linux-kernel - Possible double free in iowarrior.ko

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>] [day] [month] [year] [list]

Message-ID: <660deb7a-53d0-133c-02dc-c9b4a0de9ae6@ispras.ru>
Date:   Tue, 22 Aug 2017 17:51:14 +0300
From:   Anton Volkov <avolkov@...ras.ru>
To:     lucht@...emercs.com, johan@...nel.org
Cc:     Greg KH <gregkh@...uxfoundation.org>, linux-usb@...r.kernel.org,
        linux-kernel@...r.kernel.org, ldv-project@...uxtesting.org,
        Alexey Khoroshilov <khoroshilov@...ras.ru>
Subject: Possible double free in iowarrior.ko

Hello.

While searching for races in the Linux kernel I've come across 
"drivers/usb/misc/iowarrior.ko" module. Here are questions that I came 
up with while analyzing results. Lines are given using the info from 
Linux v4.12.

Consider the following case:

Thread 1:                    Thread 2:
iowarrior_release            iowarrior_disconnect
                                mutex_lock(&dev->mutex)
                                dev->present = 0
                                (iowarrior.c: line 889)
   mutex_lock(&dev->mutex)      mutex_unlock(&dev->mutex)
   dev->opened = 0
   (iowarrior.c: line 666)      if(dev->opened){
   if(dev->present){               //dev->opened == 0
     //dev->present ==0
   } else {                     } else {
     mutex_unlock(&dev->mutex)    iowarrior_delete(dev)
     iowarrior_delete(dev)      }
   }

In this case double free of several pointers inside iowarrior_delete 
becomes possible and no calls to usb_kill_urb() and 
wake_up_interruptible() are present. Is this feasible from your point of 
view? If so, maybe it is a good idea to move mutex_unlock(&dev->mutex) 
in iowarrior_disconnect() further down like in iowarrior_release() in 
both 'if' branches?

Thank you for your time

-- Anton Volkov
Linux Verification Center, ISPRAS
web: http://linuxtesting.org
e-mail: avolkov@...ras.ru