Syslinux - May 2006 - mboot hangs at MPI

I'm attempting to PXE boot a XEN dom0 kernel version 2.6.16, xen version
3.0.2-2.  I've built syslinux-3.20-pre8 from source and am using the
following config:

DEFAULT mboot.c32 xen-3.0.2-2.gz dom0_mem=258048 --- vmlinuz-2.6.16-xen0 ro
console=ttyS0 root=/dev/sda1 --- initrd.img

Where initrd.img is my own custom initrd that I've built to do some
pre-setup before booting to the real root on sda1.  The problem is that
everytime I try and boot I get to this line and it just hangs:

Booting: MBI=0x00010120, entry=0x00100000

I found references to this problem in the archives
(http://syslinux.zytor.com/archives/2006-March/006654.html )
and made sure that the memset of the regs_in was happening, and that is
being done in 3.20-pre8 (hence the reason I'm even using it).  I found some
references to using particular versions of gcc, I'm currently using
3.4.4could this be a potential issue?

I'm hoping that once I get past the mboot.c32 issues and actually get into
my initrd.img everything should be pretty straight forward from there but
this multiboot stuff over PXE is quite the mind-bender:)  I'm hoping
somebody has had similar problems and could at least point me in the general
direction?

Also, some other info that may be helpful.  The hardware is a Dell PowerEdge
1850 with this signature for the PXE firmware:

Initializing Intel(R) Boot Agent GE v1.2.26
PXE 2.1 Build 084 (WfM 2.0)

Thanks very much for *any* advice on this problem.

Lin Salisbury

Hi,

On Wed, 10 May 2006 23:46:41 -0600, "Lin Salisbury"
wrote:
> Where initrd.img is my own custom initrd that I've built to do some
> pre-setup before booting to the real root on sda1.  The problem is that
> everytime I try and boot I get to this line and it just hangs:
> 
> Booting: MBI=0x00010120, entry=0x00100000
> 
> I found references to this problem in the archives
> (http://syslinux.zytor.com/archives/2006-March/006654.html )
> and made sure that the memset of the regs_in was happening, and that is
> being done in 3.20-pre8 (hence the reason I'm even using it).  I found
some
> references to using particular versions of gcc, I'm currently using
> 3.4.4could this be a potential issue?
Of course, having said this isn't likely to be a compiler issue, I've
now found a bug that shows up with some compilers and not others, and
causes mboot.c32 to hang in exactly this way.  I've sent a patch to the
list in a separate email. 

Tim.

-- 
Tim Deegan                           (My opinions, not the University's)
Systems Research Group
University of Cambridge Computer Laboratory