Between r2086 and 2089 the phobos tests on win32 only have started hanging.

Last good:
    http://d.puremagic.com/test-results/test.ghtml?runid=1374

First bad:
    http://d.puremagic.com/test-results/test.ghtml?runid=1376

On 10/7/2010 9:09 PM, Brad Roberts wrote:
> Between r2086 and 2089 the phobos tests on win32 only have started hanging.
> 
> Last good:
>     http://d.puremagic.com/test-results/test.ghtml?runid=1374
> 
> First bad:
>     http://d.puremagic.com/test-results/test.ghtml?runid=1376

Until someone tells me it's been fixed (or I get a timeout to kill the stalled test coded up), I've disabled the phobos test step so that at least the rest of the testing occurs.

On 10/7/2010 9:22 PM, Brad Roberts wrote:
> On 10/7/2010 9:09 PM, Brad Roberts wrote:
>> Between r2086 and 2089 the phobos tests on win32 only have started hanging.
>>
>> Last good:
>>     http://d.puremagic.com/test-results/test.ghtml?runid=1374
>>
>> First bad:
>>     http://d.puremagic.com/test-results/test.ghtml?runid=1376
> 
> Until someone tells me it's been fixed (or I get a timeout to kill the stalled test coded up), I've disabled the phobos test step so that at least the rest of the testing occurs.

Ok, now that the linux builds are back in action, it's phobos tests also hang. But since the posix based test mechanisms are a lot nicer, it's easy to tell that it's the concurrency test that's at fault.  It seems to be a pure stall, as the app is idle, not consuming cpu time.

Strace shows no activity.

I tried attaching gdb and getting a stack trace, but gdb's stack walker got pissed off (slightly edited for brevity):

(gdb) thr apply all bt

Thread 2 (Thread 0x404d2b70 (LWP 25772)):
#0  0x4001d422 in __kernel_vsyscall ()
#1  0x40041015 in pthread_cond_wait@@GLIBC_2.3.2 () from
/lib/tls/i686/cmov/libpthread.so.0
#2  0x0809cc2a in core.sync.condition.Condition.wait() ()
buildsym.c:818: internal-error: start_symtab: Assertion `! pending_addrmap' failed.

Thread 1 (Thread 0x401d1780 (LWP 25771)):
#0  0x4001d422 in __kernel_vsyscall ()
#1  0x40041015 in pthread_cond_wait@@GLIBC_2.3.2 () from
/lib/tls/i686/cmov/libpthread.so.0
#2  0x0809cc2a in core.sync.condition.Condition.wait() ()
buildsym.c:818: internal-error: start_symtab: Assertion `! pending_addrmap' failed.

Must be related to my bug fixes.  Maybe there's a bad test. I'll check it out today.

Sent from my iPhone

On Oct 8, 2010, at 12:05 AM, Brad Roberts <braddr at puremagic.com> wrote:

> On 10/7/2010 9:22 PM, Brad Roberts wrote:
>> On 10/7/2010 9:09 PM, Brad Roberts wrote:
>>> Between r2086 and 2089 the phobos tests on win32 only have started hanging.
>>> 
>>> Last good:
>>>    http://d.puremagic.com/test-results/test.ghtml?runid=1374
>>> 
>>> First bad:
>>>    http://d.puremagic.com/test-results/test.ghtml?runid=1376
>> 
>> Until someone tells me it's been fixed (or I get a timeout to kill the stalled test coded up), I've disabled the phobos test step so that at least the rest of the testing occurs.
> 
> Ok, now that the linux builds are back in action, it's phobos tests also hang. But since the posix based test mechanisms are a lot nicer, it's easy to tell that it's the concurrency test that's at fault.  It seems to be a pure stall, as the app is idle, not consuming cpu time.
> 
> Strace shows no activity.
> 
> I tried attaching gdb and getting a stack trace, but gdb's stack walker got pissed off (slightly edited for brevity):
> 
> (gdb) thr apply all bt
> 
> Thread 2 (Thread 0x404d2b70 (LWP 25772)):
> #0  0x4001d422 in __kernel_vsyscall ()
> #1  0x40041015 in pthread_cond_wait@@GLIBC_2.3.2 () from
> /lib/tls/i686/cmov/libpthread.so.0
> #2  0x0809cc2a in core.sync.condition.Condition.wait() ()
> buildsym.c:818: internal-error: start_symtab: Assertion `! pending_addrmap' failed.
> 
> Thread 1 (Thread 0x401d1780 (LWP 25771)):
> #0  0x4001d422 in __kernel_vsyscall ()
> #1  0x40041015 in pthread_cond_wait@@GLIBC_2.3.2 () from
> /lib/tls/i686/cmov/libpthread.so.0
> #2  0x0809cc2a in core.sync.condition.Condition.wait() ()
> buildsym.c:818: internal-error: start_symtab: Assertion `! pending_addrmap' failed.
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos

Fixed.  I thought I'd dealt with this, but it must have been in test code I never committed.

