On Tue, 2013-03-05 at 20:45 -0500, Kevin O'Connor wrote:
> 
> So, I think the question isn't what does real-hardware do (there's no
> gain to be had in emulating all of this).  Instead, I think the
> question is - what makes the most sense.  

Well,... what we implement for the 'native SeaBIOS on qemu' case doesn't
need to work on real hardware, it's true.

But a bunch of other combinations *do* need to work on real hardware, so
it's worth bearing it in mind to a certain extent.

We wouldn't want Coreboot and OVMF to have to have workarounds for qemu
behaving differently, just because we've done something simpler for
SeaBIOS. But it looks like the PAM configuration is considered part of
the memory setup and isn't lost in S3, so for the case I was wondering
about it's fine. Qemu *is* behaving like real hardware, which *is* the
simple option that makes most sense.

-- 
dwmw2

Attachment: smime.p7s
Description: S/MIME cryptographic signature

Reply via email to