Sl-5500 Right Arrow Shuts Off Options

[static]

hello. my sl-5500 shuts off when i hold down the right arrow for more than a few seconds. i've come across a few posts and internet sites that i think will fix my problem but i've never made my own kernel for my zaurus and i'm afraid i might brick it.

this is a patch for the kernel that is supposed to fix the problem:

http://www.oesf.org/index.php?title=Marcan_fix

i'm using this cross compiler:

http://www.tyrannozaurus.com/feed/contribs...-2.95.3.tar.bz2

and i'm using this kernel:

http://www.tyrannozaurus.com/feed/contribs...rom3_10.tar.bz2

and i'm trying to follow these directions but they don't explain what to do after "make dep":

http://www.tyrannozaurus.com/feed/contribs...ides/kernel.txt

i did a make and wound up with a vmlinux that's about 2 megs. which sort of frightens me because the pdaxrom image that i flashed my z with was more like 1 meg. but the compilation process went smoothly. i'm pretty sure the kernel was compiled correctly.

is it safe for me to rename vmlinux to zImage and reflash with it and the pdaxrom beta1 initrd.bin?

also, and please forgive my ignorance on this, but do i need to compile new modules? the kernel versions on the old and new kernel appear to be exactly the same (2.4.18-rmk7-pxa3-embedix). is the new kernel going to be compatible with the modules that pdaxrom 1.1.0beta1 comes with?

if pdaxrom doesn't start after i flash my z, will i be able to reflash with the old zImage or will my z be bricked?

thanks, static

[static]

erm, i think i was supposed to "make bzImage" instead. my bzImage is 870k or so. sorry, i haven't compiled a kernel for a long time.

but i still don't know if it'll brick my zaurus if i flash with it.

[telemetric_au]

i got your message,

ive never gone as far as building a rom afaik , just configuring a kernel source far enough to build a 3rd party app's module against it...

afaik you wont brick it but i dont have the exp. though i have flashed alot, quite a few of them unsuccessful, ive always made it good again.

good luck, it is annoying bug i would like fixed too.

[static]

thanks for your reply. i flashed my z with the new zImage, powered it up, and it displayed a screen that said "Sharp" in red letters.
but i reflashed with the pdaxrom kernel and it starts up again. i'll try a few more kernels.

[static]

well, holy cow. i managed to make a kernel that boots with the pdaxrom initrd and then panics when it's unable to mount a root fs.

more to come later if anything interesting happens.

[static]

i used a kernel config file from pdaxrom for the sl-5500 on the kernel mentioned in the first post to get a booting kernel. it appears to point to the same ramdisk /dev/mtdblock4 (major 1f, minor 04) but it panics.

here's the panic i get:

VFS: Cannot open root device "mtdblock4" or 1f:04
Please append a correct "root=" boot option
Kernel panic: VFS: Unable to mount root fs on 1f:04

geez. how annoying. is it possible that i need to change something about the ramdisk too?

hopefully it's just something about the kernel that needs changing.

looks like i have a bit of reading up to do if no one has a simple solution to my problem.

[telemetric_au]

im not sure if it will help but i can give u my log
asap

[static]

i didn't have initrd support built into the kernel. d'oh!

but it's still borked. perhaps cos i don't have the default initrd size set correctly.

RAMDISK: Couldn't find valid RAM disk image starting at 0.
Freeing initrd memory: 3072K
VFS: Cannot open root device "mtdblock4" or 1f:04
Please append a correct "root=" boot option
Kernel panic: VFS: Unable to mount root fs on 1f:04

bed time. i'll dink around with it later.

[telemetric_au]

dmesg

CODE

Linux version 2.4.18-rmk7-pxa3-embedix-021129 (sash@pdaX86) (gcc version 2.95.3 20010315 (release)) #1 Tue, 27 Dec 2005 21:07:30 +0000
CPU: Intel StrongARM-1110 revision 9
Machine: Sharp-Collie
Security risk: creating user accessible mapping for 0x00000000 at 0xe8000000
Security risk: creating user accessible mapping for 0x08000000 at 0xea000000
Security risk: creating user accessible mapping for 0xc2000000 at 0xf2000000
Security risk: creating user accessible mapping for 0x40000000 at 0xf0000000
On node 0 totalpages: 16384
zone(0): 16384 pages.
zone(1): 0 pages.
zone(2): 0 pages.
Kernel command line: root=/dev/mtdblock4 rootfstype=jffs2 mem=64M jffs2_orphaned_inodes=delete
Relocating machine vectors to 0xffff0000
Console: colour dummy device 80x30
Calibrating delay loop... 137.21 BogoMIPS
Memory: 64MB = 64MB total
Memory: 62096KB available (1898K code, 385K data, 80K init)
Dentry-cache hash table entries: 8192 (order: 4, 65536 bytes)
Inode-cache hash table entries: 4096 (order: 3, 32768 bytes)
Mount-cache hash table entries: 1024 (order: 1, 8192 bytes)
Buffer-cache hash table entries: 4096 (order: 2, 16384 bytes)
Page-cache hash table entries: 16384 (order: 4, 65536 bytes)
POSIX conformance testing by UNIFIX
Linux NET4.0 for Linux 2.4
Based upon Swansea University Computer Society NET3.039
Initializing RT netlink socket
BlueZ Core ver 2.4 Copyright (C) 2000,2001 Qualcomm Inc
Written 2000,2001 by Maxim Krasnyansky <maxk@qualcomm.com>
GA Chip: L1
UCB1200 generic module installed
suspend temp = 489
RCSR = 2
suspend main adc = 608(608)
fatal chk = 391
apm: BIOS version 1.2 Flags 0x02 (Driver version 1.14)
apm: Current AC Status : 1
batok installed
Starting kswapd
Journalled Block Device driver loaded
devfs: v1.10 (20020120) Richard Gooch (rgooch@atnf.csiro.au)
devfs: boot_options: 0x1
JFFS2 version 2.1. (C) 2001 Red Hat, Inc., designed by Axis Communications AB.
comadj = 114,44414d43,44414d43
Console: switching to colour frame buffer device 60x53
Collie frame buffer driver initialized.
keyboard initilaized.
pty: 256 Unix98 ptys configured
Serial driver version 5.05c (2001-07-08) with no serial options enabled
COLLIE serial driver version 1.3
ttyS0 on SA1100 UART3 (irq 17)
ttyS1 on SA1100 UART1 (irq 15)ttyS2 on SA1100 UART2 (irq 16), using IRDA
touch adj= 18147,25023,-1572864,-3014656
ucb1200 touch screen driver initialized
sharpled_init: done.
SA1100 Real Time Clock driver v1.00
block: 128 slots per queue, batch=32
RAMDISK driver initialized: 16 RAM disks of 8192K size 1024 blocksize
Uniform Multi-Platform E-IDE driver Revision: 6.31
ide: Assuming 50MHz system bus speed for PIO modes; override with idebus=xx
loop: loaded (max 8 devices)
PPP generic driver version 2.4.1
PPP BSD Compression module registered
ide_cs.c 1.26 1999/11/16 02:10:49 (David Hinds)
Collie Sound Driver Installed
TC35143F audio driver initialized
MTD Sharp chip driver <ds@lineo.com>
SA1100 flash: probing 32-bit flash bus
Using static partition definition
Creating 6 MTD partitions on "SA1100 flash":
0x00000000-0x00020000 : "Angel Monitor"
0x00020000-0x00040000 : "CF Updater"
0x00040000-0x000c0000 : "Diagnostics"
0x000c0000-0x001c0000 : "kernel"
0x001c0000-0x00fe0000 : "jffs2"
0x00fe0000-0x01000000 : "angel stuff"
Linux Kernel Card Services 3.1.22
  options:  [pm]
SA-1100 PCMCIA (CS release 3.1.22)
sa1100_pcmcia_init(0)
sa1100_pcmcia_init(1)
Frontlight Driver Initialized.
BlueZ HCI UART driver ver 2.1 Copyright (C) 2000,2001 Qualcomm Inc
Written 2000,2001 by Maxim Krasnyansky <maxk@qualcomm.com>
NET4: Linux TCP/IP 1.0 for NET4.0
IP Protocols: ICMP, UDP, TCP
IP: routing cache hash table of 512 buckets, 4Kbytes
TCP: Hash tables configured (established 4096 bind 4096)
NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
BlueZ L2CAP ver 2.3 Copyright (C) 2000,2001 Qualcomm Inc
Written 2000,2001 by Maxim Krasnyansky <maxk@qualcomm.com>
BlueZ RFCOMM ver 1.0
Copyright (C) 2002 Maxim Krasnyansky <maxk@qualcomm.com>
Copyright (C) 2002 Marcel Holtmann <marcel@holtmann.org>
BlueZ BNEP ver 1.1
Copyright (C) 2001,2002 Inventel Systemes
Written 2001,2002 by Clement Moreau <clement.moreau@inventel.fr>
Written 2001,2002 by David Libault <david.libault@inventel.fr>
Copyright (C) 2002 Maxim Krasnyanskiy <maxk@qualcomm.com>
NetWinder Floating Point Emulator V0.95 (c) 1998-1999 Rebel.com
IrCOMM protocol (Dag Brattli)
RAMDISK: Couldn't find valid RAM disk image starting at 0.
VFS: Mounted root (jffs2 filesystem) readonly.
Mounted devfs on /dev
Freeing init memory: 80K

