xhci: Fix 5.12 regression of missing xHC cache clearing command after…

… a Stall

commit a7f2e92 upstream.

If endpoints halts due to a stall then the dequeue pointer read from
hardware may already be set ahead of the stalled TRB.
After commit 674f843 ("xhci: split handling halted endpoints into two
steps") in 5.12 xhci driver won't issue a Set TR Dequeue if hardware
dequeue pointer is already in the right place.

Turns out the "Set TR Dequeue pointer" command is anyway needed as it in
addition to moving the dequeue pointer also clears endpoint state and
cache.

Fixes: 674f843 ("xhci: split handling halted endpoints into two steps")
Cc: <stable@vger.kernel.org> # 5.12
Reported-by: Peter Ganzhorn <peter.ganzhorn@googlemail.com>
Tested-by: Peter Ganzhorn <peter.ganzhorn@googlemail.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20210525074100.1154090-3-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

drivers/usb/host/xhci-ring.c

@@ -934,14 +934,18 @@ static int xhci_invalidate_cancelled_tds(struct xhci_virt_ep *ep)
 			continue;
 		}
 		/*
-		 * If ring stopped on the TD we need to cancel, then we have to
+		 * If a ring stopped on the TD we need to cancel then we have to
 		 * move the xHC endpoint ring dequeue pointer past this TD.
+		 * Rings halted due to STALL may show hw_deq is past the stalled
+		 * TD, but still require a set TR Deq command to flush xHC cache.
 		 */
 		hw_deq = xhci_get_hw_deq(xhci, ep->vdev, ep->ep_index,
 					 td->urb->stream_id);
 		hw_deq &= ~0xf;
 
-		if (trb_in_td(xhci, td->start_seg, td->first_trb,
+		if (td->cancel_status == TD_HALTED) {
+			cached_td = td;
+		} else if (trb_in_td(xhci, td->start_seg, td->first_trb,
 			      td->last_trb, hw_deq, false)) {
 			switch (td->cancel_status) {
 			case TD_CLEARED: /* TD is already no-op */