Re: SetBufferCommitInfoNeedsSave and race conditions - Mailing list pgsql-hackers

From Tom Lane
Subject Re: SetBufferCommitInfoNeedsSave and race conditions
Date
Msg-id 16992.1183054566@sss.pgh.pa.us
Whole thread Raw
In response to SetBufferCommitInfoNeedsSave and race conditions  ("Pavan Deolasee" <pavan.deolasee@gmail.com>)
List pgsql-hackers
"Pavan Deolasee" <pavan.deolasee@gmail.com> writes:
> So we suspected an interaction between multiple processes each holding
> a SHARE lock and setting/checking different bits in the infomask and
> we could theoritically say that such interaction can potentially lead to
> missing hint bit updates.

Yeah.  This is in fact something that's been foreseen, but I guess it
didn't occur to anyone that those Asserts would fail.  I concur with
removing them.
        regards, tom lane


pgsql-hackers by date:

Previous
From: "Pavan Deolasee"
Date:
Subject: SetBufferCommitInfoNeedsSave and race conditions
Next
From: Patrick Welche
Date:
Subject: write past chunk end in ExprContext / to_char