lastlog

CODE

Jan  1 05:09:33 syslogd started: BusyBox v1.01 (2005.12.22-14:08+0000)
Jan  1 05:09:33 init: Entering runlevel: 5
Jan  1 05:09:35 survive: cardmgr[100]: watching 2 sockets
Jan  1 05:09:35 cardmgr[100]: watching 2 sockets
Jan  1 05:09:35 pcmcia: cardmgr startup succeeded
Jan  1 05:09:35 sd: sdmgr startup failed
Jan  1 05:09:37 survive:
Jan  1 05:09:37 survive:
Jan  1 05:09:37 ssh: sshd startup succeeded
Jan  1 05:09:38 sshd[186]: Server listening on 0.0.0.0 port 22.
Jan  1 05:09:38 usb: Loading USB modules: succeeded
Jan  1 05:10:13 login[198]: root login  on `vc/1'

[static]

thanks for the log.

it _appears_ that the root device mtdblock4 that my kernel is failing to open is created earlier up in your log:

MTD Sharp chip driver <ds@lineo.com>
SA1100 flash: probing 32-bit flash bus
Using static partition definition
Creating 6 MTD partitions on "SA1100 flash":
0x00000000-0x00020000 : "Angel Monitor" <- #0
0x00020000-0x00040000 : "CF Updater" <- #1
0x00040000-0x000c0000 : "Diagnostics" <- #2
0x000c0000-0x001c0000 : "kernel" <- #3
0x001c0000-0x00fe0000 : "jffs2" <-- jffs2 and it's #4
0x00fe0000-0x01000000 : "angel stuff"

perhaps my kernel isn't configured correctly to create this device, whatever it is.
resizing my default initrd size in the kernel didn't work.

[static]

fun stuff. i made my kernel panic right after it _should_ have done this:

TC35143F audio driver initialized
MTD Sharp chip driver <ds@lineo.com>
SA1100 flash: probing 32-bit flash bus
Using static partition definition
Creating 6 MTD partitions on "SA1100 flash":
0x00000000-0x00020000 : "Angel Monitor"
0x00020000-0x00040000 : "CF Updater"
0x00040000-0x000c0000 : "Diagnostics"
0x000c0000-0x001c0000 : "kernel"
0x001c0000-0x00fe0000 : "jffs2"
0x00fe0000-0x01000000 : "angel stuff"
Linux Kernel Card Services 3.1.22
options: [pm]
SA-1100 PCMCIA (CS release 3.1.22)

instead, it did this:

TC35143F audio driver initialized
Linux Kernel Card Services 3.1.22
options: [pm]
SA-1100 PCMCIA (CS release 3.1.22)
Kernel panic: d'oh

that would explain the lack of mtdblock4, i think.

[static]

TC35143F audio driver initialized
MTD Sharp chip driver <ds@lineo.com>
Linux Kernel Card Services 3.1.22
options: [pm]
SA-1100 PCMCIA (CS release 3.1.22)
Kernel panic: d'oh

found the sharp driver in the kernel config: mtd->ram/rom/flash chip drivers->older ... drivers..->pre CFI sharp..

[static]

./Documentation/arm/SA1100/Assabet

that document in the kernel tree has instructions that look relevant. it contains instructions for configuring the kernel. i followed the instructions. it created a .config file for me. i then did a make menuconfig, copied the changes it made to the MTD section to paper, put the old kernel config back in place, and updated it's MTD settings manually via another make menuconfig.

the result is a small improvement:

TC35143F audio driver initialized
MTD Sharp chip driver <ds@lineo.com>
SA1100 flash: probing 32-bit flash bus
Probe at base[0x0](0x00000000) past the end of the map (0xffffffff)
Probe at base[0x0](0x00000000) past the end of the map (0xffffffff)
CFI: Found no SA1100 flash device at location zero
Sharp SL series flash device: 1000000 at 0
request_module[map_rom]: Root fs not mounted
Linux Kernel Card Services 3.1.22
options: [pm]
SA-1100 PCMCIA (CS release 3.1.22)
Kernel panic: d'oh

[static]

i've managed to get it to boot but it appears that the touch screen isn't working. i haven't patched the right arrow bug yet either.

anyway, the angstrom project has a complicated .... build system... thing ... that i borrowed configuration info from to get my kernel config to work. i'm really not sure how i managed to make it boot but heck, it works. i'll just stay the hell out of the MTD configuration screen and it should work out ok.. i hope.

