From mboxrd@z Thu Jan 1 00:00:00 1970 From: Subject: RE: Re: VGA Passthrough with Nvidia GeForce 7900GS and AsusP7P55D-E, some problems Date: Mon, 21 Jun 2010 15:32:52 -0400 Message-ID: References: <4C1B648B.3080604@gmail.com><4C1D0875.2030506@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Return-path: Content-class: urn:content-classes:message In-Reply-To: List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: Thomas Richard , xen-devel@lists.xensource.com List-Id: xen-devel@lists.xenproject.org I believe the value should have some bits at the least significant end = that describe what kind of value is in the BAR, for example, = enabled/disabled, memory/io region, etc... -----Original Message----- From: xen-devel-bounces@lists.xensource.com = [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Thomas = Richard Sent: Monday, June 21, 2010 12:42 PM To: xen-devel@lists.xensource.com Subject: [Xen-devel] Re: VGA Passthrough with Nvidia GeForce 7900GS and = AsusP7P55D-E, some problems I'm sorry for bumping this but is there anyone able to tell me where i should start looking if i want to get this fixed? I waded through some code but it's a little massive for me to grasp in a short time. Could anyone give me some pointers on where I best start to look? I know the 7900GS isn't officially supported but i want to help getting it there. I'm still not sure if these off by a few bytes are an issue: pt_register_regions: IO region registered (size=3D0x10000000 base_addr=3D0xe000000c) pt_register_regions: IO region registered (size=3D0x01000000 base_addr=3D0xf5000004) pt_register_regions: IO region registered (size=3D0x00000080 base_addr=3D0x0000ac01) pt_register_regions: Expansion ROM registered (size=3D0x00020000 base_addr=3D0xf7ae0002) I tried to follow the codepath and it seems that these values come straight out of the linux pci code. The strange thing is that lspci shows other (correct) values. Also the problem with the shifting screen worries me a little. I'm still not sure my kernel config is ok because of this. Thanks Thomas Richard 2010/6/19 Thomas Richard : > Hey, > > I tested the same procedure with another vBar=3DpBar mapping ofcourse = for an > ati X300 today. It gave me the same results. I have a feeling I'm = missing > something here. Another, noteworthy, strange thing is that when i = press the > enter button on the keyboard that is attached to my dom0 kernel (the = PS/2 > keyboard) the screen shifts a little to the right. When i press the = enter > button enough times, it just makes a full round. So it seems like my = dom0 is > still able to influence the graphics card that is passed to my HVM = guest? > The card is hidden from dom0 with pciback at boottime. > > Greetings > Thomas Richard > > Op 18/06/2010 14:20, Thomas Richard schreef: >> >> Hey all, >> >> I got pretty excited when reading about the VGA Passthrough features = in >> Xen 4.0 so i decided to give it a go. I've had partial success, like = some >> other in here. I've got output on my VGA port but my windows HVM = guest >> reports that the VGA doesn't work. Windows XP says: Couldn't start = device >> (10) and Windows 7 says: Windows has stopped the device because it = reported >> problems (40). >> >> To make this work I: >> -built the kernel from Jeremy's tree (2.6.32.15) >> -extracted my VGA bios with nvflash >> -checked out xen-4.0-testing >> -applied the loadvbios, vBar=3DpBar (xen), vBar=3DpBar (qemu) and >> claimvgacycle patches >> -copied the vgabios in my xen tree >> -built xen >> >> This is the output for lspci -v -s 01:00.0: >> 01:00.0 VGA compatible controller: nVidia Corporation G71 [GeForce = 7900 >> GS] (rev a1) >> =A0 =A0 =A0 =A0Subsystem: ASUSTeK Computer Inc. Device 8224 >> =A0 =A0 =A0 =A0Flags: bus master, fast devsel, latency 0, IRQ 10 >> =A0 =A0 =A0 =A0Memory at f6000000 (32-bit, non-prefetchable) = [size=3D16M] >> =A0 =A0 =A0 =A0Memory at e0000000 (64-bit, prefetchable) = [size=3D256M] >> =A0 =A0 =A0 =A0Memory at f5000000 (64-bit, non-prefetchable) = [size=3D16M] >> =A0 =A0 =A0 =A0I/O ports at dc00 [size=3D128] >> =A0 =A0 =A0 =A0Expansion ROM at f7ee0000 [disabled] [size=3D128K] >> =A0 =A0 =A0 =A0Capabilities: >> =A0 =A0 =A0 =A0Kernel driver in use: pci-stub >> =A0 =A0 =A0 =A0Kernel modules: nvidiafb >> >> I've attached my dmesg, xm dmesg and qemu log files and my hvm config >> file. >> >> I noticed a few things which might cause problems. In de xm dmesg = file i >> found this: >> (XEN) Intel VT-d Snoop Control supported. >> (XEN) Intel VT-d DMA Passthrough not supported. >> (XEN) Intel VT-d Queued Invalidation supported. >> (XEN) Intel VT-d Interrupt Remapping not supported. >> Is DMA Passthrough or Interrupt remapping a requirement to make VGA >> Passthrough work? >> I think my motherboard should support these features though, does = this >> have to do with my kernel build or something? >> >> In my qemu log file i saw this: >> pt_iomul_init: Error: pt_iomul_init can't open file = /dev/xen/pci_iomul: No >> such file or directory: 0x1:0x0.0x0 >> >> and >> >> pt_register_regions: IO region registered (size=3D0x01000000 >> base_addr=3D0xf6000000) >> pt_register_regions: IO region registered (size=3D0x10000000 >> base_addr=3D0xe000000c) >> pt_register_regions: IO region registered (size=3D0x01000000 >> base_addr=3D0xf5000004) >> pt_register_regions: IO region registered (size=3D0x00000080 >> base_addr=3D0x0000ac01) >> pt_register_regions: Expansion ROM registered (size=3D0x00020000 >> base_addr=3D0xf7ae0002) >> >> These bytes are off by a few. I don't know why and i don't know if it >> makes any difference. >> >> Can anyone give me any pointers on where i should look to find the >> problem? Also, could anybody mail me a working (with vga passthrough) >> .config file for the 2.6.32.15 kernel? >> >> Thanks for the help >> >> Greetings >> Thomas Richard > > _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel