So I have a mtrr driver question.
The behavior I observed on my Linux box (with both 2.2.14 and 2.4.0-test2)
is as follow:
I fill all 3 memory slots with 32 Megs DIMMS (total = 96 Megs)
/proc/mtrr shows:
reg00: base=0x00000000 ( 0MB), size= 64MB: write-back, count=1
and the system is as sluggish as possible.
Now, if I type:
echo "base=0x4000000 size=0x2000000 type=write-back" > /proc/mtrr
/proc/mtrr will now shows:
reg00: base=0x00000000 ( 0MB), size= 64MB: write-back, count=1
reg01: base=0x04000000 ( 64MB), size= 32MB: write-back, count=1
and the system will run at its nominal speed.
I tried the same test on a Dell Box, and I've noticed the same /proc/mtrr
behavior. The only difference is the speed itself, the Dell box works at
nominal speed in either cases (i.e. before or after the `echo' command).
So you're gonna ask ``What kind of a system are you running ?''. That's a good
question. In fact the system (the one that's sluggish) is running a homemade
version of BIOS. What I'm suspecting here, is that our internally developped
BIOS doesn't do any setting to the mtrr registers as the Dell BIOS does.
So, even if the Linux mtrr driver doesn't see all 3 DIMMS, the system
behaves correctly anyway.
My question is: is there anything in the mtrr driver that would allow
me to force the configuration of mtrr registers ?
( Needless to say that I'm not sure to understand all the magic of the mtrr
driver ... well, actually I'm sure I don't. )
Regards,
-- Frederic.R.Roussel -o) (o- Lead Software Engineer /\\ Join the penguin force (o_ (o_ //\ Frederic.Roussel@efi.com _\_v The Linux G3N3R47!0N (/)_ (/)_ v_/_
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/