On Wed, 2012-04-11 at 17:13 -0400, Alan Stern wrote: > On Wed, 11 Apr 2012, Steven Rostedt wrote: > > > On Wed, 2012-04-11 at 15:12 -0400, Alan Stern wrote: > > > > > If so, setting the value back to 0 before suspending (or never setting > > > it to 1 in the first place) might be important. You can test this > > > easily enough. In drivers/usb/host/ehci-pci.c:ehci_pci_suspend(), add > > > a line saying > > > > > > ehci_writel(ehci, 0, &ehci->regs->configured_flag); > > > > > > just before the spin_lock_irqrestore. This will invalidate the > > > driver's criterion for determining whether or not the controller's > > > state got messed up during the suspend; we can worry about that later. > > > > I just tried the above, and it made no difference. Note, I don't even > > get to suspend. It locks up in suspend, so I haven't even tried a resume > > yet. > > What do you mean, you don't get to suspend? Is that a typo? I mean, as I hit suspend, the screen goes blank, but the backlight stays on, and the system just hangs there. It never gets to the point where the machine is in "suspend" mode. I have to press and hold the power button to force a shutdown. > > If the system locks up during the suspend procedure, it must do so > after this code runs. I know that because it runs during the "devices" > phase of suspending, and you said that "echo devices > >/sys/power/pm_test" works. > > > > You can try getting rid of the call to pci_prepare_to_sleep in > > > drivers/usb/core/hcd-pci.c:hcd_pci_suspend_noirq. This will prevent > > > the controller from being put into D3hot and might interfere with > > > wakeup detection. > > > > > > > What do I do with the retval? -EIO, 0, or other? > > 0 or -EIO, either one. I set it to 0 and it worked! (was able to suspend and resume) > > > On Thu, 12 Apr 2012, Andrey Rahmatullin wrote: > > > I replaced the hcd_pci_suspend_noirq call with retval=0 and suspend/resume > > was successful. > > This indicates that your computer doesn't like to suspend while the > EHCI controllers are in D3hot. I have no idea why not. > > If you want a really thorough test, try changing > drivers/pci/pci-driver.c:pci_pm_suspend_noirq. In the "if (!pm)" > clause, add a call to > > pci_prepare_to_sleep(pci_dev); > > just after the pci_save_state line. Then try suspending with the > script enabled (ehci-hcd unbound from the controllers). If this fails > then we'll know it is the source of the trouble. I have to go for a few hours, but I can try this when I get back. Even if Andrey does it first. I'll verify it it too. Thanks Alan and Andrey! -- Steve _______________________________________________ linux-pm mailing list linux-pm@xxxxxxxxxxxxxxxxxxxxxxxxxx https://lists.linuxfoundation.org/mailman/listinfo/linux-pm
