Linux USB
 help / color / mirror / Atom feed
From: Michal Pecio <michal.pecio@gmail.com>
To: Edward Adam Davis <eadavis@sina.com>
Cc: daniel@caiaq.org, gregkh@linuxfoundation.org,
	linux-kernel@vger.kernel.org, linux-sound@vger.kernel.org,
	linux-usb@vger.kernel.org, perex@perex.cz,
	syzbot+832ce9fa3face1b7d44d@syzkaller.appspotmail.com,
	syzkaller-bugs@googlegroups.com, tiwai@suse.com, tiwai@suse.de,
	zonque@gmail.com
Subject: Re: [PATCH v2] ALSA: caiaq: Decoupling ep1_in_urb in caiaq dev
Date: Thu, 3 Sep 2026 13:34:08 +0200	[thread overview]
Message-ID: <20260903133408.7da4c35f.michal.pecio@gmail.com> (raw)
In-Reply-To: <20260903100410.541828-1-eadavis@sina.com>

On Thu,  3 Sep 2026 18:04:10 +0800, Edward Adam Davis wrote:
> The epq_in_urb object belonging to the caiaq device is coupled within
> the struct snd_usb_caiaqdev. After usb_submit_urb(epq_in_urb, GFP_KERNEL)
> executes successfully, epq_in_urb is successfully added to the urbp_list
> queue of the dummy HCD driver (userspace specifies dummy_hcd as the HCD
> layer driver for the caiaq USB device).

Sounds like an irrelevant detail. Same crash would happen with any HCD.
 
> When init_card() calls snd_usb_caiaq_send_command() which subsequently
> fails due to a timeout, and proceeds to call snd_card_free() to release
> the card, the embedded ep1_in_urb object is also freed. When the dummy
> HCD driver detects that the URB has been unlinked, it returns the URB
> (by usb_hcd_giveback_urb()), which triggers [1].
> 
> Decouple the ep1_in_urb object from the struct snd_usb_caiaqdev and switch
> to using a pointer instead. Separately allocate and manage the memory for
> ep1_in_urb to prevent the release of the snd_card memory object from
> interfering with it.
> 
> midi_out_urb has the same issue as ep1_in_urb and is handled in the same
> way.
> 
> [1]
> BUG: KASAN: slab-use-after-free in usb_free_urb+0x24/0x120 drivers/usb/core/urb.c:96
> Write of size 4 at addr ffff88803cee1050 by task ktimers/1/29
> Call Trace:
>  usb_free_urb+0x24/0x120 drivers/usb/core/urb.c:96
>  dummy_timer+0xaac/0x4d50 drivers/usb/gadget/udc/dummy_hcd.c:2019
>  __run_hrtimer kernel/time/hrtimer.c:2067 [inline]
>  __hrtimer_run_queues+0x3eb/0xaf0 kernel/time/hrtimer.c:2124
>  hrtimer_run_softirq+0x1e1/0x2e0 kernel/time/hrtimer.c:2141
> 
> Allocated by task 36:
>  snd_card_new+0x7b/0x110 sound/core/init.c:184
>  create_card sound/usb/caiaq/device.c:429 [inline]
>  snd_probe+0x236/0x1af0 sound/usb/caiaq/device.c:544
> 
> Freed by task 36:
>  snd_card_free_when_closed sound/core/init.c:630 [inline]
>  snd_card_free+0x138/0x1d0 sound/core/init.c:662
>  snd_probe+0x162b/0x1af0 sound/usb/caiaq/device.c:553
>  
> Fixes: 523f1dce3743 ("[ALSA] Add Native Instrument usb audio device support")
> Reported-by: syzbot+832ce9fa3face1b7d44d@syzkaller.appspotmail.com
> Closes: https://syzkaller.appspot.com/bug?extid=832ce9fa3face1b7d44d
> Tested-by: syzbot+832ce9fa3face1b7d44d@syzkaller.appspotmail.com
> Signed-off-by: Edward Adam Davis <eadavis@sina.com>
> ---

Does this submission comply with the rules here?
https://docs.kernel.org/process/coding-assistants.html

> v1 -> v2: same dealwith midi_out_urb and add missing check
> 
>  sound/usb/caiaq/device.c | 39 +++++++++++++++++++++++++++------------
>  sound/usb/caiaq/device.h |  4 ++--
>  sound/usb/caiaq/midi.c   |  6 +++---
>  3 files changed, 32 insertions(+), 17 deletions(-)
> 
> diff --git a/sound/usb/caiaq/device.c b/sound/usb/caiaq/device.c
> index a16e59248480..d0fe954b8133 100644
> --- a/sound/usb/caiaq/device.c
> +++ b/sound/usb/caiaq/device.c
> @@ -192,8 +192,8 @@ static void usb_ep1_command_reply_dispatch (struct urb* urb)
>  		break;
>  	}
>  
> -	cdev->ep1_in_urb.actual_length = 0;
> -	ret = usb_submit_urb(&cdev->ep1_in_urb, GFP_ATOMIC);
> +	cdev->ep1_in_urb->actual_length = 0;

That's one of the first things usb_submit_urb() below would do,
and has done since forever.

