Closed
Bug 642150
Opened 14 years ago
Closed 14 years ago
Assertion failure: obj->containsSlot(slot)
Categories
(Core :: JavaScript Engine, defect)
Tracking
()
RESOLVED
DUPLICATE
of bug 614714
Tracking | Status | |
---|---|---|
status2.0 | --- | wanted |
People
(Reporter: bc, Unassigned)
References
Details
(Keywords: assertion, Whiteboard: [sg:critical?] )
Attachments
(1 file)
17.33 KB,
text/plain
|
Details |
See also Bug 613619 and Bug 614714
I keep getting this assertion with varying stacks for windows, linux and mac. I've not been able to reproduce them by loading the urls individually and haven't filed a bug for that reason.
Once you have the list of urls, you can test this in a debug build that is built with tests enabled using the tp loader:
firefox -tp manifest.file
Reporter | ||
Comment 2•14 years ago
|
||
FYI, I've seen this on the following domains:
1001likes.free.fr, admfilmes.com, baixae.com, clubedodownload.info, fifa08brasil.blogspot.com, j.mp, mp3hungama.com, rutube.ru, secure.accorhotels.com, shimajoker.blogsky.com, taxes.hrblock.com, translate.google.ca, translate.google.co.id, translate.google.co.in, translate.google.co.ma, translate.google.co.th, translate.google.com, translate.google.com.br, translate.google.com.eg, translate.google.com.pk, translate.google.com.tr, translate.google.com.ua, translate.google.cz, translate.google.es, translate.google.fr, translate.google.hu, translate.google.jo, translate.google.pl, translate.google.pt, translate.google.ro, translate.google.ru, translate.google.sn, uaisodownload.com, verfilmes.net, verfilmesonline.info, www.123chase.com, www.afromontreal.com, www.animefuel.com, www.baixacursos.com.br, www.baixarfilmesgratis.net, www.baixedetudo.net, www.baseballdocs.com, www.cabanaomundaogospel.net.br, www.cineturbo.com, www.degracaemaisgostoso.org, www.downgratis.com, www.elyrics.net, www.emobilez.com, www.esquemaouro.com.br, www.eurosptp.com, www.ezzal.com, www.fairytailepisodes.com, www.funrocker.com, www.futebolaovivo.net, www.g1filmes.com, www.inspirationalstories.com, www.javafile.com, www.justfilmeseseriados.org, www.kanshin.com, www.law-warrior.co.cc, www.linkagratis.net, www.mtv.com, www.ovh.co.uk, www.pcmegarapido.com, www.quick-downloads.com, www.recadosanimados.com, www.righthealth.com, www.scionnation.ca, www.sevendownloads.ws, www.softportal.com, www.songarea.com, www.sonyericsson.com, www.streamtvguide.com, www.treiber-studio.de, www.tripadvisor.co.uk, www.tripadvisor.com, www.tripadvisor.de, www.tripadvisor.in, www.tripadvisor.it, www.unscramble.net, www.urbanfonts.com, www.verfilmes.net, www.vovosalim.com, www.xerox66.com, www.y3.com, www.yooliety.de, www2.mp3raid.com
Reporter | ||
Comment 3•14 years ago
|
||
This is a very common assertion in the crash automation. It may be related to or is hiding the top js crashers for Firefox 4.0. I have it in msvc's debugger at the moment if anyone is interested.
Reporter | ||
Comment 4•14 years ago
|
||
No one appears to care about this one as no one has replied in the bug nor #jsapi, but before I blow the vm session away with the attached debugger here is the stack and auto variables.
Note the 0xde pattern.
- regs.fp 0x02480808 {flags_=6291458 exec={...} args={...} ...} JSStackFrame *
flags_ 6291458 unsigned int
+ exec {script=0x0beb6f00 fun=0x0beb6f00 } JSStackFrame::<unnamed-tag>
+ args {nactual=0 obj=0x00000000 script=0x00000000 } JSStackFrame::<unnamed-tag>
+ scopeChain_ 0x0c514410 {lastProp=0x15795bc8 map=0x15795bc8 clasp=0x03e9f724 ...} JSObject *
+ prev_ 0x024807c0 {flags_=6299650 exec={...} args={...} ...} JSStackFrame *
ncode_ 0xdededede void *
+ rval_ {data={...} } js::Value
+ prevpc_ 0x156e4e8f ":" unsigned char *
+ imacropc_ 0xdededede <Bad Ptr> unsigned char *
hookData_ 0xdededede void *
annotation_ 0xdededede void *
+ sInvalidScopeChain 0x0000beef {lastProp=??? map=??? clasp=??? ...} JSObject * const
slot 149 unsigned int
JS_ASSERT(obj->containsSlot(slot));
mozjs.dll!JS_Assert(const char * s=0x00aa5368, const char * file=0x00aa532c, int ln=5326) Line 73 C++
> mozjs.dll!js::Interpret(JSContext * cx=0x12f50028, JSStackFrame * entryFrame=0x024807c0, unsigned int inlineCallCount=1, JSInterpMode interpMode=JSINTERP_NORMAL) Line 5326 + 0x2f bytes C++
mozjs.dll!js::RunScript(JSContext * cx=0x12f50028, JSScript * script=0x156e4de0, JSStackFrame * fp=0x024807c0) Line 653 + 0x11 bytes C++
Reporter | ||
Comment 5•14 years ago
|
||
There are a 100 different stacks seen so far. The one I have in the debugger is:
JS_Assert | js::Interpret js::RunScript js::Invoke js::ExternalInvoke JS_CallFunctionValue
Updated•14 years ago
|
Comment 6•14 years ago
|
||
Is this likely to be the same as bug 614714? A different bug? A bunch of different bugs?
Reporter | ||
Comment 7•14 years ago
|
||
I tested the original 308 urls plus an additional 672 urls found since then using a mac build on tracemonkey with the patch from bug 614714 and did not reproduce the obj->containsSlot(slot) assertion.
I did see two other assertions.
Assertion failure: JSVAL_IS_DOUBLE_IMPL(data), at /work/mozilla/builds/2.0.0-tracemonkey/mozilla/js/src/jsvalue.h:705
http://www.baixarfilmesgratis.net/page/13/: EXIT STATUS: CRASHED signal 10 SIGBUS (65.559649 seconds)
This is not reproducible on tracemonkey with the patch nor on mozilla-central. :-(
Assertion failure: bit < BitCount, at /work/mozilla/builds/2.0.0-tracemonkey/mozilla/js/src/jsgc.h:210
http://www.raaga.com/channels/hindi/: EXIT STATUS: CRASHED signal 10 SIGBUS (23.502703 seconds)
Assertion failure: bit < BitCount, at /work/mozilla/builds/2.0.0-tracemonkey/mozilla/js/src/jsgc.h:210
http://www.raaga.com/channels/tamil/: EXIT STATUS: CRASHED signal 10 SIGBUS (15.763318 seconds)
These are reproducible on tracemonkey with and without the patch. I'll file a separate bug, but it I have a question first. Is that assertion something to be security sensitive?
Status: NEW → RESOLVED
Closed: 14 years ago
Resolution: --- → DUPLICATE
Reporter | ||
Comment 8•14 years ago
|
||
filed bug 649579 on the bit assert.
Updated•11 years ago
|
Group: core-security
Updated•9 years ago
|
Keywords: testcase-wanted
You need to log in
before you can comment on or make changes to this bug.
Description
•