[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH v3 2/4] x86/compat: Test both PV and PVH guests for compat mode

On 08/12/2015 02:23 AM, Jan Beulich wrote:
On 11.08.15 at 19:21, <boris.ostrovsky@xxxxxxxxxx> wrote:
On 08/11/2015 05:19 AM, Jan Beulich wrote:
On 24.07.15 at 19:54, <boris.ostrovsky@xxxxxxxxxx> wrote:
On 07/23/2015 10:07 AM, Jan Beulich wrote:
Plus - is this in line with what the tools are doing? Aren't they
assuming !PV <=> native format context? I.e. don't you need
to treat differently v->domain == current->domain and its
opposite? Roger btw. raised a similar question on IRC earlier
Not sure I understand this. You mean for copying 64-bit guest's info
into 32-bit dom0?
Basically yes - tool stack and guest invocations may need to
behave differently.
This being PVH-"classic" it follows exactly the PV path (both in tools
and the hypervisor). Wouldn't PV be broken then as well?
Note that I raised a question originally (still seen above) instead
of asking for a specific change. In the end all I'm asking for is that
you make changes in the hypervisor in a way compaible with tools
expectations, or adjust the tools accordingly.

So the tools have been adjusted (in the 32-bit path) to make PVH behave similarly to PV. This is obviously not what we need to move forward with no-dm PVH but this whole series is really tries to bring 32-bit PVH to parity with (existing) 64-bit PVH.

And of course you
should keep in mind what "no-dm" will want (i.e. perhaps sync with
Roger), such that we don't end up with guest exposed interface
behavior not suitable for the long term targets we have.

I suspect the notion of is_pvh_32bit_domain() (which is what this patch adds) will be irrelevant in no-dm world.


Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.