| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | "Simon Riggs" <simon(at)2ndquadrant(dot)com> |
| Cc: | "pgsql-hackers" <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: LWLockRelease |
| Date: | 2005-02-03 14:32:52 |
| Message-ID: | 15729.1107441172@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
"Simon Riggs" <simon(at)2ndquadrant(dot)com> writes:
> A few thoughts on LWLock data structures...
> In lwlock.c we hold a list of lwlocks held:
> held_lwlocks[MAX_SIMUL_LWLOCKS]
> where
> #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.
So? The search only examines the actually-in-use array entries.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Bruce Momjian | 2005-02-03 14:42:18 | Re: libpq API incompatibility between 7.4 and 8.0 |
| Previous Message | Tom Lane | 2005-02-03 14:31:24 | Re: LWLock cache line alignment |