On Mon, Sep 11, 2017 at 4:29 PM, Benson Leung <ble...@google.com> wrote: > Hi Oliver, > > On Mon, Sep 11, 2017 at 01:02:52PM -0700, Benson Leung wrote: >> Thanks for the pointer. I'll respin the patch with the no_resume version of >> usb_autopm_get_interface and retest. >> > > I went and tried this patch but with usb_autopm_get_interface_no_resume > instead > and the original bug I was trying to fix with a Yubikey hid security key came > back, so something is still wrong here. > > I went ahead and filed a bug to track this here: > https://bugs.chromium.org/p/chromium/issues/detail?id=764112 > > I'll find some time to dig into this deeper and understand why this device > ends up getting stuck in active instead of suspending when all handles are > closed.
Meh, it is problem in hidraw that basically does usb_autopm_put_interface() before letting usbhid_close() to run. Once I swap hid_hw_power() and hid_hw_close() it autosuspends properly. I just sent out a patch. Thanks. -- Dmitry
