Simon Horman
2009-May-01 12:03 UTC
[Xen-devel] Regression in PCI pass-through from qemu-xen changeset "Fix hvm guest cirrus_vga hvm s3 resume failure"
Apologies if this has been reported elsewhere. Changeset "Fix hvm guest cirrus_vga hvm s3 resume failure" (08251d35eafe5bafff42d226ccd56277014b16e9) in qemu-xen-unstable.git appears to cause a regression. This changeset was applied between the xen-3.4.0-rc2 and xen-3.4.0-rc3 tags of qemu-xen-unstable.git. The problem that I am seeing is that with this changeset applied I am not able to correctly initialise Intel network cards. Though I assume the problem is not specific to this hardware. In the case of an Intel e100 82557/8/9/0/1 the problem manifests as the EEPROM being declared invalid and the function is completely disabled by the guest operating system. In the case of an Intel igb 82576 the card is initialised by the guest operating system but is unable to establish a link and ethtool returning invalid values for the supported ports - it tells me the card has fibre when in fact it has copper. domU is running x86_32 2.6.30-rc4. dom0 is x86_64. It is not immediately apparent to me what a fix for this problem would look like. -- Simon Horman VA Linux Systems Japan K.K. Satellite Lab in Sydney, Australia H: www.vergenet.net/~horms/ W: www.valinux.co.jp/en _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Ian Jackson
2009-May-01 14:57 UTC
Re: [Xen-devel] Regression in PCI pass-through from qemu-xen changeset "Fix hvm guest cirrus_vga hvm s3 resume failure"
Simon Horman writes ("[Xen-devel] Regression in PCI pass-through from
qemu-xen changeset "Fix hvm guest cirrus_vga hvm s3 resume
failure""):> Apologies if this has been reported elsewhere.
Not that I''ve seen.
> Changeset "Fix hvm guest cirrus_vga hvm s3 resume failure"
> (08251d35eafe5bafff42d226ccd56277014b16e9) in qemu-xen-unstable.git
> appears to cause a regression. This changeset was applied
> between the xen-3.4.0-rc2 and xen-3.4.0-rc3 tags of qemu-xen-unstable.git.
You''ve definitely isolated it to this changeset ?
The symptoms you describe are very odd and don''t seem to correspond
meaningfully to anything in the changeset. The changeset does two
things:
* Removes a memset to 0xff of the VGA screen memory.
* Moves forward vga_bios_init which just initialises a bunch
of the cirrus_vga registers, to put it before graphic_console_init.
Ian.
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel
Simon Horman
2009-May-02 01:31 UTC
Re: [Xen-devel] Regression in PCI pass-through from qemu-xen changeset "Fix hvm guest cirrus_vga hvm s3 resume failure"
On Fri, May 01, 2009 at 03:57:49PM +0100, Ian Jackson wrote:> Simon Horman writes ("[Xen-devel] Regression in PCI pass-through from qemu-xen changeset "Fix hvm guest cirrus_vga hvm s3 resume failure""): > > Apologies if this has been reported elsewhere. > > Not that I''ve seen. > > > Changeset "Fix hvm guest cirrus_vga hvm s3 resume failure" > > (08251d35eafe5bafff42d226ccd56277014b16e9) in qemu-xen-unstable.git > > appears to cause a regression. This changeset was applied > > between the xen-3.4.0-rc2 and xen-3.4.0-rc3 tags of qemu-xen-unstable.git. > > You''ve definitely isolated it to this changeset ? > > The symptoms you describe are very odd and don''t seem to correspond > meaningfully to anything in the changeset. The changeset does two > things: > > * Removes a memset to 0xff of the VGA screen memory. > * Moves forward vga_bios_init which just initialises a bunch > of the cirrus_vga registers, to put it before graphic_console_init.Hi Ian, you are right, that does seem somewhat odd. I will re-run my tests. -- Simon Horman VA Linux Systems Japan K.K. Satellite Lab in Sydney, Australia H: www.vergenet.net/~horms/ W: www.valinux.co.jp/en _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Simon Horman
2009-May-02 01:55 UTC
Re: [Xen-devel] Regression in PCI pass-through from qemu-xen changeset "Fix hvm guest cirrus_vga hvm s3 resume failure"
On Sat, May 02, 2009 at 11:31:45AM +1000, Simon Horman wrote:> On Fri, May 01, 2009 at 03:57:49PM +0100, Ian Jackson wrote: > > Simon Horman writes ("[Xen-devel] Regression in PCI pass-through from qemu-xen changeset "Fix hvm guest cirrus_vga hvm s3 resume failure""): > > > Apologies if this has been reported elsewhere. > > > > Not that I''ve seen. > > > > > Changeset "Fix hvm guest cirrus_vga hvm s3 resume failure" > > > (08251d35eafe5bafff42d226ccd56277014b16e9) in qemu-xen-unstable.git > > > appears to cause a regression. This changeset was applied > > > between the xen-3.4.0-rc2 and xen-3.4.0-rc3 tags of qemu-xen-unstable.git. > > > > You''ve definitely isolated it to this changeset ? > > > > The symptoms you describe are very odd and don''t seem to correspond > > meaningfully to anything in the changeset. The changeset does two > > things: > > > > * Removes a memset to 0xff of the VGA screen memory. > > * Moves forward vga_bios_init which just initialises a bunch > > of the cirrus_vga registers, to put it before graphic_console_init. > > Hi Ian, > > you are right, that does seem somewhat odd. I will re-run my tests.Sorry, there was an out-by-one error in my bug report. This revision is the last good chengset. It is the following changeset "passthrough: Fix older kernel assigned device resume failure problem" (dbb8aafa702b8b4f5568e08641d98471fd04e0f8) where I see the problem. I''ll repost with the description of the problem and a fresh CC list. -- Simon Horman VA Linux Systems Japan K.K. Satellite Lab in Sydney, Australia H: www.vergenet.net/~horms/ W: www.valinux.co.jp/en _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel