Re: [Xen-devel] [PATCH v1 04/12] xen/hvmlite: Bootstrap HVMlite guest

From: "Luis R. Rodriguez" <mcgrof@suse.com>
To: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: "Andrew Cooper" <andrew.cooper3@citrix.com>,
	"Roger Pau Monné" <roger.pau@citrix.com>,
	"Juergen Gross" <jgross@suse.com>,
	"Jeremy Fitzhardinge" <jeremy@goop.org>,
	"Rusty Russell" <rusty@rustcorp.com.au>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"Andy Lutomirski" <luto@amacapital.net>,
	"David Vrabel" <david.vrabel@citrix.com>,
	"H. Peter Anvin" <hpa@zytor.com>,
	xen-devel@lists.xenproject.org, "Borislav Petkov" <bp@suse.de>,
	"X86 ML" <x86@kernel.org>
Subject: Re: [Xen-devel] [PATCH v1 04/12] xen/hvmlite: Bootstrap HVMlite guest
Date: Wed, 27 Jan 2016 01:04:35 +0100	[thread overview]
Message-ID: <20160127000435.GK20964@wotan.suse.de> (raw)
In-Reply-To: <56A7EA6A.2030502@oracle.com>

On Tue, Jan 26, 2016 at 04:51:38PM -0500, Boris Ostrovsky wrote:
> On 01/26/2016 03:30 PM, Luis R. Rodriguez wrote:

> hvmlite_start() is a 32-bit entry point [...]
>
> >4) hardware_subarch, hardware_subarch_data and future prospects
> >
> >Your patch relies on a *new* Linux entry point. Sure, we had one
> >for EFI, and sure there is another for Xen PV, but since you're just
> >rebranding PVH to HVMlite and given historic issues with any new
> >Linux entry points I'd like for us to take a breather and evaluate
> >the extent our lofty unification goals, and how the x86 boot protocol
> >could help with this already.
> 
> I am not sure I see how you can avoid having new entry point. For
> example, all HVMlite guests start in 32-bit mode. Who will switch to
> long mode?

x86 i386 entry points need to have code to do all that stuff, this can
happen for instance when you boot x86_64 from a 32-bit boot loader,
and I think other things as well are possible that trigger this as well.

> >why not just use a union and differentiate on PV subtype ? If you want to avoid
> >a lot of PV calls for HVMlite it seems you could just take advantage of
> >subarch Xen type, and differentiate on the subarch_data within Xen code
> >to make that code just PV sepecific.
> >
> >I only see gains by using the Xen subarch, so would like to know why PC is
> >being pushed.
> 
> It's not that subarch 0 is being pushed here. It's that I don't see
> how it can be useful for this particular guest type. Maybe as we add
> new features (or discover something that we've missed up till now)
> we can switch to using it. If you think we should delay initializing
> boot_params until then --- I will disagree: boot_params are used
> before we look at subarch and I don't believe it makes sense to pick
> and choose what to initialize before and what can wait.

subarch is part of boot_params, so not sure what I mean by an issue
in timing here. The question is if its set and then how early can
you possibly read the subarch from the boot_params.

> (And I am not sure it can be useful on PV neither, 

Well..

> at least the way it is used now. 

that is the issue.. If we get access to boot_params on early boot
then we can simply share the x86_64 entry point between PVH, PV,
and native x86_64 with what I'm proposing and some minor extensions.

> You will not reach the point in the (32-bit) code
> where it is tested. You will die way earlier (I think on
> startup_32()'s fourth instruction).)

Its a bit different requirement for the subarch for PV/PVH and
for HVMlite. Given what you have explained things are bit clearer
now and I see the issue.

You go:

hvmlite_start_xen() -->
	HVM stub
	startup_64() | (startup_32()

Note at the end of startup_32() though we have a neat asm
set of entries that depend on the subarch type. Perhaps we should
have a PV type early on startup_32() ? And/or I wonder if we can
work off of the EFI boot loader code. Half baked thoughts for now,
sorry have to go.

  Luis