Horton

Re: inefficient loop in StandbyReleaseLockList()

Nathan Bossart <bossartn@amazon.com>

View thread View on postgr.es
From: "Bossart, Nathan" <bossartn@amazon.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Kyotaro Horiguchi <horikyota.ntt@gmail.com>, "michael@paquier.xyz" <michael@paquier.xyz>, "andres@anarazel.de" <andres@anarazel.de>, "sulamul@gmail.com" <sulamul@gmail.com>, "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Date: 2021-11-01T17:02:49Z
Lists: pgsql-hackers 
On 11/1/21, 9:58 AM, "Tom Lane" <tgl@sss.pgh.pa.us> wrote:
> "Bossart, Nathan" <bossartn@amazon.com> writes:
>> Should there be a list_free(trgmNFA->queue) at the end of
>> transformGraph()?
>
> There could be, but that's visibly invoked only once per
> createTrgmNFAInternal call, so I didn't think it was worthwhile
> to do so (unlike the case for processState).  If we were concerned
> about leakage in that function, the hash table would be a far
> bigger issue.

Ah, I see it now.  The patch looks good to me, then.

Nathan

Commits

  1. Doc: add some notes about performance of the List functions.

  2. Avoid O(N^2) behavior in SyncPostCheckpoint().

  3. Avoid some other O(N^2) hazards in list manipulation.

  4. Avoid O(N^2) behavior when the standby process releases many locks.