Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 02 Jul 2007 12:35:43 -0400
From:      Chuck Swiger <cswiger@mac.com>
To:        Suleiman Souhlal <ssouhlal@FreeBSD.org>
Cc:        Takeharu KATO <takeharu1219@ybb.ne.jp>, current@freebsd.org
Subject:   Re: [PATCH] Machine Check Architecture on amd64
Message-ID:  <4689295F.10109@mac.com>
In-Reply-To: <27D8F2CD-156E-4962-A187-63E42CAE2B6F@FreeBSD.org>
References:  <46806B3E.2060701@FreeBSD.org> <4688EBC2.8070703@ybb.ne.jp> <27D8F2CD-156E-4962-A187-63E42CAE2B6F@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
Suleiman Souhlal wrote:
> On Jul 2, 2007, at 5:12 AM, Takeharu KATO wrote:
>> Apparently, the patch does not have memory scrub facility.
>> Do you have a plan to implement the memory scrub facility like solaris 
>> does?
>>
>> P.S. In fact, I was also trying to implement the MCE facility, so I am 
>> interesting to your patch.
> 
> While I don't know about Intel CPUs, K8 CPUs from AMD have a hardware 
> DRAM/data cache scrubber, so I didn't bother implementing it, and I 
> don't really plan to, at least in the near future. It should, however, 
> be pretty easy to implement.

Agreed, at least some BIOSes have an advanced memory tuning page somewhere 
which lets you enable or disable the BG cache memory scrubbing on AMD64 chips.

I'm at least as interested in getting FreeBSD-level support for loading CPU 
microcode into a running system, as I've got a bunch of remote HP or Dell PE 
boxes floating around in racks somewhere.  Getting them to boot into a 
DOS/Windows env to reflash their BIOS or update something like a Intel or 
Broadcomm NIC to disable their remote management or whatever is eating traffic 
to be more of a hassle than a convenience even when one is not dealing with 
remote systems.

(By this I mean to say, thanks for the work on this area.  :-)

-- 
-Chuck



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4689295F.10109>