[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[ppc] CHRP NOTE segment
From: |
Hollis Blanchard |
Subject: |
[ppc] CHRP NOTE segment |
Date: |
Sat, 25 Sep 2004 22:15:37 -0500 |
Currently we have this in crt0.S:
.section ".note"
.align 2
.note_section_header:
.long 8
.long 24
.long 0x1275
.string "PowerPC"
.note_descriptor:
.long 0x0 /* real-mode */
.long 0xffffffff /* real-base */
.long 0xffffffff /* real-size */
.long 0xffffffff /* virt-base */
.long 0xffffffff /* virt-size */
.long 0x00030000 /* load-base */
It's a nice idea, but unfortunately doesn't work. I see that the CHRP
binding is confusing here: it requires an "ELF Note section", but it
also specifies type "PT_NOTE", which if you consult elf.h is in fact a
segment type. A NOTE *segment* is required.
Also, as I understand it Apple OF becomes extremely unhappy if it finds
a NOTE segment. I don't remember how bad exactly but it's possible it
won't boot after that. Perhaps disconnecting the power and motherboard
battery to allow PRAM settings to clear will fix it (see
http://www.netbsd.org/Ports/macppc/faq.html#ofw-clear). It's possible
that Apple's OF attempts to reconfigure itself (according to the NOTE)
but ends up in an unbootable state. IBM OF, on the other hand, requires
the NOTE segment.
In any case, the code above is useless and should be removed.
-Hollis
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- [ppc] CHRP NOTE segment,
Hollis Blanchard <=