Closed Bug 589576 Opened 15 years ago Closed 15 years ago

[SeaMonkey 2.1, reftest, crashtest] timeout/crash on all platforms, plugin-related

Categories

(SeaMonkey :: Testing Infrastructure, defect)

Product:
SeaMonkey
Component:
Testing Infrastructure
Type:
defect
Priority:
Not set
Severity:
critical

Tracking

(Not tracked)

Status:
RESOLVED WORKSFORME

People

(Reporter: kairo, Unassigned)

References

(Blocks 1 open bug)

Details

(Keywords: crash, intermittent-failure, Whiteboard: [sm-perma])

We recently started to have a crash in reftest on all platforms, and it looks like this could be OOPP-related, certainly plugin-related, though. TEST-UNEXPECTED-FAIL | automation.py | application timed out after 60 seconds with no output just after some plugin messages appears on all platforms, Mac says "Can't trigger Breakpad, just killing process" while the others show crash stacks: Linux: Thread 0 (crashed) 0 linux-gate.so + 0x402 eip = 0x00fa7402 esp = 0xbf9323e4 ebp = 0xbf9323f8 ebx = 0x0a305228 esi = 0xffffffff edi = 0x04246ff4 eax = 0xfffffffc ecx = 0x00000008 edx = 0xffffffff efl = 0x00200246 Found by: given as instruction pointer in context 1 libglib-2.0.so.0.1200.3 + 0x2e282 eip = 0x00204283 esp = 0xbf932400 ebp = 0xbf932458 Found by: previous frame's frame pointer 2 libglib-2.0.so.0.1200.3 + 0x2e7b4 eip = 0x002047b5 esp = 0xbf932460 ebp = 0xbf932478 Found by: previous frame's frame pointer 3 libwidget_gtk2.so!nsAppShell::ProcessNextNativeEvent(int) [nsAppShell.cpp:29ac0213b54c : 144 + 0x12] eip = 0x05275f71 esp = 0xbf932480 ebp = 0xbf932498 Found by: previous frame's frame pointer 4 libwidget_gtk2.so!nsBaseAppShell::DoProcessNextNativeEvent(int) [nsBaseAppShell.cpp:29ac0213b54c : 156 + 0x18] eip = 0x052a4cb1 esp = 0xbf9324a0 ebp = 0xbf9324b8 Found by: previous frame's frame pointer 5 libwidget_gtk2.so!nsBaseAppShell::OnProcessNextEvent(nsIThreadInternal*, int, unsigned int) [nsBaseAppShell.cpp:29ac0213b54c : 312 + 0x11] eip = 0x052a51d5 esp = 0xbf9324c0 ebp = 0xbf932508 Found by: previous frame's frame pointer 6 libxpcom_core.so!nsThread::ProcessNextEvent(int, int*) [nsThread.cpp:29ac0213b54c : 517 + 0x61] eip = 0x010643e5 esp = 0xbf932510 ebp = 0xbf932578 Found by: previous frame's frame pointer (see e.g. http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1282473292.1282475473.6594.gz&fulltext=1 for more) Windows: Thread 18 (crashed) 0 crashinjectdll.dll!CrashingThread(void *) [crashinjectdll.cpp:29ac0213b54c : 13 + 0x3] eip = 0x0b60102e esp = 0x0fb8ffb4 ebp = 0x0fb8ffb8 ebx = 0x00000000 esi = 0x00000000 edi = 0x00000000 eax = 0x00000000 ecx = 0x0fb8ffb4 edx = 0x7c8285ec efl = 0x00010246 Found by: given as instruction pointer in context 1 kernel32.dll + 0x24828 eip = 0x77e64829 esp = 0x0fb8ffc0 ebp = 0x0fb8ffec Found by: call frame info (see e.g. http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1282482022.1282484760.28796.gz&fulltext=1 for other threads) he latter looks very much like an intentional crash, probably for OOPP to me...
Whiteboard: [sm-perma][orange]
I just discovered that the crashtest orange we're seeing is the same thing as that.
Summary: [SeaMonkey 2.1, reftest] timeout/crash on all platforms, plugin-related → [SeaMonkey 2.1, reftest, crashtest] timeout/crash on all platforms, plugin-related
