Nice side-effect of this change is simplifying the code.
Also added regression test coverage.
Bug: 570091
Change-Id: I470e9efc319632f997b02d210483988c17a7d3c8
Reviewed-on: https://chromium-review.googlesource.com/600369
Reviewed-by: Michael Moss <mmoss@chromium.org>
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Paweł Hajdan Jr. <phajdan.jr@chromium.org>
- allow variables with same names as long as values are also the same
- allow recursedeps entries referencing deps_os entries
Bug: 570091
Change-Id: I4abbd255982368a0ee33e8f3de631f3d18adf6ab
Reviewed-on: https://chromium-review.googlesource.com/583649
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Paweł Hajdan Jr. <phajdan.jr@chromium.org>
The update_depot_tools script had a line where a call out
to the test operator ([]) wasn't quoting all of the arguments,
leading to errors if a directory had spaces in the name.
R=dnj@chromium.org, phajdan.jr@chromium.org
Bug:
Change-Id: Ie1f5f51891a2582614ed3d9b16375d2ad522360b
Reviewed-on: https://chromium-review.googlesource.com/598427
Reviewed-by: Daniel Jacques <dnj@chromium.org>
Commit-Queue: Daniel Jacques <dnj@chromium.org>
This is part of preparation for gclient flatten, conditions,
and improving internal gclient workflows.
This also ensures variables have consistent values in the entire checkout,
unlike the current situation where different DEPS files can use different
values for the same variable.
We override instead of erroring out, since otherwise commonly used dependencies
such as buildtools wuld have to be rolled simultaneously in unrelated projects,
say v8 and native_client.
Another reason for overriding are possible future use cases such as is_chromium_build,
which could be set to false e.g. in v8, but chromium DEPS would override it to true.
Bug: 570091
Change-Id: If61322dd9a55e57729a4100d05ecb9c1e3ae8e38
Reviewed-on: https://chromium-review.googlesource.com/596310
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Paweł Hajdan Jr. <phajdan.jr@chromium.org>
Have the "update_depot_tools" script also do a CIPD tool sync. This will
ensure that users and systems have access to tooling at sync-time rather
than just-in-time loading them at execution time.
Update the tool boostraps to suppress any sort of syncing logs, if it
does happen. This will ensure that users who execute the tools don't see
unexpected output.
BUG=chromium:748651
TEST=local
- Tested on Mac and Windows.
Change-Id: I1aad897d885a07beeac40a372a658681720efd2a
Reviewed-on: https://chromium-review.googlesource.com/591229
Commit-Queue: Daniel Jacques <dnj@chromium.org>
Reviewed-by: Robbie Iannucci <iannucci@chromium.org>
Bump the CIPD client version to
b482bfa19d3fa79b4b40dd5ac7a34479052e3df1. This version inculdes support
for the "-log-level" flag.
Update CIPD bootstrap scripts and tool installation scripts to log at
warning level, so that standard operation doesn't produce any logs.
This version differs from the previous CL in that the "selfupdate"
command no longer uses "-log-level". This fixes an issue where the
current CIPD, which doesn't support "-log-level", is given the flag
prior to actually updating to a version that does.
BUG=chromium:748651
TEST=canary
- Tested this CIPD version on canary waterfall successfully.
Change-Id: I69dfe522d890eb307214f18d0467ce921a766446
Reviewed-on: https://chromium-review.googlesource.com/589568
Commit-Queue: Daniel Jacques <dnj@chromium.org>
Reviewed-by: Robbie Iannucci <iannucci@chromium.org>
This reverts commit 18aa2d93ed.
Reason for revert: <INSERT REASONING HERE>
Original change's description:
> [cipd] Bump version, log at warning.
>
> Bump the CIPD client version to
> e2a98b41adb3403b5b7771c4101b42a55dc2dd9e. This version inculdes support
> for the "-log-level" flag.
>
> Update CIPD bootstrap scripts and tool installation scripts to log at
> warning level, so that standard operation doesn't produce any logs.
>
> BUG=chromium:748651
> TEST=canary
> - Tested this CIPD version on canary waterfall successfully.
> - Tested "gclient" bootstrap and "vpython" / "led" on Windows system,
> no output and works.
> - Tested "vpython" / "led" on OSX system, no output and works.
