My error dump looks similar to yours, hopefully someone can give us a hint on where to fix this:

<1>Unable to handle kernel paging request at virtual address fffffffe
pgd = c212c000
[fffffffe] *pgd=80002021, *pte=00000000, *ppte=00000000
Internal error: Oops: 17 3]
Modules linked in:
CPU: 0
PC is at dput+0x28/0x218
LR is at link_path_walk+0x118/0x13c
pc : [<c009d0e0>] lr : [<c0094ad0>] Not tainted
sp : c25c3da8 ip : c25c3dc8 fp : c25c3dc4
r10: c2743000 r9 : c25c2000 r8 : c2743000
r7 : c03922a0 r6 : c25c2000 r5 : c24eb998 r4 : fffffffe
r3 : 00000000 r2 : 00000000 r1 : c030f210 r0 : fffffffe
Flags: NzCv IRQs on FIQs on Mode SVC_32 Segment user
Control: C5387D
Table: 8212C000 DAC: 00000015
Process klogd (pid: 800, stack limit = 0xc25c2250)
Stack: (0xc25c3da8 to 0xc25c4000)
3da0: c03922a0 fffffffe c24eb998 c25c3f00 c25c3e3c c25c3dc8
3dc0: c0094ad0 c009d0c4 c24eb998 c03922a0 c0164a50 c24cb8bc c24cb800 00000101
3de0: 00000001 00000000 c260f560 c2743000 c25c2000 4017a000 c25c3f64 c25c3f40
3e00: c0088040 c00a0ca4 c0093628 00000001 00000000 c03d4120 c277d6e0 ffffff9c
3e20: c03d1300 00000001 c25c3f00 00000001 c25c3e6c c25c3e40 c0094f6c c00949c4
3e40: c008b030 c030ed90 ffffff9c ffffffe9 00000001 c25c3f00 c2743000 00000001
3e60: c25c3e8c c25c3e70 c0095a68 c0094ca8 00000000 c25c3f00 ffffff9c 00000004
3e80: c25c3ea4 c25c3e90 c0095b3c c0095a20 00000001 00000000 c25c3ef4 c25c3ea8
3ea0: c0095cb0 c0095b24 00000001 c01c5ad4 000001b6 00000004 c031f7ac c277d6e0
3ec0: c24cb800 00000008 c25c3efc 00000000 c25c3f00 ffffff9c 00000004 c2743000
3ee0: c25c2000 4017a000 c25c3f64 c25c3ef8 c0088460 c0095c4c c25c3f00 c0162a30
3f00: c20bf638 c03922a0 c0164a50 c24cb8bc c24cb800 00000101 00000001 00000000
3f20: c260f560 c2743000 c25c2000 4017a000 c25c3f64 c25c3f40 c0088040 c00a0ca4
3f40: c0093628 00000001 00000000 c03d4120 00000000 000001b6 c25c3f94 c25c3f68
3f60: c00884d4 c0088440 c0039590 00000000 0000ffff 0009e318 00000008 0009e318
3f80: 00000005 c0025f64 c25c3fa4 c25c3f98 c00885a0 c008848c 00000000 c25c3fa8
3fa0: c0025de0 c0088588 0009e318 00000008 4016a5fc 00000000 000001b6 00000002
3fc0: 0009e318 00000008 0009e318 00000005 00000008 00000000 4017a000 bea4772c
3fe0: 000001b6 bea476c0 400cba64 400cb7f0 20000010 4016a5fc 00000000 00000000
Backtrace:
[<c009d0b8>] (dput+0x0/0x218) from [<c0094ad0>] (link_path_walk+0x118/0x13c)
r6 = C25C3F00 r5 = C24EB998 r4 = FFFFFFFE
[<c00949b8>] (link_path_walk+0x0/0x13c) from [<c0094f6c>] (do_path_lookup+0x2d0/0x2f8)
r8 = 00000001 r7 = C25C3F00 r6 = 00000001 r5 = C03D1300 r4 = FFFFFF9C
[<c0094c9c>] (do_path_lookup+0x0/0x2f8) from [<c0095a68>] __path_lookup_intent_open+0x54/0x94)
[<c0095a14>] (__path_lookup_intent_open+0x0/0x94) from [<c0095b3c>] (path_lookup_open+0x24/0x2c)
r7 = 00000004 r6 = FFFFFF9C r5 = C25C3F00 r4 = 00000000
[<c0095b18>] (path_lookup_open+0x0/0x2c) from [<c0095cb0>] (open_namei+0x70/0x6b0)
[<c0095c40>] (open_namei+0x0/0x6b0) from [<c0088460>] (do_filp_open+0x2c/0x4c)
[<c0088434>] (do_filp_open+0x0/0x4c) from [<c00884d4>] (do_sys_open+0x54/0xe8)
r5 = 000001B6 r4 = 00000000
[<c0088480>] (do_sys_open+0x0/0xe8) from [<c00885a0>] (sys_open+0x24/0x28)
r8 = C0025F64 r7 = 00000005 r6 = 0009E318 r5 = 00000008 r4 = 0009E318
[<c008857c>] (sys_open+0x0/0x28) from [<c0025de0>] (ret_fast_syscall+0x0/0x2c)
Code: e3cd6d7f e3c6603f e1a00004 e59f11d8 (e5943000)

Thanks you Randy for your notes.

