[U-Boot-Users] 1.1.6 at91 based board dataflash copy panic.

[U-Boot-Users] 1.1.6 at91 based board dataflash copy panic.

[Ivan Kuten]

Hello,

I had the same problem with u-boot with dataflash only setup
 (no NOR flash) and used the following patch (see attach)

cmd_mem.c doing wrong when copy from Dataflash to SDRAM:

-       if (addr_dataflash(addr) && !addr_dataflash(dest) &&
(addr2info(dest)==NULL) ){
+       if (addr_dataflash(addr) && !addr_dataflash(dest)
+#ifndef CFG_NO_FLASH
+           && (addr2info(dest)==NULL)
+#endif
+          ){

and do not forget to set CFG_NO_FLASH in your include.

------------------------------------
Best regards, Ivan
Embedded Linux Engineer
Promwad - www.promwad.com
------------------------------------

--------------------------------------------------
 Hello all,

We are trying u-boot 1.1.6 on a customised at91rm9200 board. The main
features are dataflash, 32MB and networking with DM9161.

We have been using u-boot 1.1.4 quite well now, but we would like to test
1.1.6. Kernel boots ok, but we have found a problem.

When we do a cp from the dataflash to ram like this:
cp.b $(kerneladdr) 0x21600000 0x1
where kerneladdr is 0xC0040000

we have the following:

data abort

Resetting CPU ...
-------------- next part --------------
A non-text attachment was scrubbed...
Name: u-boot-1.1.6.at91.copy_from_dataflash_to_ram.chub.diff
Type: application/octet-stream
Size: 1576 bytes
Desc: not available
Url : http://lists.denx.de/pipermail/u-boot/attachments/20070104/05f893af/attachment.obj 

[U-Boot-Users] 1.1.6 at91 based board dataflash copy panic.

[Ivan Kuten]

Hello,

> When CFG_NO_FLASH is defined the content of either flash.h or flash.c is
> not included in the code, but dataflash support depends on that so I get
> a compilation error. I have been playing around and it seems that it is
> not possible to do a exclusive dataflash binary for u-boot, am I right?

Yes, now there is quite a mess with dataflash only setup.
I was able to do "exclusive dataflash binary for u-boot" by adding
necessary defines from flash.h to dataflash.h (these changes where
included in patch I sent):
------------------------------------------------
#define ERR_OK                         0
#define ERR_TIMOUT                     1
#define ERR_NOT_ERASED                 2
#define ERR_PROTECTED                  4
#define ERR_INVAL                      8
#define ERR_ALIGN                      16
#define ERR_UNKNOWN_FLASH_VENDOR       32
#define ERR_UNKNOWN_FLASH_TYPE         64
#define ERR_PROG_ERROR                 128
/*
 * Protection Flags for flash_protect():
 */
#define FLAG_PROTECT_SET       0x01
#define FLAG_PROTECT_CLEAR     0x02
------------------------------------------------

and removing flash related commands such as CFG_CMD_FLASH CFG_CMD_IMLS
from board config. It was just quick workaround.

------------------------------------
Best regards, Ivan
Embedded Linux Engineer
Promwad - www.promwad.com
------------------------------------



> Hi.
>
> I have applied that patch and the problem is solved. Anyway I have
> needed to do some additional changes. I will try to explain.
>
> When CFG_NO_FLASH is defined the content of either flash.h or flash.c is
> not included in the code, but dataflash support depends on that so I get
> a compilation error. I have been playing around and it seems that it is
> not possible to do a exclusive dataflash binary for u-boot, am I right?
>
> As code is written CFG_NO_FLASH means that there isn't _any kind_ of
> flash. So I guess the define in your patch should be CFG_DATAFLASH_ONLY
> or similar.
>
> Thanks for your help,
>
> Regards.
>
> --
> Gerardo Mart?nez Bernat
> tlf: 91 351 11 20
> Barracuda Systems SL
> C/ Luxemburgo n?4
> 28224 Madrid
>
>
>
>
>

[U-Boot-Users] 1.1.6 at91 based board dataflash copy panic.

[Raúl Sánchez Siles]

Hello all,

We are trying u-boot 1.1.6 on a customised at91rm9200 board. The main
features are dataflash, 32MB and networking with DM9161.

We have been using u-boot 1.1.4 quite well now, but we would like to test
1.1.6. Kernel boots ok, but we have found a problem.

When we do a cp from the dataflash to ram like this:
cp.b $(kerneladdr) 0x21600000 0x1
where kerneladdr is 0xC0040000

we have the following:

data abort
 
Resetting CPU ...

and the board hangs, we have to power off and then on.

The funny thing is that if we go through the usual boot procedure, the board
boots linux as expected and then everything works ok.

I would like to debug this situation, so I tried to uncomment the -DDEBUG on
the config.mk file and rebuild u-boot again. Also tried the same adding
some lines like #define #DEBUG_xxx 1 in the board config file, but I was
unable to get more verbose messages.

I would appreciate some help to find the bug or understand what's
happenning. If you need any more information I could provide, just ask.

Thanks and regards.

-- 
Ra?l S?nchez Siles