[U-Boot-Users] help understanding bd_t passing

[U-Boot-Users] help understanding bd_t passing

[Martin Hicks]

Hi,

I've got a ppc405gp board that has been running U-boot/Linux for a
while, but using a silly hack in cmd_go to pass in our arguments.  I
haven't quite figured out how the kernel gets the bd_t pointer.  I"m
also confused by the seemingly inconsistent bd_t structures between the
kernel and u-boot.

So what the hacked version of u-boot (hacked by someone else) currently
does is declare a simple bd_t in cmd_go and pass the address of that
as the only parameter when we jump to addr.

((void (*)(void *))addr)(&bd_struct);

Obviously, I'd rather just use the normal method.  It looks to me like
the parameters to addr() ends up being the load address of the kernel,
but when we get to ppc4xx_init() in the kernel it looks like it *should*
be a pointer to bd_t.

Any hints as to how this all works?  I've looked at walnut as an
example, but the bd_t defined in the kernel looks like it starts with
bi_s_version[4] and bi_r_version[30], which look to be at some offset
inside the bd_t defined in u-boot.h

Confused,
mh

-- 
Martin Hicks || mort at bork.org || PGP/GnuPG: 0x4C7F2BEE
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://lists.denx.de/pipermail/u-boot/attachments/20060328/1639c49f/attachment.pgp 

[U-Boot-Users] help understanding bd_t passing

[Wolfgang Denk]

In message <20060328171315.GL27792@bork.org> you wrote:
> 
> I've got a ppc405gp board that has been running U-boot/Linux for a
> while, but using a silly hack in cmd_go to pass in our arguments.  I
> haven't quite figured out how the kernel gets the bd_t pointer.  I"m

"go" is not used to boot Linux. Please RTFM.

> also confused by the seemingly inconsistent bd_t structures between the
> kernel and u-boot.

Will you please read the FAQ, and maybe the U-Boot and/or the  kernel
code that deals with parameter passing?

> So what the hacked version of u-boot (hacked by someone else) currently
> does is declare a simple bd_t in cmd_go and pass the address of that
> as the only parameter when we jump to addr.

This is plain wrong/ "go" was never  designed  to  bootLinux.  Please
RTFM.


Best regards,

Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
All your people must learn before you can reach for the stars.
	-- Kirk, "The Gamesters of Triskelion", stardate 3259.2

[U-Boot-Users] help understanding bd_t passing

[Martin Hicks]

On Wed, Mar 29, 2006 at 12:16:33AM +0200, Wolfgang Denk wrote:
> 
> > So what the hacked version of u-boot (hacked by someone else) currently
> > does is declare a simple bd_t in cmd_go and pass the address of that
> > as the only parameter when we jump to addr.
> 
> This is plain wrong/ "go" was never  designed  to  bootLinux.  Please
> RTFM.

Okay, I'm *glad* that this is wrong.  It didn't look like it could work.
I just assumed that the guy who was doing this job before had clue;
obviously a poor assessment.

Thanks for the pointer Wolfgang.
mh

-- 
Martin Hicks || mort at bork.org || PGP/GnuPG: 0x4C7F2BEE
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://lists.denx.de/pipermail/u-boot/attachments/20060328/b9a4666d/attachment.pgp