RE: [Xen-devel] Video Presentation on PCI Express x16 VGA Pass Through to Xen-based Windows XP Home Edition HVM Virtual Machine

["Han, Weidong" <weidong.han@xxxxxxxxx>]

I found I still cannot watch your video, even cannot access youtube.

The output of 'xm dmesg' means it really 1:1 maps for vBAR and pBAR. It should 
be no problem. Can you post qemu log (/var/log/xen/qemu-dm-xxx.log)?

You below steps look correct. But I didn't encounter the error with build.c. 
Tim, did you encounter the error?

Regards,
Weidong

Mr. Teo En Ming (Zhang Enming) wrote:
> Hi All,
> 
> I went through the history and here are the steps I took to patch xen
> 3.5-unstable for VGA passthrough.
> 
> Based on Weidong's advice I checked out changeset earlier than 20144.
> 
> Initial trial without secondary gfx passthrough patch
> =================================
> 
>    637  hg clone http://xenbits.xensource.com/xen-unstable.hg
>    638  cd xen-unstable.hg
>    639  vi tools/firmware/hvmloader/hvmloader.c
>    640  hg checkout 20143
>    641  cp
> ~enming/intel-gfx-passthru-patches/intel-gfx-passthru-patch01.patch .
>    642  patch -p1 < intel-gfx-passthru-patch01.patch
>    643  make tools
>    644  history
>    645  cd tools/ioemu-remote/
>    646  cp
> ~enming/intel-gfx-passthru-patches/intel-gfx-passthru-patch02.patch .
>    647  patch -p1 < intel-gfx-passthru-patch02.patch
>    648  cd ..
>    649  cd ..
>    650  cp
>    ~enming/intel-gfx-passthru-patches/xen-load-vbios-file.patch . 651
>    patch -p1 < xen-load-vbios-file.patch 652  cp
>    ~enming/intel-gfx-passthru-patches/xen-vBAR-pBAR.patch . 653 
>    patch -p1 < xen-vBAR-pBAR.patch 654  cd tools/ioemu-remote/
>    655  cp
> ~enming/intel-gfx-passthru-patches/qemu-change-for-vBAR-pBAR.patch .
>    656  patch -p1 < qemu-change-for-vBAR-pBAR.patch
>    657  cd ../../
>    658  make clean
>    659  cp ~enming/vgabios-pt.bin tools/firmware/vgabios/
>    660  ls tools/firmware/vgabios/vgabios-pt.bin
>    661  make xen
>    662  make install-xen
>    663  make tools
>    664  make install-tools
> 
> In the above case, intel-gfx-passthru-patch01.patch is actually
> xen-gfx-passthrough.patch as posted by Weidong.
> 
> intel-gfx-passthru-patch02.patch is actually
> qemu-gfx-passthrough.patch as posted by Weidong.
> 
> After building and installing everything, I rebooted. When I tried to
> passthrough the nVidia Geforce 8400 GS, I got "not enough free
> resources" in Windows XP domU. This is in fact much better than my
> earlier trial when I got "device cannot start" in Windows XP domU
> device manager with the wrong approach of using onboard gma4500
> graphics as primary and nvidia geforce 8400 gs as secondary. This
> time I am using nvidia geforce 8400 gs as the primary and nothing
> else as the secondary. 
> 
> After Timothy Moore's suggestion, I proceeded to the next trial.
> 
> Next trial WITH secondary gfx passthrough patch
> ==============================
> 
>    679  cd /usr/src
>    680  mv xen-unstable.hg xen-unstable.hg-20143-no-secondary-patch
>    681  cp -R xen-unstable.hg-20143-no-secondary-patch xen-unstable.hg
>    682  cd xen-unstable.hg
>    683  history
>    684  cd tools/ioemu-remote/
>    685  cp
> ~enming/intel-gfx-passthru-patches/qemu-claim-vga-cycle-for-secondary-gfx-passthrough.patch
> .
>    686  patch -p1 <
> qemu-claim-vga-cycle-for-secondary-gfx-passthrough.patch
>    687  cd ../..
>    688  make clean
>    689  cp ~enming/vgabios-pt.bin tools/firmware/vgabios/
>    690  ls tools/firmware/vgabios/vgabios-pt.bin
>    691  make xen
>    692  make install-xen
>    693  make tools
>    694  make install-tools
> 
> At this point, I encountered an error with build.c as documented with
> my earlier postings in xen-devel mailing list.
> 
> So I applied my very own patch:
> 
> Patch created by Teo En Ming (Zhang Enming) on 29 August 2009 Saturday
> at 8:00 P.M. Singapore Time
> Email #1: enming.teo@xxxxxxxxxxxxxxx
> Email #2: space.time.universe@xxxxxxxxx
> MSN: teoenming@xxxxxxxxxxx
> Mobile Phone: +65-9648-9798
> 
> --- ssdt_pm.h 2009-08-29 19:54:52.653088000 +0800
> +++ ssdt_pm.h 2009-08-29 19:56:51.813088550 +0800
> @@ -10,7 +10,7 @@
> * C source code output
> *
> */
> -unsigned char AmlCode[] =
> +unsigned char AmlCode_PM[] =
> {
> 0x53,0x53,0x44,0x54,0xD6,0x05,0x00,0x00, /* 00000000 "SSDT...." */
> 0x02,0xB9,0x58,0x65,0x6E,0x00,0x00,0x00, /* 00000008 "..Xen..." */
> --- ssdt_tpm.h 2009-08-29 19:55:44.578738954 +0800
> +++ ssdt_tpm.h 2009-08-29 19:57:27.896638884 +0800
> @@ -10,7 +10,7 @@
> * C source code output
> *
> */
> -unsigned char AmlCode[] =
> +unsigned char AmlCode_TPM[] =
> {
> 0x53,0x53,0x44,0x54,0x4C,0x00,0x00,0x00, /* 00000000 "SSDTL..." */
> 0x02,0x2A,0x58,0x65,0x6E,0x00,0x00,0x00, /* 00000008 ".*Xen..." */
> 
> Then I checked ssdt_pm.h and ssdt_tpm.h to make sure they are correct.
> 
>    695  vi tools/firmware/hvmloader/acpi/ssdt_pm.h
>    696  vi tools/firmware/hvmloader/acpi/ssdt_tpm.h
>    697  make install-tools
> 
> 
> But I still get "not enough free resources" in Win XP domU device
> manager. So apparently
> the patch qemu-claim-vga-cycle-for-secondary-gfx-passthrough.patch
> doesn't affect my nvidia geforce 8400 gs passed through operations
> since I am using nvidia as primary and nothing as secondary. The
> secondary gfx patch is for cases when you have a primary for dom0 and
> passing through the secondary gfx to a HVM domU.


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel