Split out jsreftest and jittest files into their own packages, removing them
from the common package.
This speeds up extracting files from the common test archive for
non-jsreftest/jittest suites.
Also, remove some files from the web-platform test archive that are already
present in the common archive.
Differential Revision: https://phabricator.services.mozilla.com/D72192
Add test package mach support for gtest and hook into the custom retrigger
action. Some existing custom retrigger features, like setting gecko prefs,
are not (easily) applicable to gtest, which doesn't use mozprofile; for
this reason, use a separate action context with items suitable for gtest.
Differential Revision: https://phabricator.services.mozilla.com/D67384
--HG--
extra : moz-landing-system : lando
Adds remotegtests.py, like rungtests.py, but for Android. Unlike some other test harnesses, remotegtests is not a sub-class of rungtests: There wasn't much code that could be re-used.
The code to wait for the test app and collect and display the log is derived from and similar to existing code in remoteautomation.py: I didn't want to increase reliance on remoteautomation; I don't mind a bit of code duplication.
There is a TODO item related to libxul.so. gtest runs in the browser, but with tests compiled into a gtest-specific version of libxul.so. remotegtests.py patches the installed app by pushing the libxul.so to the installation directory of the app -- root required and fragile, but works fine on the emulator. I think the "right" way to do this is to create a gtest-specific apk; I plan to file a follow-up bug and petition build folks to help sort that out.
I have been running this locally with:
mach gtest (builds gtests, then errors out)
mach python testing/gtest/remotegtests.py --libxul=/home/gbrown/objdirs/x86_64/dist/bin/gtest/libxul.so
Work-in-progress on 'mach gtest' will make this much simpler.
Differential Revision: https://phabricator.services.mozilla.com/D26835
--HG--
extra : moz-landing-system : lando
This patch gets us building an updater binary that always embeds the dep certificates (instead of release or nightly), and builds a new tests package that includes it.
This was originally D5900, but that was backed out due to busting artifact builds. I've fixed that by removing the Makefile that Ted pointed out is unnecessary.
Differential Revision: https://phabricator.services.mozilla.com/D6418
--HG--
extra : moz-landing-system : lando
Changes the wpt manifest path to the topobjdir instead so it can be moved out of tree.
Other changes so that the manifest download and update, and |mach wpt| and |mach test <wpt-test>| work with the new path.
The manifest is also downloaded and updated when creating the tests-archive to ensure that it exists when we run tests on TC.
MozReview-Commit-ID: Fp6UsKJjhTU
Differential Revision: https://phabricator.services.mozilla.com/D5312
--HG--
extra : moz-landing-system : lando
This patch gets us building an updater binary that always embeds the dep certificates (instead of release or nightly), and builds a new tests package that includes it.
I had a lot of trouble getting the test package generated correctly due to the fact that things in "_tests" won't be included for test packages that aren't "common". My fix for that isn't ideal - I'm open to something better.
Differential Revision: https://phabricator.services.mozilla.com/D5900
--HG--
extra : moz-landing-system : lando
Changes the wpt manifest path to the topobjdir instead so it can be moved out of tree.
Other changes so that the manifest download and update, and |mach wpt| and |mach test <wpt-test>| work with the new path.
The manifest is also downloaded and updated when creating the tests-archive to ensure that it exists when we run tests on TC.
MozReview-Commit-ID: Fp6UsKJjhTU
Differential Revision: https://phabricator.services.mozilla.com/D5312
--HG--
extra : moz-landing-system : lando
This moves testing/profiles/prefs_general.js to
testing/profiles/common/user.js. It also adds an 'extensions' folder to the
common profile. Dropping extension files here will get them installed in all
test harnesses (useful for testing on try).
The idea is that all test harnesses will eventually use this 'common' profile.
We'll also create some new per harness profiles, e.g testing/profiles/mochitest
and testing/profiles/reftest. This way there will be a single location
developers can go to set preferences, both for a specific harness, and across
all harnesses.
