Horton

Re: slru.c race condition (was Re: TRAP: FailedAssertion("!((itemid)->lp_flags

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

View thread
From: Tom Lane <tgl@sss.pgh.pa.us>
To: "Jim C. Nasby" <jnasby@pervasive.com>
Cc: Alvaro Herrera <alvherre@commandprompt.com>, Greg Stark <gsstark@mit.edu>, Bruce Momjian <pgman@candle.pha.pa.us>, Gavin Sherry <swm@linuxworld.com.au>, pgsql-hackers@postgresql.org
Date: 2005-11-03T23:02:27Z
Lists: pgsql-hackers 
"Jim C. Nasby" <jnasby@pervasive.com> writes:
> Seriously, I am wondering about the performance hit of always checking
> debug_assertions.
> http://archives.postgresql.org/pgsql-hackers/2005-08/msg00389.php
> indicates that even with debug_assertions=false, --enable-cassert has a
> big performance impact.

That's because MEMORY_CONTEXT_CHECKING happens anyway --- it's not
currently switched off by the GUC variable.  I don't think we have any
recent data point about the cost of Asserts per se, except your own
report that it seems minimal.  My thought is that it would be even
more minimal if the tests of debug_assertion were removed ;-) ...
except in association with Asserts that are actually testing an
expensive-to-check condition, of which there are very few.

			regards, tom lane