679539 - test-content-proxy.testProxy triggers assertion in jscntxtinlines.h

Status: RESOLVED WORKSFORME

(Add-on SDK Graveyard :: General, defect, P2)

Description

[Myk Melez [:myk] [@mykmelez]]

Some test, which looks like test-content-proxy.testProxy, is triggering an assertion on debug builds from the central and aurora repositories for all OSes, for example in this test run <http://tinderbox.mozilla.org/showlog.cgi?log=Mozilla-Aurora/1313453895.1313454650.19515.gz&fulltext=1>:

info: executing 'test-content-proxy.testProxy'
...
info: pass: origin is null
*** Compartment mismatch 0xb305280 vs. 0xa0ca928
Assertion failure: compartment mismatched, at ../../../js/src/jscntxtinlines.h:123
Traceback (most recent call last):
  File "bin/cfx", line 29, in <module>
    cuddlefish.run()
  File "/home/cltbld/talos-slave/test/build/addon-sdk-9f45975c909d/python-lib/cuddlefish/__init__.py", line 492, in run
    test_all(env_root, defaults=options.__dict__)
  File "/home/cltbld/talos-slave/test/build/addon-sdk-9f45975c909d/python-lib/cuddlefish/__init__.py", line 323, in test_all
    test_all_packages(env_root, defaults)
  File "/home/cltbld/talos-slave/test/build/addon-sdk-9f45975c909d/python-lib/cuddlefish/__init__.py", line 379, in test_all_packages
    defaults=defaults)
  File "/home/cltbld/talos-slave/test/build/addon-sdk-9f45975c909d/python-lib/cuddlefish/__init__.py", line 790, in run
    norun=options.no_run)
  File "/home/cltbld/talos-slave/test/build/addon-sdk-9f45975c909d/python-lib/cuddlefish/runner.py", line 399, in run_app
    timeout)
Exception: Wait timeout exceeded (600s)

This seems likely to be caused by platform bug 679494.  Note that the line in jscntxtinlines.h that triggers the assertion is 120 on central but 123 on aurora (as of shortly before today's merge; it's probably the same on both now).

Comment 1

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

(Pushing all open bugs to the --- milestone for the new triage system)

Comment 2

[Myk Melez [:myk] [@mykmelez]]

Requesting re-triage, since this may be intermittent, but it is happening with frightening regularity, including all these test runs on all three OSes and multiple Firefox branches from a single SDK push <https://tbpl.mozilla.org/?usetinderbox=1&tree=Jetpack&rev=1930cc0c30f7>:

http://tinderbox.mozilla.org/showlog.cgi?log=Jetpack/1319844851.1319845472.11417.gz
http://tinderbox.mozilla.org/showlog.cgi?log=Jetpack/1319845119.1319845697.11792.gz
http://tinderbox.mozilla.org/showlog.cgi?log=Jetpack/1319844428.1319844999.10861.gz
http://tinderbox.mozilla.org/showlog.cgi?log=Jetpack/1319845675.1319846255.13210.gz
http://tinderbox.mozilla.org/showlog.cgi?log=Jetpack/1319844093.1319844558.9946.gz
http://tinderbox.mozilla.org/showlog.cgi?log=Jetpack/1319844094.1319844790.10497.gz
http://tinderbox.mozilla.org/showlog.cgi?log=Jetpack/1319844054.1319844393.9714.gz
http://tinderbox.mozilla.org/showlog.cgi?log=Jetpack/1319844054.1319844393.9713.gz
http://tinderbox.mozilla.org/showlog.cgi?log=Jetpack/1319844053.1319844393.9717.gz
http://tinderbox.mozilla.org/showlog.cgi?log=Jetpack/1319844054.1319844393.9708.gz

Comment 3

[Myk Melez [:myk] [@mykmelez]]

I just took a look at a bunch of instances of this, and it turns out they're all occurring on the beta and/or release branches, not on central nor aurora.  So perhaps this has actually already been fixed?

Comment 4

[Mark A. Hershberger (hexmode)]

Is this still happening?

Comment 5

[Alexandre Poirot [:ochameau]]

I don't think so, Wes?

Wes: Are we still facing intermittent failure due to jetpack codebase?
From what I'm seing these days, we only face tbpl specific issues,
where it fails downloading firefox tarball.

Comment 6

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

(In reply to Alexandre Poirot (:ochameau) from comment #5)
> I don't think so, Wes?
> 
> Wes: Are we still facing intermittent failure due to jetpack codebase?
> From what I'm seing these days, we only face tbpl specific issues,
> where it fails downloading firefox tarball.

I haven't seen this issue for a while, though there are still some intermittent test failures and timeouts elsewhere.

Comment 7

[Dave Townsend [:mossop]]

Sounds like this is fixed, we can reopen if not.