linuxppc_2_4_mpc5200 / 5200lite zImage

linuxppc_2_4_mpc5200 / 5200lite zImage

[Kate Alhola]

I just got Motorola5200 lite board and tried complie
linuxppc_2_4_mpc5200 from
denx to it.  When i say make bzImage it complains that __res is undefined

ppc_82xx-ld -T ../ld.script -Ttext 0x00800000 -Bstatic -o zvmlinux head.o
../common/relocate.o  ../common/misc-simple.o ../common/bootinfo.o
../common/misc-common.o ../common/string.o ../common/util.o
../common/serial_stub.o mgt5100_tty.o  image.o ../lib/zlib.a
mgt5100_tty.o: In function `serial_init':
mgt5100_tty.o(.text+0x2): undefined reference to `__res'
mgt5100_tty.o(.text+0xa): undefined reference to `__res'

__res is pointer to external scc buffer descriptors . Least it looks a
like that it is
defined in arch/ppc/platforms/icecube.c but this code looks a like complete
board initialization code so i am not sure is it right module to add
defining __res

In this setup board is running original motorola debugger, no u-boot and
this
reason i try make zimage where is uncompressinng headers included.

So, any suggestions or experiencies with 5200lite ?

Also i got Macraigor wiggler but  saddly noticed that ocd-cmdr does not
yet work with mpc5200. Any expreincies/suggestions with LOW COST
OCD debugger for 5200 that can be used loading initial code to empty
flash on my own board.

Kate

--
+==================================================================+
! Kate Marika Alhola  Internet Technologies International Oy       !
! kate@iti.fi         Phone +358 400 740701 Fax   +358 9 8138715   !
! kate@iki.fi	      http://www.iti.fi    http://www.iki.fi/kate  !
+==================================================================+


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

Re: linuxppc_2_4_mpc5200 / 5200lite zImage

[Wolfgang Denk]

In message <3FA4F3B2.6050106@iti.fi> you wrote:
>
> I just got Motorola5200 lite board and tried complie
> linuxppc_2_4_mpc5200 from
> denx to it.  When i say make bzImage it complains that __res is undefined

Please "make uImage" and use U-Boot as boot loader.

> Also i got Macraigor wiggler but  saddly noticed that ocd-cmdr does not
> yet work with mpc5200. Any expreincies/suggestions with LOW COST
> OCD debugger for 5200 that can be used loading initial code to empty
> flash on my own board.

I recommend to use a BDI2000 - not exactly low cost, but worth  every
cent of it (expecially since you can alo use it for kernel and driver
debugging).

Best regards,

Wolfgang Denk

--
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd@denx.de
If there was anything that depressed him more than his own  cynicism,
it was that quite often it still wasn't as cynical as real life.
                                 - Terry Pratchett, _Guards! Guards!_

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

Re: linuxppc_2_4_mpc5200 / 5200lite zImage

[Kate Alhola]

Kate Alhola wrote:

>
> I just got Motorola5200 lite board and tried complie
> linuxppc_2_4_mpc5200 from
> denx to it.  When i say make bzImage it complains that __res is undefined
>
> __res is pointer to external scc buffer descriptors . Least it looks a
> like that it is
> defined in arch/ppc/platforms/icecube.c but this code looks a like
> complete
> board initialization code so i am not sure is it right module to add
> defining __res

It looks a like that i have too much fixation from mpc8xx(x) programming
in this case bd didnt stand for bufferr descriptor but board u-boot
board descriptor
and in this case the mgt5100_tty.c look a like have one dependence
to ppcboot, it took the ibp bus clock frequency from ppcboot bd struct even
not ppc-boot configuration. I replaced it with just plain #define .
Now just i should do testing ...

Kate

--
+==================================================================+
! Kate Marika Alhola  Internet Technologies International Oy       !
! kate@iti.fi         Phone +358 400 740701 Fax   +358 9 8138715   !
! kate@iki.fi	      http://www.iti.fi    http://www.iki.fi/kate  !
+==================================================================+


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

Re: linuxppc_2_4_mpc5200 / 5200lite zImage continued

[Kate Alhola]

Kate Alhola wrote:

>
>
> and in this case the mgt5100_tty.c look a like have one dependence
> to ppcboot, it took the ibp bus clock frequency from ppcboot bd struct
> even
> not ppc-boot configuration. I replaced it with just plain #define .
> Now just i should do testing ...


