Author: Dieter Buerssner
Date: 03:19:05 01/07/01
Go up one level in this thread
On January 07, 2001 at 05:31:23, Gian-Carlo Pascutto wrote: >Also, have you ever figured out why Crafty's preformance drops >with higher optimization levels in GCC? If you could find this >out it could be very interesting for the GCC folks... I cannot answer for Crafty, but the same happens with Yace. Also, almost any program I have written, gets slower on x86 when using GCC with high optimzation levels. There are various reasons, that can be seen, when analysing the asm output. I.e. there is the inlining. With -O3 gcc inlines much too aggressively. It is much better to use inline manual for critical functions. Also too many indexed array accesses are atuomatically converted to pointer like code. This often needs more registers. Also it ignores the fact, that x86 architecture has quite some support for indexed array success. I found plain -O or -O2 was working best usually (plus some other flags, like -fomit-frame-pointer) On other architectures, usually higher optimization levels produces faster code. I.e. On Alpha and Sun. But also here, the performance of gcc produced programs cannot compete with the vendor compilers. -- Dieter
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.