Re: Post-mortem: final 2PC patch

Tom Lane wrote:
> Stefan Kaltenbrunner <stefan(at)kaltenbrunner(dot)cc> writes:
>
>>FYI: this commit seems to cause problems/crashes during make check on my
>>OpenBSD/Sparc64 buildfarmclient:
>
>
>>http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=spoonbill&dt=2005-06-17%2023:50:04
>
>
>>I just checked manually with a fresh checkout - and I got similiar failures.
>
>
> Can you get a stack trace from the core dump?

(gdb) bt
#0 0x0000000050377ba4 in memcpy () from /usr/lib/libc.so.34.2
#1 0x0000000000326efc in hash_search ()
#2 0x0000000000343430 in pg_tzset ()
#3 0x00000000001fbcf0 in assign_timezone ()
#4 0x0000000000330e88 in set_config_option ()
#5 0x000000000029b5b0 in PostgresMain ()
#6 0x000000000026b140 in BackendRun ()
#7 0x000000000026a9f0 in BackendStartup ()
#8 0x00000000002685bc in ServerLoop ()
#9 0x0000000000267b88 in PostmasterMain ()
#10 0x0000000000220cc8 in main ()

rebuilding with --enable-debug right now ...

>
> As best I can tell from the buildfarm report, one of the regression
> tests is causing a sig11 --- but it's *not* the prepared_xacts test,
> so I'm not sure the failure is related to this patch.

hmm - maybe one of the timezone patches ?

Stefan