| From: | Kyotaro Horiguchi <horikyota(dot)ntt(at)gmail(dot)com> |
|---|---|
| To: | k(dot)jamison(at)fujitsu(dot)com |
| Cc: | tsunakawa(dot)takay(at)fujitsu(dot)com, andres(at)anarazel(dot)de, amit(dot)kapila16(at)gmail(dot)com, tgl(at)sss(dot)pgh(dot)pa(dot)us, thomas(dot)munro(at)gmail(dot)com, robertmhaas(at)gmail(dot)com, tomas(dot)vondra(at)2ndquadrant(dot)com, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: [Patch] Optimize dropping of relation buffers using dlist |
| Date: | 2020-11-26 07:40:17 |
| Message-ID: | 20201126.164017.909760241874498640.horikyota.ntt@gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
At Thu, 26 Nov 2020 16:18:55 +0900 (JST), Kyotaro Horiguchi <horikyota(dot)ntt(at)gmail(dot)com> wrote in
> + /* Zero the array of blocks because these will all be dropped anyway */
> + MemSet(firstDelBlocks, 0, sizeof(BlockNumber) * n * (MAX_FORKNUM + 1));
>
> We don't need to prepare nforks, forks and firstDelBlocks for all
> relations before looping over relations. In other words, we can fill
> in the arrays for a relation at every iteration of relations.
Or even we could call FindAndDropRelFileNodeBuffers() for each
forks. It dones't matter in the performance perspective whether the
function loops over forks or the function is called for each forks.
regrds.
--
Kyotaro Horiguchi
NTT Open Source Software Center
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Michael Paquier | 2020-11-26 07:42:37 | Re: Online verification of checksums |
| Previous Message | Michael Paquier | 2020-11-26 07:29:47 | Re: ResourceOwner refactoring |