Hi, When xen-platform pci device is introduced, its memory and interrupt resource is not updated in ACPI dsdt. This patch fixes this. This solves several known issues with ACPI HAL Windows, including installation of ACPI windows 2000 and booting of ACPI Windows XP. Signed-off-by: Qing He <qing.he@intel.com> _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2006-Oct-12 12:53 UTC
Re: [Xen-devel] [PATCH][HVM] fix the ACPI dsdt inconsistency
On 12/10/06 10:21, "He, Qing" <qing.he@intel.com> wrote:> When xen-platform pci device is introduced, its memory and interrupt > resource is not updated in ACPI dsdt. This patch fixes this. > > This solves several known issues with ACPI HAL Windows, including > installation of ACPI windows 2000 and booting of ACPI Windows XP. > > Signed-off-by: Qing He <qing.he@intel.com>Looks like this just adds I/O memory and IRQ wire coverage ''big enough'' to happen to cover the requirements of the extra device? That sounds okay to me, just want to confirm that''s what''s going on. -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2006-Oct-12 13:03 UTC
Re: [Xen-devel] [PATCH][HVM] fix the ACPI dsdt inconsistency
On 12/10/06 13:53, "Keir Fraser" <Keir.Fraser@cl.cam.ac.uk> wrote:>> When xen-platform pci device is introduced, its memory and interrupt >> resource is not updated in ACPI dsdt. This patch fixes this. >> >> This solves several known issues with ACPI HAL Windows, including >> installation of ACPI windows 2000 and booting of ACPI Windows XP. >> >> Signed-off-by: Qing He <qing.he@intel.com> > > Looks like this just adds I/O memory and IRQ wire coverage ''big enough'' to > happen to cover the requirements of the extra device? That sounds okay to > me, just want to confirm that''s what''s going on.Also, should we add something to the PRTA section, to give a default binding for Slot5/INTA? -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser wrote:>> >> Looks like this just adds I/O memory and IRQ wire coverage ''big >> enough'' to happen to cover the requirements of the extra device? >> That sounds okay to me, just want to confirm that''s what''s going on.Yes, this is exactly what this patch does. This will provide enough MMIO and IRQ resource for xen_patform PCI device. Thus the resource conflict in windows will be resolved.> > Also, should we add something to the PRTA section, to give a default > binding for Slot5/INTA?PRTA happen to already have the extra entry: Package(){0x0004ffff, 0, 0, 11}, // Device 4 , INTA What PRTA missing is the entry for bus 0 device 0 (0x0000ffff). Fortunately, bus 0 device 0 is the host bridge, and no IRQ is needed for this device. So it doesn''t matter.> > -- Keir_______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel