netdev - Re: Memory corruption with r8169 across several device revisions and kernels

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: <9ce9eecc-345b-a742-d6f2-eb8253876e5d@01019freenet.de>
Date:   Wed, 24 Jan 2018 08:05:25 +0100
From:   Andreas Hartmann <andihartmann@...19freenet.de>
To:     Oliver Freyermuth <o.freyermuth@...glemail.com>,
        David Miller <davem@...emloft.net>
Cc:     romieu@...zoreil.com, netdev@...r.kernel.org
Subject: Re: Memory corruption with r8169 across several device revisions and
 kernels

On 01/24/2018 at 06:31 AM Andreas Hartmann wrote:
> On 01/23/2018 at 04:47 PM Oliver Freyermuth wrote:
>> Am 23.01.2018 um 16:28 schrieb David Miller:
>>> Looking at how these DMA counters are handled, there appears to be a
>>> requirement that the memory buffer is 64-byte aligned.
>>>
>>> [...]
>>>
>>> Therefore the driver needs to allocate "size + (64 - 1)" bytes and do
>>> the 64-byte alignment of the CPU pointer and the DMA address by hand.
>>
>> This is also what I wondered about as a non-expert in hardware drivers; 
>> alignment should surely be enforced here. 
>>
>> However, for the memory corruption I observed, I used an x86_64 system
>> (which I believe always has PAGE_SIZE aligned buffers). 
>> So there should be another bug, unless I am mistaken about x86_64. 
>>
>> I checked the deprecated r8168 driver by Realtek (I am not sure if this one is also affected by the issue, though)
> 
> I'm using since years this driver because r8169 is broken (it is slow
> and it misses packages - which is extremely bad for real time
> applications like asterisk, if they appear 50s later ...).
> 
> r8168-8.045.08 is an actual version which is provided by realtek on
> their homepage and which even compiles fine w/ 4.14.x.

It just *compiles* w/ 4.14 - but doesn't work here at all. It's running
fine here w/ vanilla 4.4. Realtek officially writes about support up to
4.7. at the moment - but code already knows about 4.11.


Regards,
Andreas