linux-kernel - Re: amd64 + mtrr: only 3.1gb of 8gb are covered, kernel 2.6.30 and earlier.

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <200908270313.32528.volkerarmin@googlemail.com>
Date:	Thu, 27 Aug 2009 03:13:32 +0200
From:	Volker Armin Hemmann <volkerarmin@...glemail.com>
To:	Yinghai Lu <yhlu.kernel@...il.com>, linux-kernel@...r.kernel.org
Subject: Re: amd64 + mtrr: only 3.1gb of 8gb are covered, kernel 2.6.30 and earlier.

Hi,

On Sonntag 23 August 2009, Yinghai Lu wrote:
> On Sat, Aug 22, 2009 at 12:26 PM, Volker Armin
>
> Hemmann<volkerarmin@...glemail.com> wrote:
> > when I went from 4gb to 6gb I had a performance degradation. Since the
> > 8gb are faster and the new cpu is faster, I don't see degradation. But I
> > wonder how much better performance could be with 'correct' mtrr.
>
> that is weird. with 4g should have mem hole remapping.
>
> do you still have /proc/mtrr for 6g setup?
>
> YH

some more information, mtrr of 8gb with my X2 6000:

reg00: base=0x000000000 (    0MB), size= 2048MB, count=1: write-back
reg01: base=0x080000000 ( 2048MB), size= 1024MB, count=1: write-back
reg02: base=0x0c0000000 ( 3072MB), size=  128MB, count=1: write-back
reg03: base=0x0c8000000 ( 3200MB), size=   64MB, count=1: write-back
reg04: base=0x0cc000000 ( 3264MB), size=   32MB, count=1: write-back
reg05: base=0x0ce000000 ( 3296MB), size=   16MB, count=1: write-back

from June with 2.6.30, dmesg from that incident is attached

compared that with the mtrr with my X4:

reg00: base=0x000000000 (    0MB), size= 2048MB, count=1: write-back
reg01: base=0x080000000 ( 2048MB), size= 1024MB, count=1: write-back
reg02: base=0x0c0000000 ( 3072MB), size=  128MB, count=1: write-back


just ignore the fgrlx parts - whenever I checked it did not made a difference 
if X was started or not.

View attachment "dmesg_8gb_x2_6000" of type "text/plain" (64742 bytes)

View attachment "dmesg_8gb_x4_955" of type "text/plain" (45621 bytes)