850061 - "Assertion failure: !IsPoisonedPtr(wrapped.wrapped)," with --enable-root-analysis, --enable-more-deterministic and gczeal(6)

Status: RESOLVED WORKSFORME

(Core :: JavaScript Engine, defect)

Description

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

Attachment: stack

gczeal(6)
evalcx("RegExp(\"(?!7|(l)?)\")")()

asserts js debug shell (compiled with --enable-root-analysis) on m-c changeset e6215e0357fa without any CLI arguments at Assertion failure: !IsPoisonedPtr(wrapped.wrapped),

Comment 1

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

--enable-more-deterministic builds seem to be needed here.

Comment 2

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

autoBisect shows this is probably related to the following changeset:

The first bad revision is:
changeset:   124127:05113da6e613
user:        Jon Coppeard
date:        Wed Mar 06 18:10:22 2013 +0000
summary:     Bug 848449 - GC: Remove AutoAssertNoGC and AssertCanGC - Remove use of AutoAssertNoGC r=terrence

Comment 3

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

autoBisect shows this is probably related to the following changeset:

The first good revision is:
changeset:   http://hg.mozilla.org/mozilla-central/rev/89a645d498e3
user:        Jan de Mooij
date:        Mon May 13 16:47:57 2013 -0700
summary:     Bug 857845 part 1 - rm JaegerMonkey. r=bhackett, sr=luke

This assert went away with JM, apparently. Terrence/Jan, does this seem likely?

Comment 4

[Jan de Mooij [:jandem]]

(In reply to Gary Kwong [:gkw] [:nth10sd] from comment #3)
> This assert went away with JM, apparently. Terrence/Jan, does this seem
> likely?

Hm, looking at the testcase in comment 0, JM shouldn't compile anything there (without -a) so it seems unlikely. Bug 857845 part 1 also touched non-JM code, but my guess is that removing all that code somehow changed GC or allocation behavior so that this no longer repros.

Comment 5

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

Let's resolve this WFM then - I'm sure another bug will be filed should this show up again.