1. Patchew
  2. linux
  3. View series

[PATCH] media: rtl2832_sdr: free DMA stream buffers before clearing udev on remove

Valery Borovsky posted 1 patch 1 month ago 
drivers/media/dvb-frontends/rtl2832_sdr.c | 8 ++++++++
1 file changed, 8 insertions(+)
[PATCH] media: rtl2832_sdr: free DMA stream buffers before clearing udev on remove
Posted by Valery Borovsky 1 month ago 
rtl2832_sdr_remove() runs on USB disconnect and immediately clears
dev->udev to NULL before any pending streaming teardown can run. When
the user-space application later closes its file descriptor, vb2 calls
rtl2832_sdr_stop_streaming() which in turn calls
rtl2832_sdr_free_stream_bufs(). That helper releases each coherent
buffer with:

    usb_free_coherent(dev->udev, dev->buf_size,
                      dev->buf_list[dev->buf_num],
                      dev->dma_addr[dev->buf_num]);

usb_free_coherent() returns immediately when its dev argument is NULL,
so every DMA stream buffer that was live at disconnect is silently
leaked. The URBs allocated in rtl2832_sdr_alloc_urbs() outlive the
device for the same reason.

Tear down the streaming state in rtl2832_sdr_remove() while dev->udev
is still valid: call rtl2832_sdr_kill_urbs(), rtl2832_sdr_free_urbs()
and rtl2832_sdr_free_stream_bufs() before zeroing dev->udev. The
helpers are idempotent (they clear urbs_submitted, urbs_initialized
and the URB_BUF flag), so the subsequent stop_streaming() path from
the vb2 release sequence becomes a safe no-op.

Issue identified by automated review of the INV-003 series at
https://sashiko.dev/

Fixes: 771138920eaf ("[media] rtl2832_sdr: Realtek RTL2832 SDR driver module")
Cc: stable@vger.kernel.org
Signed-off-by: Valery Borovsky <vebohr@gmail.com>
---
 drivers/media/dvb-frontends/rtl2832_sdr.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/media/dvb-frontends/rtl2832_sdr.c b/drivers/media/dvb-frontends/rtl2832_sdr.c
index 422d1a7b5456..817d91faa598 100644
--- a/drivers/media/dvb-frontends/rtl2832_sdr.c
+++ b/drivers/media/dvb-frontends/rtl2832_sdr.c
@@ -1470,6 +1470,14 @@ static void rtl2832_sdr_remove(struct platform_device *pdev)
 
