Linux kernel -stable discussions
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: Tristan Madani <tristmd@gmail.com>
Cc: Neill Kapron <nkapron@google.com>,
	linux-usb@vger.kernel.org, linux-kernel@vger.kernel.org,
	Tristan Madani <tristan@talencesecurity.com>,
	stable@vger.kernel.org
Subject: Re: [PATCH v4] usb: gadget: f_uac1_legacy: fix copy_buf lifetime on queue handoff and free
Date: Fri, 20 Nov 2026 12:16:13 +0100	[thread overview]
Message-ID: <2026112033-gummy-backward-5436@gregkh> (raw)
In-Reply-To: <20261104125346.2552742-1-tristmd@gmail.com>

On Wed, Nov 04, 2026 at 12:53:46PM +0000, Tristan Madani wrote:
> From: Tristan Madani <tristan@talencesecurity.com>
> 
> f_audio_free() tears down the f_audio structure without cancelling
> playback_work or releasing audio buffers.  The pending work can access
> the freed structure, and queued audio buffers leak.
> 
> Additionally, f_audio_set_alt() and f_audio_out_ep_complete() transfer
> copy_buf to play_queue without clearing audio->copy_buf.  When
> playback_work later consumes the buffer from the queue, audio->copy_buf
> becomes a stale reference.
> 
> Clear audio->copy_buf when its buffer is queued to play_queue, so the
> pointer reliably reflects ownership.  In f_audio_free(), cancel
> playback_work, drain the play_queue, and free copy_buf when it was not
> queued.  Use IS_ERR_OR_NULL for the copy_buf check because
> f_audio_buffer_alloc() returns ERR_PTR on failure and f_audio_set_alt()
> stores the result directly in audio->copy_buf.
> 
> Fixes: c6994e6f067c ("USB: gadget: add USB Audio Gadget driver")
> Cc: stable@vger.kernel.org
> Signed-off-by: Tristan Madani <tristan@talencesecurity.com>
> ---
> 
> Changes in v4:
> - Fix the root cause of the stale copy_buf reference: clear
>   audio->copy_buf at the two sites that transfer ownership to
>   play_queue (f_audio_out_ep_complete inside the spinlock, and
>   f_audio_set_alt alt 0) so playback_work consuming the buffer
>   can no longer leave a stale pointer (sashiko-bot)
> - Simplify f_audio_free drain: no buf == copy_buf comparison
>   needed since copy_buf is now reliably NULL when its buffer
>   is on the queue
> 
> Changes in v3:
> - Guard copy_buf free with IS_ERR_OR_NULL because
>   f_audio_buffer_alloc() returns ERR_PTR on failure (sashiko-bot)
> - Attempted to detect copy_buf during queue drain, but that
>   missed the case where playback_work already consumed it
> 
> Changes in v2:
> - Added play_queue drain and copy_buf free (sashiko-bot)
> 
>  drivers/usb/gadget/function/f_uac1_legacy.c | 11 +++++++++++
>  1 file changed, 11 insertions(+)
> 
> diff --git a/drivers/usb/gadget/function/f_uac1_legacy.c b/drivers/usb/gadget/function/f_uac1_legacy.c
> index 7c8d60c7e87d1..cba7b447b09e3 100644
> --- a/drivers/usb/gadget/function/f_uac1_legacy.c
> +++ b/drivers/usb/gadget/function/f_uac1_legacy.c
> @@ -9,6 +9,7 @@
>  #include <linux/slab.h>
>  #include <linux/kernel.h>
>  #include <linux/module.h>
> +#include <linux/err.h>

Are you sure this is needed?  If so, why here in the list?

>  #include <linux/device.h>
>  #include <linux/atomic.h>
>  
> @@ -338,6 +339,7 @@ static int f_audio_out_ep_complete(struct usb_ep *ep, struct usb_request *req)
>  	if (audio_buf_size - copy_buf->actual < req->actual) {
>  		spin_lock_irq(&audio->lock);
>  		list_add_tail(&copy_buf->list, &audio->play_queue);
> +		audio->copy_buf = NULL;
>  		spin_unlock_irq(&audio->lock);
>  		schedule_work(&audio->playback_work);
>  		copy_buf = f_audio_buffer_alloc(audio_buf_size);
> @@ -670,6 +672,7 @@ static int f_audio_set_alt(struct usb_function *f, unsigned intf, unsigned alt)
>  			if (copy_buf) {
>  				list_add_tail(&copy_buf->list,
>  						&audio->play_queue);
> +				audio->copy_buf = NULL;
>  				schedule_work(&audio->playback_work);
>  			}
>  		}
> @@ -942,7 +945,15 @@ static void f_audio_free(struct usb_function *f)
>  {
>  	struct f_audio *audio = func_to_audio(f);
>  	struct f_uac1_legacy_opts *opts;
> +	struct f_audio_buf *buf, *tmp;
>  
> +	cancel_work_sync(&audio->playback_work);
> +	list_for_each_entry_safe(buf, tmp, &audio->play_queue, list) {
> +		list_del(&buf->list);
> +		f_audio_buffer_free(buf);
> +	}
> +	if (!IS_ERR_OR_NULL(audio->copy_buf))

How can it be an error?  This is usually a sign something else is wrong,
right?

thanks,

greg k-h

      reply	other threads:[~2026-11-20 11:16 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-04 12:53 [PATCH v4] usb: gadget: f_uac1_legacy: fix copy_buf lifetime on queue handoff and free Tristan Madani
2026-11-20 11:16 ` Greg Kroah-Hartman [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=2026112033-gummy-backward-5436@gregkh \
    --to=gregkh@linuxfoundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-usb@vger.kernel.org \
    --cc=nkapron@google.com \
    --cc=stable@vger.kernel.org \
    --cc=tristan@talencesecurity.com \
    --cc=tristmd@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox