Re: [PATCH 0/5] x86_64 EFI support -v3

[Andi Kleen <andi@firstfloor.org>]

ebiederm@xmission.com (Eric W. Biederman) writes:
> 
> Since there are people actively investigating things like booting
> OpenBSD via kexec things get even worse. Nothing hardly runs
> on ia64 so that issue doesn't come up.

If you want to do a popularity contest I expect there are far more ia64 
linux users than kexec-of-openbsd users.

> As for not using EFI at all.  If we can avoid it/not use it in the
> dump kernel there is very little point in having it in the primary
> kernel.

One interesting area is to use it for saving oops data. But
that has to be simple. I'm not sure complicated context switches
are a good idea here. 

However I agree it probably doesn't make sense to do virtual
mode just for the clock services -- so far we seem to be fine
just talking to the hardware directly.

> So far there don't seem to be any compelling advantages to running
> EFI in virtual address mode and several compelling disadvantages
> included having to change the permissions on the kernels memory
> map to running EFI in virtual mode.

I don't think it's a big issue to have a few less NX bits. Just
the original patch for it was ugly.
 
> Please let's stick to a physical mode trampoline and only revisit
> the topic when users start having problems because of the performance
> hit of going through our trampoline to the EFI runtime services.

So you want to switch to new page tables when calling EFI services
after boot? 

Potential problems:
- Interrupts have to be disabled. Is that ok? 
- When EFI BIOS start crashing how do we set up exception handlers
for this? 

I guess it would get complex long term. Also doesn't really sound 
attractive.

-Andi