Closed Bug 1432213 Opened 6 years ago Closed 6 years ago

Windows mingw32 builds are failing on Gecko 58 since the uplift to Beta (Automation Error: Can't checkout https://hg.mozilla.org/build/tools! | Caught exception: could not find the robustcheckout Mercurial extension)

Categories

(Firefox Build System :: General, defect, P1)

Product:
Firefox Build System
Component:
General
Type:
defect

Tracking

(firefox59 fixed, firefox60 fixed)

Status:
RESOLVED FIXED
Milestone:
mozilla60
Tracking Flags:
Tracking Status
firefox59 --- fixed
firefox60 --- fixed

People

(Reporter: tjr, Assigned: tjr)

References

Details

(Whiteboard: [tor])

Attachments

(1 file)

Bug 1432213 Fix the MinGW build on the beta branch and re-enable it
59 bytes, text/x-review-board-request
froydnj
: review+
Details 
Alright, need to actually fix this now.

+++ This bug was initially created as a clone of Bug #1414016 +++

Filed by: rvandermeulen [at] mozilla.com

https://treeherder.mozilla.org/logviewer.html#?job_id=141726323&repo=mozilla-beta

https://queue.taskcluster.net/v1/task/Y874Dzv_SIeMFRQrpvN0dw/runs/0/artifacts/public/logs/live_backing.log

Sorry, I'd been seeing these in my Try simulations as well but had chalked it up to something odd happening with the config hackery needed to support those pushes. Looks like it's real after all.
Flags: needinfo?(tom)
Attachment #8945201 - Flags: review?(core-build-config-reviews)
Comment on attachment 8945201 [details]
Bug 1432213 Fix the MinGW build on the beta branch and re-enable it

https://reviewboard.mozilla.org/r/215432/#review221262

I think I might be willing to believe that this does what you want, but I am super-confused about the commit message vs. the code.  Am I missing something huge here?

::: commit-message-3629e:11
(Diff revision 2)
> +We avoid this override by changing the MinGW platform so it doesn't match
> +and the mozconfig doesn't get overridden.

The current situation is that we are running mingw builds on Beta, and those builds are broken, yes?  And this function is returning `False` for tasks we *don't* want to run, yes?  So the current code is actually *not* running the mingw task?  I am confused, then. :)

I am also confused by this bit of the commit message.  We're changing the platform so it doesn't match...but the patch looks like it changes the platform so it *does* match what `target_tasks.py` was looking for, so we would be disabling the task.  But then we remove the `wind32-mingw32` filter from `target_tasks.py` so it looks like we're re-enabling mingw builds?
Comment on attachment 8945201 [details]
Bug 1432213 Fix the MinGW build on the beta branch and re-enable it

https://reviewboard.mozilla.org/r/215432/#review221276

OK, RyanVM explained this to me in #build.  I should have actually looked at `branch_specifics.py` to see what was going on; for some reason I assumed that file would be in the patch--and then didn't bother to verify that assumption!
Attachment #8945201 - Flags: review+
Attachment #8945201 - Flags: review?(core-build-config-reviews) → review+
Attachment #8945201 - Flags: review?(core-build-config-reviews) → review+
Pushed by ryanvm@gmail.com:
https://hg.mozilla.org/integration/autoland/rev/47c9d061fa94
Fix the MinGW build on the beta branch and re-enable it r=froydnj
https://hg.mozilla.org/mozilla-central/rev/47c9d061fa94
Status: NEW → RESOLVED
Closed: 6 years ago
status-firefox60: --- → fixed
Resolution: --- → FIXED
Target Milestone: --- → mozilla60
Product: Core → Firefox Build System
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: