On Fri, 14 Mar 2003, Marc Aurele La France wrote: > On Fri, 14 Mar 2003, Mark Vojkovich wrote: > > > On 14 Mar 2003, Eric Anholt wrote: > > > > > On Fri, 2003-03-14 at 12:31, Mark Vojkovich wrote: > > > > Could someone put the FreeBSD int10 module from 4.3 someplace where I > > > > can get it? I'll give it a try. > > You already have it. (Or should.) The two int10 modules installed on > Linux systems are: > > /usr/X11R6/lib/modules/libint10.a > /usr/X11R6/lib/modules/linux/libint10.a > > The loader prefers the more specific, so to use the generic one, rename > the /linux/ one to something else. > > > > I put one from my system up at: > > > http://people.freebsd.org/~anholt/X/files/libint10.a > > > I just ran this module on my system and it worked fine. > > So at least on my system here: > > > Linux 2.2.17 smp > > Dual P3 system (i840 chipset) > > GeForce4 MX + GeForce2 MX using "nv" driver on both > > > both the module I built and the one at that url work fine. > > I think that rules out the FreeBSD-built module being completely > > broken. Still could be some other FreeBSD related problem or > > that the failing cases are on specific hardware that the int10 > > or related module doesn't handle as well anymore. > > > Has it been firmly established that this is a regression > > and 4.2 worked fine on those machines? > > It's not clear to me that you actually used the generic module for this > test. Check the log. >
You're right! It was still using the linux/libint10.a. The generic libint10.a doesn't work. If I switch over, the second head doesn't get posted correctly and I do see the message about the BIOS getting truncated. Mark. _______________________________________________ XFree86 mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xfree86