[prev in list] [next in list] [prev in thread] [next in thread] 

List:       oss-security
Subject:    [oss-security] Re: AMD newest ucode 0x06000832 for Piledriver-based CPUs seems to behave in a proble
From:       Robert Święcki <robert () swiecki ! net>
Date:       2016-03-18 14:28:11
Message-ID: CAP145piPtZJb=PqZT6-JJudkfryeyU4YM03D4uFFEDa5rUFF5A () mail ! gmail ! com
[Download RAW message or body]

2016-02-28 5:28 GMT+01:00 Robert Święcki <robert@swiecki.net>:

> AMD newest public ucode 0x06000832 for Piledriver-based CPUs (newer
> AMD FX, and Opteron 3300/4300/6300 series) seems to be broken. Under
> certain conditions it allows unprivileged users running under qemu VMs
> to affect the host Linux kernel in a problematic manner: the CPU
> starts to behave in an erratic way, and it leads to CPU execution flow
> of the host kernel (the one running on bare metal) to be changed.

It seems that AMD (somewhat silently) released - in
https://lkml.org/lkml/2016/3/17/43 - a new microcode for 15th family
of AMD CPUs.

I applied this patch to the previous ucode, and got this -
http://alt.swiecki.net/.a/amd-ucode-20160316.tbz2 - which resulted in:

[1634167.526985] microcode: CPU0: new patch_level=0x0600084f
[1634167.560059] microcode: CPU2: new patch_level=0x0600084f
[1634167.584795] microcode: CPU4: new patch_level=0x0600084f
[1634167.609298] microcode: CPU6: new patch_level=0x0600084f

Quick testing suggests that bugs from 0x06000832 and 0x06000836 ucode
versions are gone. Unfortunately it's not published yet on
http://www.amd64.org/microcode.html nor the new README/errata is
available, so I have no more details on that, but given that AMD
promised new ucode in March fixing this problem - as per
http://www.theregister.co.uk/2016/03/06/amd_microcode_6000836_fix/ -
this might be it.

-- 
Robert Święcki
[prev in list] [next in list] [prev in thread] [next in thread]