Skip to contentSkip to conversation

Problem starting u-boot on Xilinx V2P

U-Boot-Users Thorsten Lueck 3 participants 8 messages Jul 5, 2005 – Jul 6, 2005 1 message not in the archive

lore
Share & tools
lore

Thorsten Lueck

author
#0 lore raw

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,

I have a own board with an Virtex V2Pro (V2P30) an want to boot u-boot on the PPC.
Starting from the demo-design for the ml300 demo board, I managed to compile u-boot 1.1.1 (1.1.2 fails when running the tcl-script with EDK 7.1) and could download and run u-boot on my PPC. However, u-boot stops after printing

U-Boot 1.1.1 (Jul 5 2005 - 16:50:28)

### No HW ID - assuming ML300
DRAM: 128 MB

I get no further output afterwards. Another strange thing is, that I only have 32MB RAM on board but not 128MB as the ml300 has. However, I could not find any hardcoded RAM size in the source files for the ml300.

Does anybody has an idea? I found the comment :

/*
* WARNING: this code looks "cleaner" than the PowerPC version, but
* has the disadvantage that you either get nothing, or everything.
* On PowerPC, you might see "DRAM: " before the system hangs - which
* gives a simple yet clear indication which part of the
* initialization if failing.
*/

in the ./lib_i386/board.c, however, there is no appropriate warning in static int init_func_ram (void)
of ./lib_ppc/board.c

Thanks in advance,

Thorsten L?ck
- --
IfEN - Gesellschaft f?r Satellitennavigation mbH

Dipl. Ing. Thorsten L?ck
Tel.: +49 8121 2238-25
Fax : +49 8121 2238-11
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCyqyuQDwjNlk95KYRAmuIAJsFYHOlYn1+MJzg8tNZ1+sJQYhOFwCg5LnR 2HsWS999Jatd1N8w+3kjArg=
=kltH
-----END PGP SIGNATURE-----

#1 lore raw
Quoted 7 lines · In message <[email protected]> you wrote:

I have a own board with an Virtex V2Pro (V2P30) an want to boot u-boot on the PPC.
Starting from the demo-design for the ml300 demo board, I managed to compile u-boot 1.1.1 (1.1.2 fails when running the tcl-script with EDK 7.1) and could download and run u-boot on my PPC. However, u-boot stops after printing

You cannot just use a configuration for one board and expect it to work on a different board.

Ummm... well, actually you _can_ do that, but you should be prepared to fail.

Quoted 2 lines

### No HW ID - assuming ML300
DRAM: 128 MB

Seems your SDRAM initialization is misconfigured and/or broken.

Quoted 5 lines

I get no further output afterwards. Another strange thing is, that I only have 32MB RAM on board but not 128MB as the ml300 has. However, I could not find any hardcoded RAM size in the source files for the ml300.

Does anybody has an idea? I found the comment :

You missed to port U-Boot to your hardware. Go back to square 0. Read the README (porting guide section) and do what needs to be done.

Quoted 1 line

in the ./lib_i386/board.c, however, there is no appropriate warning in

What has the x86 code to do with that? [Nothing!]

Best regards,

Wolfgang Denk

Signature · 4 lines
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
That was the thing about deserts. They had their  own  gravity.  They
sucked you into the centre.           - Terry Pratchett, _Small Gods_
#6 lore raw

Antwort: [U-Boot-Users] Problem starting u-boot on Xilinx V2P

In reply to Thorsten Lueck

Hi

How Wolfgang says its the best that you at first read the README file for better understanding.
But you find the config for RAM
/board/xilinx/ml300/ml300.c
this functions name is initdram(int board_type)

hope could help you

best regards

Thorsten Lueck
<[email protected]>
Gesendet von: An
u-boot-users-admin at list u-boot-users at lists.sourceforge.net
s.sourceforge.net Kopie

Thema
05.07.2005 17:52 [U-Boot-Users] Problem starting u-boot on
Xilinx V2P

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,

I have a own board with an Virtex V2Pro (V2P30) an want to boot u-boot on the PPC.
Starting from the demo-design for the ml300 demo board, I managed to compile u-boot 1.1.1 (1.1.2 fails when running the tcl-script with EDK 7.1) and could download and run u-boot on my PPC. However, u-boot stops after printing

U-Boot 1.1.1 (Jul 5 2005 - 16:50:28)

### No HW ID - assuming ML300
DRAM: 128 MB

I get no further output afterwards. Another strange thing is, that I only have 32MB RAM on board but not 128MB as the ml300 has. However, I could not find any hardcoded RAM size in the source files for the ml300.

