[Xen-ia64-devel] RE: [PATCH] Patch for loading module[2of2]

["Magenheimer, Dan (HP Labs Fort Collins)" <dan.magenheimer@xxxxxx>]

> > Um, why?  Changing something this fundamental doesn't seem
> > like a good idea.  Could you please resubmit the patch
> > to change the meaning of the parameters back?
>
> Using initrd to load Linux kernel image is only a temporally 
> workround and should be corrected.  There is no backward 
> compability issue on Xen/ia64 on loading kernel image per se. 
>  Rather, "initrd" should be maintained to load "initrd" 
> maintains the backward compability with current elilo.efi.  
> The "module" add-on is to continue to maintain elilo compability
> Imaging a single elilo.conf to have multi-initrd with two 
> different meanings, this is definitely to confuse users.  

I suppose that's a reasonable argument, though I don't agree
that your changed parameter names are any less confusing than
the original way: "module" is hardly a parameter name for
the domain0 image.

Maybe the right answer is for elilo to allow a "domain0="
parameter name, which could be an alias for "module=".
I'm not sure the elilo maintainers would like that,
but it would certainly be the least confusing alternative.

This might be a good topic for a poll... we now have over
70 people on this list.  Let's ask (see next message).

> > Yes with the parameters swapped it does seem to work,
> > at least as far as trying to mount the root disk.
> > (I don't yet have a combination of a non-rhel kernel
> > and non-rhel initrd that boots.)
> As indicated in the previous mail.  The patch only validated 
> to get to initrd, but a corresponding initrd to match with 
> xenlinux is yet to be built because no initrd is built from the tree.

Yes, I meant I haven't yet gotten RHEL4 to boot (even without
Xen) with a kernel.org kernel.

Dan

_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ia64-devel