Open Bug 1118175 Opened 9 years ago Updated 1 year ago

C-C WinXP debug XPCShell tests fail, and people incorrectly think that: WindowsError: [Error 5] Access is denied is the problem

Categories

(Testing :: XPCShell Harness, defect)

Product:
Testing
Component:
XPCShell Harness
Type:
defect

Tracking

(Not tracked)

People

(Reporter: darktrojan, Unassigned)

References

Details

(Whiteboard: [windows][test][thunderbird]) 

(In reply to Geoff Lankow (:darktrojan) from comment #0)
> Here's an example log:
> https://ftp.mozilla.org/pub/mozilla.org/thunderbird/tinderbox-builds/comm-
> central-win32-debug/1420483403/comm-central_xp-ix-debug_test-xpcshell-bm110-
> tests1-windows-build37.txt.gz

Those logs are really ephemeral, and that particular one is already gone. Can you possibly provide longer snippet, or copy the log somewhere more permanent, please?
> 17:42:32     INFO -  Failed to remove directory: c:\docume~1\cltbld~1.t-x\locals~1\temp\tmpanrrgm. Waiting.

There seem to be 50 failed attempts to remove the directory, then:

> 17:47:24     INFO -  Exception in thread Thread-2674:
> 17:47:24     INFO -  Traceback (most recent call last):
> 17:47:24     INFO -    File "c:\mozilla-build\python27\Lib\threading.py", line 551, in __bootstrap_inner
> 17:47:24     INFO -      self.run()
> 17:47:24     INFO -    File "c:\mozilla-build\python27\Lib\threading.py", line 755, in run
> 17:47:24     INFO -      self.function(*self.args, **self.kwargs)
> 17:47:24     INFO -    File "C:\slave\test\build\tests\xpcshell\runxpcshelltests.py", line 642, in <lambda>
> 17:47:24     INFO -      testTimer = Timer(testTimeoutInterval, lambda: self.testTimeout(proc))
> 17:47:24     INFO -    File "C:\slave\test\build\tests\xpcshell\runxpcshelltests.py", line 305, in testTimeout
> 17:47:24     INFO -      self.postCheck(proc)
> 17:47:24     INFO -    File "C:\slave\test\build\tests\xpcshell\runxpcshelltests.py", line 273, in postCheck
> 17:47:24     INFO -      self.kill(proc)
> 17:47:24     INFO -    File "C:\slave\test\build\tests\xpcshell\runxpcshelltests.py", line 177, in kill
> 17:47:24     INFO -      return proc.kill()
> 17:47:24     INFO -    File "c:\mozilla-build\python27\Lib\subprocess.py", line 1019, in terminate
> 17:47:24     INFO -      _subprocess.TerminateProcess(self._handle, 1)
> 17:47:24     INFO -  WindowsError: [Error 5] Access is denied
> 17:47:24     INFO -  Can't trigger Breakpad, just killing process
> 17:47:24     INFO -  Can't trigger Breakpad, just killing process
> 17:47:24     INFO -  xpcshell return code: None
> 17:47:24     INFO -  Failed to kill process 1624.

There's only one log to look at right now but IIRC the other one I saw was the same thing. I've copied it to https://pastebin.mozilla.org/8169445
I'm happy to loan someone an XP slave if they want to debug this (https://wiki.mozilla.org/ReleaseEngineering/How_To/Request_a_slave), but buildduty is unlikely to have time to dig into this.
Component: Buildduty → General Automation
QA Contact: bugspam.Callek → catlee
Whiteboard: [windows][test][thunderbird]
FWIW, I've just seen the same error on my last (m-c) Try push, on Windows 8 x64, both debug and opt.
Keywords: intermittent-failure
Little difficult to tell what this bug is actually about since nobody ever seems to be willing to say "this instance is what it's about" and then copy-paste-attach enough of a log for context, but these last few things that have been starred as this are "a bunch of xpcshell tests time out on Thunderbird, and the xpcshell harness is totally broken at handling timeouts so people who don't know that think the harness bustage is the actual bustage."
Component: General Automation → XPCShell Harness
Product: Release Engineering → Testing
QA Contact: catlee
Version: unspecified → Trunk
Depends on: 1135969
Summary: C-C WinXP debug XPCShell tests fail: WindowsError: [Error 5] Access is denied → C-C WinXP debug XPCShell tests fail, and people incorrectly thing that: WindowsError: [Error 5] Access is denied is the problem
Summary: C-C WinXP debug XPCShell tests fail, and people incorrectly thing that: WindowsError: [Error 5] Access is denied is the problem → C-C WinXP debug XPCShell tests fail, and people incorrectly think that: WindowsError: [Error 5] Access is denied is the problem
Right, all this misstarring of actual failures which should be filed but have not been filed is not getting us anywhere.
Keywords: intermittent-failure
Severity: normal → S3
You need to log in before you can comment on or make changes to this bug.