Does anybody has an idea? I found the comment :

/*
* WARNING: this code looks "cleaner" than the PowerPC version, but
* has the disadvantage that you either get nothing, or everything.
* On PowerPC, you might see "DRAM: " before the system hangs - which
* gives a simple yet clear indication which part of the
* initialization if failing.
*/

in the ./lib_i386/board.c, however, there is no appropriate warning in static int init_func_ram (void)
of ./lib_ppc/board.c

Thanks in advance,

Thorsten L?ck
- --
IfEN - Gesellschaft f?r Satellitennavigation mbH

Dipl. Ing. Thorsten L?ck
Tel.: +49 8121 2238-25
Fax : +49 8121 2238-11
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCyqyuQDwjNlk95KYRAmuIAJsFYHOlYn1+MJzg8tNZ1+sJQYhOFwCg5LnR 2HsWS999Jatd1N8w+3kjArg=
=kltH
-----END PGP SIGNATURE-----

-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_idt77&alloc_id492&op=click
_______________________________________________
U-Boot-Users mailing list
U-Boot-Users at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/u-boot-users

author
#2 lore raw
In reply to Wolfgang Denk

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 2 lines · Wolfgang Denk

You cannot just use a configuration for one board and expect it to work on a different board.

My board is not as different to ml300 as one may think - indeed, it uses the same Virtex2Pro FPGA with nearly the same configuration (wrt. UART, I2C, EMAC). Therefore using ml300 as a starting point for my design really makes sense. Of course I used the BSP compiled for my board using Xilinx EDK 7.1

Quoted 2 lines

You missed to port U-Boot to your hardware. Go back to square 0. Read the README (porting guide section) and do what needs to be done.

Well - I _did_ port it (in the sense of using the appropriate BSP) - however, I modified ML300 source instead of creating a new board-directory ...

Quoted 4 lines
Quoted 2 lines

in the ./lib_i386/board.c, however, there is no appropriate warning in

What has the x86 code to do with that? [Nothing!]

Well - nothing - your are right - instead of the comment, that the PowerPC code may hang immediatley _after_ "DRAM:" printout. This led me to the assumption, that my problem may be well known ...

Best regards,

Thorsten

- --
IfEN - Gesellschaft f?r Satellitennavigation mbH

Dipl. Ing. Thorsten L?ck
Tel.: +49 8121 2238-25
Fax : +49 8121 2238-11
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCy34yQDwjNlk95KYRAm4vAJ0RGmdP6J+fu6hQdno9/FH0wBKwewCg0rIW roq+pqfElk34fcgcG++bxLw=
=uFWp
-----END PGP SIGNATURE-----

author
#8 lore raw
In reply to a message not in the archive

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 5 lines · Cedric VINCENT
Quoted 3 lines

Another strange thing is, that I
only have 32MB RAM on board but not 128MB as the ml300 has. However, I could not find any hardcoded RAM size in the source files for the ml300.

Take a look into board/xilinx/ml300/ml300.c, it *is* hardcoded !

Indeed - that was the problem - I greped for the size (in hex and dec) but _not_ for a formular ;-)

Quoted 5 lines

As Wolgang said, "you cannot just use a configuration for one board and expect it to work on a different board". But if you change your config file and U-Boot still hangs after "Now running in RAM at ....." (don't forget to #define DEBUG) it may be a U-Boot incompatibility with some (new) version of GCC. If so, please mail me !

Well, probably I can - changing the size to the real size (which is 32MB instead of 128) - u-boot boots up to the prompt. Probably some memory test addressed memory above 32MB. So it works now.
Regarding the version of gcc, I use
ppc_405-gcc (GCC) 3.3.1 (MontaVista 3.3.1-3.0.10.0300532 2003-12-24)

Thanks for your helpfull hint!

Best regards,

Thorsten L?ck

- --
IfEN - Gesellschaft f?r Satellitennavigation mbH

Dipl. Ing. Thorsten L?ck
Tel.: +49 8121 2238-25
Fax : +49 8121 2238-11
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCy3+kQDwjNlk95KYRAiMaAKDgp7miMmBTrdYvuQiTmv22J0ChhQCgr2bs +nJ4OJ7toS1avTRJ8YxkZp4=
=cSLW
-----END PGP SIGNATURE-----

#3 lore raw
In reply to Thorsten Lueck
Quoted 2 lines · In message <[email protected]> you wrote:

My board is not as different to ml300 as one may think - indeed, it uses

Even minimal diffeences can have significant effects. Feel free to ignore my advice, but please don't complain if you run into problems later.

