
freebios/src/northbridge/intel/E7500/northbridge.c In this file there is a variable called remap_high. I assume the function of this is to remap the high memory (512MB) so it is contiguous with the low 512MB. Even when remap_high is true, on -current, I end up with two 512MB segemnts, one at 0 and one at 4 GB. Anybody have any idea as to why this would happen? It's confusing 2.4.19 from what I can see. ron

"Ronald G. Minnich" <rminnich@lanl.gov> writes:
freebios/src/northbridge/intel/E7500/northbridge.c
In this file there is a variable called remap_high. I assume the function of this is to remap the high memory (512MB) so it is contiguous with the low 512MB.
Even when remap_high is true, on -current, I end up with two 512MB segemnts, one at 0 and one at 4 GB.
Anybody have any idea as to why this would happen? It's confusing 2.4.19 from what I can see.
And while I am replying the confusion of 2.4.19 if it is real, is a kernel problem. If you can confirm that the problems go away when all of memory is in 1 1GB segment that would be a very interesting data point. I suspect the kernel simply has some very confusing messages, and no problems will be solved by flipping this option. Eric

Greetings, My understanding is that it is used to accomodate PCI memory mappings. The layout is regular memory, then PCI devices, APICs and flash, then any physical ram above remap_high goes over the 4Gig mark (for PAE). G'day, sjames On Tue, 3 Dec 2002, Ronald G. Minnich wrote:
freebios/src/northbridge/intel/E7500/northbridge.c
In this file there is a variable called remap_high. I assume the function of this is to remap the high memory (512MB) so it is contiguous with the low 512MB.
Even when remap_high is true, on -current, I end up with two 512MB segemnts, one at 0 and one at 4 GB.
Anybody have any idea as to why this would happen? It's confusing 2.4.19 from what I can see.
ron
_______________________________________________ Linuxbios mailing list Linuxbios@clustermatic.org http://www.clustermatic.org/mailman/listinfo/linuxbios
-- -------------------------steven james, director of research, linux labs ... ........ ..... .... 230 peachtree st nw ste 701 the original linux labs atlanta.ga.us 30303 -since 1995 http://www.linuxlabs.com office 404.577.7747 fax 404.577.7743 -----------------------------------------------------------------------

On Wed, 4 Dec 2002, steven james wrote:
My understanding is that it is used to accomodate PCI memory mappings. The layout is regular memory, then PCI devices, APICs and flash, then any physical ram above remap_high goes over the 4Gig mark (for PAE).
yea, I read the sense of that remap_high variable backwards: I thought it meant "remap high memory to low" but it mean "remap low memory to high". Very cool. Anybody remember when 32 bits seemed like a lot of memory? ron

Greetings, Sorta Like when my machine w/ 4 MEG of ram and 120 MEG HD seemed imposible to ever fill up :-) These days, 32 bits is too cramped for some single processes. G'day, sjames On Wed, 4 Dec 2002, Ronald G. Minnich wrote:
On Wed, 4 Dec 2002, steven james wrote:
My understanding is that it is used to accomodate PCI memory mappings. The layout is regular memory, then PCI devices, APICs and flash, then any physical ram above remap_high goes over the 4Gig mark (for PAE).
yea, I read the sense of that remap_high variable backwards: I thought it meant "remap high memory to low" but it mean "remap low memory to high".
Very cool.
Anybody remember when 32 bits seemed like a lot of memory?
ron
-- -------------------------steven james, director of research, linux labs ... ........ ..... .... 230 peachtree st nw ste 701 the original linux labs atlanta.ga.us 30303 -since 1995 http://www.linuxlabs.com office 404.577.7747 fax 404.577.7743 -----------------------------------------------------------------------
participants (3)
-
ebiederman@lnxi.com
-
Ronald G. Minnich
-
steven james