Horton

Re: Use pg_pwritev_with_retry() instead of write() in dir_open_for_write() to avoid partial writes?

Alvaro Herrera <alvherre@alvh.no-ip.org>

View thread View on postgr.es
From: Alvaro Herrera <alvherre@alvh.no-ip.org>
To: Michael Paquier <michael@paquier.xyz>
Cc: Nathan Bossart <nathandbossart@gmail.com>, Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>, Thomas Munro <thomas.munro@gmail.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2022-10-30T14:44:32Z
Lists: pgsql-hackers

Attachments

On 2022-Oct-27, Michael Paquier wrote:

> On Thu, Sep 29, 2022 at 08:09:56PM -0700, Nathan Bossart wrote:
> > Looks reasonable to me.
> 
> 0001, to move pg_pwritev_with_retry() to a new home, seems fine, so
> applied.

Maybe something a bit useless, but while perusing the commits I noticed
a forward struct declaration was moved from one file to another; this is
claimed to avoid including port/pg_iovec.h in common/file_utils.h.  We
do that kind of thing in a few places, but in this particular case it
seems a bit of a pointless exercise, since pg_iovec.h doesn't include
anything else and it's a quite simple header.

So I'm kinda proposing that we only do the forward struct initialization
dance when it really saves on things -- in particular, when it helps
avoid or reduce massive indirect header inclusion.

-- 
Álvaro Herrera               48°01'N 7°57'E  —  https://www.EnterpriseDB.com/

Commits

  1. Revise pg_pwrite_zeros()

  2. Use pg_pwrite_zeros() in walmethods.c

  3. Introduce pg_pwrite_zeros() in fileutils.c

  4. Move pg_pwritev_with_retry() to src/common/file_utils.c

  5. Restore pg_pread and friends.

  6. Remove dead pread and pwrite replacement code.