On Wed, 22 Dec 2010, Sarah Sharp wrote: > On Wed, Dec 22, 2010 at 09:31:49PM -0500, Alan Stern wrote: > > On Wed, 22 Dec 2010, Sarah Sharp wrote: > > > > > Alan, > > > > > > Please take a look at patch 1 if you have time. > > > > I still don't understand the reason for adding that test to > > suspend_common(). (And I would prefer it if the comment above the test > > were more grammatical.) > > If the xHCI driver has enabled MSI-X, then the legacy PCI irq is no > longer associated with our hardware. But the USB core still thinks it > is, and will sync pci_dev->irq. I've been told it would be harmless to > synchronize someone else's IRQ, but I'd rather not wait for them. It's > better to let the xHCI driver synchronize all the MSI-X irqs in its > suspend method. Then instead of guessing about whether or not pdev->irq should be the same as hcd->irq, and what it might mean if they are different, just add a new "don't synchronize irqs" flag to the usb_hcd structure. Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
