Horton

Re: [HACKERS] ginInsertCleanup called from vacuum could still miss tuples to be deleted

Robert Haas <robertmhaas@gmail.com>

View thread View on postgr.es
From: Robert Haas <robertmhaas@gmail.com>
To: Masahiko Sawada <sawada.mshk@gmail.com>
Cc: PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2017-11-16T02:24:49Z
Lists: pgsql-hackers 
On Mon, Nov 13, 2017 at 3:25 AM, Masahiko Sawada <sawada.mshk@gmail.com> wrote:
> In ginInsertCleanup(), we lock the GIN meta page by LockPage and could
> wait for the concurrent cleaning up process if stats == NULL. And the
> source code comment says that this happen is when ginINsertCleanup is
> called by [auto]vacuum/analyze or gin_clean_pending_list(). I agree
> with this behavior. However, looking at the callers the stats is NULL
> only either if pending list exceeds to threshold during insertions or
> if only analyzing is performed by an autovacum worker or ANALYZE
> command. So I think we should inVacuum = (stats != NULL) instead.

Argh.  Yeah, that looks wrong.

Instead of relying on this indirect method, how about passing an
explicit inVacuum argument to that function?  And while we're at it,
how about renaming inVacuum to forceCleanup?

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

Commits

  1. Fix broken cleanup interlock for GIN pending list.

  2. Prevent multiple cleanup process for pending list in GIN.

  3. Add pages deleted from pending list to FSM

  4. Fix race condition in GIN posting tree page deletion.