OH. the angstrom build system also has like 10 or 15 patches for the 2.4 kernel on the collie. many of these patches look important for getting all the 5500s hardware to work right. the biggest one is "cacko.patch".. it looks like it contains mtd stuff so i bet it's required if you want a kernel that boots. i applied it. it also provides the tiny console font that pdaxrom uses... and a ton of other stuff too, i'm sure. i'll attach the other patches to this thread once i've decided if they are needed or not.

i've attached cacko.patch, and my kernel config.

Attached File(s)

 cacko.patch.gz ( 287.47K ) Number of downloads: 1
 CONFIG_GOLDEN.gz ( 6.33K ) Number of downloads: 2

 

[static]

this is a list of the patches that angstrom appears to apply to the kernel. the openswan patch didn't quite apply correctly so i chose not to apply it in my instructions below.

the touchscreen still doesn't work after applying all these patches. either i am leaving something out of the kernel config or there is some other patch that pdaxrom uses that i don't have... this fixed the right arrow bug, though

who knows what else fails to work in this kernel. i haven't done much testing.

all of these patches are in an attachment to this post along with the kernel config file i used.

cacko.patch
unb0rk-rightarrow.patch
unb0rk-apm.patch
battery.patch
bluetooth-2.4.18-mh15.patch
iw_handlers.w13-5.diff
iw_handlers.w14-5.diff
iw240_we15-6.diff
iw249_we16-6.diff
iw249_we17-13.diff
iw240_we18-5.diff
idecs.patch
logo.patch
initsh.patch
keymap-more-sane.patch
mkdep.patch
disable-pcmcia-probe.patch
linux-2.4.18-list_move.patch
openswan-2.2.0-kernel-2.4-klips.patch.gz
1764-1.patch
module_licence.patch
ir240_sys_max_tx-2.diff
ir241_qos_param-2.diff

put the contents of the cross sdk under /usr/local/arm (see my first post for a link to the sdk).

the compiler binaries should wind up in /usr/local/arm/2.95.3/bin once you uncompress the cross sdk tarball.

iirc, there are some hardwired paths in the cross compiler that require it to be in that directory.

create a directory to work in.
put the bziped kernel tarball (my first post links to it), the kernel.config file and all patches in this directory.
cd into the directory.

tar -xvjf linux-sl5500-20030509-rom3_10.tar.bz2
ln -s linux linux-2.4.18
ln -s linux linux-2.6-bkpxa

patch -p0 < cacko.patch
patch -p0 < unb0rk-rightarrow.patch
patch -p0 < unb0rk-apm.patch
patch -p0 < battery.patch
patch -p0 < bluetooth-2.4.18-mh15.patch
patch -p0 < iw_handlers.w13-5.diff
patch -p0 < iw_handlers.w14-5.diff
patch -p0 < iw240_we15-6.diff
patch -p0 < iw249_we16-6.diff
patch -p0 < iw249_we17-13.diff
patch -p0 < iw240_we18-5.diff
patch -p0 < idecs.patch
patch -p0 < logo.patch
patch -p0 < initsh.patch
patch -p0 < keymap-more-sane.patch
patch -p0 < mkdep.patch
patch -p0 < disable-pcmcia-probe.patch
patch -p0 < linux-2.4.18-list_move.patch
patch -p0 < 1764-1.patch
patch -p0 < module_licence.patch
patch -p0 < ir240_sys_max_tx-2.diff
patch -p0 < ir241_qos_param-2.diff

cp kernel.config linux/.config

cd linux
export PATH=$PATH:/usr/local/arm/2.95.3/bin/

for some reason, i need to do a make menuconfig and then immediately exit choosing to save my kernel config at this point. otherwise, the kernel will fail to compile. dunno why.

make menuconfig
select "Exit"
select "Yes" at the prompt

make dep && make bzImage

copy the newly created zImage from arch/arm/boot/ to your cf card along with the pdaxrom beta1 initrd.bin

reflash.

you might try leaving out the logo.patch if you don't like the openzaurus logo appearing at the top of the screen. lol.

Attached File(s)

 allpatches.tar.gz ( 954.16K ) Number of downloads: 2