Re: [PATCH 23/25] [PATCH] paravirt hooks for arch initialization

Jeremy Fitzhardinge escreveu:

Glauber de Oliveira Costa wrote:

quoted

On 8/9/07, Alan Cox [off-list ref] wrote:
  

quoted

quoted

What's the EBDA actually used for?  The only place which seems to use
ebda_addr is in the e820 code to avoid that area as RAM.
      

It belongs to the firmware.
    

Wouldn't it be better, then, to just skip this step unconditionally if
we are running a paravirtualized guest? What do we from doing it?
  

It's better to make discover_ebda() quietly cope with a missing ebda for
whatever reason.  We could add an explicit interface to paravirt_ops to
handle this one little corner, but it isn't very important, not very
general and really its just clutter.  Its much better to have things
cope with being virtualized quietly on their own rather than hit them
all with the pv_ops hammer.   pv_ops is really for things where the
hypervisor-specific code really has to get actively involved.

I think the idea you gave me earlier of using probe_kernel_address could
work. Xen/lguest/put_yours_here that won't use an ebda would then have 
to unmap the page, to make sure a read would fault.

Glauber de Oliveira Costa wrote:

I think the idea you gave me earlier of using probe_kernel_address could
work. Xen/lguest/put_yours_here that won't use an ebda would then have
to unmap the page, to make sure a read would fault.

Hm, the memory might be mapped anyway, but we could make sure its all
zero.  discover_ebda should be able to deal with that OK.

    J