Re: Bus error in pg_logical_slot_get_changes (9.4.7, sparc)

Hi,

On 2016-04-13 11:41:17 +0200, Christoph Berg wrote:
> I've just uploaded 9.4.7 for Debian's wheezy-backports suite.
> Unfortunately it's now crashing twice in the contrib/test_decoding
> regression tests on sparc. I could reproduce the error on the porter
> box:
>
> LOG: server process (PID 22534) was terminated by signal 10: Bus error
> DETAIL: Failed process was running: SELECT data
> FROM pg_logical_slot_get_changes('regression_slot', NULL, NULL, 'include-xids', '0', 'skip-empty-xacts', '1')
> WHERE data ~ 'UPDATE';
>
> LOG: server process (PID 22558) was terminated by signal 10: Bus error
> DETAIL: Failed process was running: SELECT substr(data, 1, 200) FROM pg_logical_slot_get_changes('regression_slot', NULL, NULL, 'include-xids', '0', 'skip-empty-xacts', '1');
>
>
> Reading symbols from /home/myon/postgresql-9.4-9.4.7/build/src/backend/postgres...done.
> [New LWP 22558]
> [Thread debugging using libthread_db enabled]
> Using host libthread_db library "/lib/sparc-linux-gnu/libthread_db.so.1".
> Core was generated by `postgres: myon regression [local] SELECT '.
> Program terminated with signal 10, Bus error.
> #0 ReorderBufferRestoreChange (data=0x5de0f7 "", txn=0x622660, rb=0x6225d8)
> at /home/myon/postgresql-9.4-9.4.7/build/../src/backend/replication/logical/reorderbuffer.c:2327
> 2327 Size tuplelen = ((HeapTuple) data)->t_len;
> (gdb) l
> 2322 data += tuplelen;
> 2323 }
> 2324
> 2325 if (change->data.tp.newtuple)
> 2326 {
> 2327 Size tuplelen = ((HeapTuple) data)->t_len;

Hrmpf, I guess this is an alignment issue. We should probably replace
that ->t_len lookup with a memcpy. IIRC it's not entirely trivial to
access porter boxes anymore - could you perhaps just replace those two
t_len lookups with a memcpy (i.e. memcpy(&tuplelen, &(HeapTuple)
data)->t_len, sizeof(uint32))?

Greetings,

Andres Freund