcevic <sokcevic@google.com>
The tryserver.require_is_tryserver method will create an infra-failing
step that raises an exception if there is no CL associated with the
build. In the case of an LED task, it will only fail so as not to cause
the removal of the builder cache in the case of a minor user error.
Change-Id: I05d3a86cc727577f067fb3eae49e46398345f672
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3413971
Auto-Submit: Garrett Beaty <gbeaty@google.com>
Reviewed-by: Gavin Mak <gavinmak@google.com>
Commit-Queue: Gavin Mak <gavinmak@google.com>
Those paremeters are deprected for over 4 years now. I found no usages
while searching for output_manifest in source.chromium.org.
R=machenbach@google.com
Bug: 1289274
Change-Id: I7d8f8ee29fdb046b255c96ad26f8dc7129b1f83a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3402741
Auto-Submit: Josip Sokcevic <sokcevic@google.com>
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
"Publish Change Edit" is not a synchronized API call on Gerrit.
After the call, we need to explicitly check if the new patch exists
in Gerrit. Or, we may label +1 on an old patch.
BUG=1286212
TEST=train
Recipe-Nontrivial-Roll: chrome_release
Change-Id: I1443e2b4cf702f7cb4feebb43e34923f11383f35
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3384422
Reviewed-by: Josip Sokcevic <sokcevic@google.com>
Commit-Queue: Xinan Lin <linxinan@chromium.org>
The blamelist pins are produced by iterating over a dictionary which
yields a different order between python2 and python3. The blamelist pins
will now be produced with a consistent ordering between python versions
by sorting on the revision name.
Change-Id: Ieb6c4dfe0aacea83c4a86a3a10380cc2f3839114
Recipe-Nontrivial-Roll: build
Recipe-Nontrivial-Roll: build_limited
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3352250
Reviewed-by: Josip Sokcevic <sokcevic@google.com>
Commit-Queue: Garrett Beaty <gbeaty@google.com>
These whould use XCode13
Change-Id: Idb1f0a30ed1c37e421f2d7fb05f5e6f1e017c431
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3340948
Reviewed-by: Jeff Yoon <jeffyoon@google.com>
Auto-Submit: Olivier Robin <olivierrobin@chromium.org>
Commit-Queue: Olivier Robin <olivierrobin@chromium.org>
Calling git number requires that that depot tools is on the path so that
git-number can be found and requires the CHROME_HEADLESS environment
variable to be set.
Change-Id: Iabb8ec6732ca2bcc0f3209449e23a7d0a4f629e3
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3314033
Auto-Submit: Garrett Beaty <gbeaty@google.com>
Reviewed-by: Gavin Mak <gavinmak@google.com>
Commit-Queue: Gavin Mak <gavinmak@google.com>
SetEnv json entries in cipd packages for more recent Windows SDKs are
changed from:
[
"..",
"..",
"win_sdk",
"Include",
"10.0.17134.0",
"um"
],
to
[
"Windows Kits",
"10",
"Include",
"10.0.19041.0",
"um"
],
so we should accept these paths now.
R=iannucci@chromium.org
Bug: 1250098
Change-Id: I56c84bc5ed46dda354e018b0abac2fae0bad8199
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3296663
Reviewed-by: Robbie Iannucci <iannucci@chromium.org>
Commit-Queue: Alex Gough <ajgo@chromium.org>
The step API now supports the same functionality that can_fail_build
provides, so standardize on that argument rather than maintaining a
duplicate mechanism.
Bug: 1274644
Change-Id: Idcc3c0dacbab8d2d786b27ec277713063da6d334
Recipe-Manual-Change: build_limited
Recipe-Manual-Change: infra
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3301439
Reviewed-by: Josip Sokcevic <sokcevic@google.com>
Commit-Queue: Garrett Beaty <gbeaty@google.com>
In order to enforce that the test data provided for placeholders is of
the appropriate type (text vs binary), existing mismatches must be
resolved. This will prevent type mismatch issues from occurring when
switching builders to use python3.
Bug: 1273548
Change-Id: I83da25432b67853992d7771aba558fd38520a24e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3300310
Reviewed-by: Gavin Mak <gavinmak@google.com>
Commit-Queue: Garrett Beaty <gbeaty@google.com>
This is an automated CL created by the recipe roller. This CL rolls
recipe changes from upstream projects (recipe_engine) into this repository.
The build that created this CL was
https://ci.chromium.org/b/8829745138191541841
Please review the expectation changes, and LGTM+CQ.
recipe_engine:
206cd301ad
206cd30 (gbeaty@chromium.org)
Remove the log from write_raw.
More info is at https://goo.gl/zkKdpD. Use https://goo.gl/noib3a to file a bug.
R=dpranke@google.com
Recipe-Tryjob-Bypass-Reason: Autoroller
Bugdroid-Send-Email: False
Change-Id: Ifb9fe37d3d17fdb9b13ddab317c3357bf4e014b1
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3298693
Auto-Submit: Recipe Roller <recipe-mega-autoroller@chops-service-accounts.iam.gserviceaccount.com>
Commit-Queue: Garrett Beaty <gbeaty@google.com>
Reviewed-by: Garrett Beaty <gbeaty@google.com>
Previous CL[0] had the paths inside the SDK wrong so did not
correctly find the SetEnv.x64.json files in recent cipd packaged
SDKs. The 'win_sdk' portion was not necessary.
