public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Erez Doron <erez@savan.com>
To: "Peter Wächtler" <pwaechtler@loewe-komp.de>
Cc: linux kernel <linux-kernel@vger.kernel.org>
Subject: Re: non volatile ram disk
Date: 22 Jan 2002 11:39:50 +0200	[thread overview]
Message-ID: <1011692390.5447.2.camel@hal.savan.com> (raw)
In-Reply-To: <3C4C3201.8956660F@loewe-komp.de>
In-Reply-To: <1011618928.2825.5.camel@hal.savan.com>  <3C4C1A96.3504174D@loewe-komp.de>  <1011620576.2978.0.camel@hal.savan.com> <1011622206.2978.3.camel@hal.savan.com>  <3C4C3201.8956660F@loewe-komp.de>

On Mon, 2002-01-21 at 17:21, Peter Wächtler wrote:
> Erez Doron schrieb:
> > 
> > the exact log i get:
> > 
> > Creating 3 MTD partitions on "SA1100 flash":
> > 0x00000000-0x00040000 : "bootldr"
> > mtd: Giving out device 0 to bootldr
> > 0x00040000-0x02000000 : "root"
> > mtd: Giving out device 1 to root
> > 0xc2000000-0xc4000000 : "rd"
> > mtd: partition "rd" is out of reach -- disabled
> > 
> > notes:
> > 1. the flash is at physical adress 0-0x1ffffff (32mb)
> > 2. the ram is at physical adress 0xc0000000-0xc3ffffff
> > i tried to map an mtd device to the second part of the ram, but got
> > "partition is out of reach"
> > 
> > any idea ?
> 
> you are confusing the address space of the kernel (starting with 3GB [0xc0000000])
> with the addresses of the RAM (the upper 32MB starts at 0x2000000)
> 
no, i do not ( compaq document says the ram starts at 0xc0000000, i also
verified it by doing peek and poke via the bootloader)

> Then there will be some CS (chip selects) to distinguish what "bus address space"
> to use. Like CS1 for Flash and CS0 for RAM (don't know details of SA1100/iPAQ)
> 
> You have to use the special MTD device that talks to system RAM.
> Again, I don't know the details, but menuconfig gives you:
> 
> RAM/ROM/Flash chip drivers
> 	Support for RAM chips in bus mapping
> 
> Self-contained MTD device drivers
>  	Uncached system RAM (NEW)
> 
> The former sounds like RAM on expansion bus, the latter seems to be
> what you are looking for.
> 
> Now: is the RAM battery backed, or not? If not, you want to use
> the flash as nonvolatile disk - otherwise there would be no need 
> to restart the system and hence a normal RAM disk would be nonvolatile.
> 
yes it is

thanks anyway
erez.
> 
> > 
> > On Mon, 2002-01-21 at 15:42, Erez Doron wrote:
> > > hi
> > >
> > > thanks for replying,
> > >
> > > I already tried to map an MTD to physical memory, but got an error and
> > > an mtd with size 0
> > >
> > > dou you know why ?
> > >
> > > regards
> > > erez
> > >
> > > On Mon, 2002-01-21 at 15:41, Peter Wächtler wrote:
> > > > Erez Doron schrieb:
> > > > >
> > > > > hi
> > > > >
> > > > > I'm looking for a way to make a ramdisk which is not erased on reboot
> > > > > this is for use with ipaq/linux.
> > > > >
> > > > > i tought of booting with mem=32m and map a block device to the rest of
> > > > > the 32M ram i have.
> > > > >
> > > > > the probelm is that giving mem=32m to the kernel will cause the kernel
> > > > > to map only the first 32m of physical memory to virtual one, so using
> > > > > __pa(ptr) on the top 32m causes a kernel oops.
> > > > >
> > > > > any idea ?
> > > > >
> > > >
> > > > a MTD is the way to go, which uses the "reserved" mem area.
> > > > I assume that the RAM is battery backed
> > > > -


  reply	other threads:[~2002-01-22 10:04 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2002-01-21 13:15 non volatile ram disk Erez Doron
2002-01-21 13:41 ` Peter Wächtler
2002-01-21 13:42   ` Erez Doron
2002-01-21 14:10     ` Erez Doron
2002-01-21 15:21       ` Peter Wächtler
2002-01-22  9:39         ` Erez Doron [this message]
2002-01-21 23:54     ` David Woodhouse
2002-01-22 13:53 ` solved: ( was Re: non volatile ram disk) Erez Doron

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1011692390.5447.2.camel@hal.savan.com \
    --to=erez@savan.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pwaechtler@loewe-komp.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox