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] [day] [month] [year] [list]
Date:   Thu, 15 Sep 2022 22:03:33 -0700
From:   Hyunwoo Kim <imv4bel@...il.com>
To:     Dominik Brodowski <linux@...inikbrodowski.net>
Cc:     Arnd Bergmann <arnd@...db.de>, laforge@...monks.org,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>,
        linux-kernel@...r.kernel.org, Paul Fulghum <paulkf@...rogate.com>,
        akpm@...l.org, Lubomir Rintel <lkundrak@...sk>, imv4bel@...il.com
Subject: Re: [PATCH] pcmcia: synclink_cs: Fix use-after-free in mgslpc_ioctl()

The previous mailing list is here:
https://lore.kernel.org/lkml/20220913052020.GA85241@ubuntu/#r


Dear all,


I reported that the scr24x_cs driver does not cause UAF, which is incorrect.

UAFs can also occur in the scr24x_cs driver in the following order:
```
                cpu0                                                cpu1
       1. open()
                                                             2. scr24x_remove()
                                                                device_destroy()
                                                                cdev_del()
                                                                kref_put()
       3. scr24x_open()
          kref_get()   <- refcount_t: addition on 0;
                                                             4. scr24x_delete()
                                                                kfree(dev);
       5. scr24x_read()    <- UAF
```
Because this driver uses kref_init, it looks as if no UAF is happening, but it does.

Since there is no lock between .open and .remove, kref_get() is called after kref_put() 
is called, so "refcount_t: addition on 0;" This is what happens.


So I submitted a patch for all drivers in this path:
- synclink_cs.c "v2" patch : https://lore.kernel.org/lkml/20220916045734.GA187909@ubuntu/T/#u
- cm4040_cs.c : https://lore.kernel.org/lkml/20220916045834.GA188033@ubuntu/T/#u
- cm4000_cs.c : https://lore.kernel.org/lkml/20220916045929.GA188153@ubuntu/T/#u
- scr24x_cs.c : https://lore.kernel.org/lkml/20220916050006.GA188273@ubuntu/T/#u


Best Regards,
Hyunwoo Kim.

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