Unfortunatly for me, I boot up and get uncompress........

and it then just hangs. How can I debug this?

Yes, sometimes I see that too. What I do is reset the device and try again (and again
and again if need be).

what I've noticed is that sometimes when I compile it doesn't seem to like the bits. But if I recompile w/o changing anything it will suddenly be happy and boot up every time. Very touchy (and difficult to engineer/debug)

I noticed other folks don't boot the same way I do, I am not sure how they are doing it so perhaps someone with more linux knowledge can chime in here and suggest a different way of loading the kernel that is more robust.

I also would like to boot with my rootfs in nand, but as you can probably see from my other posts my nand is all trashed for some reason and that makes it impractical. I am not sure how to recover from this or what is the cause. I also have a 2GB type1 compact flash card that I would love to instlal the rootfs on and boot from there (thus avoiding NFS entirely) but it's not clear to me how to go about doing that from the documentation.

As for your problem, make sure your loader compiled correctly with the right values and it's being loaded correctly (right size, etc) that got me for a while until I realized the simple mistake I was making.

Randy

I am booting the way you described earlier and I'm having the same problem (sometimes it just doesn't boot)

Hi folks

I havent seen such errors in my system but I suggest you follow the linux-arm-kernel@lists.arm.linux.org.uk mailing list. there are a lots of experienced people. you can post your questions and get answers.

I have a similar problem too. All I can get it to do is get to the end of uncompressing. Are there any patches available that apply cleanly to the latest Freescale BSP, or a working kernel zImage so that I can see if it's just me being stupid?

anyone could get a stable system?

I am still getting kernel panics.

please reach me at thehexa@gmail.com, since PMs seems to be disabled here

Posted By hexa on 23 May 2007 9:19 PM
anyone could get a stable system?

I am still getting kernel panics.

please reach me at thehexa@gmail.com, since PMs seems to be disabled here

maybe you can log the boot message to find the last complete kernel message before kernel panic happen

first time when I follow the patch step written by randyc , I also got the kernel panic, and I found that it was failed after PMIC driver, so I disable PMIC driver at kernel config, and then boot seems success.

A little experience to share.

Posted By alvinchaung on 23 May 2007 9:19 PM

Posted By hexa on 23 May 2007 9:19 PM
anyone could get a stable system?

I am still getting kernel panics.

please reach me at thehexa@gmail.com, since PMs seems to be disabled here

maybe you can log the boot message to find the last complete kernel message before kernel panic happen

first time when I follow the patch step written by randyc , I also got the kernel panic, and I found that it was failed after PMIC driver, so I disable PMIC driver at kernel config, and then boot seems success.

A little experience to share.

I also got to that conclusion and disabled the PMIC drivers, i can boot the system without problems.

After the system boots, everything seems to be normal, but then, with no particular reason, the kernel panics.

Are you not getting that behaviour?

regards,
Kamakura

Posted By randyc on 23 Apr 2007 11:42 AM

1. Gather the bits needed

download ltib package and mount iso or burn a CD. This came
from the Freescale site:
imx31_ads_20070128-rel3-ltib.iso

Does somebody still have this iso file:
imx31_ads_20070128-rel3-ltib.iso

on freescale there is now imx31_ads_20070511-rel4-ltib.iso and I get a lot of rejects if I try to apply the patch on it.

thanks,
Vali

Hi Ben,

I am facing similare problem as yours. After exec command it hangs. No messages after that. Can you please let me know what procedure you followed to make it work.
Thanks
VD

Hai

We just downloaded unpatched linux-2.6.19 kernel from http://www.kernel.org and patch with http://caiaq.org/download...9.2-mx3lite.patch.gz (~1.1MB)

It works fine

Hello everyone,

First I'd like to thank everyone working on this for all their help, special thanks to Randy for his great instructions.

I've managed to get to the loader/zImage download step, then once I run



exec 0x800d03a8 -



It hangs.



I'll figure it out i expect and post the possible solution but I was curious if someone was able to do it quicker?

Anyone else experience this issue and sort it out?

here's a dump from the terminal



losh> ifconfig sm0 192.168.33.102 255.255.255.0 192.168.33.1

losh> load raw 0x800d03a8 35342 /tftp/192.168.33.105:/tftpboot/loader

loading from /tftp/192.168.33.105:/tftpboot/loader:

loading raw binary to 0x800d03a8 (ram) len 00008a0e:

........done

file loaded

losh> load raw 0x81000000 1629660 /tftp/192.168.33.105:/tftpboot/zImage

loading from /tftp/192.168.33.105:/tftpboot/zImage:

loading raw binary to 0x81000000 (ram) len 0018dddc:

................................................................................

................................................................................

..........................................done

file loaded

losh> ls

R : load 1629660

D : tftp 0

D : etc 0

D : dev 0

losh> exec 0x800d03a8 -

> losh> load raw 0x800d03a8 35342 /tftp/192.168.33.105:/tftpboot/loader

If loader is the executable file as produced by gcc you can use this command instead:
load elf /tftp/192.168.33.105:/tftpboot/loader

You might also want to verify that the start address of loader is indeed 0x800d03a8 by using:
objdump -f /tftpboot/loader

Regards

Posted By mlilja on 13 Feb 2008 9:35 AM
> losh> load raw 0x800d03a8 35342 /tftp/192.168.33.105:/tftpboot/loader

If loader is the executable file as produced by gcc you can use this command instead:
load elf /tftp/192.168.33.105:/tftpboot/loader

You might also want to verify that the start address of loader is indeed 0x800d03a8 by using:
objdump -f /tftpboot/loader

Regards

Hi Magnus and others,
I am unable to build loader as per the directions provided by randyc in posting.php?mode=quote&f=29&p=1074. I am getting "could not find -lc" error.
Can you tell which might cause this problem? LTIB, that i am using is linux-2.6.22.6 (as suggested by magnus, in "yet another linux thread" forum

linux-xfm5:/home/sr/linux_new/bootloader/bootloader # vi Makefile
linux-xfm5:/home/sr/linux_new/bootloader/bootloader # make
/opt/freescale/usr/local/gcc-4.1.2-glibc-2.5-nptl-3/arm-none-linux-gnueabi/bin/arm-none-linux-gnueabi-gcc -c bootloaderstg2.c -I./include
bootloaderstg2.c: In function 'setup_tags':
bootloaderstg2.c:223: warning: passing argument 1 of 'setup_core_tag' makes pointer from integer without a cast
bootloaderstg2.c: In function 'main':
bootloaderstg2.c:250: warning: assignment makes pointer from integer without a cast
/opt/freescale/usr/local/gcc-4.1.2-glibc-2.5-nptl-3/arm-none-linux-gnueabi/bin/arm-none-linux-gnueabi-ld -o loader bootloaderstg2.o -static -T config.lds -lc -L /opt/freescale/usr/local/gcc-4.1.2-glibc-2.5-nptl-3/arm-none-linux-gnueab/lib
/opt/freescale/usr/local/gcc-4.1.2-glibc-2.5-nptl-3/arm-none-linux-gnueabi/bin/arm-none-linux-gnueabi-ld: cannot find -lc
make: *** [loader] Error 1

- Srini

Hi Magnus and others,
I am unable to build loader as per the directions provided by randyc in posting.php?mode=quote&f=29&p=1074. I am getting "could not find -lc" error.
Can you tell which might cause this problem? LTIB, that i am using is linux-2.6.22.6 (as suggested by magnus, in "yet another linux thread" forum

linux-xfm5:/home/sr/linux_new/bootloader/bootloader # vi Makefile
linux-xfm5:/home/sr/linux_new/bootloader/bootloader # make
/opt/freescale/usr/local/gcc-4.1.2-glibc-2.5-nptl-3/arm-none-linux-gnueabi/bin/arm-none-linux-gnueabi-gcc -c bootloaderstg2.c -I./include
bootloaderstg2.c: In function 'setup_tags':
bootloaderstg2.c:223: warning: passing argument 1 of 'setup_core_tag' makes pointer from integer without a cast
bootloaderstg2.c: In function 'main':
bootloaderstg2.c:250: warning: assignment makes pointer from integer without a cast
/opt/freescale/usr/local/gcc-4.1.2-glibc-2.5-nptl-3/arm-none-linux-gnueabi/bin/arm-none-linux-gnueabi-ld -o loader bootloaderstg2.o -static -T config.lds -lc -L /opt/freescale/usr/local/gcc-4.1.2-glibc-2.5-nptl-3/arm-none-linux-gnueab/lib
/opt/freescale/usr/local/gcc-4.1.2-glibc-2.5-nptl-3/arm-none-linux-gnueabi/bin/arm-none-linux-gnueabi-ld: cannot find -lc
make: *** [loader] Error 1

- Srini[/quote]

Hi,
I could resolve this issue. For the benefit of others, I could resolve this by following the lib path mentioned in the following thread

viewtopic.php?f=29&t=1445

Thanks everyone
Srini

What is the second stage bootloader doing? Why do we need it?




I got a kernel image from LinuxLink through a free trial. It gets up and running fine. They also provided a rootfs. Those images are here (they don't provide the source):

https://linuxlink.timesys.com/webshare/2/index.pt/timesys/factory/logic_i.mx31lite_debugging/3/output/

You may have to sign up for a trial.

Here's a getting started guide..I followed most of it. I already had a tftp-server and nfs service

https://linuxlink.timesys.com/node/2722