[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20260122142156.GEaXIyhNQeGvOR5yyu@fat_crate.local>
Date: Thu, 22 Jan 2026 15:21:56 +0100
From: Borislav Petkov <bp@...en8.de>
To: Hou Wenlong <houwenlong.hwl@...group.com>
Cc: linux-kernel@...r.kernel.org, Thomas Gleixner <tglx@...nel.org>,
Ingo Molnar <mingo@...hat.com>,
Dave Hansen <dave.hansen@...ux.intel.com>, x86@...nel.org,
"H. Peter Anvin" <hpa@...or.com>, Juergen Gross <jgross@...e.com>,
Boris Ostrovsky <boris.ostrovsky@...cle.com>,
Peter Zijlstra <peterz@...radead.org>,
"Rafael J. Wysocki" <rafael.j.wysocki@...el.com>,
K Prateek Nayak <kprateek.nayak@....com>,
Li Chen <chenl311@...natelecom.cn>, Brian Gerst <brgerst@...il.com>,
Sohil Mehta <sohil.mehta@...el.com>,
Tim Chen <tim.c.chen@...ux.intel.com>,
Patryk Wlazlyn <patryk.wlazlyn@...ux.intel.com>,
Eric Dumazet <edumazet@...gle.com>, xen-devel@...ts.xenproject.org
Subject: Re: [PATCH 1/2] x86/smp: Move the static call update for 'smp_ops'
into smp_prepare_boot_cpu()
On Thu, Jan 22, 2026 at 10:15:43PM +0800, Hou Wenlong wrote:
> The commit 1f60230cdc63 ("x86/smp: Use static_call for
> arch_send_call_function_ipi()") changed to use a static call for
> arch_send_call_function_ipi(), which causes two problems:
>
> First, the KVM guest also changes 'smp_ops.send_call_func_ipi' when the
> PV sched yield feature is available. However, the missing
> static_call_update() breaks the PV sched yield feature.
>
> Additionally, xen_smp_init() is called before static_call_init() during
> the booting of the XENPV guest, which triggers a warning in
> __static_call_update().
>
> To simplify, move the static call update for 'smp_ops' into
> smp_prepare_boot_cpu() to address these two problems together.
>
> Fixes: 1f60230cdc63 ("x86/smp: Use static_call for arch_send_call_function_ipi()")
> Signed-off-by: Hou Wenlong <houwenlong.hwl@...group.com>
> ---
> I'm not sure if the XEN part is okay or not. I think there should be no
> IPI before smp_prepare_boot_cpu(), and even if there is, it's okay for
> the KVM guest to use the native version before smp_prepare_boot_cpu().
All three commits zapped from tip:x86/core.
We can try again once this is resolved.
Thx.
--
Regards/Gruss,
Boris.
https://people.kernel.org/tglx/notes-about-netiquette
Powered by blists - more mailing lists