Horton

Re: Spurious "apparent wraparound" via SimpleLruTruncate() rounding

x4mmm@yandex-team.ru

View thread View on postgr.es
From: Andrey Borodin <x4mmm@yandex-team.ru>
To: Noah Misch <noah@leadboat.com>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Tomas Vondra <tomas.vondra@2ndquadrant.com>, pgsql-hackers@postgresql.org
Date: 2021-01-01T18:05:29Z
Lists: pgsql-hackers 
Hi Noah!

I've found this thread in CF looking for something to review.

> 9 нояб. 2020 г., в 09:53, Noah Misch <noah@leadboat.com> написал(а):
> 
> Rebased both patches, necessitated by commit c732c3f (a repair of commit
> dee663f).  As I mentioned on another branch of the thread, I'd be content if
> someone reviews the slru-truncate-modulo patch and disclaims knowledge of the
> slru-truncate-insurance patch; I would then abandon the latter patch.
> <slru-truncate-modulo-v5.patch><slru-truncate-t-insurance-v4.patch>

Commit c732c3f adds some SYNC_FORGET_REQUESTs.
slru-truncate-modulo-v5.patch fixes off-by-one error in functions like *PagePrecedes(int page1, int page2).
slru-truncate-t-insurance-v4.patch ensures that off-by-one errors do not inflict data loss.

While I agree that fixing error is better than hiding it, I could not figure out how c732c3f is connected to these patches.
Can you please give me few pointers how to understand this connection?

Best regards, Andrey Borodin.

Commits

  1. Prevent excess SimpleLruTruncate() deletion.

  2. Fix unlinking of SLRU segments.

  3. Defer flushing of SLRU files.

  4. Change XID and mxact limits to warn at 40M and stop at 3M.