Nick Jones writes: > > Python 2.2.2 (#1, Jan 22 2003, 20:44:19) > > >>> int(10.2) > > Traceback (most recent call last): > > File "<stdin>", line 1, in ? > > OverflowError: float too large to convert > > > I have: > > CFLAGS='-mcpu=pentium4 -O2 -pipe' > The pentium4 would be the problem. It's not friendly. > Drop down to pentium2 and recompile it. Pentium4 opts aren't > recommended for the time being. See the comments portage-2.0.46-r12's > make.conf.
Thanks Nick. I dropped down to i686 and recompiled glibc-2.3.1-r2 and that worked! Now I will try pentium2 as you suggest. For anyone else interested, the problem is coming from the glibc modf() function as pointed out in this forum discussion that I came across yesterday (complete with a tiny C program to demonstrate): http://forums.gentoo.org/viewtopic.php?t=27211&highlight=python+float Nick, maybe I've gone completely stupid, but I can't find any comments in /etc/make.conf that warn against using pentium4. I had thought my choice of CFLAGS safe based upon http://www.freehackers.org/gentoo/gccflags/flag_gcc3.html Oh, one more thing: Great job Nick and all Gentoo-ers! -- Jeff Kowing [EMAIL PROTECTED] -- [EMAIL PROTECTED] mailing list
