VxWorks.c ??

2003-06-12 Thread Flavio Pereira

Hi Mark

After that, I need to set up a development environment.  I have the
ppc kernel source, but I'm not able to get libc and u-boot to compile.
Should I give up and just use the ELDK?

Definitely go with ELDK, it will save you quite a few grey hairs and a
considerable amount of time.

Also, what JTAG Emulators and programmers do you use that are
compatible with Linux and MPC8xx?

Abatron's BDI2000 is your best option. Well that's a personal preference
anyway :-)

Regards
Flavio Pereira

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





System crash in kernel_thread

2003-06-04 Thread Flavio Pereira

I experienced a similar problem with my 860 based board.
The solution was in DER (Debug Enable Register) configured by the BDI.
I had to remove a bit, I just don't remember which one.


Thanks, that solved it, I had to clear the SYSIE in the Debug Enable
register.

Regards
Flavio Pereira

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





System crash in kernel_thread

2003-06-03 Thread Flavio Pereira

Hi Guys

I'm trying to port Linux to a proprietary board. I got u-boot working
good, and now that I start the kernel I find that the system _halts_
when it tries to run init using the kernel_thread routine in
init/main.c/rest_init(). Can someone suggest why this could be
happening?

This is my output:

Linux version 2.4.20 (fpereira at rigel) (gcc version 2.95.4 20010319
(prerelease/f
ranzo/20011204)) #17 Tue Jun 3 11:02:30 SAST 2003
On node 0 totalpages: 8192
zone(0): 8192 pages.
zone(1): 0 pages.
zone(2): 0 pages.
Kernel command line:
Decrementer Frequency = 18750/60
Warning: real time clock seems stuck!
Calibrating delay loop... 49.76 BogoMIPS
Memory: 30892k available (1048k kernel code, 380k data, 52k init, 0k
highmem)
Dentry cache hash table entries: 4096 (order: 3, 32768 bytes)
Inode cache hash table entries: 2048 (order: 2, 16384 bytes)
Mount-cache hash table entries: 512 (order: 0, 4096 bytes)
Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes)
Page-cache hash table entries: 8192 (order: 3, 32768 bytes)
POSIX conformance testing by UNIFIX

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





mmap with flash

2003-04-30 Thread Flavio Pereira

I want to implement mmap with flash device!

You can use the kernel flash module for this. See
/linux/drivers/char/flash.c. Instead of mmaping memory (mem) you can
simply open and then mmap the actual flash _device_.

Regards
Flavio Pereira


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





consistent_alloc

2003-04-02 Thread Flavio Pereira

Hi All

I'm using Denx Linux 2.4.20 running on am MPC8xx processor. I've found
that consistent_alloc does not seem to allocate memory from a cache-free
region. I've found that even though I'm using consistent_alloc to
allocate memory, I still have to flush the relevant area before letting
the CPM read it. Is there any alternative to consistent_alloc, or should
it work and possibly I am using it incorrectly?

Regards
Flavio Pereira

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





Cramfs root file system

2002-11-22 Thread Flavio Pereira

Hi all / Dennis

1)   in kernel file init/main.c 's dev_name_struct,  add entry for
 mtdblock, 0x1f00
if it's not there already.

2) in ppcboot =setenv bootargs init=/linuxrc root=/dev/mtdblockx

I added the mtdblock entry and, yes it seems to have helped. The kernel
however now returns...

mtdblock_open
ok
mtdblock: read on initrd at 0x400, size 0x400
mtdblock: read on initrd at 0x0, size 0x1000
mtdblock: read on initrd at 0x1000, size 0x1000
mtdblock: read on initrd at 0x2000, size 0x1000
mtdblock: read on initrd at 0x3000, size 0x1000
wrong magic
VFS: Mounted root (jffs filesystem) readonly.

Ive looked at the kernel code, and it looks like it doesn't support a
cramfs system only jffs (ROM systems anyway). Can anyone confirm whether
Linux 2.4.4 supports cramfs for root file systems?

Thanks
Flavio Pereira


-Original Message-
From: Dennis Khoo [mailto:[EMAIL PROTECTED]
Sent: Thursday, November 21, 2002 9:18 PM
To: Flavio Pereira; LinuxPPC
Subject: Re: Cramfs root file system

Flavio,

 Please append a correct root= boot option
 Kernel panic: VFS: Unable to mount root fs on 02:00

 Now, is it possible to do what Im trying to do, if so are there any
 suggestions as to what I'm doing wrong?


