On 16/03/17 09:44, Andres Freund wrote: > On 2017-03-16 09:40:48 +0100, Petr Jelinek wrote: >> On 16/03/17 04:42, Andres Freund wrote: >>> On 2017-03-15 20:28:33 -0700, Andres Freund wrote: >>>> Hi, >>>> >>>> I just unstuck a bunch of my buildfarm animals. That triggered some >>>> spurious failures (on piculet, calliphoridae, mylodon), but also one >>>> that doesn't really look like that: >>>> https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=culicidae&dt=2017-03-16%2002%3A40%3A03 >>>> >>>> with the pertinent point being: >>>> >>>> ================== stack trace: >>>> pgsql.build/src/test/regress/tmp_check/data/core ================== >>>> [New LWP 1894] >>>> [Thread debugging using libthread_db enabled] >>>> Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". >>>> Core was generated by `postgres: bgworker: logical replication launcher >>>> '. >>>> Program terminated with signal SIGSEGV, Segmentation fault. >>>> #0 0x000055e265bff5e3 in ?? () >>>> #0 0x000055e265bff5e3 in ?? () >>>> #1 0x000055d3ccabed0d in StartBackgroundWorker () at >>>> /home/andres/build/buildfarm-culicidae/HEAD/pgsql.build/../pgsql/src/backend/postmaster/bgworker.c:792 >>>> #2 0x000055d3ccacf4fc in SubPostmasterMain (argc=3, argv=0x55d3cdbb71c0) >>>> at >>>> /home/andres/build/buildfarm-culicidae/HEAD/pgsql.build/../pgsql/src/backend/postmaster/postmaster.c:4878 >>>> #3 0x000055d3cca443ea in main (argc=3, argv=0x55d3cdbb71c0) at >>>> /home/andres/build/buildfarm-culicidae/HEAD/pgsql.build/../pgsql/src/backend/main/main.c:205 >>>> >>>> it's possible that me killing things and upgrading caused this, but >>>> given this is a backend running EXEC_BACKEND, I'm a bit suspicous that >>>> it's more than that. The machine is a bit backed up at the moment, so >>>> it'll probably be a while till it's at that animal/branch again, >>>> otherwise I'd not have mentioned this. >>> >>> For some reason it ran again pretty soon. And I'm afraid it's indeed an >>> issue: >>> https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=culicidae&dt=2017-03-16%2003%3A30%3A02 >>> >> >> Hmm, I tried with EXEC_BACKEND (and with --disable-spinlocks) and it >> seems to work fine on my two machines. I don't see anything else >> different on culicidae though. Sadly the backtrace is not that >> informative either. I'll try to investigate more but it will take time... > > I can give you a login to that machine, it doesn't do anything but run > buildfarm animals... Will have to be my tomorrow however. >
That would be helpful, thanks. -- Petr Jelinek http://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Training & Services -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
