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

Re: [PATCH 13/17] libxl: Add xen-platform device for Q35 machine


  • To: Thierry Escande <thierry.escande@xxxxxxxxxx>
  • From: Anthony PERARD <anthony.perard@xxxxxxxxxx>
  • Date: Fri, 19 Jun 2026 16:58:24 +0200
  • Authentication-results: eu.smtp.expurgate.cloud; dkim=pass header.s=selector1 header.d=vates.tech header.i="@vates.tech" header.h="From:Subject:Date:Message-ID:To:Cc:MIME-Version:Content-Type:In-Reply-To:References:Feedback-ID"
  • Cc: xen-devel@xxxxxxxxxxxxxxxxxxxx, Juergen Gross <jgross@xxxxxxxx>
  • Delivery-date: Fri, 19 Jun 2026 14:58:34 +0000
  • Feedback-id: default:8631fc262581453bbf619ec5b2062170:Sweego
  • List-id: Xen developer discussion <xen-devel.lists.xenproject.org>

On Fri, Mar 13, 2026 at 04:35:04PM +0000, Thierry Escande wrote:
> Current Xen/QEMU method to control Xen Platform device is done by
> setting the 'xen_platform_device' option value

You mean `xen_platform_pci` ?

This is still the case with the new type of machine ;-).

> that modifies QEMU
> emulated machine type, namely xenfv <--> pc.

> In order to avoid multiplying machine types, this patch supplies
> '-device xen-platform' directly to Qemu. To maintain backward
> compatibility with existing Xen/QEMU setups, this is currently only
> applicable to q35 machine.

We can already use `-device xen-platform` with QEMU's machine `pc` /
`pc-i440fx-*` but this would break migration from previous version of
libxl, PCI devices would move around while the guest is running.

So instead of "currently only applicable", something like "we will only
apply this to q35" would be a bit more accurate.

> i440 emulation uses the old method (xenfv/pc
> machine) to control Xen Platform device.

To be honest, the change between `xenfv` and `pc` is bigger than just
that device, it's also a different machine version.

`-machine xenfv` is kind of equivalent to:
`-machine pc-i440fx-3.1,accel=xen,suppress-vmdesc=on -device xen-platform`

But `xenfv` can also have `igd-passthru=on` added to `-machine` which
has an influence on the machine setup.

If you manage to rework the commit that would be nice, in anycase,
overall the patch looks fine: Reviewed-by: Anthony PERARD 
<anthony.perard@xxxxxxxxxx>

Cheers,


--
Anthony Perard | Vates XCP-ng Developer

XCP-ng & Xen Orchestra - Vates solutions

web: https://vates.tech

 


Rackspace

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