I saw the same error when I was trying to mount jffs2 as root.  Maybe
your
problem is similar to mine.

My solution:

1)   in kernel file init/main.c 's dev_name_struct,  add entry for
 mtdblock, 0x1f00
if it's not there already.

2) in ppcboot =setenv bootargs init=/linuxrc root=/dev/mtdblockx

Hope this helps.

Dennis Khoo
Software Engineer
Adapcom, Inc


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





Cramfs root file system

2002-11-21 Thread Flavio Pereira

Hi All

Im trying to use a cramfs image as a root-file system at kernel boot
time.

I've built a cramfs image, copied it into my flash device (mtd3), and
can successfully mount it at runtime.

However at boot time I use root=dev/mtdblock3 (or dev/mtd3 im not sure
which one to use) and I get the following errors...

NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
Root-NFS: No NFS server available, giving up.
VFS: Unable to mount root fs via NFS, trying floppy.
request_module[block-major-2]: Root fs not mounted
VFS: Cannot open root device mtdblock3 or 02:00
Please append a correct root= boot option
Kernel panic: VFS: Unable to mount root fs on 02:00

Now, is it possible to do what Im trying to do, if so are there any
suggestions as to what I'm doing wrong?

Thanks
Flavio Pereira

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





Bug in sharp.c flash driver

2002-11-04 Thread Flavio Pereira

Hi Wolfgang

I finally sorted out my MTD problems. It turns out that there is a
problem in drvivers/mtd/sharp.c. In function sharp_erase

if (instr-callback) {
instr-callback(instr);
}

instr-state never gets set so the mtd driver never exits. We need to
set the state before this if statement:

instr-state = MTD_ERASE_DONE;

Regards
Flavio Pereira


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





MTD Driver (denx Linux 2.4.4)

2002-10-31 Thread Flavio Pereira

Hi All

I'm having difficulty in getting the MTD driver working. I'm using a
non-CFI complaint Sharp chip so I have defined CONFIG_MTD_SHARP. However
I cannot locate anywhere in the code where the kernel will call
sharp_probe_init. In any case I forced init_mtd (mtdcore.c) to call
sharp_probe_init and it seems to load the driver okay as I see the
following on my serial console

JFFS version 1.0, (C) 1999, 2000  Axis Communications AB
JFFS2 version 2.1. (C) 2001 Red Hat, Inc., designed by Axis
Communications AB.
MTD Sharp chip driver ds at lineo.com
MTD Sharp chip driver ds at lineo.com

Now my question is, do I need to force this driver to load or am I
missing some compile flag? How do I go about partitioning the flash, ie
how does this driver link to a flash map?

Any help would be appreciated

Thanks in advance
Flavio Pereira


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





Linker problem

2002-10-22 Thread Flavio Pereira

Hi All

Im trying to cross compile some c++ code for the 8260 chip. However when
the linker executes it return the following error:

undefined reference to `__throw'

Can anyone please give some advice on how to tackle this problem? What
does this mean? Does it have anything to do with c++ libraries??? Im
using the cross compile environment set up by the ELDK Kit from DENX.

Regards
Flavio Pereira

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





8260 configuration

2002-10-02 Thread Flavio Pereira

Hi Bo

1. Yes the IMMR needs to be configured at least above 0xC000. The
kernel executes code at 0xC000, and an IMMR lower that this will
cause the kernel to crash as soon as it starts executing code. I
recommend that you stay with the default value 0xF000.

2. The start address is fixed to 0x0

3. ?

4. No, not that I know of.

5. Try getting ELDK from the Denx site. (www.denx.de)



-Original Message-
From: Yu Bo-BOYU1 [mailto:[EMAIL PROTECTED]
Sent: Wednesday, October 02, 2002 3:08 AM
To: linuxppc embedded
Subject: 8260 configuration


I am new to embedded linux but I had good experience in vxWorks.
Right now I had a custom 8260 board with SMC1 connected to serial
port and FCC3 for fast ethernet. I would like to try running linux on
this
custom board. I downloaded the latest monta vista preview kit 2.1
as my base release.

I had boot code ( with vxWorks and it is based on EST8260 bsp) in my
custom
board
already. I decide to download the linux object via tftp. But I have the
following
questions:

1. Do I have to configure IMMR as 0xF_ or higher. The default
configuration
in our boot is 0x0F00_ ? Is this a hard requirement or not ? If
so,
why can't linux
init code do so by writing to IMMR+0x101a8 ?

2. How to specify the start address where linux runs when I build the
zImage
?

3. When I make xconfig, what configuration I must have in order to run ?
I
certainly
   do not need VGA/HARD DISK/PCI (this board does not have pci devices).

4. Is there any good website/paper/documentation about linux for 8260
platform.

5. Can anyone access ftp.leo.org to get ELDK ? I can ping it but I
cannot
ftp it today.

Thanks.

Bo


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





RT-Linux vs RTAI

2002-09-18 Thread Flavio Pereira

Hi All

Can anyone explain the difference between RT-Linux and the RTAI
patches available from the Denx site...? Is there a difference?

Thanks in advance
Flavio Pereira


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





Simple drivers....denx ftp

2002-09-10 Thread Flavio Pereira

Hi All

Can someone please e-mail the DENX example drivers that are available
whose sources can be found on:

ftp://ftp.denx.de/pub/LinuxPPC/usr/src/drivers.tar.gz

I'm having trouble accessing the ftp site for a few days now...

Thanks in advance
Flavio


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





Linux on 8260 using ELDK..Help?!?

2002-09-05 Thread Flavio Pereira

Hi Wolfgang

 I recommend to return to the original code, and get this working. You
 can probably use the driver in our  CVS  as  reference;  we  fixed  a
 couple of issues for example in the area of BRG assignment.

I have returned to the original uart.c. Although it still kills my
serial port it has nothing to do with the unable to open initial
console. Will the original uart.c I still get logging in
/var/log/messages and I can see it is still doing the same thing. Yes,
uart.c is probably something that I will still have to look at, but it
is not the root cause of my original problem. Any more suggestions would
be appreciated since I feel I've hit the brick wall quite severely now.
Besides file permissions (which Im conviced are fine now) what else
could cause the kernel to not be able to open a console?

Regards
Flavio


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/





Linux on 8260 using ELDK..Help?!?

2002-09-04 Thread Flavio Pereira

Hi Pete

Yes, both the read and write permissions are enabled for console and
tty0 in dev...

Flavio

-Original Message-
From: Pete McCormick [mailto:[EMAIL PROTECTED]
Sent: Wednesday, September 04, 2002 4:40 PM
To: Flavio Pereira
Cc: linuxppc-embedded at lists.linuxppc.org
Subject: RE: Linux on 8260 using ELDK..Help?!?


Flavio,

What do your file permissions look like on /dev/ttyS0?  Do you have read
and
write access enabled?

Pete

P.S.  I am also using the linux 2.4.18 kernel on the MPC8260ADS board,
so I
might be able to help on further issues...

--- Flavio Pereira fpereira at azisa.com wrote:

 Hi Wolfgang

 Firstly let me say thanks for your response.

 Did you fix the IMMR mapping in PPCBoot? The  CFG_DEFAULT_IMMR  defi-
 nition  will  NOT  work with LInux; see the Linux memory map require-
 ments - You _must_ map the IMMR at 0xF000 or above.

 Yes I have fixed to IMMR mappings in PPCBoot to correspond to
0xF000

  Are you sure this was done well?

 Well _no_ I'm not sure this was done well but I can tell you exactly
 what I have done. For uart.c I've completely removed SMC1 and SMC2
from
 serial_state rs_table and rs_brg_map[]. If I don't remove this my
serial
 port does not come up at all initially. For tty_io.c I simply to do
not
 call rs_8xx_init because it kills my serial port, i.e. if I let it
this
 code execute there is no more output on my serial terminal.

  Did you run the ELDK_MAKEDEV and ELDK_FIXPERM scripts as documented?
 I've run ELDK_MAKEDEV, so the dev/console file does exist and I've
also
 run ELDK_FIXOWNER. I'm guessing that FIXOWNER = FIXPERM.

