Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.35 as permitted sender) client-ip=23.128.96.35;
MIME-Version: 1.0
References: <CABCjUKBXPLNaBjELHqz=Oke_JsxXLQauTWrsNmTp3KToW3-9MQ@mail.gmail.com>
In-Reply-To: <CABCjUKBXPLNaBjELHqz=Oke_JsxXLQauTWrsNmTp3KToW3-9MQ@mail.gmail.com>
From:   Suleiman Souhlal <suleiman@google.com>
Date:   Wed, 13 Sep 2023 23:31:49 +0900
Message-ID: <CABCjUKCRXeT2eExm6+y7cZCX4yEbj5MbjFnCHv4xvRPL5ce13g@mail.gmail.com>
Subject: Re: NOHZ interaction between IPI-less kick_ilb() and nohz_csd_func().
To:     Vincent Guittot <vincent.guittot@linaro.org>
Cc:     Ingo Molnar <mingo@redhat.com>, Juri Lelli <juri.lelli@redhat.com>,
        Linux Kernel <linux-kernel@vger.kernel.org>,
        Steven Rostedt <rostedt@google.com>,
        Joel Fernandes <joelaf@google.com>,
        Vineeth Pillai <vineethrp@google.com>,
        Youssef Esmat <youssefesmat@google.com>,
        Peter Zijlstra <peterz@infradead.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

(I forgot to also add Vincent...)

On Wed, Sep 13, 2023 at 9:49=E2=80=AFPM Suleiman Souhlal <suleiman@google.c=
om> wrote:
>
> Hello,
>
> I noticed that on x86 machines that have MWAIT, with NOHZ, when the
> kernel decides to kick the idle load balance on another CPU in
> kick_ilb(), there's an optimization that makes it avoid using an IPI
> and instead exploit the fact that the remote CPU is MWAITing on the
> thread_info flags, by just setting TIF_NEED_RESCHED, in
> call_function_single_prep_ipi().
> However, on the remote CPU, in nohz_csd_func(), we end up not raising
> the sched softirq due to NEED_RESCHED being set, so the ILB doesn't
> end up getting done.
>
> Is this intended?
>
> Thanks,
> -- Suleiman