pg_xlogdump - Mailing list pgsql-hackers

From Alvaro Herrera
Subject pg_xlogdump
Date
Msg-id 20130213150937.GE4546@alvh.no-ip.org
Whole thread Raw
Responses Re: pg_xlogdump  (Andres Freund <andres@2ndquadrant.com>)
relpath() to src/common  (Alvaro Herrera <alvherre@2ndquadrant.com>)
List pgsql-hackers
Here's an updated version of pg_xlogdump.  This is rebased on top of the
committed xlogreader, palloc restructuring and libpgcommon, PG_RMGR
stuff, and is basically a revamped version of what Andres submitted in
http://www.postgresql.org/message-id/1357672187-7693-5-git-send-email-andres@2ndquadrant.com

I also attach a patch to move the relpathbackend() function to
src/common.  On top of that, it's trivial to change pg_xlogdump and
remove the "uninplemented" stub we're currently using.  (I also tried it
with a real implementation of relpath() that was mostly a duplicate of
the backend's relpath(), but I didn't like the duplication at all even
though I stripped the unnecessary bits).

(The more adventorous might think about moving timestamp_to_str to
src/common, as well, but this isn't a simple task: it depends on some
backend global state variables such as GUC vars, so it requires detailed
surgery.  I think it's a worthy goal nonetheless, because it'd allow us
to reduce useless duplication in ECPG, but it's not a 9.3 project)

--
Álvaro Herrera                http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services

Attachment

pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: 9.2.3 crashes during archive recovery
Next
From: Pavel Stehule
Date:
Subject: Re: Re: proposal: a width specification for s specifier (format function), fix behave when positional and ordered placeholders are used