linux-kernel - Re: A reliable kernel panic (3.6.2) and system crash when visiting a particular website

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <s5hvcdgfv49.wl%tiwai@suse.de>
Date:	Thu, 08 Nov 2012 07:48:22 +0100
From:	Takashi Iwai <tiwai@...e.de>
To:	Alan Stern <stern@...land.harvard.edu>
Cc:	Christof Meerwald <cmeerw@...erw.org>,
	Daniel Mack <zonque@...il.com>,
	"Artem S. Tashkinov" <t.artem@...os.com>,
	Kernel development list <linux-kernel@...r.kernel.org>,
	USB list <linux-usb@...r.kernel.org>
Subject: Re: A reliable kernel panic (3.6.2) and system crash when visiting a particular website

At Wed, 7 Nov 2012 15:37:17 -0500 (EST),
Alan Stern wrote:
> 
> On Wed, 7 Nov 2012, Takashi Iwai wrote:
> 
> > > What is the right solution for this problem?
> > 
> > How about the patch below?  (It's for 3.6, and won't be applied cleanly
> > to 3.7, but easy to adapt.)
> 
> I simplified your patch a little.

You can't drop the check of stopping endpoint.  As Daniel pointed,
endpoints might be still running when it's called.  I already did a
similar failure in the past, so this patch is a revised version with
the check for pending operations.

>  This is for 3.7, not 3.6.  I 
> verified that it does fix the problem raised by the test program.
> 
> If you think this is okay, I'll submit it officially.

Don't worry, my patch is also based on 3.7, too :)  3.6 patch was
provided just for convenience, testers seemed to have 3.6 systems.


thanks,

Takashi

> 
> Alan Stern
> 
> 
> 
> Index: usb-3.7/sound/usb/endpoint.h
> ===================================================================
> --- usb-3.7.orig/sound/usb/endpoint.h
> +++ usb-3.7/sound/usb/endpoint.h
> @@ -19,6 +19,7 @@ int snd_usb_endpoint_set_params(struct s
>  int  snd_usb_endpoint_start(struct snd_usb_endpoint *ep, int can_sleep);
>  void snd_usb_endpoint_stop(struct snd_usb_endpoint *ep,
>  			   int force, int can_sleep, int wait);
> +void snd_usb_endpoint_sync_stop(struct snd_usb_endpoint *ep);
>  int  snd_usb_endpoint_activate(struct snd_usb_endpoint *ep);
>  int  snd_usb_endpoint_deactivate(struct snd_usb_endpoint *ep);
>  void snd_usb_endpoint_free(struct list_head *head);
> Index: usb-3.7/sound/usb/pcm.c
> ===================================================================
> --- usb-3.7.orig/sound/usb/pcm.c
> +++ usb-3.7/sound/usb/pcm.c
> @@ -576,6 +576,11 @@ static int snd_usb_pcm_prepare(struct sn
>  		subs->need_setup_ep = false;
>  	}
>  
> +	if (subs->sync_endpoint)
> +		snd_usb_endpoint_sync_stop(subs->sync_endpoint);
> +	if (subs->data_endpoint)
> +		snd_usb_endpoint_sync_stop(subs->data_endpoint);
> +
>  	/* some unit conversions in runtime */
>  	subs->data_endpoint->maxframesize =
>  		bytes_to_frames(runtime, subs->data_endpoint->maxpacksize);
> Index: usb-3.7/sound/usb/endpoint.c
> ===================================================================
> --- usb-3.7.orig/sound/usb/endpoint.c
> +++ usb-3.7/sound/usb/endpoint.c
> @@ -481,7 +481,7 @@ __exit_unlock:
>  /*
>   *  wait until all urbs are processed.
>   */
> -static int wait_clear_urbs(struct snd_usb_endpoint *ep)
> +void snd_usb_endpoint_sync_stop(struct snd_usb_endpoint *ep)
>  {
>  	unsigned long end_time = jiffies + msecs_to_jiffies(1000);
>  	unsigned int i;
> @@ -502,8 +502,6 @@ static int wait_clear_urbs(struct snd_us
>  	if (alive)
>  		snd_printk(KERN_ERR "timeout: still %d active urbs on EP #%x\n",
>  					alive, ep->ep_num);
> -
> -	return 0;
>  }
>  
>  /*
> @@ -556,7 +554,7 @@ static void release_urbs(struct snd_usb_
>  
>  	/* stop urbs */
>  	deactivate_urbs(ep, force, 1);
> -	wait_clear_urbs(ep);
> +	snd_usb_endpoint_sync_stop(ep);
>  
>  	for (i = 0; i < ep->nurbs; i++)
>  		release_urb_ctx(&ep->urb[i]);
> @@ -833,7 +831,7 @@ int snd_usb_endpoint_start(struct snd_us
>  	/* just to be sure */
>  	deactivate_urbs(ep, 0, can_sleep);
>  	if (can_sleep)
> -		wait_clear_urbs(ep);
> +		snd_usb_endpoint_sync_stop(ep);
>  
>  	ep->active_mask = 0;
>  	ep->unlink_mask = 0;
> @@ -917,7 +915,7 @@ void snd_usb_endpoint_stop(struct snd_us
>  		ep->prepare_data_urb = NULL;
>  
>  		if (wait)
> -			wait_clear_urbs(ep);
> +			snd_usb_endpoint_sync_stop(ep);
>  	}
>  }
>  
> @@ -940,7 +938,7 @@ int snd_usb_endpoint_deactivate(struct s
>  		return -EINVAL;
>  
>  	deactivate_urbs(ep, 1, 1);
> -	wait_clear_urbs(ep);
> +	snd_usb_endpoint_sync_stop(ep);
>  
>  	if (ep->use_count != 0)
>  		return 0;
> 
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/