usb-storage: fix deadlock when a scsi command timeouts more than once

commit a398d5e upstream.

With faulty usb-storage devices, read/write can timeout, in that case
the SCSI layer will abort and re-issue the command. USB storage has no
internal timeout, it relies on SCSI layer aborting commands via
.eh_abort_handler() for non those responsive devices.

After two consecutive timeouts of the same command, SCSI layer calls
.eh_device_reset_handler(), without calling .eh_abort_handler() first.

With usb-storage, this causes a deadlock:

  -> .eh_device_reset_handler
    -> device_reset
      -> mutex_lock(&(us->dev_mutex));

mutex already by usb_stor_control_thread(), which is waiting for
command completion:

  -> usb_stor_control_thread (mutex taken here)
    -> usb_stor_invoke_transport
      -> usb_stor_Bulk_transport
        -> usb_stor_bulk_srb
	  -> usb_stor_bulk_transfer_sglist
	    -> usb_sg_wait

Make sure we cancel any pending command in .eh_device_reset_handler()
to avoid this.

Signed-off-by: Maxime Bizon <mbizon@freebox.fr>
Cc: linux-usb@vger.kernel.org
Cc: stable <stable@kernel.org>
Link: https://lore.kernel.org/all/ZEllnjMKT8ulZbJh@sakura/
Reviewed-by: Alan Stern <stern@rowland.harvard.edu>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Link: https://lore.kernel.org/r/20230505114759.1189741-1-mbizon@freebox.fr
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

drivers/usb/storage/scsiglue.c

@@ -406,22 +406,25 @@ static DEF_SCSI_QCMD(queuecommand)
  ***********************************************************************/
 
 /* Command timeout and abort */
-static int command_abort(struct scsi_cmnd *srb)
+static int command_abort_matching(struct us_data *us, struct scsi_cmnd *srb_match)
 {
-	struct us_data *us = host_to_us(srb->device->host);
-
-	usb_stor_dbg(us, "%s called\n", __func__);
-
 	/*
 	 * us->srb together with the TIMED_OUT, RESETTING, and ABORTING
 	 * bits are protected by the host lock.
 	 */
 	scsi_lock(us_to_host(us));
 
-	/* Is this command still active? */
-	if (us->srb != srb) {
+	/* is there any active pending command to abort ? */
+	if (!us->srb) {
 		scsi_unlock(us_to_host(us));
 		usb_stor_dbg(us, "-- nothing to abort\n");
+		return SUCCESS;
+	}
+
+	/* Does the command match the passed srb if any ? */
+	if (srb_match && us->srb != srb_match) {
+		scsi_unlock(us_to_host(us));
+		usb_stor_dbg(us, "-- pending command mismatch\n");
 		return FAILED;
 	}
 
@@ -444,6 +447,14 @@ static int command_abort(struct scsi_cmnd *srb)
 	return SUCCESS;
 }
 
+static int command_abort(struct scsi_cmnd *srb)
+{
+	struct us_data *us = host_to_us(srb->device->host);
+
+	usb_stor_dbg(us, "%s called\n", __func__);
+	return command_abort_matching(us, srb);
+}
+
 /*
  * This invokes the transport reset mechanism to reset the state of the
  * device
@@ -455,6 +466,9 @@ static int device_reset(struct scsi_cmnd *srb)
 
 	usb_stor_dbg(us, "%s called\n", __func__);
 
+	/* abort any pending command before reset */
+	command_abort_matching(us, NULL);
+
 	/* lock the device pointers and do the reset */
 	mutex_lock(&(us->dev_mutex));
 	result = us->transport_reset(us);