It does the appropriate pushes but gives a perf compare url of:
https://treeherder.mozilla.org/perfherder/compare?originalProject=try&originalRevision=None&newProject=try&newRevision=None

and

https://hg.mozilla.org/try/rev/c132bb0472bc36dd96c1b1dad3a6f83366790aec

has "PERF_BASE_REVISION": null, which seems to prevent things from building

:jrmuizel, could you post the log that you get when you make a try run with ./mach try perf? I'm curious to see if there's something different about it versus what we see in HG.

$ ./mach try perf --rebuild 5
Gathering tasks for Speedometer 3 desktop category
Executing raptor queries: 'browsertime 'speedometer3, !android 'shippable !-32 !clang, !bytecode, !live, !profil, !chrom, !safari
Note: switching to '396238772940f981924b8055c3cdb936c38136ef'.

You are in 'detached HEAD' state. You can look around, make experimental
changes and commit them, and you can discard any commits you make in this
state without impacting any branches by switching back to a branch.

If you want to create a new branch to retain commits you create, you may
do so (now or later) by using -c with the switch command. Example:

  git switch -c <new-branch-name>

Or undo this operation with:

  git switch -

Turn off this advice by setting config variable advice.detachedHead to false

HEAD is now at 396238772940f Bug 1825867 - Make browser_sma_show_migration_wizard.js pass with both the old and new migration wizards. r=aminomancer
estimates: Runs 16 tasks (3 selected, 13 dependencies)
estimates: Total task duration 0:58:40
estimates: In the top 94% of durations
estimates: Should take about 0:47:39 (Finished around 2023-04-03 17:37)
Bundling 1 changesets
Bundling 1 manifests
Bundling 1 revisions of 1 files
remote: waiting for lock on working directory of /repo/hg/mozilla/try held by process '21712' on host 'hgssh1.dmz.mdc1.mozilla.com/effffffc'
remote: got lock after 36 seconds
remote: adding changesets
remote: adding manifests
remote: adding file changes
remote: recorded push in pushlog
remote: added 1 changesets with 0 changes to 1 files (+1 heads)
remote:
remote: View your change here:
remote:   https://hg.mozilla.org/try/rev/d5d1bec772d5147e54198008bf559abb470ca81d
remote:
remote: Follow the progress of your build on Treeherder:
remote:   https://treeherder.mozilla.org/jobs?repo=try&revision=d5d1bec772d5147e54198008bf559abb470ca81d
remote: recorded changegroup in replication log in 0.014s
To hg::ssh://hg.mozilla.org/try
 * [new branch]                  HEAD -> branches/default/tip
Previous HEAD position was 396238772940f Bug 1825867 - Make browser_sma_show_migration_wizard.js pass with both the old and new migration wizards. r=aminomancer
Switched to branch 'sp3'
estimates: Runs 16 tasks (3 selected, 13 dependencies)
estimates: Total task duration 0:58:40
estimates: In the top 94% of durations
estimates: Should take about 0:47:39 (Finished around 2023-04-03 17:40)
Bundling 2 changesets
Bundling 2 manifests
Bundling 2 revisions of 2 files
remote: waiting for lock on working directory of /repo/hg/mozilla/try held by process '22642' on host 'hgssh1.dmz.mdc1.mozilla.com/effffffc'
remote: got lock after 150 seconds
remote: adding changesets
remote: adding manifests
remote: adding file changes
remote: recorded push in pushlog
remote: added 1 changesets with 0 changes to 2 files (+1 heads)
remote:
remote: View your change here:
remote:   https://hg.mozilla.org/try/rev/f98a4bef73f9829e3554cb788bbcdfc048f41f97
remote:
remote: Follow the progress of your build on Treeherder:
remote:   https://treeherder.mozilla.org/jobs?repo=try&revision=f98a4bef73f9829e3554cb788bbcdfc048f41f97
remote: recorded changegroup in replication log in 0.014s
To hg::ssh://hg.mozilla.org/try
 * [new branch]                  HEAD -> branches/default/tip

!!!NOTE!!!
 You'll be able to find a performance comparison here once the tests are complete (ensure you select the right framework): https://treeherder.mozilla.org/perfherder/compare?originalProject=try&originalRevision=None&newProject=try&newRevision=None

If you need any help, you can find us in the #perf-help Matrix channel:
https://matrix.to/#/#perf-help:mozilla.org

For more information on the performance tests, see our PerfDocs here:
https://firefox-source-docs.mozilla.org/testing/perfdocs/

I was about to report this, hah :)

:emilio/:jrmuizel, would either of you have some instructions for how to setup the git repo? We're going to need to debug this locally.

Just wanted to follow up with both :emilio & :jrmuizel and understand how you setup the git repo?

I'm using a git worktree with a git cinnabar clone.

:andrej, maybe these instructions could help? It looks like there's an argument you can pass to change the vcs: https://firefox-source-docs.mozilla.org/setup/macos_build.html#bootstrap-a-copy-of-the-firefox-source-code

Yeah, my setup is just that: A git cinnabar clone, and a bunch of git worktrees out of that.

Only weirdness is that my remote is called mozilla rather than origin because that's what the original git cinnabar instructions used back in the day.

What version of git cinnabar and git are you using? Also what operating system are you using?

• git 2.40.0
• git-cinnabar 0.6.0-c224e301ad05f4e337b0833a57fde97d41154d7d
• Linux (though I think this also failed on Windows)

• git 2.31.1
• git-cinnabar 0.5.2
• macOS

I have narrowed down the issue to inconsistencies with the push_to_try function stdoutputting in mercurial vs git-cinnibar
Will provide updates as I get them

Changes work on 2 devices and have been merged, let me know if you still encounter the issues :jrmuizel and :emilio

https://hg.mozilla.org/mozilla-central/rev/94a716371015