I wrote:
> I'm thinking we should change the code and the README to specify that
> you must mark the buffer dirty before you can END_CRIT_SECTION().
While looking at this I realized that in fact we need to, and do,
mark the buffer dirty even earlier than that: look at bufmgr.c
LockBuffer and SyncO
On Fri, 2006-03-31 at 09:36 -0500, Tom Lane wrote:
> Simon Riggs <[EMAIL PROTECTED]> writes:
> > Problem for indexes only. heap xlrecs don't specify exact insert points
>
> Sure they do. They had better, else (for example) the associated index
> insertions will be wrong.
Yep, you're right.
Best
"Qingqing Zhou" <[EMAIL PROTECTED]> writes:
> It may be not good but not harmful either. On step2, the transaction will
> abort and leave a page that has been changed but not marked dirty. There are
> two situtations could happen after that. One is step 3, the other is the
> page is still in the bu
Simon Riggs <[EMAIL PROTECTED]> writes:
> Problem for indexes only. heap xlrecs don't specify exact insert points
Sure they do. They had better, else (for example) the associated index
insertions will be wrong.
> Accesses to local buffers don't need to be critical sections either.
True, but in
On Thu, 2006-03-30 at 13:51 -0500, Tom Lane wrote:
> This is pretty much what heapam and btree currently do, but on looking
> at it I think it's got a problem: we really ought to mark the buffer
> dirty before releasing the critical section. Otherwise, if there's an
> elog(ERROR) before the Write
"Tom Lane" <[EMAIL PROTECTED]> wrote
>
> This is pretty much what heapam and btree currently do, but on looking
> at it I think it's got a problem: we really ought to mark the buffer
> dirty before releasing the critical section. Otherwise, if there's an
> elog(ERROR) before the WriteBuffer call
