linux-kernel - Re: [PATCH v2] Bluetooth: vhci, fix open

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: <56F16A2C.6080303@suse.cz>
Date:	Tue, 22 Mar 2016 16:52:12 +0100
From:	Jiri Slaby <jslaby@...e.cz>
To:	Takashi Iwai <tiwai@...e.de>
Cc:	johan.hedberg@...il.com, linux-kernel@...r.kernel.org,
	Marcel Holtmann <marcel@...tmann.org>,
	Gustavo Padovan <gustavo@...ovan.org>,
	linux-bluetooth@...r.kernel.org,
	Dmitry Vyukov <dvyukov@...gle.com>,
	"stable 3 . 13+" <stable@...r.kernel.org>
Subject: Re: [PATCH v2] Bluetooth: vhci, fix open_timeout vs. hdev race

On 03/22/2016, 03:00 PM, Takashi Iwai wrote:
> int main()
> {
> 	int i;
> 	long fd;
>         pthread_t th[32];
> 
> 	fd = open("/dev/vhci", O_RDWR);
> 	if (fd < 0)
> 		err(1, "open");
>         for (i = 0; i < 32; i++)
> 		pthread_create(&th[i], 0, test, (void*)fd);
>         for (i = 0; i < 32; i++)
> 		pthread_join(th[i], NULL);
> 	return 0;
> }
> ================================================================
> 
> For fixing these, we need a proper mutex protection.  A patch like
> below seems helping, at least for the test case above.

I didn't fix this and kept it simple as I believe concurrent accesses
from a single opener are not allowed by design. But if we want to
support this, we of course need locking like this.

Anyway, I would appreciate any input from the BT fellows. There are more
races like this, apparently.

thanks,
-- 
js
suse labs