All of lore.kernel.org
 help / color / mirror / Atom feed
From: Takashi Iwai <tiwai@suse.de>
To: Alan Stern <stern@rowland.harvard.edu>
Cc: Brahmajit Das <listout@listout.xyz>,
	syzbot+f02665daa2abeef4a947@syzkaller.appspotmail.com,
	clemens@ladisch.de, linux-kernel@vger.kernel.org,
	linux-sound@vger.kernel.org, linux-usb@vger.kernel.org,
	perex@perex.cz, syzkaller-bugs@googlegroups.com, tiwai@suse.com
Subject: Re: [PATCH 1/1] ALSA: usb-audio: Avoid NULL dereference in snd_usbmidi_do_output()
Date: Tue, 23 Sep 2025 08:16:45 +0200	[thread overview]
Message-ID: <87plbhn16a.wl-tiwai@suse.de> (raw)
In-Reply-To: <43ab3d0e-4b56-4292-aa51-2473e766dca5@rowland.harvard.edu>

On Tue, 23 Sep 2025 04:40:13 +0200,
Alan Stern wrote:
> 
> On Tue, Sep 23, 2025 at 04:47:20AM +0530, Brahmajit Das wrote:
> > Syzkaller reported a general protection fault in snd_usbmidi_do_output(),
> > caused by dereferencing a NULL URB pointer when accessing
> > ep->urbs[urb_index].urb.
> > 
> > This can happen in rare race conditions where the URB was not initialized
> > or was already freed (e.g. during disconnect or after errors), and the
> > output timer or other path tries to reuse it.
> > 
> > Fix this by checking if the URB is NULL before accessing it, and skipping
> > the current slot if it is.
> > 
> > Reported-by: syzbot+f02665daa2abeef4a947@syzkaller.appspotmail.com
> > Link: https://syzkaller.appspot.com/bug?extid=f02665daa2abeef4a947
> > 
> > Signed-off-by: Brahmajit Das <listout@listout.xyz>
> > ---
> >  sound/usb/midi.c | 5 +++++
> >  1 file changed, 5 insertions(+)
> > 
> > diff --git a/sound/usb/midi.c b/sound/usb/midi.c
> > index acb3bf92857c..7919a39decb4 100644
> > --- a/sound/usb/midi.c
> > +++ b/sound/usb/midi.c
> > @@ -307,6 +307,10 @@ static void snd_usbmidi_do_output(struct snd_usb_midi_out_endpoint *ep)
> >  	for (;;) {
> >  		if (!(ep->active_urbs & (1 << urb_index))) {
> >  			urb = ep->urbs[urb_index].urb;
> > +			if (!urb) {
> > +				// Skip this urb
> > +				goto next_urb;
> > +			}
> 
> What prevents the URB from being freed right here?  If this happens, 
> the code below would access memory that was deallocated.
> 
> To prevent races, you have to use some sort of lock or other 
> synchronization mechanism.  A simple test won't work.

Yes.  The timer instance itself should have been killed before
releasing the resources.

I guess the patch below could cover it, but since I'm traveling, I
can't check it for now.  Will check later once after I'm back.


thanks,

Takashi

--- a/sound/usb/midi.c
+++ b/sound/usb/midi.c
@@ -1511,6 +1511,7 @@ static void snd_usbmidi_free(struct snd_usb_midi *umidi)
 {
 	int i;
 
+	timer_shutdown_sync(&umidi->error_timer);
 	for (i = 0; i < MIDI_MAX_ENDPOINTS; ++i) {
 		struct snd_usb_midi_endpoint *ep = &umidi->endpoints[i];
 		if (ep->out)
@@ -1519,7 +1520,6 @@ static void snd_usbmidi_free(struct snd_usb_midi *umidi)
 			snd_usbmidi_in_endpoint_delete(ep->in);
 	}
 	mutex_destroy(&umidi->mutex);
-	timer_shutdown_sync(&umidi->error_timer);
 	kfree(umidi);
 }
 

  reply	other threads:[~2025-09-23  6:16 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-09-22 16:54 [syzbot] [sound?] [usb?] general protection fault in snd_usbmidi_do_output syzbot
2025-09-22 23:17 ` [PATCH 1/1] ALSA: usb-audio: Avoid NULL dereference in snd_usbmidi_do_output() Brahmajit Das
2025-09-23  2:40   ` Alan Stern
2025-09-23  6:16     ` Takashi Iwai [this message]
2025-09-23  0:48 ` [syzbot] [sound?] [usb?] general protection fault in snd_usbmidi_do_output Hillf Danton
2025-09-23  2:22   ` syzbot
2025-09-23  7:39 ` Hillf Danton
2025-09-23  9:48   ` syzbot

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=87plbhn16a.wl-tiwai@suse.de \
    --to=tiwai@suse.de \
    --cc=clemens@ladisch.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-sound@vger.kernel.org \
    --cc=linux-usb@vger.kernel.org \
    --cc=listout@listout.xyz \
    --cc=perex@perex.cz \
    --cc=stern@rowland.harvard.edu \
    --cc=syzbot+f02665daa2abeef4a947@syzkaller.appspotmail.com \
    --cc=syzkaller-bugs@googlegroups.com \
    --cc=tiwai@suse.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.