Linux USB
 help / color / mirror / Atom feed
* [PATCH v2] usb: musb: fix crash with highmen PIO and usbmon
@ 2020-03-09 15:26 Mans Rullgard
  2020-03-09 16:16 ` Bin Liu
  0 siblings, 1 reply; 2+ messages in thread
From: Mans Rullgard @ 2020-03-09 15:26 UTC (permalink / raw)
  To: Bin Liu; +Cc: Greg Kroah-Hartman, linux-usb, linux-kernel

When handling a PIO bulk transfer with highmem buffer, a temporary
mapping is assigned to urb->transfer_buffer.  After the transfer is
complete, an invalid address is left behind in this pointer.  This is
not ordinarily a problem since nothing touches that buffer before the
urb is released.  However, when usbmon is active, usbmon_urb_complete()
calls (indirectly) mon_bin_get_data() which does access the transfer
buffer if it is set.  To prevent an invalid memory access here, reset
urb->tranfer_buffer to NULL when finished (musb_host_rx()), or do not
set it at all (musb_host_tx()).

Fixes: 8e8a55165469 ("usb: musb: host: Handle highmem in PIO mode")
Signed-off-by: Mans Rullgard <mans@mansr.com>
---
Changed in v2:
- Do not set urb->transfer_buffer in musb_host_tx()
---
 drivers/usb/musb/musb_host.c | 17 +++++------------
 1 file changed, 5 insertions(+), 12 deletions(-)

diff --git a/drivers/usb/musb/musb_host.c b/drivers/usb/musb/musb_host.c
index 1c813c37462a..8736f4251a22 100644
--- a/drivers/usb/musb/musb_host.c
+++ b/drivers/usb/musb/musb_host.c
@@ -1436,10 +1436,7 @@ void musb_host_tx(struct musb *musb, u8 epnum)
 	 * We need to map sg if the transfer_buffer is
 	 * NULL.
 	 */
-	if (!urb->transfer_buffer)
-		qh->use_sg = true;
-
-	if (qh->use_sg) {
+	if (!urb->transfer_buffer) {
 		/* sg_miter_start is already done in musb_ep_program */
 		if (!sg_miter_next(&qh->sg_miter)) {
 			dev_err(musb->controller, "error: sg list empty\n");
@@ -1447,9 +1444,8 @@ void musb_host_tx(struct musb *musb, u8 epnum)
 			status = -EINVAL;
 			goto done;
 		}
-		urb->transfer_buffer = qh->sg_miter.addr;
 		length = min_t(u32, length, qh->sg_miter.length);
-		musb_write_fifo(hw_ep, length, urb->transfer_buffer);
+		musb_write_fifo(hw_ep, length, qh->sg_miter.addr);
 		qh->sg_miter.consumed = length;
 		sg_miter_stop(&qh->sg_miter);
 	} else {
@@ -1458,11 +1454,6 @@ void musb_host_tx(struct musb *musb, u8 epnum)
 
 	qh->segsize = length;
 
-	if (qh->use_sg) {
-		if (offset + length >= urb->transfer_buffer_length)
-			qh->use_sg = false;
-	}
-
 	musb_ep_select(mbase, epnum);
 	musb_writew(epio, MUSB_TXCSR,
 			MUSB_TXCSR_H_WZC_BITS | MUSB_TXCSR_TXPKTRDY);
@@ -1977,8 +1968,10 @@ void musb_host_rx(struct musb *musb, u8 epnum)
 	urb->actual_length += xfer_len;
 	qh->offset += xfer_len;
 	if (done) {
-		if (qh->use_sg)
+		if (qh->use_sg) {
 			qh->use_sg = false;
+			urb->transfer_buffer = NULL;
+		}
 
 		if (urb->status == -EINPROGRESS)
 			urb->status = status;
-- 
2.25.1


^ permalink raw reply related	[flat|nested] 2+ messages in thread

* Re: [PATCH v2] usb: musb: fix crash with highmen PIO and usbmon
  2020-03-09 15:26 [PATCH v2] usb: musb: fix crash with highmen PIO and usbmon Mans Rullgard
@ 2020-03-09 16:16 ` Bin Liu
  0 siblings, 0 replies; 2+ messages in thread
From: Bin Liu @ 2020-03-09 16:16 UTC (permalink / raw)
  To: Mans Rullgard; +Cc: Greg Kroah-Hartman, linux-usb, linux-kernel

On Mon, Mar 09, 2020 at 03:26:53PM +0000, Mans Rullgard wrote:
> When handling a PIO bulk transfer with highmem buffer, a temporary
> mapping is assigned to urb->transfer_buffer.  After the transfer is
> complete, an invalid address is left behind in this pointer.  This is
> not ordinarily a problem since nothing touches that buffer before the
> urb is released.  However, when usbmon is active, usbmon_urb_complete()
> calls (indirectly) mon_bin_get_data() which does access the transfer
> buffer if it is set.  To prevent an invalid memory access here, reset
> urb->tranfer_buffer to NULL when finished (musb_host_rx()), or do not

WARNING: 'tranfer' may be misspelled

I have fixed it locally.

> set it at all (musb_host_tx()).
> 
> Fixes: 8e8a55165469 ("usb: musb: host: Handle highmem in PIO mode")
> Signed-off-by: Mans Rullgard <mans@mansr.com>

Applied. Thanks.
-Bin.

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2020-03-09 16:11 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2020-03-09 15:26 [PATCH v2] usb: musb: fix crash with highmen PIO and usbmon Mans Rullgard
2020-03-09 16:16 ` Bin Liu

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox