[Buildroot] booting JFFS2 or UBIFS created with buildroot

[Charles Krinke]

Dear Thomas:

This gives me a couple of clues.

I also see that there is no mention at all of JFFS2 in
u-boot-2010.06/include/configs/MPC8323ERDB.h, so I added:

#define CONFIG_JFFS2_CMDLINE
#define MTDIDS_DEFAULT "nor0"
#define MTDPARTS_DEFAULT "mtdparts=nor:1152k(U-Boot),2176k(kernel),-(JFFS2)

And this gets me to a buildroot question.

When I change this file in the output/build/u-boot-2010.06 directory, I
would normally expect 'make' to find it, but it does not make a new u-boot.

"What is the proper incantation from the top of the buildroot tree to make a
new u-boot and can I do a clean in just the u-boot directory also".

I am trying to avoid make clean at the top of the buildroot tree so I dont
get all the sources again from the repositories *and* build the toolchain
again.

Additionally, on the Linux question, I *believe* there is indeed no
description of the NOR flash in the device tree source (.dts file) from
Freescale and maybe that is at the root of the problem.

I did look at the dotconfig and I can see that CONFIG_MTD=y and
CONFIG_MTD_PARTITIONS=y are both defined.

Unfortunately, I am only a tiny bit familiar with device tree and dont yet
know what the exact syntax is for adding NOR flash to the .dts file, but
that becomes my next area of research.

Charles

On Mon, Feb 21, 2011 at 2:42 PM, Thomas Petazzoni <
thomas.petazzoni at free-electrons.com> wrote:

> Hello,
>
> On Mon, 21 Feb 2011 14:29:35 -0800
> Charles Krinke <charles.krinke at gmail.com> wrote:
>
> > First of all, I appreciate greatly the time you are taking to discuss
> this
> > issue. Thank you very much.
>
> You're welcome!
>
> > My flash layout is:
> >
> > 0xFE340000 to top of flash rootfs.jffs2 programmed here
> > 0xFE120000 uImage here
> > 0xFE080000 deviceTree here (1 sector)
> > 0xFE000000 u-boot here
> >
> > Now with that said, here are my recipes assuming I counted the number of
> > 128k sectors properly for the offset to get to JFFS2 in the bootargs
> line.
>
> Ok, sounds correct.
>
> > Linux version 2.6.36.1 (ckrinke at Yin) (gcc version 4.3.5 (Buildroot
> 2010.11)
> > ) #1 Thu Feb 17 19:04:48 PST 2011
>
> The problem seems to be in your kernel: there are no lines indicating
> that a MTD device has been detected. So maybe the MTD device is not
> described in your device tree, or improperly described, or the
> corresponding kernel driver is missing.
>
> Here are the messages you typically see when a NAND device is being
> detected by the kernel (on an ARM AT91 platform) :
>
> NAND device: Manufacturer ID: 0xec, Chip ID: 0xd3 (Samsung NAND 1GiB 3,3V
> 8-bit)
> Scanning device for bad blocks
> Bad eraseblock 464 at 0x03a00000
> Bad eraseblock 1694 at 0x0d3c0000
> Bad eraseblock 3101 at 0x183a0000
> Bad eraseblock 3199 at 0x18fe0000
> Bad eraseblock 3866 at 0x1e340000
> Bad eraseblock 5478 at 0x2acc0000
> Bad eraseblock 6473 at 0x32920000
> Bad eraseblock 6606 at 0x339c0000
> 3 cmdlinepart partitions found on MTD device atmel_nand
> Creating 3 MTD partitions on "atmel_nand":
> 0x00000000-0x00100000 : "bootloader"
> 0x00100000-0x00400000 : "kernel"
> 0x00400000-0x40000000 : "rootfs"
>
> Thomas
> --
> Thomas Petazzoni, Free Electrons
> Kernel, drivers, real-time and embedded Linux
> development, consulting, training and support.
> http://free-electrons.com
> _______________________________________________
> buildroot mailing list
> buildroot at busybox.net
> http://lists.busybox.net/mailman/listinfo/buildroot
>



-- 
Charles Krinke
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20110221/e8eca693/attachment-0002.html>