Hi, On Wed, Sep 28, 2011 at 03:18:21PM -0400, Alan Stern wrote: > On Wed, 28 Sep 2011, Felipe Balbi wrote: > > > > > first hunk will prevent that flag from ever being cleared, no ? > > > > According to documentation > > > > include/linux/usb/gadget.h::usb_gadget_set_wedge(), it should prevent > > > > ClearFeature(HALT) from succeding, but gadget driver still should be > > > > able to clear that flag. > > > > > > > > Unless I'm missing something, I think your first hunk is also wrong, as > > > > it would prevent WEDGE and STALL flags from ever being cleared after > > > > they are set. > > > > > > No, that is correct. Once the WEDGE flag is set, the only way to clear > > > it is by doing something drastic, like resetting the gadget or > > > disabling and re-enabling the endpoint (as in Set-Interface or > > > Set-Config). > > > > > > This is the required behavior for the Mass Storage bulk-only transport > > > class, as described in section 6.6.1 of that spec. > > > > Sorry but what's correct ? My rationale or original patch ? > > Sorry for being unclear. I'm not sure about the first hunk; it depends > on how other parts of the driver are written. I meant that the WEDGE > and STALL _should_ be prevented from ever being cleared (by a > Clear-Feature request from the host) after WEDGE is set. I understand. Thanks for clearing that out. -- balbi
Attachment:
signature.asc
Description: Digital signature
