Hi, On Monday 24 April 2006 23:29, David Brownell wrote: > I've noticed a bunch of problem reports that go like this: > > - boot system with some USB devices attached > - echo disk > /sys/power/state > - ... later resume ... > - now those USB devices don't work right > - unplug them/replug them, all is OK > > I recently observed this myself and tracked down one problem. The solution > involves what kexec() does in much the same situation: before starting a > new kernel, most hardware needs to be reset. Today, swsusp will suspend it > instead, which is the root cause of the problem. > > This seems like something that might need to sit in the MM tree for > a while, as it's a longstanding bug and the fix could shake loose driver > goofs. Luckily there's a partial workaround already in wide use: link > drivers as modules if they use true suspend states (like all USB HCDs), > don't link them statically into the kernel image. > > That workaround is partial because when BIOS takes over a controller, the > resume() method has to handle that nasty case too! Last I checked, all > the USB HCDs do handle that case; and few drivers other than those HCDs > would ever need to worry about such issues. I agree with this change but I think it'll cause some strange things to happen for some time. Greetings, Rafael
