Michael Ellerman wrote:
> On Fri, 2009-05-15 at 13:59 +0900, Hidetoshi Seto wrote:
>> @@ -445,8 +445,10 @@ static int msix_capability_init(struct pci_dev *dev,
>> /* MSI-X Table Initialization */
>> for (i = 0; i < nvec; i++) {
>> entry = alloc_msi_entry(dev);
>> - if (!entry)
>> - break;
>> + if (!entry) {
>> + msi_free_irqs(dev);
>> + return i;
>> + }
>>
>> j = entries[i].entry;
>> entry->msi_attrib.is_msix = 1;
>
> That looks right to me. Have you tested it? By faking allocation failure
> in alloc_msi_entry().
Oops, my test found a bug... we should return -ENOMEM if it fails with i==0.
I'll post v2 with my Tested-by soon. Sorry, please review it again.
> I guess this would be a nice fix for 30, but no one's hit it yet so it's
> not urgent and could wait for 31.
I agree with it.
Thanks,
H.Seto
--
To unsubscribe from this list: send the line "unsubscribe linux-pci" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