 	mutex_lock(&dev->vb_queue_lock);
 	mutex_lock(&dev->v4l2_lock);
+	/*
+	 * Release URBs and coherent DMA stream buffers while dev->udev
+	 * is still valid; once it is cleared, usb_free_coherent() silently
+	 * returns and any later stop_streaming() leaks the DMA memory.
+	 */
+	rtl2832_sdr_kill_urbs(dev);
+	rtl2832_sdr_free_urbs(dev);
+	rtl2832_sdr_free_stream_bufs(dev);
 	/* No need to keep the urbs around after disconnection */
 	dev->udev = NULL;
 	v4l2_device_disconnect(&dev->v4l2_dev);
-- 
2.51.0
Re: [PATCH] media: rtl2832_sdr: free DMA stream buffers before clearing udev on remove
Posted by Hans Verkuil 3 weeks, 2 days ago 
On 13/05/2026 07:57, Valery Borovsky wrote:
> rtl2832_sdr_remove() runs on USB disconnect and immediately clears
> dev->udev to NULL before any pending streaming teardown can run. When
> the user-space application later closes its file descriptor, vb2 calls
> rtl2832_sdr_stop_streaming() which in turn calls
> rtl2832_sdr_free_stream_bufs(). That helper releases each coherent
> buffer with:
> 
>     usb_free_coherent(dev->udev, dev->buf_size,
>                       dev->buf_list[dev->buf_num],
>                       dev->dma_addr[dev->buf_num]);
> 
> usb_free_coherent() returns immediately when its dev argument is NULL,
> so every DMA stream buffer that was live at disconnect is silently
> leaked. The URBs allocated in rtl2832_sdr_alloc_urbs() outlive the
> device for the same reason.
> 
> Tear down the streaming state in rtl2832_sdr_remove() while dev->udev
> is still valid: call rtl2832_sdr_kill_urbs(), rtl2832_sdr_free_urbs()
> and rtl2832_sdr_free_stream_bufs() before zeroing dev->udev. The
> helpers are idempotent (they clear urbs_submitted, urbs_initialized
> and the URB_BUF flag), so the subsequent stop_streaming() path from
> the vb2 release sequence becomes a safe no-op.
> 
> Issue identified by automated review of the INV-003 series at
> https://sashiko.dev/
> 
> Fixes: 771138920eaf ("[media] rtl2832_sdr: Realtek RTL2832 SDR driver module")
> Cc: stable@vger.kernel.org
> Signed-off-by: Valery Borovsky <vebohr@gmail.com>
> ---
>  drivers/media/dvb-frontends/rtl2832_sdr.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/drivers/media/dvb-frontends/rtl2832_sdr.c b/drivers/media/dvb-frontends/rtl2832_sdr.c
> index 422d1a7b5456..817d91faa598 100644
> --- a/drivers/media/dvb-frontends/rtl2832_sdr.c
> +++ b/drivers/media/dvb-frontends/rtl2832_sdr.c
> @@ -1470,6 +1470,14 @@ static void rtl2832_sdr_remove(struct platform_device *pdev)
>  
>  	mutex_lock(&dev->vb_queue_lock);
>  	mutex_lock(&dev->v4l2_lock);
> +	/*
> +	 * Release URBs and coherent DMA stream buffers while dev->udev
> +	 * is still valid; once it is cleared, usb_free_coherent() silently
> +	 * returns and any later stop_streaming() leaks the DMA memory.
> +	 */
> +	rtl2832_sdr_kill_urbs(dev);
> +	rtl2832_sdr_free_urbs(dev);
> +	rtl2832_sdr_free_stream_bufs(dev);
>  	/* No need to keep the urbs around after disconnection */
>  	dev->udev = NULL;
>  	v4l2_device_disconnect(&dev->v4l2_dev);

This isn't the right solution.

What needs to happen is that the video_unregister_device(&dev->vdev);
calls is replaced by vb2_video_unregister_device(&dev->vdev); and that
that line is moved up to before dev->udev = NULL.

vb2_video_unregister_device ensures that all streaming stops at once,
so it calls stop_streaming which will free all streaming resources.

You were on the right track, but it's not quite the right solution.

Regards,

	Hans
[PATCH v2] media: rtl2832_sdr: use vb2_video_unregister_device() on remove to fix DMA leak
Posted by Valery Borovsky 2 weeks, 6 days ago 
rtl2832_sdr_remove() runs on USB disconnect and clears dev->udev to
NULL before any pending streaming teardown has run. When user space
later closes its file descriptor, vb2 calls rtl2832_sdr_stop_streaming()
which in turn calls rtl2832_sdr_free_stream_bufs(). That helper releases
each coherent buffer with:

    usb_free_coherent(dev->udev, dev->buf_size,
                      dev->buf_list[dev->buf_num],
                      dev->dma_addr[dev->buf_num]);

usb_free_coherent() returns immediately when its dev argument is NULL,
so every DMA stream buffer that was live at disconnect is silently
leaked. The URBs allocated in rtl2832_sdr_alloc_urbs() outlive the
device for the same reason.

