On Fri, 9 Sep 2011, Paul Zimmerman wrote: > Adding Sarah to CC, I just noticed she was the one who added this code. > > > -----Original Message----- > > From: Paul Zimmerman > > Sent: Friday, September 09, 2011 1:16 PM > > To: alan stern > > Cc: 'linux-usb@xxxxxxxxxxxxxxx' > > Subject: EHCI isoc scheduling > > > > Hi Alan, > > > > Can you give me a clue about this code in ehci-sched.c/iso_stream_schedule()? > > > > /* For high speed devices, allow scheduling within the > > * isochronous scheduling threshold. For full speed devices > > * and Intel PCI-based controllers, don't (work around for > > * Intel ICH9 bug). > > */ > > if (!stream->highspeed && ehci->fs_i_thresh) > > next = now + ehci->i_thresh; > > else > > next = now; > > > > Why does the driver allow scheduling within the threshold? I though the whole > > point of the threshold value in the HCCPARAMS register was to tell the driver > > the earliest time when it is safe to do scheduling. As I recall, people observed that an old version of the driver which always ignored fs_i_thresh had worked well. But then problems starting occurring with newer Intel controllers when communicating with full-speed devices, so the test was added. The comment could stand to be phrased a little better. > > Plus, the code doesn't seem to be doing what the comment says for the ICH9 > > workaround case (ehci->fs_i_thresh = 1). It seems the && should be an || > > instead. That's because the comment is so easy to misinterpret. > > The driver seems to be violating the EHCI spec here. Or am I misunderstanding > > what the code is doing? It _is_ violating the spec. Or rather, it's trying to do something which the spec doesn't guarantee will work. The only justification is that so far nobody has complained. Alan Stern -- 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
