Failed assert ((data - start) == data_size) in heaptuple.c

Hi folks,

I am running a 9.0.3 Hot Standy + Streaming Replication slave which
occasionally segfaults (every 1-2 days). I rebuilt Postgres with
--enable-cassert and --enable-debug, switched on core dumping and
waited for some results.

The first crash since enabling debugging was a failed assert in heaptuple.c:

TRAP: FailedAssertion("!((data - start) == data_size)", File:
"heaptuple.c", Line: 255)
2011-04-07 04:20:20 EST LOG: server process (PID 32195) was
terminated by signal 6: Aborted
2011-04-07 04:20:20 EST LOG: terminating any other active server processes

For context, the only things running on this server are the slave
database, and a tomcat instance. The tomcat instance is the only
connection into this database, which continually runs through a series
of SELECTs (100ms sleep between each run).

The slave database is basically stock standard 9.0.3 config, apart
from the replication setup and shared_buffers increased to 2GB.

Here's the backtrace:

Core was generated by `postgres: backend surecast 127.0.0.1(37155)
SELECT '.
Program terminated with signal 6, Aborted.
#0 0x00007f40a93aba75 in *__GI_raise (sig=<value optimised out>)
at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
64 ../nptl/sysdeps/unix/sysv/linux/raise.c: No such file or directory.
in ../nptl/sysdeps/unix/sysv/linux/raise.c
(gdb) bt
#0 0x00007f40a93aba75 in *__GI_raise (sig=<value optimised out>)
at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
#1 0x00007f40a93af5c0 in *__GI_abort () at abort.c:92
#2 0x00000000006f861d in ExceptionalCondition (conditionName=<value
optimised out>,
errorType=<value optimised out>, fileName=<value optimised out>,
lineNumber=<value optimised out>) at assert.c:57
#3 0x0000000000459b07 in heap_form_minimal_tuple
(tupleDescriptor=0xf6bdd0, values=0x8,
isnull=0xf6c680 "") at heaptuple.c:1459
#4 0x0000000000580d12 in ExecFetchSlotMinimalTuple (slot=0xf6bb90) at
execTuples.c:684
#5 0x0000000000588d10 in ExecHashTableInsert (hashtable=0xf4c3b0, slot=0x7dc3,
hashvalue=6) at nodeHash.c:697
#6 0x0000000000589bf6 in MultiExecHash (node=<value optimised out>)
at nodeHash.c:123
#7 0x000000000058a9ab in ExecHashJoin (node=0xf24008) at nodeHashjoin.c:154
#8 0x00000000005788a8 in ExecProcNode (node=0xf24008) at execProcnode.c:427
#9 0x000000000058fb21 in ExecNestLoop (node=0xf8eb98) at nodeNestloop.c:120
#10 0x00000000005788c8 in ExecProcNode (node=0xf8eb98) at execProcnode.c:419
#11 0x000000000057756d in ExecutePlan (queryDesc=0xf8bc30,
direction=32195, count=0)
at execMain.c:1187
#12 standard_ExecutorRun (queryDesc=0xf8bc30, direction=32195,
count=0) at execMain.c:280
#13 0x0000000000642e28 in PortalRunSelect (portal=0xf11f88,
forward=<value optimised out>,
count=0, dest=0xe00120) at pquery.c:952
#14 0x00000000006442e9 in PortalRun (portal=<value optimised out>,
count=<value optimised out>, isTopLevel=<value optimised out>,
dest=<value optimised out>, altdest=<value optimised out>,
completionTag=<value optimised out>) at pquery.c:796
#15 0x00000000006419e3 in exec_execute_message (argc=<value optimised out>,
argv=<value optimised out>, username=<value optimised out>) at
postgres.c:2003
#16 PostgresMain (argc=<value optimised out>, argv=<value optimised out>,
username=<value optimised out>) at postgres.c:3988
#17 0x0000000000606a07 in BackendRun () at postmaster.c:3555
#18 BackendStartup () at postmaster.c:3242
#19 ServerLoop () at postmaster.c:1431
#20 0x000000000060931d in PostmasterMain (argc=14918336, argv=0xdfb160)
at postmaster.c:1092
#21 0x00000000005a9310 in main (argc=5, argv=0xdfb140) at main.c:188

Let me know if there is any additional information I can provide.

Cheers,
BJ