>
> Change-Id: Iae316908283c43c504f293aafdf897df83dccf8d
> Reviewed-on: https://chromium-review.googlesource.com/588889
> Reviewed-by: Robbie Iannucci <iannucci@chromium.org>
> Commit-Queue: Daniel Jacques <dnj@chromium.org>
TBR=iannucci@chromium.org,dnj@chromium.org
Change-Id: Ia611c1aeb204c609ff3469076b8256c54e702230
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: chromium:748651
Reviewed-on: https://chromium-review.googlesource.com/589527
Reviewed-by: Daniel Jacques <dnj@chromium.org>
Commit-Queue: Daniel Jacques <dnj@chromium.org>
Bump the CIPD client version to
e2a98b41adb3403b5b7771c4101b42a55dc2dd9e. This version inculdes support
for the "-log-level" flag.
Update CIPD bootstrap scripts and tool installation scripts to log at
warning level, so that standard operation doesn't produce any logs.
BUG=chromium:748651
TEST=canary
- Tested this CIPD version on canary waterfall successfully.
- Tested "gclient" bootstrap and "vpython" / "led" on Windows system,
no output and works.
- Tested "vpython" / "led" on OSX system, no output and works.
Change-Id: Iae316908283c43c504f293aafdf897df83dccf8d
Reviewed-on: https://chromium-review.googlesource.com/588889
Reviewed-by: Robbie Iannucci <iannucci@chromium.org>
Commit-Queue: Daniel Jacques <dnj@chromium.org>
This is an exact reland of https://chromium-review.googlesource.com/583617 .
One of the main use cases is making it clear which revision hashes
need to be changed together. The way it's usually done is one variable
referenced several times. With this CL, we preserve the references
from original DEPS, as opposed to evaluating them and losing some info.
This CL actually makes Var() emit a variable placeholder
instead of its value, and adds support for these placeholders
to gclient.
One of possible next steps might be to deprecate Var().
Bug: 570091, 748486
Change-Id: Id47e3771b7163149a4cd427b84f84ece52772f34
Reviewed-on: https://chromium-review.googlesource.com/586594
Reviewed-by: Andrii Shyshkalov <tandrii@chromium.org>
Reviewed-by: William Hesse <whesse@google.com>
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Paweł Hajdan Jr. <phajdan.jr@chromium.org>
The only user of this method needs to apply two patches from two
different repos in a try job. Extending the method to handle this.
BUG=628259
R=agable@chromium.org
Change-Id: Ia119780d1323add0094631e5c291b0e7a573e6e5
Reviewed-on: https://chromium-review.googlesource.com/588008
Commit-Queue: Sergey Berezin <sergeyberezin@chromium.org>
Reviewed-by: Aaron Gable <agable@chromium.org>
This reverts commit 4d92fe4300.
Reason for revert: This moved the logging from invoking `led` or `vpython` to whenever someone would run something that would end up invoking `update_depot_tools`. It's good that we are calling this when we run update_depot_tools, but we probably should've suppressed the logging there in at least the success case, because now things are even more confusing. See crbug.com/748651.
Original change's description:
> [bootstraps] Sync at gclient, suppress output.
>
> Have the "update_depot_tools" script also do a CIPD tool sync. This will
> ensure that users and systems have access to tooling at sync-time rather
> than just-in-time loading them at execution time.
>
> Update the tool boostraps to suppress any sort of syncing logic, if it
> does happen. This will ensure that users who execute the tools don't se
> unexpected output.
>
> BUG=None
> TEST=local
> - Tested on Mac and Windows.
>
> R=dpranke@chromium.org, iannucci@chromium.org
>
> Change-Id: I8efce8c73cc4e82ffdf5067ba9b917119a81e843
> Reviewed-on: https://chromium-review.googlesource.com/581494
> Commit-Queue: Daniel Jacques <dnj@chromium.org>
> Reviewed-by: Robbie Iannucci <iannucci@chromium.org>
TBR=iannucci@chromium.org,dpranke@chromium.org,dnj@chromium.org
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: None
Change-Id: I2485c9dd2e48a8dbdeebfff5da9d4c708e0edcb7
Reviewed-on: https://chromium-review.googlesource.com/585867
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Dirk Pranke <dpranke@chromium.org>
This reverts commit e79ddeaabf.
Reason for revert: https://bugs.chromium.org/p/chromium/issues/detail?id=748486
Original change's description:
> gclient flatten: preserve variable placeholders
>
> One of the main use cases is making it clear which revision hashes
> need to be changed together. The way it's usually done is one variable
> referenced several times. With this CL, we preserve the references
> from original DEPS, as opposed to evaluating them and losing some info.
>
> This CL actually makes Var() emit a variable placeholder
> instead of its value, and adds support for these placeholders
> to gclient.
