On 2011-10-18 15:49, Michael S. Tsirkin wrote:
> On Tue, Oct 18, 2011 at 03:20:43PM +0200, Michael S. Tsirkin wrote:
>> Ah. Right. Good catch. It's probably best to just make that
>> state be !enabled || masked. Like the below? (compiled only):
>>
>> --->
>>
>> msix: track function masked in pci device state
>>
>> Only go over the table when function is masked.
>> This optimization is not really important for qemu.git but helps
>> qemu-kvm.git.
>>
>> Signed-off-by: Michael S. Tsirkin <mst@xxxxxxxxxx>
>
> ...
>
>> diff --git a/hw/pci.h b/hw/pci.h
>> index 86a81c8..da4c9c3 100644
>> --- a/hw/pci.h
>> +++ b/hw/pci.h
>> @@ -178,6 +178,8 @@ struct PCIDevice {
>> unsigned *msix_entry_used;
>> /* Region including the MSI-X table */
>> uint32_t msix_bar_size;
>> + /* MSIX function mask set or MSIX disabled */
>> + bool msix_function_masked;
>> /* Version id needed for VMState */
>> int32_t version_id;
>
> BTW, since code checks msix_function_masked consistently,
> this also covers the case your other patch addressed,
> where there's a table access while msix is disabled, right?
Looks like. Not sure about potential side effects of this change to
msi_is_masked, but conceptually there should be none.
Jan
--
Siemens AG, Corporate Technology, CT T DE IT 1
Corporate Competence Center Embedded Linux
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
