Author: Pallav Nawani
Date: 11:36:23 01/08/05
Go up one level in this thread
On January 08, 2005 at 03:47:33, Fabien Letouzey wrote: >On January 07, 2005 at 14:28:13, Pallav Nawani wrote: > >>Hi, > >>I recall seeing this same problem but cannot remember the exact cause. However, >>I do recall that reducing the optimization level seemed to help. >>If you are using -O3 try -O2 or just -O or compile without any optimization at >>all. > >>Best, >>Pallav > >Hm, never heard of this ... > >Please show me raw output from a faulty binary (no GUI). > >Thanks, > >Fabien. That was from a Fruit 15 binary that I tried to compile. Maybe I can reproduce it. How do I show a raw output? This is the makefile I used. Note that it may _not_ reproduce the problem because I remember that I had managed to fix it somehow by tinkering with compile flags, so this may be a fixed one. CC="/c/MinGW31/bin/g++.exe" fruit: $(CC) -O3 -Wall -o fruit -fomit-frame-pointer -fschedule-insns -fschedule-insns2 -fmove-all-movables -fno-rtti -funroll-loops -march=i686 *.cpp strip fruit.exe clean: rm fruit.exe Also, it is possible that I may have used these flags initially (As these are the flags I use to compile Natwarlal): -O3 -Wall -o natwarlal -fforce-mem -fgcse -mpreferred-stack-boundary=2 -fforce-addr -fomit-frame-pointer -finline-functions -foptimize-register-move -ffast-math -fstrength-reduce -frerun-loop-opt -frerun-cse-after-loop -fschedule-insns -fschedule-insns2 -fmove-all-movables -fno-rtti -fpeephole -falign-functions=4 -funroll-loops -march=i686 Best regards, Pallav
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.