| From: | Konstantin Knizhnik <k(dot)knizhnik(at)postgrespro(dot)ru> |
|---|---|
| To: | Zhihong Yu <zyu(at)yugabyte(dot)com>, PostgreSQL Developers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
| Subject: | Re: Double partition lock in bufmgr |
| Date: | 2020-12-19 12:50:30 |
| Message-ID: | 335c3a13-74d1-26c0-584c-73ceed3a9ffc@postgrespro.ru |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On 19.12.2020 10:53, Zhihong Yu wrote:
> Hi,
> w.r.t. the code in BufferAlloc(), the pointers are compared.
>
> Should we instead compare the tranche Id of the two LWLock ?
>
> Cheers
As far as LWlocks are stored in the array, comparing indexes in this
array (tranche Id) is equivalent to comparing element's pointers.
So I do not see any problem here.
Just as experiment I tried a version of BufferAlloc without double
locking (patch is attached).
I am not absolutely sure that my patch is correct: my main intention was
to estimate influence of this buffer reassignment on performance.
I just run standard pgbench for database with scale 100 and default
shared buffers size (256Mb). So there are should be a lot of page
replacements.
I do not see any noticeable difference:
vanilla: 13087.596845
patch: 13184.442130
| Attachment | Content-Type | Size |
|---|---|---|
| bufmgr.patch | text/x-patch | 1.7 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andrew Dunstan | 2020-12-19 16:19:07 | Re: multi-install PostgresNode |
| Previous Message | Alastair Turner | 2020-12-19 11:45:15 | Re: Proposed patch for key managment |