MozReview-Commit-ID: 8sqBqLiypgU
--HG--
rename : testing/profiles/prefs_general.js => testing/profiles/common/user.js
extra : rebase_source : 72a4a4b691e93c77479c7e70647b0ca373862c51
This moves testing/profiles/prefs_general.js to
testing/profiles/common/user.js. It also adds an 'extensions' folder to the
common profile. Dropping extension files here will get them installed in all
test harnesses (useful for testing on try).
The idea is that all test harnesses will eventually use this 'common' profile.
We'll also create some new per harness profiles, e.g testing/profiles/mochitest
and testing/profiles/reftest. This way there will be a single location
developers can go to set preferences, both for a specific harness, and across
all harnesses.
MozReview-Commit-ID: 8sqBqLiypgU
--HG--
rename : testing/profiles/prefs_general.js => testing/profiles/common/user.js
extra : rebase_source : 7599913e547533f2f57b597ad0f238c6cd391c82
We currently generate these to dist/ and then use upload.py to copy
them into UPLOAD_PATH. This is wasteful. Let's generate the test
archives into UPLOAD_PATH directly, if the variable is defined.
MozReview-Commit-ID: BHoxZYdxzq4
--HG--
extra : rebase_source : f6d80adb4e1eb7333f0cd3f847fd92cbb75c33bc
This target is a prereq for all test archives. Why we only remove a
single test archive, I don't know. We always build this test archive.
So it should always get overwritten and we don't need to remove an
existing file.
MozReview-Commit-ID: 1qd1T3w6kVT
--HG--
extra : rebase_source : ab27910a45961da636a4077952a2d67819ad701b
It is a one-liner and doesn't need to exist in a separate make file.
MozReview-Commit-ID: 5w3eLWD0CuB
--HG--
extra : rebase_source : 15ebb63156f8b4e8d3288dfaf0d9ec2c328b5914
The browser JUnit 3 test suite has never run in automation, and no
test has been added or updated since 2016. Half of the existing
"tests" are broken on my local device. This allows to clean several
Makefile's related to this undeveloped testing approach, which won't
be pursued post-Gradle.
MozReview-Commit-ID: 1EbGMLntuyR
--HG--
extra : rebase_source : b4751f96851a8df61c0333a3e0ade3805b280dfe
Several years ago there was a single zip file for all test files. Clients
would only extract the files they needed. Thus, zip was a reasonable
archive format because it allowed direct access to members without
having to decompress the entirety of the stream.
We have since split up that monolithic archive into separate,
domain-specific archives. e.g. 1 archive for mochitests and one
for xpcshell tests. This drastically cut down on network I/O
required on testers because they only fetched archives/data that
was relevant. It also enabled parallel generation of test archives,
we shaved dozens of seconds off builds due to compression being
a long pole.
Despite the architectural changes to test archive management, we
still used zip files. This is not ideal because we no longer access
specific files in test archives and thus don't care about single/partial
member access performance.
This commit implements support for generating tar.gz test archives.
And it switches the web-platform archive to a tar.gz file.
The performance implications for archive generation are significant:
before: 48,321,250 bytes; 6.05s
after: 31,844,267 bytes; 4.57s
The size is reduced because we have a single compression context
so data from 1 file can benefit compression in a subsequent file.
CPU usage is reduced because the compressor has to work less with
1 context than it does with N. While I didn't measure it, decompression
performance should also be improved for the same reasons. And of course
network I/O will be reduced.
mozharness consumers use a generic method for handling unarchiving.
This method automagically handles multiple file extensions. So as long
as downstream consumers aren't hard coding ".zip" this change should
"just work."