I just fixed it by comenting out setting baudrate, then it will leave it as
it was after dBug. Now i get kernel booting up and initial strings
coming out
but kernel crashes after initMMU ( early boot prit enabled ),
I think that i have something wrong in .config.
I did not find goog example gor 5200lite from denx tree so i made
my best quess and it seems not to bee so good.


Kate
-- continues hacking ....

dBUG> dn -i zImage.icecube
Address:  0x00020000
Downloading Image 'zImage.icecube' from 195.218.81.140
..............
939645 bytes read via TFTP
dBUG> go 20000
loaded at:     00020000 001071B8
relocated to:  00800000 008E71B8
zimage at:     00805830 008E3C88
avail ram:     00400000 00800000

Linux/PPC load: console=ttyS0,9600 root=/dev/sda2
Uncompressing Linux...done.
Now booting the kernel
id mach(): doneMMU:enterMMU:hw initMMU:mapin


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

Re: linuxppc_2_4_mpc5200 / 5200lite zImage continued

[Wolfgang Denk]

In message <3FA5235B.7080007@iti.fi> you wrote:
>
> I just fixed it by comenting out setting baudrate, then it will leave it as

This is not a good idea.

> it was after dBug. Now i get kernel booting up and initial strings
> coming out
> but kernel crashes after initMMU ( early boot prit enabled ),
> I think that i have something wrong in .config.

No. It's much more likely that you missed to pass  someother  of  the
required  information.  You  have  several options: manually set up a
bd_info structure with _all_ of the required  information;  modifying
U-Boot  so  that  you  can load and start U-Boot from RAM; installing
U-Boot into flash.

You can always get a tested and working configuration  by  performing
the standard kernel build procedure:

	bash$ make mrproper
	bash$ make icecube_5200_config
	bash$ make oldconfig
	bash$ make dep
	bash$ make uImage

[The "LITE5200" can  be  considered  identical  to  the  "IceCube"  /
"IceCube+" as far as the Linux kernel is affected.]

Best regards,

Wolfgang Denk

--
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd@denx.de
Don't you know anything? I should have thought anyone knows that  who
knows anything about anything...      - Terry Pratchett, _Soul Music_

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

Re: linuxppc_2_4_mpc5200 / 5200lite zImage now working

[Kate Alhola]

Thank you about usefull help, it looks a like doing things in newer way
may be
faster than do older way ;)

Wolfgang Denk wrote:

>No. It's much more likely that you missed to pass  someother  of  the
>required  information.  You  have  several options: manually set up a
>bd_info structure with _all_ of the required  information;  modifying
>U-Boot  so  that  you  can load and start U-Boot from RAM; installing
>U-Boot into flash.
>
>
This worked with u-boot, setting CFG_RAMBOOT made it bootting from
RAM and allowed also programming u-boot to same flash as dbug so now
i can start uboot from dbug and then boot linux up

>You can always get a tested and working configuration  by  performing
>the standard kernel build procedure:
>
>	bash$ make mrproper
>	bash$ make icecube_5200_config
>	bash$ make oldconfig
>	bash$ make dep
>	bash$ make uImage
>
>
Tanks, that was also usefull

Kate


--
+==================================================================+
! Kate Marika Alhola  Internet Technologies International Oy       !
! kate@iti.fi         Phone +358 400 740701 Fax   +358 9 8138715   !
! kate@iki.fi	      http://www.iti.fi    http://www.iki.fi/kate  !
+==================================================================+


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

RE: linuxppc_2_4_mpc5200 / 5200lite zImage now working

[Tord Andersson]

Kate,

Thanks for sharing your experiences with the MPC5200Lite card.
Would it be possible for you to explain the steps you took to replace
dbug with u-boot. Did you use dbug or a bdm?
I am about to do the same thing, and fear that I might end up with a
"dead" card if
things go wrong : )

Kind regards,

Tord Andersson

-----Original Message-----
From: Kate Alhola [mailto:kate@iti.fi]
Sent: den 2 november 2003 20:13
To: Wolfgang Denk
Cc: linuxppc-embedded@lists.linuxppc.org
Subject: Re: linuxppc_2_4_mpc5200 / 5200lite zImage now working



Thank you about usefull help, it looks a like doing things in newer way
may be faster than do older way ;)

