On Thu, 3 Nov 2011, Tejun Heo wrote: > Hello, > > On Thu, Nov 3, 2011 at 8:39 AM, Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> wrote: > >> I don't know. I haven't really thought about it but as it's currently > >> designed, I don't think it'll be possible to avoid going through EH to > >> put ATA ports into suspend mode. > > > > Why not? > > It's just how things are designed now. There's no reason why it's > fundamentally impossible but going around that would require some > amount of hackery or preferably re-design. Synchronization against > command processing, interrupts and all are built around EH. Then it sounds like the best idea is something like what Ming proposed earlier: Have the error handler make some sort of test to determine whether it has been invoked for suspend/resume handling, and skip the runtime-PM calls on the host if it has. This new test, whatever it is, should apply both to system suspend and runtime suspend. By the way, what would happen if a system suspend occurred while the ATA port was already runtime-suspended? Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
