Luke Lonergan wrote:
> Can you try this with just "-O3" versus "-O2"?
Thanks to Daniel for doing these tests.
I happen to have done the same tests about 3/4 years ago,
and concluded that gcc flags did *not* influence performance.
Moved by curiosity, I revamped those tests now on a test
machine (single P4 @ 3.2 Ghz, with 2Mb cache and 512 Mb Ram).
Here are the results:
http://www.streppone.it/cosimo/work/pg/gcc.png
In short: tests executed with postgresql 8.2.0,
gcc version 3.4.3 20041212 (Red Hat 3.4.3-9.EL4),
tps figures computed as average of 9 pgbench runs (don't ask why 9... :-),
with exactly the same commands given by Daniel:
"-O0" ~ 957 tps
"-O1 -mcpu=pentium4 -mtune=pentium4" ~ 1186 tps
"-O2 -mcpu=pentium4 -mtune=pentium4" ~ 1229 tps
"-O3 -mcpu=pentium4 -mtune=pentium4" ~ 1257 tps
"-O6 -mcpu=pentium4 -mtune=pentium4" ~ 1254 tps
I'm curious now to get the same tests run with
a custom-cflags-compiled glibc.
--
Cosimo