Mailing List Archive

Max. Physical Memory Amount
With a friend I've setup a 4 CPU Compaq Proliant 5500 (Don't pin me down
on that number, I'd have to recheck; it's just some rack mounted high
availability hardware) using a standard 2.0.36 based linux distrib. Well,
we added the new Compaq diskarray driver (had to use a little bit older
diskarray controller, as that is the only one supported), and the machine
now nicely works under linux standalone on a part (48gig) of an external
Raid-5 drive bay.
Problem is, the kernel explicitly refuses to accept more than 1Gig of
main memory.
Now, my question is: Is there a very good reason for this (I can see you
might not get 4G of physical memory for some reasons, but I'd have
expected at least 3G to work.) ?
I checked the source, and it appears to be a simple sanity check in
head.c. Can I savely remove it, what is the highest memory amount I can
hope to work w/o a crash?
I noticed the decent development kernels handel this different. It seems
they would go almost up to 4GB. Is that right?
Even if so, I'd be happy if I could avoid installing yet another kernel on
that machine (I don't want be forced to modify the rest of the
installation due to 2.0 2.2 incompats), so if hacking the 2.0.36 would
work I'd be more happy.
Is there anyone out with an advice to us?
--
Michael Weller: eowmob@exp-math.uni-essen.de, eowmob@ms.exp-math.uni-essen.de,
or even mat42b@spi.power.uni-essen.de. If you encounter an eowmob account on
any machine in the net, it's very likely it's me.
-
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/
Re: Max. Physical Memory Amount [ In reply to ]
Sorry, correction:
> I checked the source, and it appears to be a simple sanity check in
> head.c. Can I savely remove it, what is the highest memory amount I can
> hope to work w/o a crash?
The sanity check is actually in arch/i386/kernel/setup.c .
Michael.
--
Michael Weller: eowmob@exp-math.uni-essen.de, eowmob@ms.exp-math.uni-essen.de,
or even mat42b@spi.power.uni-essen.de. If you encounter an eowmob account on
any machine in the net, it's very likely it's me.
-
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/
Re: Max. Physical Memory Amount [ In reply to ]
> With a friend I've setup a 4 CPU Compaq Proliant 5500 (Don't pin me down
> on that number, I'd have to recheck; it's just some rack mounted high
> availability hardware) using a standard 2.0.36 based linux distrib. Well,
> we added the new Compaq diskarray driver (had to use a little bit older
> diskarray controller, as that is the only one supported), and the machine
> now nicely works under linux standalone on a part (48gig) of an external
> Raid-5 drive bay.
>
> Problem is, the kernel explicitly refuses to accept more than 1Gig of
> main memory.
>
> Now, my question is: Is there a very good reason for this (I can see you
> might not get 4G of physical memory for some reasons, but I'd have
> expected at least 3G to work.) ?
>
> I checked the source, and it appears to be a simple sanity check in
> head.c. Can I savely remove it, what is the highest memory amount I can
> hope to work w/o a crash?
>
> I noticed the decent development kernels handel this different. It seems
> they would go almost up to 4GB. Is that right?
>
> Even if so, I'd be happy if I could avoid installing yet another kernel on
> that machine (I don't want be forced to modify the rest of the
> installation due to 2.0 2.2 incompats), so if hacking the 2.0.36 would
> work I'd be more happy.
>
> Is there anyone out with an advice to us?
>
As far as i've understood, the problem is that Linux uses a single
memory descriptor and 32 bits pointeurs. You then have 4Go total
memory space, and you have to dissociate Physical memory and
Virtual memory. Programs are using virtual memory.
The more you have physical memory, the less you have virtual
memory. In order to let programs have 3Go virtual memory, limit of
physical memory has been set to 1Go.
If you want more memory, you have to review memory managing in
Linux.
christophe
-
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/
Re: Max. Physical Memory Amount [ In reply to ]
On Fri, Jan 15, 1999 at 11:57:32AM +0100, Dr. Michael Weller wrote:
> Now, my question is: Is there a very good reason for this (I can see you
> might not get 4G of physical memory for some reasons, but I'd have
> expected at least 3G to work.) ?
>
> I checked the source, and it appears to be a simple sanity check in
> head.c. Can I savely remove it, what is the highest memory amount I can
> hope to work w/o a crash?
No. The kernel implements a 1GB/3GB phys.mem./virt.mem. split. Max. phys.
mem. supported is 960MB, this way.
> I noticed the decent development kernels handel this different. It seems
> they would go almost up to 4GB. Is that right?
No. But you can read linux/include/asm-i386/page.h from a 2.1/2.2 kernel and
do the changes to this file and linux/arch/i386/vmlinux.lds and recompile
your kernel to support more phys. mem. However, the virtual mem, a process
can use, will shrink. My kernel runs happily with 0x70000000 ...
Regards,
--
Kurt Garloff <kurt@garloff.de> [Dortmund, FRG]
Plasma physics, high perf. computing [Linux-ix86,-axp, DUX]
PGP key on http://www.garloff.de/kurt/ [Linux SCSI driver: DC390]
-
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/