On Út 27-12-05 17:39:00, Dmitry Torokhov wrote: > On 12/27/05, Pavel Machek <pavel@xxxxxx> wrote: > > On Út 27-12-05 17:21:28, Dmitry Torokhov wrote: > > > On 12/27/05, Pavel Machek <pavel@xxxxxx> wrote: > > > > On Út 27-12-05 16:45:17, Dmitry Torokhov wrote: > > > > > > > > > > Hmm, I am confused. What should ->resume() method do when it fails to > > > > > resume a device? I would say kill it, espacially if it is > > > > > bus-generic > > > > > > > > That should basically never happen. You can kill the device, or maybe > > > > even panic() whole machine. It just should not happen. > > > > > > It can happen, it did happen and it will continue to happen. You hit > > > your suspend button and stand up and accidentially kick your USB hub > > > or your PS/2 mouse gets loose and kernel fails to resume it. It is no > > > big deal, we just have to be prepared to handle it. > > > > Ok, in case of hotpluggable devices, you just handle it in a similar > > way to surprise removal. > > Yes, and we go back to beginning - currently the driver does not know > whether it is allowed to handle surprise removal or not because it > does not know if it is resumed for real or just to be able to write > suspend image. Would a flag passed down to the driver, telling it if it is "for real" or only as a part of suspend help? Pavel -- Thanks, Sharp!
