592491 - Intermittent Assertion failure: (l.asBits >> 32) == 0, at /builds/slave/tracemonkey-macosx64-debug/build/js/src/jsvalue.h:262 in test_xhrAbort.html

Status: RESOLVED WORKSFORME

(Core :: JavaScript Engine, defect)

Description

[Phil Ringnalda (:philor)]

Saw this once yesterday in JM, and now in TM post-merge

http://tinderbox.mozilla.org/showlog.cgi?log=TraceMonkey/1283294661.1283295228.10950.gz
Rev3 MacOSX Snow Leopard 10.6.2 tracemonkey debug test mochitests-2/5 on 2010/08/31 15:44:21
s: talos-r3-snow-016

2931 INFO TEST-START | /tests/dom/src/threads/test/test_xhrAbort.html
--DOMWINDOW == 14 (0x114d26248) [serial = 1131] [outer = 0x111758cc0] [url = http://mochi.test:8888/tests/dom/src/threads/test/test_throwingOnerror.html]
--DOMWINDOW == 13 (0x1123240f8) [serial = 1130] [outer = 0x111758cc0] [url = http://mochi.test:8888/tests/dom/src/threads/test/test_threadTimeouts.html]
Assertion failure: (l.asBits >> 32) == 0, at /builds/slave/tracemonkey-macosx64-debug/build/js/src/jsvalue.h:262
NEXT ERROR TEST-UNEXPECTED-FAIL | /tests/dom/src/threads/test/test_xhrAbort.html | Exited with code 1 during test run

Comment 1

[Luke Wagner [:luke]]

FYI, this assertion relates to the encoding of uint32 privates, which is different than a normal int32 value: a private uint32, like private void*, is a double and has the high bits all zero.  If JM is writing this value as a normal int32 value, that would trip the assertion.

Comment 2

[Phil Ringnalda (:philor)]

Doesn't seem to have ever managed to happen again.