On Mon, 28 Jul 2014, Adam Borowski wrote: > On Wed, Jul 23, 2014 at 11:19:06AM -0400, Alan Stern wrote: > > > > Your system wakes up okay from suspend, but not from hibernation. Was > > > > there ever any kernel version where USB wakeup from hibernation did work? > > > > > > Any version before c1db30a2a79eb59997b13b8cabf2a50bea9f04e1. That very > > > commit also happens to fix a far more severe bug during _suspend_ on my > > > hardware, in return for this regression during hibernation. > > > > What happens with hibernation if you try the patch in > > > > http://marc.info/?l=linux-usb&m=140259672425512&w=2 > > > > using both the vanilla kernel and one with commit c1db30a2a79e > > reverted? > > Reverting c1db30a2a79e does fix hibernation issues (and brings back suspend > breakage), the patch you mentioned above seems to have no effect whatsoever, > with c1db30a2a79e reverted or not. I don't get it. With that patch in that URL applied, the code added by c1db30a2a79e doesn't do anything -- the if ((temp & (RH_PS_PES | RH_PS_PSS)) == RH_PS_PES) test will always be false, and so the following ohci_writel() will never be called. You should be able to test that easily enough. I suppose it's barely possible that the preceding ohci_readl() call might have an effect. You can try experimenting with those two calls; comment out the ohci_writel() and see what happens, then also comment out the ohci_readl() and see what happens. You might as well make these tests without that other patch applied, to avoid confusion. 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
