freebsd security - Apr 2006 - cvs commit: src/sys/amd64/amd64 mp_machdep.c src/sys/i386/i386 mp_machdep.c

Am 24.04.2006 um 23:17 schrieb Colin Percival:
> cperciva    2006-04-24 21:17:02 UTC
>
>   FreeBSD src repository
>
>   Modified files:
>     sys/amd64/amd64      mp_machdep.c
>     sys/i386/i386        mp_machdep.c
>   Log:
>   Adjust dangerous-shared-cache-detection logic from "all shared data
>   caches are dangerous" to "a shared L1 data cache is
dangerous".
> This
>   is a compromise between paranoia and performance: Unlike the L1  
> cache,
>   nobody has publicly demonstrated a cryptographic side channel which
>   exploits the L2 cache -- this is harder due to the larger size,  
> lower
>   bandwidth, and greater associativity -- and prohibiting shared L2
>   caches turns Intel Core Duo processors into Intel Core Solo  
> processors.
>
>   As before, the 'machdep.hyperthreading_allowed' sysctl will allow
> even
>   the L1 data cache to be shared.
I do not pretend to understand the background, but from your  
description it sounds like performance on Core Duo machines will be  
bad unless this change is made, or the potentially dangerous sysctl  
is active. If that is indeed the case, will this change make it into  
6.1?


Thanks,
Stefan

-- 
Stefan Bethke <stb@lassitu.de>   Fon +49 170 346 0140

Stefan Bethke wrote:
> Am 24.04.2006 um 23:17 schrieb Colin Percival:
>>   FreeBSD src repository
>>
>>   Modified files:
>>     sys/amd64/amd64      mp_machdep.c
>>     sys/i386/i386        mp_machdep.c
> 
> I do not pretend to understand the background, but from your description
> it sounds like performance on Core Duo machines will be bad unless this
> change is made, or the potentially dangerous sysctl is active. If that
> is indeed the case, will this change make it into 6.1?
I'm not part of the release engineering team, but I think they're hoping
to merge this before the release, yes.

If anyone has a Core Duo system running -CURRENT and can tell me what
# sysctl machdep.hyperthreading_allowed
# sysctl machdep.hlt_cpus
says before and after this change, it would probably speed up the process
of merging this into other branches.

Colin Percival