> +	ret = usb_submit_urb(cdev->ep1_in_urb, GFP_ATOMIC);
>  	if (ret < 0)
>  		dev_err(dev, "unable to submit urb. OOM!?\n");
>  }
> @@ -408,6 +408,10 @@ static void card_free(struct snd_card *card)
>  #endif
>  	snd_usb_caiaq_audio_free(cdev);
>  	usb_put_dev(cdev->chip.dev);
> +	usb_free_urb(cdev->ep1_in_urb);
> +	cdev->ep1_in_urb = NULL;
> +	usb_free_urb(cdev->midi_out_urb);
> +	cdev->midi_out_urb = NULL;

Is clearing this necessary? (I'm not familiar with ALSA)

>  }
>  
>  static int create_card(struct usb_device *usb_dev,
> @@ -457,22 +461,33 @@ static int init_card(struct snd_usb_caiaqdev *cdev)
>  		return -EIO;
>  	}
>  
> -	usb_init_urb(&cdev->ep1_in_urb);
> -	usb_init_urb(&cdev->midi_out_urb);
> +	cdev->ep1_in_urb = usb_alloc_urb(0, GFP_KERNEL);
> +	if (!cdev->ep1_in_urb) {
> +		dev_err(dev, "alloc ep1_in_urb failed.\n");

AFAIK logging on allocation failure is frowned upon.

> +		return -ENOMEM;
> +	}
> +	cdev->midi_out_urb = usb_alloc_urb(0, GFP_KERNEL);
> +	if (!cdev->midi_out_urb) {
> +		usb_free_urb(cdev->ep1_in_urb);
> +		dev_err(dev, "alloc midi_out_urb failed.\n");
> +		return -ENOMEM;
> +	}
> +	usb_init_urb(cdev->ep1_in_urb);
> +	usb_init_urb(cdev->midi_out_urb);

And what would be the point of that?