Downloading the cipd packages and verifies the old and new paths:
gci . -File -Recurse | Where-Object {$_.Name -match 'SetEnv.x64.json'}
CPID-----------------\
C:\src\dt\tmp\new-sdk\ Windows Kits\10\bin\SetEnv.x64.json
C:\src\dt\tmp\old-sdk\ win_sdk\bin\SetEnv.x64.json
[0] https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3283800R=iannucci@chromium.org
Bug: 1250098
Change-Id: I579fd288824ca9014176b7b1c221affd966fb953
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3290913
Commit-Queue: Robbie Iannucci <iannucci@chromium.org>
Reviewed-by: Robbie Iannucci <iannucci@chromium.org>
This will make it easier to call new or uncommon Gerrit APIs, without
needing to add dedicated methods for each one. This also makes it easier
to debug API calls by exposing the raw response data, which is often
processed or filtered in the existing wrapper methods.
BUG=1052179
R=linxinan@chromium.org
Change-Id: I0b594a9fdaabae4b0ebb379dc5473f4d6de67a13
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3235714
Auto-Submit: Michael Moss <mmoss@chromium.org>
Reviewed-by: Dirk Pranke <dpranke@google.com>
Reviewed-by: Xinan Lin <linxinan@chromium.org>
Commit-Queue: Dirk Pranke <dpranke@google.com>
If one uses update_files() method to file a chain of CLs and wants
to submit them together, they have to set submit=False to wait for the
last CL. As a result, users have to run `set Bot-Commit+1` for each CL
at their side, which is annoying.
This change makes the gerrit module to handle that, and callers only
need to submit the chain.
BUG=1261724
TEST=train
Change-Id: If7d420d8d3c40b1d462284395cdc87ac680d0918
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3235388
Commit-Queue: Xinan Lin <linxinan@chromium.org>
Reviewed-by: Dirk Pranke <dpranke@google.com>
Reviewed-by: Michael Moss <mmoss@chromium.org>
Only hanging process recorded so far is gsutil. Sending sigterm doesn't
result in successfully terminating the process. Instead, use sigkill.
R=gavinmak@google.com
Bug: 1255228
Change-Id: Iad354284422676df804c68cb6f18c7db723bf862
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3232250
Auto-Submit: Josip Sokcevic <sokcevic@google.com>
Reviewed-by: Gavin Mak <gavinmak@google.com>
Commit-Queue: Gavin Mak <gavinmak@google.com>
This change removes the need for the Gerrit plugin recipe to clone the
Gerrit repo itself.
Recipe-Nontrivial-Roll: infra
Bug: 1196790
Change-Id: I3d6bb71a2e95c50c4dc29345ad24e482177233d5
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3224087
Commit-Queue: Gavin Mak <gavinmak@google.com>
Reviewed-by: Josip Sokcevic <sokcevic@google.com>
It's generaly not useful and makes it harder to understand bot_update
logs. If we need to debug specific builders, we can set environment
variables in the recipe itself.
R=apolito@google.com
Bug: 1255228
Change-Id: I18fc6ee8cb15b30bc3a4422ddd04979189bacf64
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3221250
Commit-Queue: Anthony Polito <apolito@google.com>
Reviewed-by: Anthony Polito <apolito@google.com>
Patch failures shouldn't be treated as infra_failures (as long as patch
was successfully downloaded).
Prior to https://crbug.com/1207685, bot_update patch failurs were set as
successful step to work around buildbucket plugin requirements.
One alternative approach is to add necessary information into patch
failure step, by editing its presentation.
R=gavinmak@google.com
Bug: 1207685, 1255228
Change-Id: I696356d40a80536c3efbca99091536b8b25deb40
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3219890
Reviewed-by: Gavin Mak <gavinmak@google.com>
Commit-Queue: Josip Sokcevic <sokcevic@google.com>
This change starts observers as soon as possible, and wraps the entire
block that can cause an exception in try..finally. This ensures that we
stop threads are stopped.
If we notice the same behavior, it could be worth to put a cap on number
of executions per observer, and shut the thread down once reached.
R=apolito@google.com
Bug: 1255228
Change-Id: I1f165267460da02b3bbba39022c262e6c29fe86b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/tools/depot_tools/+/3213451
Reviewed-by: Anthony Polito <apolito@google.com>
Commit-Queue: Josip Sokcevic <sokcevic@google.com>
Stephanie Kim
Xinan Lin
Aravind Vasudevan
Ravi Mistry
Chidera Olibie
Garrett Beaty
Andrew Grieve
Josip Sokcevic
Gavin Mak
Brian Ryner
Olivier Robin
Alex Gough
Anthony Polito
recipe-roller
Robert Iannucci
Scott Lee
Michael Moss