I noticed that when using the preloaded kernel, the kernel has 4 bytes
memory corruption at address 0xf001000c (physical 0x1000c). The corrupting
value is 0x7ffffffff, MAXINT. If I restore the original value manually from
gdb session, Linux boots much better. I haven't yet found where the
corruption happens.
I made a dirty (don't commit this) version which takes the hardcoded prom
tree and tree browsing functions from Proll. If OLD_OBP is enabled, the
version can boot Linux as well as Proll/Qemu can at the moment, meaning that
we can reach the userland installer without any faults. There could be bugs
in the Forth-based tree, or maybe calling Forth is not working so well after
Linux has started doing more things.
_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
My 2 pennies,
I know EFI device driver developers here at work and
I know for sure that among other things EFI's debugging
framework is not as powerful as OpenFirmware. Heck! you
can literally _interactively_ develope an FCode driver in
OpenFirmware. This you cannot do in EFI.
Perhaps OpenBIOS will be to EFI/x86Bios, what Linux is
to Windows today. We'll find out within the next ~5 years.
_____________
Asif Haswarey
C1069W
QLogic Corporation
26650 Aliso Viejo Parkway
Aliso Viejo, CA 92656
Phone: 949.389.7607
Fax: 949.389.6308
-----Original Message-----
From: openbios-bounces(a)openbios.org
[mailto:openbios-bounces@openbios.org] On Behalf Of Ronald G Minnich
Sent: Tuesday, May 23, 2006 9:42 AM
To: The OpenBIOS Mailinglist
Subject: Re: [OpenBIOS] old shoes vs new shoes.
Stefan Reinauer wrote:
> Hi,
>
> some voices whisper that EFI programs are so much easier and simpler
> than Open Firmware. I am not making any statement but let you, much
> valued readers, decide here:
>
> http://www.osxbook.com/book/bonus/chapter4/efiprogramming/
yes, but EFI is such a powerful operating system! And, fortunately, the
API is totally new and incompatible with EVERYTHING, and it's not a very
good API, so we get the benefit of having to learn a whole new api, with
a special toolchain, that is useless for everything else we do.
I am sure I have not listed all the benefits of this new EFI operating
system :-) ron
--
OpenBIOS http://openbios.org/
Mailinglist: http://lists.openbios.org/mailman/listinfo
Free your System - May the Forth be with you
Author: stepan
Date: 2006-05-23 00:37:46 +0200 (Tue, 23 May 2006)
New Revision: 29
Modified:
openbios-devel/arch/ppc/kernel.c
Log:
forgot this one
Modified: openbios-devel/arch/ppc/kernel.c
===================================================================
--- openbios-devel/arch/ppc/kernel.c 2006-05-22 22:36:01 UTC (rev 28)
+++ openbios-devel/arch/ppc/kernel.c 2006-05-22 22:37:46 UTC (rev 29)
@@ -36,12 +36,6 @@
/* F U N C T I O N S */
/************************************************************************/
-void
-exception( int errcode )
-{
- /* no-op */
-}
-
int
forth_segv_handler( char *segv_addr )
{
I added handling of Qemu parameters in nvram. Now kernel arguments, default
boot device etc. can be set in the Qemu command line. Interrupt controller
and timer are initialised.
No progress with Linux.
_________________________________________________________________
Don't just search. Find. Check out the new MSN Search!
http://search.msn.com/