>
> One of possible next steps might be to deprecate Var().
>
> Bug: 570091
> Change-Id: I9b13a691b5203cc284c33a59438720e31c9ebf7a
> Reviewed-on: https://chromium-review.googlesource.com/583617
> Commit-Queue: Paweł Hajdan Jr. <phajdan.jr@chromium.org>
> Reviewed-by: Dirk Pranke <dpranke@chromium.org>
TBR=phajdan.jr@chromium.org,dpranke@chromium.org
Change-Id: If9c52ebfa78aba8041ce797ff842d09952d0e2ce
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 570091, 748486
Reviewed-on: https://chromium-review.googlesource.com/584907
Reviewed-by: Paweł Hajdan Jr. <phajdan.jr@chromium.org>
Commit-Queue: Paweł Hajdan Jr. <phajdan.jr@chromium.org>
One of the main use cases is making it clear which revision hashes
need to be changed together. The way it's usually done is one variable
referenced several times. With this CL, we preserve the references
from original DEPS, as opposed to evaluating them and losing some info.
This CL actually makes Var() emit a variable placeholder
instead of its value, and adds support for these placeholders
to gclient.
One of possible next steps might be to deprecate Var().
Bug: 570091
Change-Id: I9b13a691b5203cc284c33a59438720e31c9ebf7a
Reviewed-on: https://chromium-review.googlesource.com/583617
Commit-Queue: Paweł Hajdan Jr. <phajdan.jr@chromium.org>
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
With VS 2017 Update 3 Preview 4 the behavior of vswhere was changed so
that it only reports on non-prerelease versions by default. This can be
overridden by passing -prerelease. This broke our packaging setup. A
temporary fix was used to package Preview 4 and this is the permanent.
When -prerelease is passed then vswhere will report on all installed
versions of VS, whether prerelease or not. The script will package the
first one that it encounters. For best results you should only install
one copy of VS when packaging it.
Trivia: the original RTW version of VS 2017 was incorrectly tagged as
isPrerelease: 1 which means that without the -prerelease flag it
doesn't show up either!
BUG=683729
Change-Id: I98c1acb671dccef7ede4443fbbf498796946c52b
Reviewed-on: https://chromium-review.googlesource.com/578767
Reviewed-by: Scott Graham <scottmg@chromium.org>
Commit-Queue: Bruce Dawson <brucedawson@chromium.org>
Have the "update_depot_tools" script also do a CIPD tool sync. This will
ensure that users and systems have access to tooling at sync-time rather
than just-in-time loading them at execution time.
Update the tool boostraps to suppress any sort of syncing logic, if it
does happen. This will ensure that users who execute the tools don't se
unexpected output.
BUG=None
TEST=local
- Tested on Mac and Windows.
R=dpranke@chromium.org, iannucci@chromium.org
Change-Id: I8efce8c73cc4e82ffdf5067ba9b917119a81e843
Reviewed-on: https://chromium-review.googlesource.com/581494
Commit-Queue: Daniel Jacques <dnj@chromium.org>
Reviewed-by: Robbie Iannucci <iannucci@chromium.org>
Added a regression test. Simplified some logic - if we don't add os-specific
deps and hooks to |dependencies|, we don't need to keep separate original values.
Bug: 570091
Change-Id: I5bdd0b6a66df6b3a2b99d0ad9c6e54ee7114f09b
Reviewed-on: https://chromium-review.googlesource.com/581687
Commit-Queue: Paweł Hajdan Jr. <phajdan.jr@chromium.org>
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Reviewed-by: Michael Moss <mmoss@chromium.org>
Roll the bleeding-edge entry batch file into production. This follows
successful tested deployment of the batch file on canary systems.
BUG=chromium:746602
TEST=canary
R=iannucci@chromium.org, smut@chromium.org, vadimsh@chromium.org
Change-Id: Iaee2a9b4d178ffff3da9a73f4538465bb5367e84
Reviewed-on: https://chromium-review.googlesource.com/580307
Reviewed-by: Vadim Shtayura <vadimsh@chromium.org>
Reviewed-by: Robbie Iannucci <iannucci@chromium.org>
Commit-Queue: Daniel Jacques <dnj@chromium.org>
Using "set /p" to read a file fails when STDOUT is not connected.
Replace it with a method of reading using "for".
In the Python batch file, we eat up some of our sled buffer in order to
accommodate currently-running Python scripts. They should now resume
past the end of the file without issue. The sled was oversized, so this
shouldn't affect upgrade paths.
