Quinn wrote:
                Anyone having any luck with booting jJOS from GRUB?  (I can
boot
        it, but it dies almost immediately with a corrupt VM error.  I
suspect
        this is a timing problem with respect to static initialization, but
I
        could be wrong -- I haven't looked into it, because I'm running it
with
        the new interrupt code currently.)

I'm getting the exact same behaviour when booting from GRUB when compiling
with -DNEW_INTERUPT_HANDLING.  When I compile without it, the run ends with
something like:
************00000009***************

I've done System.out.println's such that I know decaf is initializing (ie
the constructor is run) but I haven't had time to delve any deeper.

Avery J. Regier



> -----Original Message-----
> From: Todd L. Miller [SMTP:[EMAIL PROTECTED]]
> Sent: Wednesday, December 01, 1999 9:23 AM
> To:   [EMAIL PROTECTED]
> Subject:      [JOS-Kernel] [JOS-Arch] [DistributionGroup] Planning for
> next release (fwd)
> 
> ---------- Forwarded message ----------
> Date: Wed, 01 Dec 1999 08:49:29 -0500
> From: Gilbert Carl Herschberger II <[EMAIL PROTECTED]>
> To: [EMAIL PROTECTED]
> Subject: [JOS-Arch] [DistributionGroup] Planning for next release
> 
> Let's think about the next distribution of JOS.
> 
> When the kernel group reaches a concensus that jJOS is GRUB-bootable, it
> gives us an opportunity to create the next distribution of JOS.
> 
> ...
> 
> We also need to refine the definition of a technical distribution. A
> technical distribution requires Linux (doesn't it?). A technical
> distribution includes tools (or links to tools) so that you can (re)build
> the JOS runtime distribution from the JOS source distribution. Are there
> new tools required to build JOS? The tools.tgz archive does not include
> GRUB and other tools recently mentioned by the kernel group. And yet, a
> GRUB-bootable runtime edition needs something from GRUB.
> 
> _________________________
> 
>       Anyone having any luck with booting jJOS from GRUB?  (I can boot
> it, but it dies almost immediately with a corrupt VM error.  I suspect
> this is a timing problem with respect to static initialization, but I
> could be wrong -- I haven't looked into it, because I'm running it with
> the new interrupt code currently.)
> 
>       The current tool chain is NASM and one of etherboot+mknbi-jos and
> GRUB, right?  Where GRUB can be distributed as that pair of images (or a
> single floppy image) that George has lying around, right?  (I.E. no need
> for a source distribution for the 'runtime' version.)  Repackacking this
> would just split NASM off from etherboot+mknbi-jos and package up GRUB.
> 
> -_Quinn
> 
> 
> _______________________________________________
> Kernel maillist  -  [EMAIL PROTECTED]
> http://jos.org/mailman/listinfo/kernel

_______________________________________________
Kernel maillist  -  [EMAIL PROTECTED]
http://jos.org/mailman/listinfo/kernel

Reply via email to