On Wednesday, July 25, 2012 at 9:50:34 PM UTC-4, Steven Collins wrote: > Problem solved. I've learned more about nVidia's Optimus technology in > the last day than I ever wanted to know. The solution, somewhat > ironically, is to ENABLE Optimus in the BIOS, even though the help > text there says only to do so when running Windows. When it's enabled, > the OS sees the Intel integrated graphics as the primary graphics > card, which is fine by me and probably preferable for Qubes. The ACPI > thing was a red herring (a separate and known problem, with a separate > and known solution). > > The install went fine and I am indeed writing this very email from my > newly installed Qubes RC1 system. Thanks for all your hard work! > > In summary, for the HCL: > > Dell Latitude E6520 with i7-2760QM CPU. VT-d works fine (has to be > enabled in BIOS). If you have nVidia graphics you may need to enable > Optimus in BIOS also, even though help text says not to for > non-Windows OSes. (Note that this essentially DISABLES the nVidia GPU > for dom0.) > > On Wed, Jul 25, 2012 at 11:43 AM, Steven Collins <stevencoll...@acm.org> > wrote: > > I can't get very far with the RC1 installer because every time, X > > fails to start and anaconda aborts. My hardware is a Dell Latitude > > E6520 with an i7-2760QM CPU and nVidia NVS 4200M discrete graphics. > > Optimus is disabled in the BIOS (so the OS sees only the nVidia GPU > > and not the integrated graphics), and this problem didn't occur with > > the Beta 3 installer (or Beta 3 installed, or a previous installation > > of bare-metal Debian). > > > > I'm not in front of the machine right now but the errors in the X log > > are always of the form > > > > (WW) System lacks support for changing MTRRs > > (EE) VESA(0): V_BIOS address [something] out of range > > > > (The GPU is too new for nouveau to handle it, so it ends up using VESA > > regardless of the install option chosen.) > > > > The [something] seems to be a different non-zero hex number every time. > > > > I tried pretty much every permutation of starting the installer: the > > regular install and the "basic video driver" install, with and without > > nopat, with and without various mtrr-related hypervisor and kernel > > command-line parameters. One thing I haven't had a chance to try yet: > > booting Xen with ACPI off seemed promising (without it, the installer > > always hangs after it fails, I hit Ctrl-Alt-Delete, and it says > > "Restarting system"; with "acpi=0", it actually restarted), but I was > > trying the install from USB, and passing acpi=off to the kernel seemed > > to render it unable to find its own USB drive after initial boot. So > > next step is to try installing from DVD with acpi=off, unless someone > > else has a better idea. > > > > Speaking of DVD, I tried to re-download the ISO several times > > yesterday evening (central USA time) but the signature would no longer > > verify. > > > > I realize that nVidia is known to be problematic (and the Optimus > > configuration, even if "disabled", is probably not helping things), > > but it's what I have to work with at the moment, and I thought you'd > > appreciate the report. > > > > Regards, > > Steven Collins
This didn't fix it for me. I had Optimus on the whole time, still got a black screen after a couple of seconds of text from Xen. -- You received this message because you are subscribed to the Google Groups "qubes-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to qubes-devel+unsubscr...@googlegroups.com. To post to this group, send email to qubes-devel@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/qubes-devel/a49aeeb1-1c8a-44f2-9c9a-b07a3fa66662%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.