login
Header Space

 
 

Re: x86_64 ten times slower than i386

Score:
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
To: Andi Kleen <andi@...>
Cc: Ray Lee <ray-lk@...>, Bo Brantén <bosse@...>, Andrew Morton <akpm@...>, Jesse Barnes <jesse.barnes@...>, <linux-kernel@...>
Date: Monday, November 5, 2007 - 9:19 pm

Andi Kleen wrote:

There really are only two ways to deal with this -- drop the memory 
(which should be automated, and a warning printed) or adjust the MTRRs. 
  The problem is that at some point we run out of MTRRs, partially 
because they're masks instead of base/limit.

Even use of PAT doesn't trivially resolve this issue with less than 
doing MTRR emulation via PAT (setting the default MTRR to WB); however, 
that is bound to cause trouble with SMM.

	-hpa
-
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

Messages in current thread:
Re: x86_64 ten times slower than i386, Andi Kleen, (Mon Nov 5, 8:26 pm)
Re: x86_64 ten times slower than i386, Jesse Barnes, (Wed Nov 7, 2:38 pm)
Re: x86_64 ten times slower than i386, H. Peter Anvin, (Mon Nov 5, 9:19 pm)
Re: x86_64 ten times slower than i386, Willy Tarreau, (Tue Nov 6, 3:40 pm)
Re: x86_64 ten times slower than i386, H. Peter Anvin, (Tue Nov 6, 3:50 pm)
Re: x86_64 ten times slower than i386, Willy Tarreau, (Tue Nov 6, 3:53 pm)
speck-geostationary