From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Jack Kaufman <jack(dot)kaufman(at)sanmina(dot)com> |
Cc: | pgsql-novice(at)postgresql(dot)org |
Subject: | Re: Determine the name of the calling function |
Date: | 2013-01-18 22:38:19 |
Message-ID: | 6727.1358548699@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-novice |
Jack Kaufman <jack(dot)kaufman(at)sanmina(dot)com> writes:
> The output from PG_EXCEPTION_CONTEXT looks parseable, with line 1
> containing the current function, and line 2, the calling function. The
> lines are variable length but the LF or CR should be identifiable. Add in
> the fact of the variable number of lines you got from
> PG_EXCEPTION_CONTEXT--where line 1 become 2, and line 2 becomes 3--and the
> project is a pain but doable.
Cool. FYI, I intend to see about fixing the bug --- IMO there should
not be a dependency on the number of executions here, and probably the
form with the extra context line is preferable.
Another thing that you should be aware of before plunging into this is
that GET STACKED DIAGNOSTICS is new as of 9.2; if you're hoping that
this code might be back-portable onto older PG versions, you'd better
not depend on it.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Keith Ouellette | 2013-01-21 13:54:45 | WAL replication question |
Previous Message | Casey Allen Shobe | 2013-01-18 22:34:56 | Re: Determine the name of the calling function |