[RFC 2/3] xhci: Fix command completion after a drop endpoint.

[Sarah Sharp <sarah.a.sharp@xxxxxxxxxxxxxxx>]

The xHCI driver issues a Configure Endpoint command for two reasons:
 - a new configuration or alternate interface setting is selected
 - a quirky Fresco Logic prototype requires the command after a Reset
   Endpoint command.
The xHCI driver only waits on the command in the first case.

When a configure endpoint command completes, the driver needs to know why
the command was generated.  When the driver only supported selecting an
initial configuration, the check was simple.  Unfortunately that check
doesn't work now that the driver supports alternate interfaces.  If an
endpoint must be dropped (because it's not in the new alternate setting)
and no new endpoints are added, the math involving
xhci_last_valid_endpoint() will assign -1 to an unsigned integer and cause
an out-of-bounds array access.

Move the check for the quirky hardware sooner and avoid the bad array
access.

Signed-off-by: Sarah Sharp <sarah.a.sharp@xxxxxxxxxxxxxxx>
---
 drivers/usb/host/xhci-ring.c |   36 ++++++++++++++++++++----------------
 1 files changed, 20 insertions(+), 16 deletions(-)

diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c
index 224f373..64cbbb9 100644
--- a/drivers/usb/host/xhci-ring.c
+++ b/drivers/usb/host/xhci-ring.c
@@ -903,28 +903,32 @@ static void handle_cmd_completion(struct xhci_hcd *xhci,
 				virt_dev->in_ctx);
 		/* Input ctx add_flags are the endpoint index plus one */
 		ep_index = xhci_last_valid_endpoint(ctrl_ctx->add_flags) - 1;
-		ep_ring = xhci->devs[slot_id]->eps[ep_index].ring;
-		if (!ep_ring) {
-			/* This must have been an initial configure endpoint */
-			xhci->devs[slot_id]->cmd_status =
-				GET_COMP_CODE(event->status);
-			complete(&xhci->devs[slot_id]->cmd_completion);
-			break;
-		}
-		ep_state = xhci->devs[slot_id]->eps[ep_index].ep_state;
-		xhci_dbg(xhci, "Completed config ep cmd - last ep index = %d, "
-				"state = %d\n", ep_index, ep_state);
+		/* A usb_set_interface() call directly after clearing a halted
+		 * condition may race on this quirky hardware.
+		 * Not worth worrying about, since this is prototype hardware.
+		 */
 		if (xhci->quirks & XHCI_RESET_EP_QUIRK &&
-				ep_state & EP_HALTED) {
+				ep_index != (unsigned int) -1 &&
+				ctrl_ctx->add_flags - SLOT_FLAG ==
+					ctrl_ctx->drop_flags) {
+			ep_ring = xhci->devs[slot_id]->eps[ep_index].ring;
+			ep_state = xhci->devs[slot_id]->eps[ep_index].ep_state;
+			if (!(ep_state & EP_HALTED))
+				goto bandwidth_change;
+			xhci_dbg(xhci, "Completed config ep cmd - "
+					"last ep index = %d, state = %d\n",
+					ep_index, ep_state);
 			/* Clear our internal halted state and restart ring */
 			xhci->devs[slot_id]->eps[ep_index].ep_state &=
 				~EP_HALTED;
 			ring_ep_doorbell(xhci, slot_id, ep_index);
-		} else {
-			xhci->devs[slot_id]->cmd_status =
-				GET_COMP_CODE(event->status);
-			complete(&xhci->devs[slot_id]->cmd_completion);
+			break;
 		}
+bandwidth_change:
+		xhci_dbg(xhci, "Completed config ep cmd\n");
+		xhci->devs[slot_id]->cmd_status =
+			GET_COMP_CODE(event->status);
+		complete(&xhci->devs[slot_id]->cmd_completion);
 		break;
 	case TRB_TYPE(TRB_EVAL_CONTEXT):
 		virt_dev = xhci->devs[slot_id];
-- 
1.6.0.4

--
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