Wolfgang Denk wrote:

>No. It's much more likely that you missed to pass  someother  of  the
>required  information.  You  have  several options: manually set up a
>bd_info structure with _all_ of the required  information;  modifying
>U-Boot  so  that  you  can load and start U-Boot from RAM; installing
>U-Boot into flash.
>
>
This worked with u-boot, setting CFG_RAMBOOT made it bootting from RAM
and allowed also programming u-boot to same flash as dbug so now i can
start uboot from dbug and then boot linux up

>You can always get a tested and working configuration  by  performing
>the standard kernel build procedure:
>
>	bash$ make mrproper
>	bash$ make icecube_5200_config
>	bash$ make oldconfig
>	bash$ make dep
>	bash$ make uImage
>
>
Tanks, that was also usefull

Kate


--
+==================================================================+
! Kate Marika Alhola  Internet Technologies International Oy       !
! kate@iti.fi         Phone +358 400 740701 Fax   +358 9 8138715   !
! kate@iki.fi	      http://www.iti.fi    http://www.iki.fi/kate  !
+==================================================================+


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

Re: linuxppc_2_4_mpc5200 / 5200lite zImage now working

[Kate Alhola]

Tord Andersson wrote:

>Kate,
>
>Thanks for sharing your experiences with the MPC5200Lite card.
>Would it be possible for you to explain the steps you took to replace
>dbug with u-boot. Did you use dbug or a bdm?
>I am about to do the same thing, and fear that I might end up with a
>"dead" card if
>things go wrong : )
>
I actually did not eplace it all. I just compiled and programmed u-boot
to flash in diferent
address than 0xfff00000 I used 0xff800000 but it is not a good choice
because it
leaved big hole between dbug and uboot.May be the 0xfff40000 will be
more sensible choice.

Then i just loaded s-record image to RAM and copied it to flash in right
address.
To load it in diferent address that it is intended you should use -o option
because dbug can't load directly to flash.

Now, when i am more familiar  with u-boot and if i wish replace dbug
i recommend to first compile u-boot to ram address and then use ram-running
u-boot to erase dbug and program u-boot on the place. Of cource this kind of
procedures allways have risk to make dead board and then you will need
a WERY EXPENSIVE bdm tools reprogram it. So, loading u-boot
to alternate address is safest way.
Only inconvicienc of ths is that u-buut must be started grom dbug and so
one
extra step is needed.


Kate

--
+==================================================================+
! Kate Marika Alhola  Internet Technologies International Oy       !
! kate@iti.fi         Phone +358 400 740701 Fax   +358 9 8138715   !
! kate@iki.fi	      http://www.iti.fi    http://www.iki.fi/kate  !
+==================================================================+


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

Re: linuxppc_2_4_mpc5200 / 5200lite zImage now working

[Wolfgang Denk]

In message <3FAB51D2.104@iti.fi> you wrote:
>
> >Thanks for sharing your experiences with the MPC5200Lite card.
> >Would it be possible for you to explain the steps you took to replace
> >dbug with u-boot. Did you use dbug or a bdm?
> >I am about to do the same thing, and fear that I might end up with a
> >"dead" card if
> >things go wrong : )
> >
> I actually did not eplace it all. I just compiled and programmed u-boot
> to flash in diferent
> address than 0xfff00000 I used 0xff800000 but it is not a good choice

This is not an optimal way to solve this.

I just checked in patches  for  U-Boot  to  allow  for  a  "low-boot"
configuration,  i.  e.  you  can configure and built U-Boot using the
commands "make MPC5200LITE_LOWBOOT_config;make all" and then  program
it at the low end of the flash (0xFF000000). After switching the boot
high/low  jumper  ("B  H/L") the board will boot directly into U-Boot
and all you need to switch back to dBUG is to switch the jumper back.

It may take a while until the new version of U-Boot shows up  on  the
CVS  server; the patch against U-Boot version 1.0.0 is also available
at ftp://ftp.denx.de/pub/u-boot/lowboot-1.0.0.patch.gz


Best regards,

Wolfgang Denk

--
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd@denx.de
: ... and it's got weird formatting - Notepad, Write, Works  3  can't
: decipher it, and it's too big to go in DOS Edit. Help!
Install an operating system. :-)                  -- Tom Christiansen

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