Re: [Fastboot] [CFT] kexec syscall for 2.5.43 (linux booting linux)

[Eric W. Biederman]

Digging further into the one failure I can reproduced, I have
found a very weird failure case.  The kernel code dies after switching
into 32bit mode.  I found this boot setting the setup.S hooks and printing
a character to the serial port whenever they were encountered.

I will release another version of kexec-tools shortly with a -debug switch
to enable this debugging, and anything else I can think of.  For the most
part I have avoided printing messages out the serial port because not
everyone has one, or has it setup as a serial console.

But if I enable it just on a debugging switch it should be o.k. and help
quite a bit with figuring out why some machines fail, and others do not.

Eric