On Mon, Feb 18, 2013 at 03:34:07PM +0100, Stefan Tauner wrote: > This allows user space to retrieve the current frame number on a USB > as returned by usb_get_current_frame_number(...) via sysfs. > > Signed-off-by: Stefan Tauner <stefan.tauner@xxxxxxxxxxxxxxxxxxxx> > --- > That's sadly not necessarily the raw value seen on the bus because > the individual driver functions called by usb_get_current_frame_number(...) > seem to limit the possible range to the "schedule horizon" of isochronous > packets. IMHO the function name is a misnomer and I would like to hear your > opinion on this, a possible new name for it and a better way to retrieve the > real/raw value. > > The rationale for this patch can be found in the thread with the subject > "Correlating SOF with host system time" from last december > (201212042020.qB4KKt0N008541@xxxxxxxxxxxxxxxxxxxxxxxxxx). > My whole use case/idea was kinda frowned upon then, but there was little > discussion about how it should/could be implemented in the kernel code in > case one really wants to try it. For those of us with bad memories, care to expand on why you want this in the changelog body of the patch so that the world will remember it as well? > Adding the sysfs attribute was the easiest way for me to communicate > this to user space, but at least for my use case a non-polling approach > would have been better. Any suggestions on how this could be done by a > newbie like me are very welcome. Whenever you add/modify/remove a sysfs attribute, you also need to document it in Documentation/ABI/. Care to do that and resend this patch? thanks, greg k-h -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
