All of lore.kernel.org
 help / color / mirror / Atom feed
From: Alejandro Vallejo <alejandro.garciavallejo@amd.com>
To: Teddy Astie <teddy.astie@vates.tech>, <xen-devel@lists.xenproject.org>
Cc: Anthony PERARD <anthony.perard@vates.tech>,
	Juergen Gross <jgross@suse.com>,
	Andrew Cooper <andrew.cooper3@citrix.com>,
	"Jan Beulich" <jbeulich@suse.com>,
	Grygorii Strashko <grygorii_strashko@epam.com>
Subject: Re: [RFC PATCH v2] x86/hvm: Allow pre-enabling x2apic mode on BSP
Date: Wed, 21 Jan 2026 11:14:53 +0100	[thread overview]
Message-ID: <DFU6SLZE4AMS.35ZPC6NEOJZ9N@amd.com> (raw)
In-Reply-To: <0cb4d1f91212a65baf924ed0ef825d8adb4b5423.1762958551.git.teddy.astie@vates.tech>

On Wed Nov 12, 2025 at 3:51 PM CET, Teddy Astie wrote:
> Introduce a new option to start the BSP vCPU in x2APIC mode instead
> of xAPIC mode. Expose this in xl through a new "x2apic_mode" option.
>
> Signed-off-by: Teddy Astie <teddy.astie@vates.tech>
> ---
> Cc: Andrew Cooper <andrew.cooper3@citrix.com>
> Cc: Jan Beulich <jbeulich@suse.com>
> Cc: Alejandro Vallejo <alejandro.garciavallejo@amd.com>
> Cc: Grygorii Strashko <grygorii_strashko@epam.com>
>
> Later on, we could consider with this option to use x2APIC ACPI
> tables instead of xAPIC ones.
>
> There is also some room into introducing a new Kconfig option to
> only support x2apic mode, which would change how the "Xen default"
> would behave.
>
> changed in v2:
>  - only pre-enable instead of forcing
>  - use domain builder to pre-enable instead of introducing a new domain creation flag

Hmmm. For dom0less/Hyperlaunch and CPU hotplug it'd be beneficial to actually
have it in the domain creation struct, I think, annoying as this might sound
(because it's a circle back to what you had before) could we keep the misc_flag
with the different meaning of "preenable x2APIC" rather than "force"?

Then on each vCPU creation (or even hotplug) we could check whether they need to
be preinitialised as x2APIC or not. Otherwise hotplug needs different treatment.

On the plus side, Hyperlaunch would merely have a new trivial binding rather
than an ad-hoc solution.

Cheers,
Alejandro


      parent reply	other threads:[~2026-01-21 10:15 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-11-12 14:51 [RFC PATCH v2] x86/hvm: Allow pre-enabling x2apic mode on BSP Teddy Astie
2026-01-21  9:24 ` [PING][RFC " Teddy Astie
2026-01-21 10:14 ` Alejandro Vallejo [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=DFU6SLZE4AMS.35ZPC6NEOJZ9N@amd.com \
    --to=alejandro.garciavallejo@amd.com \
    --cc=andrew.cooper3@citrix.com \
    --cc=anthony.perard@vates.tech \
    --cc=grygorii_strashko@epam.com \
    --cc=jbeulich@suse.com \
    --cc=jgross@suse.com \
    --cc=teddy.astie@vates.tech \
    --cc=xen-devel@lists.xenproject.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.