RE: [PATCH 1/7] bootwrapper: Add a firmware-independent "raw" target.

=20

-----Original Message-----
From: Milton Miller [mailto:miltonm@bga.com]=20
Sent: Friday, December 14, 2007 12:06 AM
To: Stephen Neuendorffer
Cc: ppcdev; Grant Likely
Subject: Re: [PATCH 1/7] bootwrapper: Add a=20
firmware-independent "raw" target.
=20
On Fri Dec 14 10:43:27 EST 2007,  Stephen Neuendorffer wrote:
=20

quoted

From: Grant Likely <grant.likely at secretlab.ca>

This target produces a flat binary rather than an ELF file,
fixes the entry point at the beginning of the image, and takes
a complete device tree with no fixups needed.

The device tree must have labels on /#address-cells, the timebase
frequency, and the memory size.

Signed-off-by: Grant Likely <grant.likely at secretlab.ca>
---

=20
=20
You indicated in the intro in 0/ that this was not ready, and you=20
didn't include your own s-o-b, but you did not put any statements to=20
that effect in the header.  The intro is not copied into patchwork,=20
which maintainers often use when deciding what to push.

Sorry... Still trying to figure out the process.

Now on to why this should not be merged:
=20
In addition to the above, it changes the build rules.  It tries to=20
change wrapper to assemble the .dtb into a .o from a .S file, but=20
doesn't set any flags to force the assembler into the right mode.  In=20
contrast the linker is controlled by the .lds linker script.
=20
In addition, the requirement for assembly labels can easily be=20
eliminated.  As mentioned above, they are used for 3=20
properties.  With=20
the existing library (in 2.6.24 and earlier), call simple_malloc_init=20
with a small bss array (like BSS_STACK does to allocate stack), and=20
then read the properties out of the device tree.  At that point, call=20
simple_malloc_init a second time using the found memory size.   As I=20
said the last time this was posted, my patches to boot from kexec=20
implemented this strategy.
=20
However, with the new libfdt, which is already in for-2.6.25,=20
we should=20
no longer need malloc() to simple read the tree.   At least that is=20
what was advertised.

Yes, I agree, I just haven't had a chance to go back and write that code
yet...

Thanks for the comments,
Steve