kernel boot on mpc8245

kernel boot on mpc8245

[emanuel stiebler]

Hi,
I have a board which is close to Sandpoints mpc8245 cpu board.
(using internal DUARTs, EPIC_MODE_SERIAL)

Trying to get it working, I got until:

Filename 'pImage'.
Load address: 0x1000000
Loading: ################################T ###################T
##############
          ###T
##############################################################
          ##T #########################
done
Bytes transferred = 801369 (c3a59 hex)
=> bootm 1000000
## Booting image at 01000000 ...
    Image Name:   Linux-2.4.20
    Created:      2003-03-04  23:57:29 UTC
    Image Type:   PowerPC Linux Kernel Image (gzip compressed)
    Data Size:    801305 Bytes = 782 kB = 0 MB
    Load Address: 00000000
    Entry Point:  00000000
    Verifying Checksum ... OK
    Uncompressing Kernel Image ... OK
id mach(): done
MMU:enter
MMU:hw init
MMU:mapin
MMU:mapin_ram done
MMU:setio
MMU:exit
setup_arch: enter
setup_arch: bootmem
mpc10x:enter
mpc10x:exit
arch: exit
openpic: enter
openpic: timer
openpic: external
openpic: spurious
openpic: exit


then it stops. Any idea what it is waiting for ?

And another question, I don't see any of the "printk ()" outputs I put
in the code, did I missed a flag ?

Cheers & thanks for any help


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

Re: kernel boot on mpc8245

[Yuji Nishimura]

Hi,

I was some experience for similar problem on past time. It was happened
on kernel porting process (to other board).
For my problem, It caused by ppc_md.calibrate_decr =
todc_calibrate_decr; . i forgot to initialize external real time clock(
so rtc was stopped). that means it does not work and stop on same
point.... If you use another calibrate_decr (based on bus clock or
another clock), maybe it caused by other problem...

Best Regards,
Yuji Nishimura - Tokyo, Japan.




emanuel stiebler wrote:

>
> Hi,
> I have a board which is close to Sandpoints mpc8245 cpu board.
> (using internal DUARTs, EPIC_MODE_SERIAL)
>
> Trying to get it working, I got until:
>
> Filename 'pImage'.
> Load address: 0x1000000
> Loading: ################################T ###################T
> ##############
> ###T
> ##############################################################
> ##T #########################
> done
> Bytes transferred = 801369 (c3a59 hex)
> => bootm 1000000
> ## Booting image at 01000000 ...
> Image Name: Linux-2.4.20
> Created: 2003-03-04 23:57:29 UTC
> Image Type: PowerPC Linux Kernel Image (gzip compressed)
> Data Size: 801305 Bytes = 782 kB = 0 MB
> Load Address: 00000000
> Entry Point: 00000000
> Verifying Checksum ... OK
> Uncompressing Kernel Image ... OK
> id mach(): done
> MMU:enter
> MMU:hw init
> MMU:mapin
> MMU:mapin_ram done
> MMU:setio
> MMU:exit
> setup_arch: enter
> setup_arch: bootmem
> mpc10x:enter
> mpc10x:exit
> arch: exit
> openpic: enter
> openpic: timer
> openpic: external
> openpic: spurious
> openpic: exit
>
>
> then it stops. Any idea what it is waiting for ?
>
> And another question, I don't see any of the "printk ()" outputs I put
> in the code, did I missed a flag ?
>
> Cheers & thanks for any help
>
>
>
>


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

Re: kernel boot on mpc8245

[Mark A. Greer]

My recommendation for a first step is to look at your printk data. You
didn't miss a flag, you're simply crapping out before the console is set
up to dump out the printk data.

Go into the System.map file associated with the kernel your running,
search for 'log_buf', subtract 0xc0000000 to get the physical addr of
'log_buf', hook up a BDI or something similar (and hopefully, your board
has an appropriate connector), and dump the [printk] data at that address.

Mark
--

Yuji Nishimura wrote:

