Ani Joshi wrote: > > Michel Danzer wrote: > > > The problem is that the vgaHW layer now tries to do ISA I/O using a new > > syscall to get the iobase. Unfortunately, it doesn't seem to work on > > UniNorth > > machines. For now you'll have to either use Option "UseFBDev" or hack the > > syscall (something with sysconfig_iobase) out of your kernel or X server. > > I don't know who started relating this to a UniNorth-specific problem
Me, who else? :) I just tried to deduce the cause from the facts that you submitted the patch to make the X server use the new syscall (so it seemed to work for you) and it didn't work for me. > but that is wrong. The crash occurs because vgaHW tries to read the DAC > which in theory doesn't exist on "Mac" boards, or just doesn't respond. > This is not a UniNorth issue but an issue on *all* Macs. Now, please enlighten me: Which Mac cards need to have the VGA mode saved? I guess I'm only starting to see the whole picture. I guess The Right Thing would be to change vgaHW so that it decides _at runtime_ what needs to be done, regardless of the architecture - after all, people might try Mac cards in i386 boxen. I don't know if this is even possible on i386 though. > > BenH, do you have an idea why this doesn't work or how we could make that > > syscall fail for machines where it doesn't work? > > The syscall has *nothing* to do with this situation. I beg to differ as everything works fine without it, but let's not start a fight about this. -- Earthling Michel Dänzer (MrCooper) \ Debian GNU/Linux (powerpc) developer CS student, Free Software enthusiast \ XFree86 and DRI project member