Windows last good (reftest): rev:061a69146c3e / moz:2dbb1278a15c Windows first bad (reftest): rev:f734adc19f35 / moz:b211364cfa25 Mac last good (reftest): rev:061a69146c3e / moz:34737d9895f8 Mac first bad (reftest): rev:95ac9dd1a272 / moz:9f434423bdf9 Linux last good (reftest): rev:061a69146c3e / moz:2dbb1278a15c Linux first bad (reftest): rev:95ac9dd1a272 / moz:9f434423bdf9 Linux had other failure before and after, but this is when this one came up. Crashtest has a similar turn on Mac but the others have been orange before, and mac also had orange cycles before, so it's a bit more unclear.
Hrm, and nothing sticks out in those ranges :(
Severity: normal → critical
blocking-seamonkey2.1: --- → ?
Keywords: crash
OS: Linux → All
Hardware: x86 → All
(In reply to comment #1) Fwiw, though I checked Windows (crashtest) only, crashtest sometimes does not crash (but has 1-4+ random oranges). Examples: http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1282456507.1282459776.31022.gz WINNT 5.2 comm-central-trunk debug test crashtest on 2010/08/21 22:55:07 http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1282482822.1282484572.27922.gz WINNT 5.2 comm-central-trunk debug test crashtest on 2010/08/22 06:13:42 http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1283233699.1283235062.27117.gz WINNT 5.2 comm-central-trunk debug test crashtest on 2010/08/30 22:48:19 http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1283324450.1283325955.10941.gz WINNT 5.2 comm-central-trunk debug test crashtest on 2010/09/01 00:00:50
Crashtest again: Linux: crash in 457362-1.xhtml (= Bug 591898) MacOSX: green Windows: time out (= crash!?) randomly in 121533-1.html or 265999-1.html for example Reftest crash current status seems to be "Linux only": Linux: perma-orange (a few failure then crash) MacOSX: random-orange (a few "assertion count 1 is more than expected 0 assertions ") Windows: green It looks like either this initial bug has been fixed or has morphed or should be a meta or...
(In reply to comment #6) > Windows: time out (= crash!?) randomly in 121533-1.html or 265999-1.html for > example If it's a crash, it has a crash stack in the log, it's as easy as that. Windows always has a crash stack there if it crashes.
(In reply to comment #7) Then Windows crashtest is not crashing (anymore), just timing out. And comment 0 Linux reftest crash disappeared between http://tinderbox.mozilla.org/showbuilds.cgi?tree=SeaMonkey&maxdate=1282654579&hours=24&legend=0&norules=1 http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1282609897.1282611875.30732.gz Linux comm-central-trunk debug test reftest on 2010/08/23 17:31:37 http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1282616260.1282618125.27224.gz Linux comm-central-trunk debug test reftest on 2010/08/23 19:17:40 R.WorksForMe NB: I'll file bugs about the other comment 6 issues.
Status: NEW → RESOLVED
blocking-seamonkey2.1: ? → ---
Closed: 15 years ago
Resolution: --- → WORKSFORME
(In reply to comment #6) I filed bugs: > Crashtest again: > Windows: time out (= crash!?) randomly in 121533-1.html or 265999-1.html for > example Bug 593257. > Reftest crash current status seems to be "Linux only": > Linux: perma-orange (a few failure then crash) Bug 593245. > MacOSX: random-orange (a few "assertion count 1 is more than expected 0 > assertions > ") Bug 593278.
Whiteboard: [sm-perma][orange] → [sm-perma]
You need to log in before you can comment on or make changes to this bug.