Re: [Patch] Optimize dropping of relation buffers using dlist

On Tue, Dec 22, 2020 at 8:12 AM Kyotaro Horiguchi
<horikyota.ntt@gmail.com> wrote:
>
> At Tue, 22 Dec 2020 08:08:10 +0530, Amit Kapila <amit.kapila16@gmail.com> wrote in
>
> > Why would all client backends wait for AccessExclusive lock on this
> > relation? Say, a client needs a buffer for some other relation and
> > that might evict this buffer after we release the lock on the
> > partition. In StrategyGetBuffer, it is important to either have a pin
> > on the buffer or the buffer header itself must be locked to avoid
> > getting picked as victim buffer. Am I missing something?
>
> I think exactly like that.  If we acquire the bufHdr lock before
> releasing the partition lock, that steal doesn't happen but it doesn't
> seem good as a locking protocol.
>

Right, so let's keep the code as it is but I feel it is better to add
some comments explaining the rationale behind this code.

-- 
With Regards,
Amit Kapila.