Re: WIP patch for Oid formatting in printf/elog strings

Mark Dilger wrote:
> I found a few places in the code where a variable of
> type Oid is printed using "%d" rather than "%u" and
> changed them in the attached patch.
>
> In src/backend/replication/logical/reorderbuffer.c,
> circa line 2494, chunk_seq is of type Oid, but
> ent->last_chunk_seq is of type int32, leading me
> to question if perhaps the use of %d for chunk_seq
> is correct, but the use of Oid for the type of chunk_seq
> is in error. If neither is in error, perhaps someone
> can provide a short code comment explaining the
> logic of the signed/unsigned discrepancy.

tuptoaster defines chunk_seq as signed int32; ReorderBufferToastAppendChunk
is wrong in declaring it Oid, and so this part of your patch is bogus.
The others seem correct.

> diff --git a/src/backend/replication/logical/reorderbuffer.c b/src/backend/replication/logical/reorderbuffer.c
> index 6e75398..41a4896 100644
> --- a/src/backend/replication/logical/reorderbuffer.c
> +++ b/src/backend/replication/logical/reorderbuffer.c
> @@ -2487,11 +2487,11 @@ ReorderBufferToastAppendChunk(ReorderBuffer *rb, ReorderBufferTXN *txn,
> dlist_init(&ent->chunks);
>
> if (chunk_seq != 0)
> - elog(ERROR, "got sequence entry %d for toast chunk %u instead of seq 0",
> + elog(ERROR, "got sequence entry %u for toast chunk %u instead of seq 0",
> chunk_seq, chunk_id);
> }
> else if (found && chunk_seq != ent->last_chunk_seq + 1)
> - elog(ERROR, "got sequence entry %d for toast chunk %u instead of seq %d",
> + elog(ERROR, "got sequence entry %u for toast chunk %u instead of seq %d",
> chunk_seq, chunk_id, ent->last_chunk_seq + 1);
>
> chunk = DatumGetPointer(fastgetattr(&newtup->tuple, 3, desc, &isnull));

--
Álvaro Herrera http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services