On Oct 8, 2010, at 6:43 AM, Sean Kelly wrote:

> Must be related to my bug fixes.  Maybe there's a bad test. I'll check it out today.
> 
> Sent from my iPhone
> 
> On Oct 8, 2010, at 12:05 AM, Brad Roberts <braddr at puremagic.com> wrote:
> 
>> On 10/7/2010 9:22 PM, Brad Roberts wrote:
>>> On 10/7/2010 9:09 PM, Brad Roberts wrote:
>>>> Between r2086 and 2089 the phobos tests on win32 only have started hanging.
>>>> 
>>>> Last good:
>>>>   http://d.puremagic.com/test-results/test.ghtml?runid=1374
>>>> 
>>>> First bad:
>>>>   http://d.puremagic.com/test-results/test.ghtml?runid=1376
>>> 
>>> Until someone tells me it's been fixed (or I get a timeout to kill the stalled test coded up), I've disabled the phobos test step so that at least the rest of the testing occurs.
>> 
>> Ok, now that the linux builds are back in action, it's phobos tests also hang. But since the posix based test mechanisms are a lot nicer, it's easy to tell that it's the concurrency test that's at fault.  It seems to be a pure stall, as the app is idle, not consuming cpu time.
>> 
>> Strace shows no activity.
>> 
>> I tried attaching gdb and getting a stack trace, but gdb's stack walker got pissed off (slightly edited for brevity):
>> 
>> (gdb) thr apply all bt
>> 
>> Thread 2 (Thread 0x404d2b70 (LWP 25772)):
>> #0  0x4001d422 in __kernel_vsyscall ()
>> #1  0x40041015 in pthread_cond_wait@@GLIBC_2.3.2 () from
>> /lib/tls/i686/cmov/libpthread.so.0
>> #2  0x0809cc2a in core.sync.condition.Condition.wait() ()
>> buildsym.c:818: internal-error: start_symtab: Assertion `! pending_addrmap' failed.
>> 
>> Thread 1 (Thread 0x401d1780 (LWP 25771)):
>> #0  0x4001d422 in __kernel_vsyscall ()
>> #1  0x40041015 in pthread_cond_wait@@GLIBC_2.3.2 () from
>> /lib/tls/i686/cmov/libpthread.so.0
>> #2  0x0809cc2a in core.sync.condition.Condition.wait() ()
>> buildsym.c:818: internal-error: start_symtab: Assertion `! pending_addrmap' failed.
>> _______________________________________________
>> phobos mailing list
>> phobos at puremagic.com
>> http://lists.puremagic.com/mailman/listinfo/phobos
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos

Thanks Sean.  I reenabled the phobos tests and they're passing once again on both Win_32 and Linux_32 (still need to get macos and freebsd added).

On Fri, 8 Oct 2010, Sean Kelly wrote:

> Date: Fri, 8 Oct 2010 10:36:39 -0700
> From: Sean Kelly <sean at invisibleduck.org>
> Reply-To: Discuss the phobos library for D <phobos at puremagic.com>
> To: Discuss the phobos library for D <phobos at puremagic.com>
> Subject: Re: [phobos] unit test hangs
> 
> Fixed.  I thought I'd dealt with this, but it must have been in test code I never committed.
> 
> On Oct 8, 2010, at 6:43 AM, Sean Kelly wrote:
> 
> > Must be related to my bug fixes.  Maybe there's a bad test. I'll check it out today.
> > 
> > Sent from my iPhone
> > 
> > On Oct 8, 2010, at 12:05 AM, Brad Roberts <braddr at puremagic.com> wrote:
> > 
> >> On 10/7/2010 9:22 PM, Brad Roberts wrote:
> >>> On 10/7/2010 9:09 PM, Brad Roberts wrote:
> >>>> Between r2086 and 2089 the phobos tests on win32 only have started hanging.
> >>>> 
> >>>> Last good:
> >>>>   http://d.puremagic.com/test-results/test.ghtml?runid=1374
> >>>> 
> >>>> First bad:
> >>>>   http://d.puremagic.com/test-results/test.ghtml?runid=1376
> >>> 
> >>> Until someone tells me it's been fixed (or I get a timeout to kill the stalled test coded up), I've disabled the phobos test step so that at least the rest of the testing occurs.
> >> 
> >> Ok, now that the linux builds are back in action, it's phobos tests also hang. But since the posix based test mechanisms are a lot nicer, it's easy to tell that it's the concurrency test that's at fault.  It seems to be a pure stall, as the app is idle, not consuming cpu time.
> >> 
> >> Strace shows no activity.
> >> 
> >> I tried attaching gdb and getting a stack trace, but gdb's stack walker got pissed off (slightly edited for brevity):
> >> 
> >> (gdb) thr apply all bt
> >> 
> >> Thread 2 (Thread 0x404d2b70 (LWP 25772)):
> >> #0  0x4001d422 in __kernel_vsyscall ()
> >> #1  0x40041015 in pthread_cond_wait@@GLIBC_2.3.2 () from
> >> /lib/tls/i686/cmov/libpthread.so.0
> >> #2  0x0809cc2a in core.sync.condition.Condition.wait() ()
> >> buildsym.c:818: internal-error: start_symtab: Assertion `! pending_addrmap' failed.
> >> 
> >> Thread 1 (Thread 0x401d1780 (LWP 25771)):
> >> #0  0x4001d422 in __kernel_vsyscall ()
> >> #1  0x40041015 in pthread_cond_wait@@GLIBC_2.3.2 () from
> >> /lib/tls/i686/cmov/libpthread.so.0
> >> #2  0x0809cc2a in core.sync.condition.Condition.wait() ()
> >> buildsym.c:818: internal-error: start_symtab: Assertion `! pending_addrmap' failed.
> >> _______________________________________________
> >> phobos mailing list
> >> phobos at puremagic.com
> >> http://lists.puremagic.com/mailman/listinfo/phobos
> > _______________________________________________
> > phobos mailing list
> > phobos at puremagic.com
> > http://lists.puremagic.com/mailman/listinfo/phobos
> 
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos
>