Horton

Re: SegFault on 9.6.14

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

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Amit Kapila <amit.kapila16@gmail.com>
Cc: Robert Haas <robertmhaas@gmail.com>, Alvaro Herrera <alvherre@2ndquadrant.com>, Thomas Munro <thomas.munro@gmail.com>, vignesh C <vignesh21@gmail.com>, Jerry Sievers <gsievers19@comcast.net>, Tomas Vondra <tomas.vondra@2ndquadrant.com>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2019-08-14T04:52:18Z
Lists: pgsql-hackers 
Amit Kapila <amit.kapila16@gmail.com> writes:
> Another point which I am wondering is why can't we use the existing
> REWIND flag to solve the current issue, basically if we have access to
> that information in nodeLimit.c (ExecLimit), then can't we just pass
> down that to ExecShutdownNode?

The existing REWIND flag tells subnodes whether they should *optimize*
for getting rewound or not.  I don't recall right now (well past
midnight) why that seemed like a useful definition, but if you grep for
places that are paying attention to that flag, I'm sure you'll find out.

We probably don't want to give up that distinction --- if it had been
equally good to define the flag as a hard yes-or-no, I'm sure we would
have taken that definition, because it's simpler.

			regards, tom lane

Commits

  1. Don't shut down Gather[Merge] early under Limit.

  2. Prohibit shutting down resources if there is a possibility of back up.

  3. Fix buffer usage stats for parallel nodes.

  4. Fix crashes on plans with multiple Gather (Merge) nodes.

  5. Separate reinitialization of shared parallel-scan state from ExecReScan.