On Tuesday 13 January 2009, Alan Stern wrote: > This patch (as1199) changes the initial wakeup settings for PCI USB > host controllers. The controllers are marked as capable of waking the > system, but wakeup is not enabled by default. > > It turns out that enabling wakeup for USB host controllers has a lot > of bad consequences. As the simplest example, if a USB mouse or > keyboard is unplugged immediately after the computer is put to sleep, > the unplug will cause the system to wake back up again! We are better > off marking them as wakeup-capable and leaving wakeup disabled. > > Signed-off-by: Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> > Reported-by: Rafael J. Wysocki <rjw@xxxxxxx> > CC: David Brownell <david-b@xxxxxxxxxxx> Thanks for the patch, I'll test it later today, although it looks obviously correct. Please have a look at the patch I sent here: http://marc.info/?l=linux-kernel&m=123184641714487&w=4 Thanks, Rafael > --- > > This patch should also be queued for 2.6.28.stable. > > > Index: usb-2.6/drivers/usb/core/hcd-pci.c > =================================================================== > --- usb-2.6.orig/drivers/usb/core/hcd-pci.c > +++ usb-2.6/drivers/usb/core/hcd-pci.c > @@ -128,7 +128,6 @@ int usb_hcd_pci_probe(struct pci_dev *de > } > > pci_set_master(dev); > - device_set_wakeup_enable(&dev->dev, 1); > > retval = usb_add_hcd(hcd, dev->irq, IRQF_DISABLED | IRQF_SHARED); > if (retval != 0) > Index: usb-2.6/drivers/usb/host/ehci-pci.c > =================================================================== > --- usb-2.6.orig/drivers/usb/host/ehci-pci.c > +++ usb-2.6/drivers/usb/host/ehci-pci.c > @@ -230,7 +230,7 @@ static int ehci_pci_setup(struct usb_hcd > pci_read_config_word(pdev, 0x62, &port_wake); > if (port_wake & 0x0001) { > dev_warn(&pdev->dev, "Enabling legacy PCI PM\n"); > - device_init_wakeup(&pdev->dev, 1); > + device_set_wakeup_capable(&pdev->dev, 1); > } > } > > Index: usb-2.6/drivers/usb/host/ohci-hcd.c > =================================================================== > --- usb-2.6.orig/drivers/usb/host/ohci-hcd.c > +++ usb-2.6/drivers/usb/host/ohci-hcd.c > @@ -593,12 +593,10 @@ static int ohci_run (struct ohci_hcd *oh > * to be checked in case boot firmware (BIOS/SMM/...) has set up > * wakeup in a way the bus isn't aware of (e.g., legacy PCI PM). > * If the bus glue detected wakeup capability then it should > - * already be enabled. Either way, if wakeup should be enabled > - * but isn't, we'll enable it now. > + * already be enabled; if so we'll just enable it again. > */ > - if ((ohci->hc_control & OHCI_CTRL_RWC) != 0 > - && !device_can_wakeup(hcd->self.controller)) > - device_init_wakeup(hcd->self.controller, 1); > + if ((ohci->hc_control & OHCI_CTRL_RWC) != 0) > + device_set_wakeup_capable(hcd->self.controller, 1); > > switch (ohci->hc_control & OHCI_CTRL_HCFS) { > case OHCI_USB_OPER: > > > -- Everyone knows that debugging is twice as hard as writing a program in the first place. So if you're as clever as you can be when you write it, how will you ever debug it? --- Brian Kernighan _______________________________________________ linux-pm mailing list linux-pm@xxxxxxxxxxxxxxxxxxxxxxxxxx https://lists.linux-foundation.org/mailman/listinfo/linux-pm
