[GSoC] Question about "Explicitly support predicate locks in index access methods besides btree"

From: Dong Yuan <doffery20(at)gmail(dot)com>
To: pgsql-hackers(at)postgresql(dot)org
Subject: [GSoC] Question about "Explicitly support predicate locks in index access methods besides btree"
Date: 2017-04-02 14:54:50
Message-ID: CAFbcfvRFYxjRRZGiERsoqG3+DrbsKyE65cie8wcDeURu7HikBA@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi guys,

I'm confused about the this function, CheckForSerializableConflictOut(...).
Can anyone help me out?

It seems like this function is used to check the rw-antidependencies out
edge. This should be done when reading a written tuple. But btree does not
call this function at all. The heapam used predicatelockpage(...)...

I use to think the in and out edge will be built when rw-antidependencies
detected. CheckForSerializableConflictIn(...) can do this job. So
CheckForSerializableConflictOut(...) function really confused me.

Best Wishes!
---
Dong

Browse pgsql-hackers by date

  From Date Subject
Next Message Fabien COELHO 2017-04-02 14:58:38 Re: Suggested fix for \p and \r in psql
Previous Message Fabien COELHO 2017-04-02 14:38:16 Re: Variable substitution in psql backtick expansion