Re: Booting Hurd-L4 under qemu

[Rian Hunter]

Sam Mason wrote:

> Bas Wijnen wrote:
>  
>
> > I haven't actually looked into it at all, but a 0 SP doesn't sound 
> > really strange to me.  I'd expect sigma0's ia32-crt0.s (or whatever it's 
> > called) to set the stack pointer and call main (possibly via cmain), 
> > just like wortel does that.
> >    
>
> OK, I just wasn't sure where this message was coming from before.
> Not having a stack pointer when you go into a C function is probably
> a bad thing, but as it's just the intermediate result of loading
> the module I would agree that this probably isn't "a bad thing".
>
> Cheers,
>  Sam
>
>  
The stack pointer being 0x0 at the time that sigma0 screws up (in 
init_mempool()) does seem like a bad thing to me, especially in the 
middle of a C function. Anyway, the stack pointer being 0x0 is probably 
a result of some other real problem.

If you have bochs booting the same binaries that qemu isn't booting, 
then that binary of qemu that you have is buggy or something. FWIW 
though I have qemu running under Mac OSX. If bochs is working for you, 
then I wouldn't worry about qemu, they are more or less doing the same 
thing. (except qemu is faster).
-rian