>Hi,
>
>I was some experience for similar problem on past time. It was happened
>on kernel porting process (to other board).
>For my problem, It caused by ppc_md.calibrate_decr =
>todc_calibrate_decr; . i forgot to initialize external real time clock(
>so rtc was stopped). that means it does not work and stop on same
>point.... If you use another calibrate_decr (based on bus clock or
>another clock), maybe it caused by other problem...
>
>Best Regards,
>Yuji Nishimura - Tokyo, Japan.
>
>
>
>
>emanuel stiebler wrote:
>
>
>
>>Hi,
>>I have a board which is close to Sandpoints mpc8245 cpu board.
>>(using internal DUARTs, EPIC_MODE_SERIAL)
>>
>>Trying to get it working, I got until:
>>
>>Filename 'pImage'.
>>Load address: 0x1000000
>>Loading: ################################T ###################T
>>##############
>>###T
>>##############################################################
>>##T #########################
>>done
>>Bytes transferred = 801369 (c3a59 hex)
>>=> bootm 1000000
>>## Booting image at 01000000 ...
>>Image Name: Linux-2.4.20
>>Created: 2003-03-04 23:57:29 UTC
>>Image Type: PowerPC Linux Kernel Image (gzip compressed)
>>Data Size: 801305 Bytes = 782 kB = 0 MB
>>Load Address: 00000000
>>Entry Point: 00000000
>>Verifying Checksum ... OK
>>Uncompressing Kernel Image ... OK
>>id mach(): done
>>MMU:enter
>>MMU:hw init
>>MMU:mapin
>>MMU:mapin_ram done
>>MMU:setio
>>MMU:exit
>>setup_arch: enter
>>setup_arch: bootmem
>>mpc10x:enter
>>mpc10x:exit
>>arch: exit
>>openpic: enter
>>openpic: timer
>>openpic: external
>>openpic: spurious
>>openpic: exit
>>
>>
>>then it stops. Any idea what it is waiting for ?
>>
>>And another question, I don't see any of the "printk ()" outputs I put
>>in the code, did I missed a flag ?
>>
>>Cheers & thanks for any help
>>
>>
>>
>>
>>
>>
>
>
>
>
>
>
>


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

Re: kernel boot on mpc8245

[emanuel stiebler]

Mark A. Greer wrote:
> My recommendation for a first step is to look at your printk data. You
> didn't miss a flag, you're simply crapping out before the console is set
> up to dump out the printk data.
>
> Go into the System.map file associated with the kernel your running,
> search for 'log_buf', subtract 0xc0000000 to get the physical addr of
> 'log_buf', hook up a BDI or something similar (and hopefully, your board
> has an appropriate connector), and dump the [printk] data at that address.

I have the connector, but not a BDI2000 :(

And the log_buf ends with:

0019f544: 20617263 682c2062 65666f72 65206669     arch, before fi
0019f554: 6e645f62 72696467 6573200a 3c363e45    nd_bridges .<6>E
0019f564: 63756275 65726f20 50383234 35205369    cubuero P8245 Si
0019f574: 6e676c65 20426f61 72642043 6f6d7075    ngle Board Compu
0019f584: 7465720a 3c363e50 6f727420 62792065    ter.<6>Port by e
0019f594: 6d750a3c 343e4f6e 206e6f64 65203020    mu.<4>On node 0
0019f5a4: 746f7461 6c706167 65733a20 33323736    totalpages: 3276
0019f5b4: 380a3c34 3e7a6f6e 65283029 3a203332    8.<4>zone(0): 32
0019f5c4: 37363820 70616765 732e0a3c 343e7a6f    768 pages..<4>zo
0019f5d4: 6e652831 293a2030 20706167 65732e0a    ne(1): 0 pages..
0019f5e4: 3c343e7a 6f6e6528 32293a20 30207061    <4>zone(2): 0 pa
0019f5f4: 6765732e 0a3c343e 4b65726e 656c2063    ges..<4>Kernel c
0019f604: 6f6d6d61 6e64206c 696e653a 20636f6e    ommand line: con
0019f614: 736f6c65 3d747479 53302c31 31353230    sole=ttyS0,11520
0019f624: 3020636f 6e736f6c 653d7474 79302072    0 console=tty0 r
0019f634: 6f6f743d 2f646576 2f72616d 2072770a    oot=/dev/ram rw.
0019f644: 00000000 00000000 00000000 00000000    ................

Cheers & thanks


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

Re: kernel boot on mpc8245, internal UART

[emanuel stiebler]

> emanuel stiebler wrote:
>>I have a board which is close to Sandpoints mpc8245 cpu board.
>>(using internal DUARTs, EPIC_MODE_SERIAL)

Trying to figure out what is going wrong, I tried to locate the driver
for the internal DUART of the MPC8245 ...

So, probably a stupid question, but where is the link/information, or
where do I provide it, that the MPC8245 should use the internal UART as
a console on linux ?

Thanks a lot in advance, but I guess I got lost ...


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

RE: kernel boot on mpc8245, internal UART

[Callebaut, Benoit]

To know witch physical uart is mapped to which STTY, you must see in the
kernel the sanpoint_serial.h file in /arch/platform.
And to know whitch one is used by Linux at startup /etc/inittab.

-----Original Message-----
From: emanuel stiebler [mailto:emu@ecubics.com]
Sent: Wednesday, March 05, 2003 10:36 PM
To: linux-emb
Subject: Re: kernel boot on mpc8245, internal UART



> emanuel stiebler wrote:
>>I have a board which is close to Sandpoints mpc8245 cpu board.
>>(using internal DUARTs, EPIC_MODE_SERIAL)

Trying to figure out what is going wrong, I tried to locate the driver
for the internal DUART of the MPC8245 ...

So, probably a stupid question, but where is the link/information, or
where do I provide it, that the MPC8245 should use the internal UART as
a console on linux ?

Thanks a lot in advance, but I guess I got lost ...


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