"Qingqing Zhou" <zhouqq@cs.toronto.edu> wrote
>
> The overall performance improvement might be marginal but why not if it is
> right. What I cares is the correctness. As I understand, the orginal code
> puts a shared lock (1) to prevent the vacuum process to move tuples around
> so the hint bits change may happen in a wrong place; (2) to prevent other
> operations holding EXCLUSIVE lock to change bits at the same time.
>
I realized I made an aweful mistake. The shared lock also (3) to prevent
other operations holding EXCLUSIVE lock to change the xid fields at the
same. So the final conclusion is: the original code is right and my patch is
terriblly wrong :-(
Regards,
Qingqing