Quoted 3 lines
Quoted 2 lines

You missed to port U-Boot to your hardware. Go back to square 0. Read the README (porting guide section) and do what needs to be done.

Well - I _did_ port it (in the sense of using the appropriate BSP) -

No, you did not. For example, you missed to adapt the SDRAM initialization.

Quoted 2 lines

PowerPC code may hang immediatley _after_ "DRAM:" printout. This led me to the assumption, that my problem may be well known ...

It is. It is actually a FAQ.

Best regards,

Wolfgang Denk

Signature · 4 lines
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
Don't put off for tomorrow what you can  do  today,  because  if  you
enjoy it today you can do it again tomorrow.
author
#4 lore raw

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 13 lines · Wolfgang Denk
Quoted 1 line · In message <[email protected]> you wrote:

My board is not as different to ml300 as one may think - indeed, it uses

Even minimal diffeences can have significant effects. Feel free to ignore my advice, but please don't complain if you run into problems later.

Quoted 3 lines
Quoted 2 lines

You missed to port U-Boot to your hardware. Go back to square 0. Read the README (porting guide section) and do what needs to be done.

Well - I _did_ port it (in the sense of using the appropriate BSP) -

No, you did not. For example, you missed to adapt the SDRAM initialization.

Quoted 9 lines
Quoted 2 lines

PowerPC code may hang immediatley _after_ "DRAM:" printout. This led me to the assumption, that my problem may be well known ...

It is. It is actually a FAQ.

Best regards,

Wolfgang Denk

Whatsoever - I started using Xilinx application note XAPP542, which advised to start using ml300 design and generate a new BSP using EDK. Unfortunately the DRAM size was hardcoded in the board driver source ml300.c instead of using (the available) hash-defines. I fixed this and can therefore provide a patch if wanted.

I realized, that the uboot-1.1.1 version supplied by Xilinx within the attachment of XAPP542 is not identical to the original version. Several patches and improvements has been made by Xilinx, which I cannot find in the acutal source (even 1.1.2 or cvs; indeed, 1.1.2 and cvs are almost the same for the xilinx board support package).

One problem when using EDK is, that variables within the tcl-script uboot_v2_1_0.tcl are declared "global" instead of "variable", which leads to errors of "unknown variables" in EDK. Xilinx fixed this to "variable" in their distributed version. Are there any reasons (wrt ELDK) why to use "global" instead of "variable"?

However, it seems, that the Xilinx version runs almost well (after some minor adoptations).

Thanks for your help and worthwhile hints,

Thorsten

- --
IfEN - Gesellschaft f?r Satellitennavigation mbH

Dipl. Ing. Thorsten L?ck
Tel.: +49 8121 2238-25
Fax : +49 8121 2238-11
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iD8DBQFCy/W4QDwjNlk95KYRAhPZAJwLqJqNJw/qVbBSs6M7x6f0Ycib5QCglMjf XBlu7e1SmHSzHDH5fYsL76Q=
=FZsE
-----END PGP SIGNATURE-----

#5 lore raw
Quoted 6 lines · In message <[email protected]> you wrote:

Whatsoever - I started using Xilinx application note XAPP542, which advised to start using ml300 design and generate a new BSP using EDK. Unfortunately the DRAM size was hardcoded in the board driver source ml300.c instead of using (the available) hash-defines. I fixed this and can therefore provide a patch if wanted.

Please do. But BTW: no matter if your hard-code a number or use a preprocessor define - both ways are a poor solution and not the original U-Boot style you should strive for. Normally you shoult auto-detect the type and size of memory in your system using the get_ram_size() function [in common/memsize.c].

Quoted 5 lines

I realized, that the uboot-1.1.1 version supplied by Xilinx within the attachment of XAPP542 is not identical to the original version. Several patches and improvements has been made by Xilinx, which I cannot find in the acutal source (even 1.1.2 or cvs; indeed, 1.1.2 and cvs are almost the same for the xilinx board support package).

We live with what we get contributed back...

Quoted 5 lines

One problem when using EDK is, that variables within the tcl-script uboot_v2_1_0.tcl are declared "global" instead of "variable", which leads to errors of "unknown variables" in EDK. Xilinx fixed this to "variable" in their distributed version. Are there any reasons (wrt ELDK) why to use "global" instead of "variable"?

Is this a typo? Our ELDK has nothing to do with this. From the U-Boot code point of view (which is ANSI C) I have not the faintest idea what "variable" or "global" means.

Best regards,

Wolfgang Denk

Signature · 4 lines
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
What about WRITING it first and rationalizing it afterwords?  :-)
                       - Larry Wall in <[email protected]>