[lkml]   [2001]   [Jul]   [11]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRE: [ixp1200] questions on memory in linux
a) For memory, one easy solution is to map this memory as unprotected
uncached memory in kernel space. We did exactly this. Look at the
ixp1200 linux kernel on

Look at mm-ixp1200.c, and include/asm/arch/hardware.h.

We mapped the packet memory into virtual memory 0xd8000000 with
cache off and memory protection off. It is accessible to all processes
at that address.

b) This is doable. Cygmon puts the boot string into 0xc0000000. But
currently the linux port does not read this string. You need to change
the head.S in arch/arm/boot/compressed



-----Original Message-----
From: Josh Fryman []
Sent: Tuesday, July 10, 2001 11:44 AM
To: ixp1200@CS.Princeton.EDU;
Subject: [ixp1200] questions on memory in linux


we've got some intel IXP 1200 (spin B0) eval boards, and are running
linux in them. we have some questions about how to do specific things
with (a) memory and (b) the cygmon mini-boot loader...

(a) memory:
the IXP 1200 boards have 32M of SDRAM in two 16MB banks. the
bottom 24MB are used for linux, the rest are free...
however, we want to allocate some memory in that first bank
and *tie* it down so that it's always available at that address
physically to the IXP microengines as well as virtually to the
linux system. can anyone provide pointers to FAQs or other
documents for handling this? eg, some way to do a
malloc( 1MB ) and then get the true physical address of the
malloc results?

(b) the IXP 1200 has a SA-1100 core running the cygmon mini loader from
flash. is there any way to specify kernel-load arguments,
similar to how LILO works, with this system? any pointers to
where we can find out more details for this? the process is to
boot the cygmon from flash, give it the parameters for where to
TFTP the kernel / ramdisk from, and then let it run. we'd like
to force the kernel to run with a "mem=4M" type of setting so
it doesn't suck down our very meager memory for buffers we don't
want/need. (note, if we can do this, the part (a) above doesn't
necessarily go away... it's something we still need to know how
to do...)

thanks for any input,


To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
More majordomo info at
Please read the FAQ at

 \ /
  Last update: 2005-03-22 12:56    [W:0.018 / U:0.316 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site