option memmap

option memmap

[Reinhold Jordan]

Hello,

is there any documentation for this option better than this in
linux/Documentation/kernel-parameters.txt ?

I have a laptop with a defect memory soldered on the main-board.
128KB of 128MB are defect started at 7936KB

As I read from kernel-parameters.txt the option
memmap=128K$7936K
reserve this area. But this seems to be simply ignored. Even, if I
fade out 64MB, Knoppix still create a RAM disk, which is larger
as the remaining memory...

Does someone know advice?

Thanks, regards, Reinhold

-- 

                         Reinhold Jordan
WWW: http://reinhold.bachrain.de          Mail: reinhold@bachrain.de
qmail-spam-protection with intelligent-greylisting, reverse-smtp....
qmail.spamfilter.051215.076.tar.bz2 actual rejects 99.995% of spam!!

Re: option memmap

[Reinhold Jordan]

Hello,

is there any problem with this question? Who should I ask
kernel dependent questions beside the kernel developers?

Regards, Reinhold

Reinhold Jordan wrote:
> Hello,
> 
> is there any documentation for this option better than this in
> linux/Documentation/kernel-parameters.txt ?
> 
> I have a laptop with a defect memory soldered on the main-board.
> 128KB of 128MB are defect started at 7936KB
> 
> As I read from kernel-parameters.txt the option
> memmap=128K$7936K
> reserve this area. But this seems to be simply ignored. Even, if I
> fade out 64MB, Knoppix still create a RAM disk, which is larger
> as the remaining memory...
> 
> Does someone know advice?

-- 
ASC telecom AG                   Research & Development
Seibelstr. 2                     F: +49-6021-5001-309
D-63768 Hösbach                  E: r.jordan@asc.de
      Visit us on http://www.asctelecom.com

Re: option memmap

[Ram Gupta]

On 1/12/06, Reinhold Jordan <r.jordan@asc.de> wrote:
> Hello,
>
> is there any problem with this question? Who should I ask
> kernel dependent questions beside the kernel developers?
>
> Regards, Reinhold
>
> Reinhold Jordan wrote:
> > Hello,
> >
> > is there any documentation for this option better than this in
> > linux/Documentation/kernel-parameters.txt ?
> >
> > I have a laptop with a defect memory soldered on the main-board.
> > 128KB of 128MB are defect started at 7936KB
> >

Did you use memmap option in combination with mem option.mem option is
used to specify the amount of memory to be used & memmap specifies the
region map.

regards
Ram Gupta

Re: option memmap

[Randy.Dunlap]

On Wed, 11 Jan 2006 15:48:29 +0100 Reinhold Jordan wrote:

> Hello,
> 
> is there any documentation for this option better than this in
> linux/Documentation/kernel-parameters.txt ?
> 
> I have a laptop with a defect memory soldered on the main-board.
> 128KB of 128MB are defect started at 7936KB
> 
> As I read from kernel-parameters.txt the option
> memmap=128K$7936K
> reserve this area. But this seems to be simply ignored. Even, if I
> fade out 64MB, Knoppix still create a RAM disk, which is larger
> as the remaining memory...
> 
> Does someone know advice?

Please give the complete list of kernel boot options that
you used.

Documentation/kernel-parameters.txt does say (for IA-32)
to use mem= and memmap= together in some cases.

And there are some comments in the source code
(arch/i386/kernel/setup.c) but they probably don't help
any more than the Doc file does:

  * "mem=XXX[kKmM]" defines a memory region from HIGH_MEM
  * to <mem>, overriding the bios size.
  * "memmap=XXX[KkmM]@XXX[KkmM]" defines a memory region from
  * <start> to <start>+<mem>, overriding the bios size.


---
~Randy

Re: option memmap

[Reinhold Jordan]

Hello,

> Did you use memmap option in combination with mem option.mem option is
> used to specify the amount of memory to be used & memmap specifies the
> region map.

thanks for this hint. Now I see a difference between
memmap=128K$7936K mem=130944K
and
mem=130944K memmap=128K$7936K
With mem first, the memmap region is listed as user defined. With memmap
first, it isn't.

But I have to do more tests. In both cases my system crash on heavy load.
After trying the badram patch I'm not sure, that my address ranges are
correct...

Thanks, Reinhold

-- 
ASC telecom AG                   Research & Development
Seibelstr. 2                     F: +49-6021-5001-309
D-63768 Hösbach                  E: r.jordan@asc.de
      Visit us on http://www.asctelecom.com