lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:   Tue, 24 Oct 2017 15:06:59 +0200
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     linux-kernel@...r.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        stable@...r.kernel.org, Lu Baolu <baolu.lu@...ux.intel.com>,
        Mathias Nyman <mathias.nyman@...ux.intel.com>
Subject: [PATCH 4.13 25/85] usb: xhci: Reset halted endpoint if trb is noop

4.13-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Lu Baolu <baolu.lu@...ux.intel.com>

commit 810a624bd1b64b13ddcc2eb5c1880526a750a870 upstream.

When a URB is cancled, xhci driver turns the untransferred trbs
into no-ops.  If an endpoint stalls on a no-op trb that belongs
to the cancelled URB, the event handler won't reset the endpoint.
Hence, it will stay halted.

Link: http://marc.info/?l=linux-usb&m=149582598330127&w=2

Signed-off-by: Lu Baolu <baolu.lu@...ux.intel.com>
Signed-off-by: Mathias Nyman <mathias.nyman@...ux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>

---
 drivers/usb/host/xhci-ring.c |   20 +++++++++++++-------
 1 file changed, 13 insertions(+), 7 deletions(-)

--- a/drivers/usb/host/xhci-ring.c
+++ b/drivers/usb/host/xhci-ring.c
@@ -2578,15 +2578,21 @@ static int handle_tx_event(struct xhci_h
 				(struct xhci_generic_trb *) ep_trb);
 
 		/*
-		 * No-op TRB should not trigger interrupts.
-		 * If ep_trb is a no-op TRB, it means the
-		 * corresponding TD has been cancelled. Just ignore
-		 * the TD.
+		 * No-op TRB could trigger interrupts in a case where
+		 * a URB was killed and a STALL_ERROR happens right
+		 * after the endpoint ring stopped. Reset the halted
+		 * endpoint. Otherwise, the endpoint remains stalled
+		 * indefinitely.
 		 */
 		if (trb_is_noop(ep_trb)) {
-			xhci_dbg(xhci,
-				 "ep_trb is a no-op TRB. Skip it for slot %u ep %u\n",
-				 slot_id, ep_index);
+			if (trb_comp_code == COMP_STALL_ERROR ||
+			    xhci_requires_manual_halt_cleanup(xhci, ep_ctx,
+							      trb_comp_code))
+				xhci_cleanup_halted_endpoint(xhci, slot_id,
+							     ep_index,
+							     ep_ring->stream_id,
+							     td, ep_trb,
+							     EP_HARD_RESET);
 			goto cleanup;
 		}
 


Powered by blists - more mailing lists