Re: PL/R Median Busts Commit (Postgres 9.1.6 + plr 8.3.0.13 on Ubuntu 12.10 64 bit)

From: Mark Kirkwood <mark(dot)kirkwood(at)catalyst(dot)net(dot)nz>
To: pgsql-bugs(at)postgresql(dot)org
Subject: Re: PL/R Median Busts Commit (Postgres 9.1.6 + plr 8.3.0.13 on Ubuntu 12.10 64 bit)
Date: 2013-01-24 22:41:51
Message-ID: 5101B8AF.3090307@catalyst.net.nz
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

On 25/01/13 10:36, Tom Lane wrote:
> Mark Kirkwood <mark(dot)kirkwood(at)catalyst(dot)net(dot)nz> writes:
>> Doh! Yes of course, sorry for the noise. I was busy thinking that the
>> issue could be tied up with sinval and plan caching (if there is any) in
>> plr and got excited about seeing something in gdb...and didn't think
>> carefully about why what I was seeing was not a bug at all :-( Thanks
>> for clarifying!
>
> FWIW, presumably the failure comes from something throwing elog(ERROR)
> after RecordTransactionCommit(). So what I'd try is setting a
> breakpoint at errfinish and investigating what's causing the first error
> call.

If I have done this right, then this is the trace for the 1st message...
from my wandering through the calls here it looks like a normal commit,
and something goes a bit weird as SI messages are being processed...

0x00007f4e2a538425 in raise () from /lib/x86_64-linux-gnu/libc.so.6
(gdb) bt
#0 0x00007f4e2a538425 in raise () from /lib/x86_64-linux-gnu/libc.so.6
#1 0x00007f4e2a53bb8b in abort () from /lib/x86_64-linux-gnu/libc.so.6
#2 0x00000000006e0a70 in errfinish (dummy=dummy(at)entry=0) at elog.c:523
#3 0x00000000006e3031 in elog_finish (elevel=elevel(at)entry=22,
fmt=fmt(at)entry=0x714d58 "cannot abort transaction %u, it was already
committed") at elog.c:1202
#4 0x0000000000481d87 in RecordTransactionAbort (
isSubXact=isSubXact(at)entry=0 '\000') at xact.c:1366
#5 0x0000000000481e89 in AbortTransaction () at xact.c:2272
#6 0x00000000004851c5 in AbortOutOfAnyTransaction () at xact.c:3807
#7 0x00000000006eddc9 in ShutdownPostgres (code=<optimized out>,
arg=<optimized out>) at postinit.c:976
#8 0x0000000000610b7d in shmem_exit (code=code(at)entry=-1) at ipc.c:221
#9 0x0000000000610bf5 in proc_exit_prepare (code=-1) at ipc.c:181
#10 0x00007f4e2a53d901 in ?? () from /lib/x86_64-linux-gnu/libc.so.6
#11 0x00007f4e2a53d985 in exit () from /lib/x86_64-linux-gnu/libc.so.6
#12 0x00007f4e272b951a in ?? () from /usr/lib/libR.so
#13 <signal handler called>
#14 0x00007f4e2a538707 in kill () from /lib/x86_64-linux-gnu/libc.so.6
#15 0x00000000006152e5 in SICleanupQueue (
callerHasWriteLock=callerHasWriteLock(at)entry=1 '\001',
minFree=minFree(at)entry=4) at sinvaladt.c:672
#16 0x00000000006153bd in SIInsertDataEntries (data=0x3943b70, n=0)
at sinvaladt.c:437
#17 0x00000000006d1371 in ProcessInvalidationMessagesMulti (
hdr=hdr(at)entry=0x204a280, func=0x6149f0 <SendSharedInvalidMessages>)
at inval.c:402
#18 0x00000000006d2020 in AtEOXact_Inval (isCommit=isCommit(at)entry=1 '\001')
at inval.c:940
#19 0x0000000000482bdd in CommitTransaction () at xact.c:1893
#20 0x0000000000484975 in CommitTransactionCommand () at xact.c:2562
#21 0x00000000006276d3 in finish_xact_command () at postgres.c:2452
#22 finish_xact_command () at postgres.c:2441
#23 0x000000000062a6f5 in exec_simple_query (query_string=0x20e8d70
"COMMIT;")
at postgres.c:1037
#24 PostgresMain (argc=<optimized out>, argv=argv(at)entry=0x20476b8,
username=<optimized out>) at postgres.c:3968
#25 0x00000000005ee92f in BackendRun (port=0x2068f40) at postmaster.c:3617
#26 BackendStartup (port=0x2068f40) at postmaster.c:3302
#27 ServerLoop () at postmaster.c:1466
#28 0x00000000005ef4dc in PostmasterMain (argc=argc(at)entry=1,
argv=argv(at)entry=0x2044640) at postmaster.c:1127
#29 0x0000000000453edb in main (argc=1, argv=0x2044640) at main.c:199

In response to

Responses

Browse pgsql-bugs by date

  From Date Subject
Next Message Alvaro Herrera 2013-01-24 23:43:37 Re: BUG #6528: pglesslog still referenced in docs, but no 9.1 support
Previous Message Bruce Momjian 2013-01-24 22:21:41 Re: BUG #6528: pglesslog still referenced in docs, but no 9.1 support