Horton

Re: Refactoring the checkpointer's fsync request queue

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Thomas Munro <thomas.munro@gmail.com>
Cc: Shawn Debnath <sdn@amazon.com>, Andres Freund <andres@anarazel.de>, Dmitry Dolgov <9erthalion6@gmail.com>, Robert Haas <robertmhaas@gmail.com>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2019-03-05T16:53:16Z
Lists: pgsql-hackers 
Thomas Munro <thomas.munro@gmail.com> writes:
> +#include "fmgr.h"
> +#include "storage/block.h"
> +#include "storage/relfilenode.h"
> +#include "storage/smgr.h"
> +#include "storage/sync.h"

> Why do we need to include fmgr.h in md.h?

More generally, any massive increase in an include file's inclusions
is probably a sign that you need to refactor.  Cross-header inclusions
are best avoided altogether if you can --- obviously that's not always
possible, but we should minimize them.  We've had some very unfortunate
problems in the past from indiscriminate #includes in headers.

			regards, tom lane

Commits

  1. Fix bugs in mdsyncfiletag().

  2. Refactor the fsync queue for wider use.

  3. Don't forget about failed fsync() requests.

  4. PANIC on fsync() failure.

  5. Move LockClauseStrength, LockWaitPolicy into new file nodes/lockoptions.h.

  6. Add new file for checkpointer.c

  7. Split work of bgwriter between 2 processes: bgwriter and checkpointer.

  8. Install infrastructure for shared-memory free space map. Doesn't actually