On 29 Jun 2014 at 8:54, Michael D. Setzer II wrote:

From:                      "Michael D. Setzer II" <[email protected]>
To:                          [email protected]
Date sent:              Sun, 29 Jun 2014 08:54:38 +1000
Subject:                  Re: LZO security bug might affect Busybox
Priority:                   normal

> On 28 Jun 2014 at 6:33, Isaac Dunham wrote:
>
> Date sent:            Sat, 28 Jun 2014 06:33:35 -0700
> From:                   Isaac Dunham <[email protected]>
> To:                       [email protected]
> Subject:               LZO security bug might affect Busybox
>
> > There's an integer overflow in LZO (LMS-2014-06-16-1):
> > http://www.openwall.com/lists/oss-security/2014/06/26/20
> >
> > I suspect that this affects Busybox; the code would be in
> > archival/libarchive/lzo1x_d.c
> > But I wouldn't be able to verify that or to fix it.
> >
>
> Couple of items.
> From http://www.oberhumer.com/opensource/lzo/
>
> Version 2.07
> 25 Jun 2014
>
> Copyright (C) 1996 - 2014
> Markus F.X.J. Oberhumer
>
> News
>
>     LZO 2.07 has been released:
>
>     Fixed a potential integer overflow condition in the "safe"
>     decompressor
> variants which could result in a possible buffer overrun when
> processing maliciously crafted compressed input data.
>
>     Fortunately this issue only affects 32-bit systems and also can
>     only
> happen if you use uncommonly huge buffer sizes where you have to
> decompress more than 16 MiB (> 2^24 bytes) untrusted compressed bytes
> within a single function call, so the practical implications are
> limited.
>
>     POTENTIAL SECURITY ISSUE. But then, I personally do not know about
>    
> any client program that actually is affected.
>
>     TL;DR: the Linux kernel is *not* affected; media hype.
>
> From my understanding, someone has to make a file to produce the
> problem. I did download and compile the new LZO 2.07, and it creates a
> new library file, but it has the .la extention instead of the .so
> extension that the standard lzop is linked with.
>
> My G4L project uses the lzop within busybox as the default compression
> program, and have never had an issue, but am interested in elimanating
> any potential issues.
>

Just got two replies from the author of lzo.

lzop 1.03 uses a hard-coded BLOCK_SIZE of 256 KiB, so it is not vulnerable.

I don't know about the busybox implementation, though.

Best regards,
Markus

The other one was to create the new libraries.
./configure --enable-shared
make
make install
Seems the libraries are in src/.libs directory
and it installs them in /usr/local/lib
but my distro has the libraries in /usr/lib64
So would probable need to copy or link the new ones.

Believe the busybox lzop is version 1.02, so don't know if it is not affected or not?


>
>
> > Thanks,
> > Isaac Dunham
> > _______________________________________________
> > busybox mailing list
> > [email protected]
> > http://lists.busybox.net/mailman/listinfo/busybox
>
>
> +----------------------------------------------------------+
>   Michael D. Setzer II -  Computer Science Instructor     
>   Guam Community College  Computer Center                 
>   mailto:[email protected]                           
>   mailto:[email protected]
>   http://www.guam.net/home/mikes
>   Guam - Where America's Day Begins                       
>   G4L Disk Imaging Project maintainer
>   http://sourceforge.net/projects/g4l/
> +----------------------------------------------------------+
>
> http://setiathome.berkeley.edu (Original)
> Number of Seti Units Returned:  19,471
> Processing time:  32 years, 290 days, 12 hours, 58 minutes
> (Total Hours: 287,489)
>
> BOINC@HOME CREDITS
> ROSETTA     16875087.457129   |   SETI        28014256.354595
> ABC         16613838.513356   |   EINSTEIN    26934147.994939
>
> _______________________________________________
> busybox mailing list
> [email protected]
> http://lists.busybox.net/mailman/listinfo/busybox

  
_______________________________________________
busybox mailing list
[email protected]
http://lists.busybox.net/mailman/listinfo/busybox

Reply via email to