The rtl2832_sdr driver uses vb2_fop_release() in its file_operations,
so replace video_unregister_device(&dev->vdev) with
vb2_video_unregister_device(&dev->vdev) and move it before clearing
dev->udev. vb2_video_unregister_device() releases the vb2 queue, which
synchronously runs rtl2832_sdr_stop_streaming() if streaming is active,
so URBs and coherent DMA stream buffers are freed while dev->udev is
still valid.

vb2_video_unregister_device() locks vdev->queue->lock (vb_queue_lock)
internally, and stop_streaming() locks v4l2_lock, so the previous outer
mutex_lock(&dev->vb_queue_lock) / mutex_lock(&dev->v4l2_lock) pair
around the unregister sequence would self-deadlock and has been removed.
A short v4l2_lock critical section around dev->udev = NULL remains so
any ioctl path that still holds the file descriptor sees coherent state.

Issue identified by automated review of the INV-003 series at
https://sashiko.dev/

Fixes: 771138920eaf ("[media] rtl2832_sdr: Realtek RTL2832 SDR driver module")
Cc: stable@vger.kernel.org
Suggested-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Signed-off-by: Valery Borovsky <vebohr@gmail.com>
---
Changes since v1
(https://lore.kernel.org/linux-media/20260513055745.146998-1-vebohr@gmail.com/):
- Rewritten per Hans Verkuil's review
  (https://lore.kernel.org/linux-media/3c427dde-54c5-4a63-bcab-dd0079593ba1@kernel.org/):
  replace video_unregister_device() with vb2_video_unregister_device()
  and move it before clearing dev->udev, instead of open-coding the
  URB/DMA teardown. vb2_video_unregister_device() releases the queue,
  which synchronously calls stop_streaming() while dev->udev is still
  valid, so the explicit rtl2832_sdr_kill_urbs() /
  rtl2832_sdr_free_urbs() / rtl2832_sdr_free_stream_bufs() block from
  v1 is no longer needed.
- Dropped the outer mutex_lock(&dev->vb_queue_lock) /
  mutex_lock(&dev->v4l2_lock) around the unregister sequence:
  vb2_video_unregister_device() acquires vb_queue_lock internally and
  stop_streaming() acquires v4l2_lock, so holding either of those
  while calling the unregister helper self-deadlocks.
- Rebased on media-committers/next.

 drivers/media/dvb-frontends/rtl2832_sdr.c | 18 +++++++++++++-----
 1 file changed, 13 insertions(+), 5 deletions(-)

diff --git a/drivers/media/dvb-frontends/rtl2832_sdr.c b/drivers/media/dvb-frontends/rtl2832_sdr.c
index c564485e3bbb..c1f5f07c42a8 100644
--- a/drivers/media/dvb-frontends/rtl2832_sdr.c
+++ b/drivers/media/dvb-frontends/rtl2832_sdr.c
@@ -1477,14 +1477,22 @@ static void rtl2832_sdr_remove(struct platform_device *pdev)
 
 	dev_dbg(&pdev->dev, "\n");
 
-	mutex_lock(&dev->vb_queue_lock);
+	/*
+	 * vb2_video_unregister_device() releases the vb2 queue, which
+	 * triggers rtl2832_sdr_stop_streaming() if streaming is active.
+	 * stop_streaming() uses dev->udev to free URBs and coherent DMA
+	 * stream buffers via usb_free_coherent(), so it must run before
+	 * dev->udev is cleared. vb2_video_unregister_device() locks
+	 * vb_queue_lock internally and stop_streaming() locks v4l2_lock,
+	 * so neither may be held by the caller.
+	 */
+	v4l2_device_disconnect(&dev->v4l2_dev);
+	vb2_video_unregister_device(&dev->vdev);
+
 	mutex_lock(&dev->v4l2_lock);
-	/* No need to keep the urbs around after disconnection */
 	dev->udev = NULL;
-	v4l2_device_disconnect(&dev->v4l2_dev);
-	video_unregister_device(&dev->vdev);
 	mutex_unlock(&dev->v4l2_lock);
-	mutex_unlock(&dev->vb_queue_lock);
+
 	v4l2_device_put(&dev->v4l2_dev);
 	module_put(pdev->dev.parent->driver->owner);
 }
-- 
2.51.0

Patchew 2.1-devel-ea86937

© 2016 - 2026 Red Hat, Inc.