On Wed, 2013-08-28 at 17:12 +0200, Gustav Sorenson wrote: > Hi Alex, > > thanks for your answer. > > Do you happen to know someone who might be able to judge whether or > not the radeon driver could be patched to support passthrough? Also, > could you please point me to the fixes to i915 you mentioned? I think > I might write to the radeon list to see whether they have ideas how to > tackle this issue, and I assume it may help them if they know what > fixed VGA arbitration on Haswell. > > Thanks again! VGA Arbiter: http://lists.freedesktop.org/archives/intel-gfx/2013-August/031963.html i915: http://lists.freedesktop.org/archives/intel-gfx/2013-August/032295.html The changes are very device dependent. In the case of newer Intel HD graphics the i915 driver does not need VGA memory access and only occasionally needs VGA I/O access. We can therefore disable VGA memory completely so that it doesn't need to participate in arbitration of that. Then the arbiter will disable the I/O bit in the PCI command register when another device requests VGA. Perhaps if you can send `dmesg | grep -i vga` from your system we can see how the arbiter is getting setup. On an Intel system, a simple test to disable VGA to IGD is to disable I/O and memory in the PCI command register: setpci -s 00:02.0 4.w=0 (of course you really want to read it first to restore it later, you only need to clear the bottom 2 bits) Of course even if you get VGA working, you're still likely to get a Code 43 trying to enable the Nvidia driver, but I appreciate any efforts to help figure that out. Thanks, Alex -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
