Also add rc-{google-play-track,rollout-percentage} for RC pushapk.
One nice side effect of using the same push-apk kind: we don't re-run push-apk during the ship_fennec relpro flavor if we've run the ship_fennec_rc flavor with the same build. (Google Play would reject the same buildid.)
This is really for bug 1433536, but MozReview is forcing me to include this patch with the others for reasons.
MozReview-Commit-ID: 69ymqVLv9E2
--HG--
extra : rebase_source : b87d4dd2394788a5452ff3f52a8ca5022a15b9ee
extra : intermediate-source : 7a826d274a4828018a836cf1149df29d403a7c11
extra : source : a3c60b0370a3e08ce765f87c1d7d5dad24879881
This used to only be relevant to Devedition and Firefox releases.
In bug 1433536 we're going to add RC Fennec releases. Let's rename
the parameter now, for less parameters churn.
MozReview-Commit-ID: 28e1Y5FG4On
--HG--
extra : rebase_source : 59d41887d856481ab85bb8d2221dfcebca4430b0
extra : source : 4c96e312d88a3f7037c1eb39a3031d4baf997015
also clean up and move more config to the promotion config.
MozReview-Commit-ID: FmTWNNPcEaZ
--HG--
extra : rebase_source : 40431217fafb6796dbd65c7dfeab0e891ac1bbd4
extra : source : 0f5418a83477c1b6b221e4d28515792410e504d0
Add support for the three firefox and devedition relpro flavors (we
could probably reduce devedition to 2).
Also, instead of defining which kinds to use from the previous graph
in `previous_graph_kinds`, specify which kinds to rebuild (ignore)
from the previous graph in `rebuild_kinds`. This list will be much
smaller (currently empty).
MozReview-Commit-ID: 5rH1TW7GbAD
--HG--
extra : rebase_source : b4294a0d17a99b2ffd48f5d62821c724324b242c
extra : histedit_source : b3fe87101e4595f5fc5c7daaa2d4a0bd88418667
Instead of relying on environment variables, pass these in as parameters.
MozReview-Commit-ID: An58Bu2kd1g
--HG--
extra : rebase_source : 9c7b6beb13b676a3376a897f2c8143cc042b8276
This patch adds the `release_deps` transform, which adds every
kind-dependency task that has the same product as a task dependency
(with some exceptions).
This patch made it clear that we need a standard way of defining
product.
MozReview-Commit-ID: 4xOJRQSCTgF
--HG--
extra : rebase_source : 4b98b040ac5d8777394cff6f3249ab354fff641b
This patch adds the beetmover-cdns kind, and adds it to
`publish_fennec`.
This was the first non-buildbot-bridge, non-dummy relpro task, so this
needed a new transform.
This patch also updates the `previous_graph_kinds` and updates the
beetmover scopes in scriptworker.py.
MozReview-Commit-ID: 3rpkjuLjjXz
--HG--
extra : rebase_source : d491f1ce5d10ee8f2602698236106007e203c358
We currently use `previous_graph_kinds` as a pointer to use a previously
built graph's task instead of generating a new task. We were re-running
`release-bouncer-sub` in the `publish_fennec` release promotion flavor
because we hadn't specified `release-bouncer-sub` as a
`previous_graph_kind`; this patch fixes that error.
MozReview-Commit-ID: 92MAVsG8Jrk
--HG--
extra : rebase_source : cf8411d017de20ef05ba9ac547e94797c88901dd
This patch adds the initial `release-version-bump` kind, and adds it to
the `publish_fennec` `target_tasks_method`.
It also adds support for `next_version`.
MozReview-Commit-ID: 9YRswddeuZ3
--HG--
extra : rebase_source : 991cbf89b40c4ef980c42251001e81be5cedaf00
This hooks up the `target_tasks_method`s and the
`release_promotion_flavor`s for these four release promotion flavors.
We also add some maple support and add `version_display.txt` to the
decision task sparse checkout so we can read the version number during
decision/action task time.
MozReview-Commit-ID: CdxUUXZtXO0
--HG--
extra : rebase_source : 99e4f060a964d34a377aea6a6e36fbcef66c6aec
Add a release promotion custom action for releng's TC release promotion migration work.
This action generates a graph dependent on previously built tasks. To track these, we add the `do_not_optimize` and `existing_tasks` parameters. The `do_not_optimize` parameter specifies tasks that we want to explicitly exclude from taskgraph optimization. The `existing_tasks` parameter specifies a label-to-taskid map for tasks from previous graphs.
MozReview-Commit-ID: 1vKrNUavM4V
--HG--
extra : rebase_source : b8ba95d270aafe1464c2b3bfc318b9568500a7a1
Using /home/worker is the build directory has a 30% talos performance
loss, because test machines has a /home mount directory.
MozReview-Commit-ID: 554IPMRWgzK
--HG--
extra : rebase_source : 00827d3f6bd705419bc801eb05b543af1ddc274f
Tom Prince
Aki Sasaki
Rail Aliiev
Brian Stack
Johan Lorenzo
Ben Hearsum
Simon Fraser
Rail Aliiev
Rok Garbas
Dustin J. Mitchell
Wes Kocher
Wander Lairson Costa