Re: [Xen-devel] [PATCH v2 2/3] VT-d: suppress UR signaling for desktop chipsets

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 16.04.14 at 15:48,  wrote:
> @@ -454,5 +456,33 @@ void __init pci_vtd_quirk(struct pci_dev
>          printk(XENLOG_INFO "Masked UR signaling on %04x:%02x:%02x.%u\n",
>                 seg, bus, dev, func);
>          break;
> +
> +    case 0x100: case 0x104: case 0x108: /* Sandybridge */
> +    case 0x150: case 0x154: case 0x158: /* Ivybridge */
> +    case 0xa04: /* Haswell ULT */
> +    case 0xc00: case 0xc04: case 0xc08: /* Haswell */

Sadly this set of PCI IDs seems to be incomplete: While trying to find
an as realistic as possible way to test the backport of this change to
4.2, I realized my laptop should come reasonably close (but it being
1st generation Core-i7 based it doesn't fall in any of the above groups).
Starting to suspect that it may simply be missing, I dumped the values
from the couple of items surrounding offset 0x1c8 which have a
meaning in at least one of the datasheets for the chipsets above, and
as kind of expected the values seen match up with the reset values
specified for the newer versions. Hence I'm rather confident that
ID 0x0044 is missing here; based on the patterns above one might
guess that 0x0040 (Core-i5-600 and -i3-500 Desktop, datasheet
matching the pattern) and 0x0048 (no documentation found) would
also need adding.

In any event this puts up the question again - how (and when) can
we be certain to have a complete list?

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel