A few thoughts on LWLock data structures...
In lwlock.c we hold a list of lwlocks held:
#define MAX_SIMUL_LWLOCKS 100
The code for LWLockRelease assumes that the last acquired lock will
always be the first one to be released, and uses an O(N) loop to search
for the lock to release.
Setting MAX_SIMUL_LWLOCKS to this fairly high number doesn't seem to
match the optimistic use of the O(N) algorithm.
Any thoughts on reducing the size of that array and/or reducing the lock
Best Regards, Simon Riggs
pgsql-hackers by date
|Next:||From: Tom Lane||Date: 2005-02-03 06:32:03|
|Subject: Re: Crash when inserting gist records, or creating index on ( int, geom ) |
|Previous:||From: Tom Lane||Date: 2005-02-03 05:17:19|
|Subject: Re: subselects in the target list |