Am 09.03.2017 um 18:09 schrieb Stefan Eissing:
> Thanks for running our patches with many changes! Not a mistake to have it
> running fine for four days. ;-)
>
> Hmm, so we hunt a rare beast. After much effort from all sides we made it
> rare. So, nothing to be ashamed of. Need to keep looking..
Thanks for running our patches with many changes! Not a mistake to have it
running fine for four days. ;-)
Hmm, so we hunt a rare beast. After much effort from all sides we made it rare.
So, nothing to be ashamed of. Need to keep looking...
> Am 09.03.2017 um 17:32 schrieb Stefan Priebe - Profi
Hi Stefan,
while doing longer testing i can say that also the version which was
working for 4 days segfaults. So it was never solved ;-( Sorry about
that. Testing was just too short.
Greets,
Stefan
Am 05.03.2017 um 14:49 schrieb Stefan Priebe - Profihost AG:
> Hi Stefan,
>
> yes this seems to c
Hi Stefan,
yes this seems to correct BUT i'm not sure i the test was long enough
where i hadn't a segfault. I'll rerun a test with that version. To
verify this was no just a "fault".
Greets,
Stefan
Am 05.03.2017 um 14:34 schrieb Stefan Eissing:
> Thanks. I try to summarize:
>
> - we did see this
Thanks. I try to summarize:
- we did see this rarely with unpatched v1.9.2 and Yann's mpm patch v??
- we still see this rarely with the patch adding mutex to the main conn
allocator (so this seems not to change anything)
- we did not see this with v1.9.2 and Yann's patch on ptrans version ??
Is
Hi Stefan,
current trace - not sure whether this is http2 related at all:
Core was generated by `/usr/local/apache/bin/httpd -DFOREGROUND'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0 allocator_free (node=0x0, allocator=0x7fbf4c05a410)
at memory/unix/apr_pools.c:381
#0 al
Hi Stefan,
live. Sorry for the late reply.
Stefan
Am 28.02.2017 um 11:49 schrieb Stefan Eissing:
> Meh, my mistake. Sorry about that. Did not cleanup properly. Dare you with an
> improved version:
>
>
>
>
>
>> Am 28.02.2017 um 08:41 schrieb Stefan Priebe - Profihost AG
>> :
>>
>> That one
Meh, my mistake. Sorry about that. Did not cleanup properly. Dare you with an
improved version:
h2_192_alloc_mutex_v2.diff
Description: Binary data
> Am 28.02.2017 um 08:41 schrieb Stefan Priebe - Profihost AG
> :
>
> That one breaks everyting. Multiple crashes per second.
>
> [Thread debu
That one breaks everyting. Multiple crashes per second.
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Core was generated by `/usr/local/apache/bin/httpd -DFOREGROUND'.
Program terminated with signal SIGABRT, Aborted.
#0 0x
Damnit. Can you try the attached patch on top of mod_http2 v1.9.2? This one
sets a mutex on the main connection allocator if missing and is pretty close to
the version we ran successfully with on your site for days.
Thanks again!
-Stefan
h2_192_alloc_mutex.diff
Description: Binary data
> Am
Hi Stefan,
two new crashes here.
Reading symbols from /usr/local/apache/bin/httpd...Reading symbols from
/usr/lib/debug//usr/local/apache2/bin/httpd...done.
done.
(gdb) (gdb) quit
Reading symbols from /usr/local/apache/bin/httpd...Reading symbols from
/usr/lib/debug//usr/local/apache2/bin/httpd..
Hi Stefan,
currently everything fine. No segfaults.
Stefan
Am 25.02.2017 um 20:40 schrieb Stefan Priebe - Profihost AG:
> Hi Stefan,
> Am 25.02.2017 um 13:51 schrieb Stefan Eissing:
>> Stefan,
>>
>> whenever you have time, please deploy
>> https://github.com/icing/mod_h2/releases/tag/v1.9.2
>>
12 matches
Mail list logo