BUG=chromium:746602
TEST=local
- On Windows machine, tested upgrade from pre-rework and current
bootstraps. This code works on both, and no running Python
processes were disrupted on exit, suggesting the sled change is
sound.
R=iannucci@chromium.org, smut@chromium.org
Change-Id: Ica51999d672dd4e448fdad797bffc06ec9e9febf
Reviewed-on: https://chromium-review.googlesource.com/578725
Commit-Queue: Daniel Jacques <dnj@chromium.org>
Reviewed-by: Robbie Iannucci <iannucci@chromium.org>
Reviewed-by: smut <smut@google.com>
Using self.GetDescription() uses the description from the web.
Given that the user is purposefully disassociating their current
branch from the review on the web, they may have already changed
the commit message. We shouldn't set it back.
Bug: 742730
Change-Id: I0545cb6288c332fd475d1de7fb302f71ee41a415
Reviewed-on: https://chromium-review.googlesource.com/578229
Reviewed-by: Andrii Shyshkalov <tandrii@chromium.org>
Commit-Queue: Aaron Gable <agable@chromium.org>
The query parameters are parsed out of a string using
urlparse.parse_qs, which produces a dictionary whose values
are lists of values.
The query parameters are then stuffed back into a string
using urlencode.urlencode, which uses percent-encoding to
encode lists as literal sequences of brackets and commas,
resulting in parse_qs and urlencode not being inverse
operations.
This change adds the doseq flag to urlencode, thus making
it encode each value from the list of values as a separate
key=value pair, thus making it a true inverse of parse_qs.
R=dnj@chromium.org
Change-Id: Ic4631cc672f9e42ba371306efdda71892d625119
Reviewed-on: https://chromium-review.googlesource.com/575398
Reviewed-by: Daniel Jacques <dnj@chromium.org>
Commit-Queue: Aaron Gable <agable@chromium.org>
This was removed in downstream recipes after it was turned on by default.
Bug: 740456
Change-Id: Idae1905af97333970c8b31508634f8b71550c6f9
Reviewed-on: https://chromium-review.googlesource.com/574030
Reviewed-by: Andrii Shyshkalov <tandrii@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
This adds angle to the got_revision_mapping to take account of the checked
out revision and use it correctly in the without-patch application.
This just adds the got_revision property to all test cases in the output
and fixes the second bot_update run of two angle test cases.
Bug: 740456
Change-Id: I6ae6bf627653280bc2a94230ec77949411dc4331
Reviewed-on: https://chromium-review.googlesource.com/573741
Reviewed-by: Andrii Shyshkalov <tandrii@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Update the CIPD recipe module to generate platform suffixes for
arbitrary OS, architecture, and bitness.
BUG=None
TEST=expectations
Change-Id: Ic713adc403dc135895a1cfaaa6ba1546742125d9
Reviewed-on: https://chromium-review.googlesource.com/573412
Commit-Queue: Daniel Jacques <dnj@chromium.org>
Reviewed-by: Robbie Iannucci <iannucci@chromium.org>
Previously, running git crrev-parse {crrev_number} on Mac doesn't
produce any meaningful output, and the reason is that parameter
substitutions resulted in extra back slashes ('\'):
ref="refs/heads/master"
remote_ref="${ref/refs\/heads/refs\/remotes\/origin}"
expected: remote_ref="refs/remotes/origin/master"
actual: remote_ref="refs\/remotes\/origin/master"
This CL fixes the issue by removing the extra back slashes.
Bug:
Change-Id: I1ee1b4702f09c96fb65518ee1465d0de927e95cc
Reviewed-on: https://chromium-review.googlesource.com/572052
Reviewed-by: Dirk Pranke <dpranke@chromium.org>
Commit-Queue: Yuke Liao <liaoyuke@chromium.org>
The actual flag is --send-mail. Also includes a fix
to publish these changes if a comment is included.
Bug: 740950
Change-Id: I38ca0a35c1364c8364eb0ef301137c04daede40b
Reviewed-on: https://chromium-review.googlesource.com/572033
Reviewed-by: Andrii Shyshkalov <tandrii@chromium.org>
Commit-Queue: Aaron Gable <agable@chromium.org>
recipe-roller
Sergey Berezin
Paweł Hajdan, Jr
Varun Khaneja
Dirk Pranke
Takeshi Yoshino
Clemens Hammacher
Aaron Gable
Dan Jacques
Daniel Jacques
Bruce Dawson
Robert Iannucci
John Budorick
Michael Achenbach
Yuke Liao