Thom Brown wrote:
> Hit another issue with exactly the same procedure:
>
> pgbench=# create index minmaxtest on pgbench_accounts using minmax (aid);
> ERROR: lock 176475 is not held
That's what I get for restructuring the way buffers are acquired to use
the FSM, and then neglecting to test creation on decently-sized indexes.
Fix attached.
I just realized that xlog replay is also broken.
--
Álvaro Herrera http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
diff --git a/src/backend/access/minmax/minmax.c b/src/backend/access/minmax/minmax.c
index 3b41100..47cb05e 100644
--- a/src/backend/access/minmax/minmax.c
+++ b/src/backend/access/minmax/minmax.c
@@ -1510,10 +1510,8 @@ mm_getinsertbuffer(Relation irel, Buffer *buffer, Size itemsz)
}
if (!BufferIsInvalid(*buffer))
- {
- MarkBufferDirty(*buffer);
- UnlockReleaseBuffer(*buffer);
- }
+ ReleaseBuffer(*buffer);
+
*buffer = buf;
}
else
--
Sent via pgsql-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers
