1123342 - Use a separate tooltool manifest for clang static analysis builds

Status: RESOLVED FIXED

(Release Engineering :: General, defect)

Description

[(no longer active)]

This will allow us to upgrade the clang we use on those builds separately than what we use for production builds.

Comment 1

[(no longer active)]

Attachment: Add a new tooltool manifest for OS X static analysis builds

This is currently a copy of releng.manifest.

Comment 2

[(no longer active)]

Attachment: Use the new tooltool manifest on OS X static analysis builds

Comment 3

[(no longer active)]

Comment on attachment 8551317 [details] [diff] [review]
Add a new tooltool manifest for OS X static analysis builds

https://hg.mozilla.org/mozilla-central/rev/80b221f602c3

Comment 4

[(no longer active)]

https://hg.mozilla.org/build/mozharness/rev/f481508606ec

Comment 5

[Chris Cooper [:coop] (he/him)]

In production: https://hg.mozilla.org/build/mozharness/rev/f481508606ec

Comment 6

[Justin Wood (:Callek)]

There is an OSX static analysis bustage on b2g-inbound that strangely is not affecting other trees.

See the irc snippet thrown into Bug 1123025 c#1 as a refrence point.

I have retriggered the static-analysis job that was burning and those directly before and after. no idea how to explain it though so far.

Comment 7

[Justin Wood (:Callek)]

fwiw both of those earlier jobs failed, fast. The merge that was done on b2g-i from m-c included https://hg.mozilla.org/integration/b2g-inbound/rev/80b221f602c3 and its retrigger is still running.  I'm suspecting that this is a culprit of simply not having that cset everywhere it was needed when this mozharness change went live.

Comment 8

[(no longer active)]

That is very weird.  I'm 99% sure that I did merge that cset to b2g-inbound yesterday, together with central/inbound/fx-team.

Comment 9

[Phil Ringnalda (:philor)]

It's the 1% that kills you - your merge to b-i merged back all the stuff Tomcat forgot to merge back to it, but *not* your cset, which didn't get there until I merged to it later on.

Comment 10

[(no longer active)]

Oops, sorry!