On Wednesday 28 July 2010 17:21:38 Alan Stern wrote: > On Wed, 28 Jul 2010, Hans Petter Selasky wrote: > > I grepped the EHCI specification for the word short, and in section 3.5.2 > > it says about the alternate next pointer that this will alaways be used > > when a short packet is received. If the alternate next pointer is set to > > the terminate value, shouldn't the QH be retired aswell? > > No. When "using" the altnext pointer, the first thing the controller > does is check the Terminate bit. If that bit is set, the controller > ignores the altnext field and uses the regular "next" pointer instead. > From section 4.10.2: > > If the field Bytes to Transfer is not zero and the T-bit in the > Alternate Next qTD Pointer is set to zero, then the host > controller uses the Alternate Next qTD Pointer. Otherwise, the > host controller uses the Next qTD Pointer. If Next qTD > Pointer's T-bit is set to a one, then the host controller exits > this state and uses the horizontal pointer to the next schedule > data structure. > > By making altnext point to an inactive qTD, you can force the queue to > stop: > > Using the selected pointer the host controller fetches the > referenced qTD. If the fetched qTD has its Active bit set to > a one, the host controller moves the pointer value used to > reach the qTD (Next or Alternate Next) to the Current qTD > Pointer field, then performs the overlay. If the fetched qTD > has its Active bit set to a zero, the host controller aborts > the queue advance and follows the queue head's horizontal > pointer to the next schedule data structure. > > > Is a ZLP not regarded a short packet? > > It is. That's not the issue; you didn't understand what the spec > meant when it said the pointer was "used". > > > If your view is correct, than you cannot setup more than 16K of TD's when > > doing mass storage, risking that on IN transactions the Command Status > > Wrapper block will get part of the data buffer. > > Not at all. You can set up as much as you like. Just make sure that > in all the qTD's except possibly the last, the altnext field points to > a qTD that has the Active bit clear and the Halted bit set. (Some > controllers disobey the spec and perform the overlay even when Active > is clear.) That makes sense. I will look into this issue when I find some time and try it out. Thanks for your explanation. --HPS -- 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