>  
> -	usb_fill_bulk_urb(&cdev->ep1_in_urb, usb_dev,
> +	usb_fill_bulk_urb(cdev->ep1_in_urb, usb_dev,
>  			  usb_rcvbulkpipe(usb_dev, 0x1),
>  			  cdev->ep1_in_buf, EP1_BUFSIZE,
>  			  usb_ep1_command_reply_dispatch, cdev);
>  
> -	usb_fill_bulk_urb(&cdev->midi_out_urb, usb_dev,
> +	usb_fill_bulk_urb(cdev->midi_out_urb, usb_dev,
>  			  usb_sndbulkpipe(usb_dev, 0x1),
>  			  cdev->midi_out_buf, EP1_BUFSIZE,
>  			  snd_usb_caiaq_midi_output_done, cdev);
>  
>  	/* sanity checks of EPs before actually submitting */
> -	if (usb_urb_ep_type_check(&cdev->ep1_in_urb) ||
> -	    usb_urb_ep_type_check(&cdev->midi_out_urb)) {
> +	if (usb_urb_ep_type_check(cdev->ep1_in_urb) ||
> +	    usb_urb_ep_type_check(cdev->midi_out_urb)) {
>  		dev_err(dev, "invalid EPs\n");
>  		return -EINVAL;
>  	}
> @@ -480,7 +495,7 @@ static int init_card(struct snd_usb_caiaqdev *cdev)
>  	init_waitqueue_head(&cdev->ep1_wait_queue);
>  	init_waitqueue_head(&cdev->prepare_wait_queue);
>  
> -	if (usb_submit_urb(&cdev->ep1_in_urb, GFP_KERNEL) != 0)
> +	if (usb_submit_urb(cdev->ep1_in_urb, GFP_KERNEL) != 0)
>  		return -EIO;
>  
>  	err = snd_usb_caiaq_send_command(cdev, EP1_CMD_GET_DEVICE_INFO, NULL, 0);
> @@ -530,7 +545,7 @@ static int init_card(struct snd_usb_caiaqdev *cdev)
>  	return 0;
>  
>   err_kill_urb:
> -	usb_kill_urb(&cdev->ep1_in_urb);
> +	usb_kill_urb(cdev->ep1_in_urb);
>  	return err;
>  }
>  
> @@ -576,8 +591,8 @@ static void snd_disconnect(struct usb_interface *intf)
>  #endif
>  	snd_usb_caiaq_audio_disconnect(cdev);
>  
> -	usb_kill_urb(&cdev->ep1_in_urb);
> -	usb_kill_urb(&cdev->midi_out_urb);
> +	usb_kill_urb(cdev->ep1_in_urb);
> +	usb_kill_urb(cdev->midi_out_urb);
>  
>  	snd_card_free_when_closed(card);
>  }
> diff --git a/sound/usb/caiaq/device.h b/sound/usb/caiaq/device.h
> index 743eb0387b5f..1c6f34693fa8 100644
> --- a/sound/usb/caiaq/device.h
> +++ b/sound/usb/caiaq/device.h
> @@ -60,8 +60,8 @@ struct snd_usb_caiaq_cb_info;
>  struct snd_usb_caiaqdev {
>  	struct snd_usb_audio chip;
>  
> -	struct urb ep1_in_urb;
> -	struct urb midi_out_urb;
> +	struct urb *ep1_in_urb;
> +	struct urb *midi_out_urb;
>  	struct urb **data_urbs_in;
>  	struct urb **data_urbs_out;
>  	struct snd_usb_caiaq_cb_info *data_cb_info;
> diff --git a/sound/usb/caiaq/midi.c b/sound/usb/caiaq/midi.c
> index c656d0162432..18529484c8dc 100644
> --- a/sound/usb/caiaq/midi.c
> +++ b/sound/usb/caiaq/midi.c
> @@ -43,7 +43,7 @@ static int snd_usb_caiaq_midi_output_close(struct snd_rawmidi_substream *substre
>  {
>  	struct snd_usb_caiaqdev *cdev = substream->rmidi->private_data;
>  	if (cdev->midi_out_active) {
> -		usb_kill_urb(&cdev->midi_out_urb);
> +		usb_kill_urb(cdev->midi_out_urb);
>  		cdev->midi_out_active = 0;
>  	}
>  	return 0;
> @@ -64,9 +64,9 @@ static void snd_usb_caiaq_midi_send(struct snd_usb_caiaqdev *cdev,
>  		return;
>  
>  	cdev->midi_out_buf[2] = len;
> -	cdev->midi_out_urb.transfer_buffer_length = len+3;
> +	cdev->midi_out_urb->transfer_buffer_length = len+3;
>  
> -	ret = usb_submit_urb(&cdev->midi_out_urb, GFP_ATOMIC);
> +	ret = usb_submit_urb(cdev->midi_out_urb, GFP_ATOMIC);
>  	if (ret < 0)
>  		dev_err(dev,
>  			"snd_usb_caiaq_midi_send(%p): usb_submit_urb() failed,"
> -- 
> 2.43.0
> 

  reply	other threads:[~2026-09-03 11:34 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <6a6e9502.f794c993.27aeb.0018.GAE@google.com>
2026-09-02  5:09 ` [syzbot] [usb?] KASAN: slab-use-after-free Write in usb_free_urb syzbot
2026-09-03  8:47   ` [PATCH] ALSA: caiaq: Decoupling ep1_in_urb in caiaq dev Edward Adam Davis
2026-09-03  9:32     ` Takashi Iwai
2026-09-03  9:41       ` Edward Adam Davis
2026-09-03 10:03         ` Takashi Iwai
2026-09-03 11:56           ` [PATCH v3] " Edward Adam Davis
2026-09-03 13:05             ` [PATCH v4] " Edward Adam Davis
2026-09-03 10:04         ` [PATCH v2] " Edward Adam Davis
2026-09-03 11:34           ` Michal Pecio [this message]
2026-09-03  9:24   ` [syzbot] [usb?] KASAN: slab-use-after-free Write in usb_free_urb Michal Pecio
2026-09-03  9:44     ` Takashi Iwai
2026-09-03 11:07       ` DIY allocation or embedding of URBs in larger structures Michal Pecio
2026-09-03 14:37         ` Takashi Iwai
2026-09-03  9:55     ` [syzbot] [usb?] KASAN: slab-use-after-free Write in usb_free_urb Hillf Danton
2026-09-03 10:01     ` 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=20260903133408.7da4c35f.michal.pecio@gmail.com \
    --to=michal.pecio@gmail.com \
    --cc=daniel@caiaq.org \
    --cc=eadavis@sina.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-sound@vger.kernel.org \
    --cc=linux-usb@vger.kernel.org \
    --cc=perex@perex.cz \
    --cc=syzbot+832ce9fa3face1b7d44d@syzkaller.appspotmail.com \
    --cc=syzkaller-bugs@googlegroups.com \
    --cc=tiwai@suse.com \
    --cc=tiwai@suse.de \
    --cc=zonque@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