On Fri, Aug 16, 2013 at 04:01:50AM +0000, Wang, Yu Y wrote: > > On Fri, Aug 16, 2013 at 02:22:43AM +0000, Wang, Yu Y wrote: > > > > On Thu, Aug 15, 2013 at 06:43:59PM -0700, Sarah Sharp wrote: > > > > > The xHCI platform driver calls into usb_add_hcd to register the > > > > > irq for its platform device. It does not want the xHCI generic > > > > > driver to register an interrupt for it at all. The original code > > > > > did that by setting the XHCI_BROKEN_MSI quirk, which tells the > > > > > xHCI driver to not enable MSI or MSI-X for a PCI host. > > > > > > > > > > Unfortunately, if CONFIG_PCI is enabled, and CONFIG_USB_DW3 is > > > > > enabled, the xHCI generic driver will attempt to register a legacy > > > > > PCI interrupt for the xHCI platform device in > > > > > xhci_try_enable_msi(). This will result in a bogus irq being > > > > > registered, since the underlying device is a platform_device, not > > > > > a pci_device, and thus the pci_device->irq pointer will be bogus. > > > > > > > > What shipping hardware has this problem today? > > > > > > > > In other words, doesn't seem like late -rc material to me. > > > > > > > > thanks, > > > > > > > > greg k-h > > > > > > [Yu:] I met this issue on Intel mobile SOC(Merrifield). But to my > > > understanding, this issue should not depend on hardware. This is one > > software issue. > > > > > > When system enabled CONFIG_PCI and xHCI driver registered as platform > > > device driver. Then xHCI-plat driver will be met initialization failed. > > > > Yes, but what "normal" system ever registers the xHCI driver as a platform > > driver? Does this happen today with systems? Which ones? > > [Yu:] Yes. I think there is no scenario for xHCI driver register as a platform driver. > Then can't reproduce on normal system. Thanks for confirming. Then you have no objection for this waiting until 3.12-rc1 to be merged? > > And is Merrifield shipping in devices today (sorry, I don't know the Intel mobile > > codenames anymore, it's been almost a year since I saw my last Intel > > powerpoint presentation... > > > > [Yu:] It should not only on Merrifield. Are any Merrifield systems "in the wild" yet? Can I buy one? > But should be also met for ARM SOC which use dwc3 controller as xHCI > controller. Because DWC3 driver register xHCI as platform driver by > default. So this issue can be reproduce. Which system can I reproduce this on? Do you know of a board with a dwc3 controller running in xHCI mode that I can get to test with (I would _love_ to get an ARM board running in xHCI mode for testing.) thanks, greg k-h -- 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