MozReview-Commit-ID: LQa5MIHLsms
--HG--
extra : rebase_source : 100092c2f2ff609362a724fff60f46dd6e49c94e
extra : intermediate-source : d10f5ccd882b965fcad39914f7c3c930d1301a41
extra : source : a0e257e346ccf3c1db332ec5903241f4eeb9a7ee
Several years ago there was a single zip file for all test files. Clients
would only extract the files they needed. Thus, zip was a reasonable
archive format because it allowed direct access to members without
having to decompress the entirety of the stream.
We have since split up that monolithic archive into separate,
domain-specific archives. e.g. 1 archive for mochitests and one
for xpcshell tests. This drastically cut down on network I/O
required on testers because they only fetched archives/data that
was relevant. It also enabled parallel generation of test archives,
we shaved dozens of seconds off builds due to compression being
a long pole.
Despite the architectural changes to test archive management, we
still used zip files. This is not ideal because we no longer access
specific files in test archives and thus don't care about single/partial
member access performance.
This commit implements support for generating tar.gz test archives.
And it switches the web-platform archive to a tar.gz file.
The performance implications for archive generation are significant:
before: 48,321,250 bytes; 6.05s
after: 31,844,267 bytes; 4.57s
The size is reduced because we have a single compression context
so data from 1 file can benefit compression in a subsequent file.
CPU usage is reduced because the compressor has to work less with
1 context than it does with N. While I didn't measure it, decompression
performance should also be improved for the same reasons. And of course
network I/O will be reduced.
mozharness consumers use a generic method for handling unarchiving.
This method automagically handles multiple file extensions. So as long
as downstream consumers aren't hard coding ".zip" this change should
"just work."
MozReview-Commit-ID: LQa5MIHLsms
--HG--
extra : rebase_source : cd029cdbbcccc1d16f03d63a5f1fdf60be5db4fd
extra : source : a0e257e346ccf3c1db332ec5903241f4eeb9a7ee
Several years ago there was a single zip file for all test files. Clients
would only extract the files they needed. Thus, zip was a reasonable
archive format because it allowed direct access to members without
having to decompress the entirety of the stream.
We have since split up that monolithic archive into separate,
domain-specific archives. e.g. 1 archive for mochitests and one
for xpcshell tests. This drastically cut down on network I/O
required on testers because they only fetched archives/data that
was relevant. It also enabled parallel generation of test archives,
we shaved dozens of seconds off builds due to compression being
a long pole.
Despite the architectural changes to test archive management, we
still used zip files. This is not ideal because we no longer access
specific files in test archives and thus don't care about single/partial
member access performance.
This commit implements support for generating tar.gz test archives.
And it switches the web-platform archive to a tar.gz file.
The performance implications for archive generation are significant:
before: 48,321,250 bytes; 6.05s
after: 31,844,267 bytes; 4.57s
The size is reduced because we have a single compression context
so data from 1 file can benefit compression in a subsequent file.
CPU usage is reduced because the compressor has to work less with
1 context than it does with N. While I didn't measure it, decompression
performance should also be improved for the same reasons. And of course
network I/O will be reduced.
mozharness consumers use a generic method for handling unarchiving.
This method automagically handles multiple file extensions. So as long
as downstream consumers aren't hard coding ".zip" this change should
"just work."
MozReview-Commit-ID: LQa5MIHLsms
--HG--
extra : rebase_source : 19ec875917546abc147b234a815e1a64c204b92a
sutagent is no longer built or used; devicemanagerSUT is completely
unused. After this change, devicemanagerADB is the only implementation of
devicemanager, and test harness options like --dm_trans are eliminated.
Nils Ohlmeier [:drno]
Ricky Stewart
Bogdan Tara
Chris AtLee
Geoff Brown
Mike Shal
Daniel Varga
Edwin Takahashi
Jonathan Kew
Gurzau Raul
Tarek Ziadé
Tom Ritter
Christian Holler
Ben Hearsum
Ahilya Sinha
shindli
Rob Wood
Andrew Halberstadt
Gregory Szorc
Nick Alexander
Kris Maglione
Justin Wood
Wes Kocher