From: Dan K. <da...@ke...> - 2003-06-09 15:26:50
|
Yeah... worth a shot, though. Most of the gcc 3.3 problems I've seen so far are for other architectures (like sh4). - Dan Joshua Moore-Oliva wrote: > Actually I think it is fixed in gcc 3.3 however I'm nto running it yet cause > it's not marked stable yet :) > > Just thought I'd give youa heads up. > > On June 9, 2003 11:01 am, Dan Kegel wrote: > >>Which version of gcc? gcc-3.3 fixes a whole bunch of pentium 4 bugs. >>- Dan >> >>Joshua Moore-Oliva wrote: >> >>>Just a warning, but march=pentium 4 is broken in gcc. >>> >>>I've had everything from sound problems to floating point conversion >>>problems due to it... If you have anything compiled with march=pentium4 >>>I'd definately recommend downgrading to at least march=pentium3 since >>>pentium4 has a floating point conversion bug that's really nasty. >>> >>>On June 9, 2003 10:30 am, Olivier Chapuis wrote: >>> >>>>On Sun, Jun 08, 2003 at 09:56:49AM -0400, Joshua Moore-Oliva wrote: >>>> >>>>>What system are you running? >>>>> >>>>>The only time I got that error was when I was running gentoo and ran the >>>>>march=pentium[34] gcc flags.. >>>>> >>>>>Currently I have compiled my system with >>>>> >>>>>march=i686 =O2 pipe -fomit-frame-pointer >>>>> >>>>>and it works well. >>>>> >>>>>Josh. >>>> >>>>Yes I compiled X with these options (as I've a pentium4): >>>> >>>>-O2 -march=pentium4 -fno-strength-reduce -pipe >>>> >>>>will try with march=i686. >>>> >>>>Thanks, Olivier -- Dan Kegel http://www.kegel.com http://counter.li.org/cgi-bin/runscript/display-person.cgi?user=78045 |