 Also in my var\log\messages I get the following...so init is trying to
 execute:

 Dec 31 19:00:13 192 kernel: Warning: unable to open an initial
console.
 Dec 31 19:00:13 192 /sbin/mingetty[133]: /dev/ttyS0: Operation not
 permitted
 Dec 31 19:00:06 192 rc.sysinit: Mounting proc filesystem:  succeeded
 Dec 31 19:00:06 192 sysctl: net.ipv4.ip_forward = 0
 Dec 31 19:00:06 192 sysctl: net.ipv4.conf.all.rp_filter = 1
 Dec 31 19:00:07 192 sysctl: error: 'net.ipv4.ip_always_defrag' is an
 unknown key
 Dec 31 19:00:07 192 sysctl: error: 'kernel.sysrq' is an unknown key
 Dec 31 19:00:07 192 rc.sysinit: Configuring kernel parameters:
 succeeded
 Dec 31 19:00:07 192 date: Wed Dec 31 19:00:06 EST 1969
 Dec 31 19:00:07 192 rc.sysinit: Setting clock : Wed Dec 31 19:00:06
EST
 1969 succeeded
 Dec 31 19:00:07 192 swapon: swapon: cannot open /etc/fstab: No such
file
 or directory
 Dec 31 19:00:07 192 rc.sysinit: Activating swap partitions:  failed
 Dec 31 19:00:07 192 rc.sysinit: Setting hostname 192.168.1.246:
 succeeded
 Dec 31 19:00:10 192 rc.sysinit: Checking filesystems succeeded
 Dec 31 19:00:10 192 fsck: Checking all file systems.
 Dec 31 19:00:10 192 rc.sysinit: Mounting local filesystems:  failed
 Dec 31 19:00:10 192 fsck: WARNING: couldn't open /etc/fstab: No such
 file or directory
 Dec 31 19:00:10 192 mount: can't create lock file /etc/mtab~62:
 Permission denied (use -n flag to override)
 Dec 31 19:00:11 192 rc.sysinit: Enabling swap space:  succeeded
 Dec 31 19:00:11 192 init: can't open(/etc/ioctl.save, O_WRONLY):
 Permission denied
 Dec 31 19:00:13 192 /sbin/mingetty[134]: /dev/ttyS0: Operation not
 permitted
 Dec 31 19:00:11 192 init: Entering runlevel: 3
 Dec 31 19:00:14 192 /sbin/mingetty[135]: /dev/ttyS0: Operation not
 permitted
 Dec 31 19:00:14 192 /sbin/mingetty[136]: /dev/ttyS0: Operation not
 permitted
 Dec 31 19:00:14 192 /sbin/mingetty[137]: /dev/ttyS0: Operation not
 permitted
 Dec 31 19:00:14 192 /sbin/mingetty[138]: /dev/ttyS0: Operation not
 permitted
 Dec 31 19:00:14 192 /sbin/mingetty[139]: /dev/ttyS0: Operation not
 permitted
 Dec 31 19:00:14 192 /sbin/mingetty[140]: /dev/ttyS0: Operation not
 permitted

 Regards,
 Flavio

 -Original Message-
 From: Wolfgang Denk [mailto:wd at denx.de]
 Sent: Wednesday, September 04, 2002 3:41 PM
 To: Flavio Pereira
 Cc: linuxppc-embedded at lists.linuxppc.org
 Subject: Re: Linux on 8260 using ELDK..Help?!?

 Dear flavio,

 in message 1E354700654661418593A0D6B5A9D41008943B at mail.azisa.co.za
you
 wrote:
 
  I'm a _newbie_ to Linux and am busy porting Linux to a MPC8260ADS
  board. PPCBoot is working fine and the Linux kernel seems to boot up
  okay up and till the point where it begins to load init, it then
  crashes. The trace is shown below. I've also tried forcing sh and

 Did you fix the IMMR mapping in PPCBoot? The  CFG_DEFAULT_IMMR  defi-
 nition  will  NOT  work with LInux; see the Linux memory map require-
 ments - You _must_ map the IMMR at 0xF000 or above.

  bash as the start-up but these also fail with similar errors. I'm
  using the binaries provided by the ELDK development package. Also my
  changes to the base kernel code (2.4.4) are minimal and are limited
  to uart.c (force SCC1 as serial port) and tty_io.c (don't call
  rs_8xx_init). Can anyone please point me in the right direction...?

 Are you sure this was done well