Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20230410023041.49857-1-kerneljasonxing@gmail.com> <87y1lxzmc1.ffs@tglx>
In-Reply-To: <87y1lxzmc1.ffs@tglx>
From:   Jason Xing <kerneljasonxing@gmail.com>
Date:   Tue, 9 May 2023 21:25:19 +0800
Message-ID: <CAL+tcoDyCYjdW1E7gs0EsKGAN1Esv524EeA_=ORt+Chyd5Sj5w@mail.gmail.com>
Subject: Re: [PATCH] softirq: let the userside tune the SOFTIRQ_NOW_MASK with sysctl
To:     Thomas Gleixner <tglx@linutronix.de>
Cc:     paulmck@kernel.org, peterz@infradead.org, bigeasy@linutronix.de,
        frederic@kernel.org, linux-kernel@vger.kernel.org,
        netdev@vger.kernel.org, Jason Xing <kernelxing@tencent.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Tue, May 9, 2023 at 9:05=E2=80=AFPM Thomas Gleixner <tglx@linutronix.de>=
 wrote:
>
> On Mon, Apr 10 2023 at 10:30, Jason Xing wrote:
> > From: Jason Xing <kernelxing@tencent.com>
> >
> > Currently we have two exceptions which could avoid ksoftirqd when
> > invoking softirqs: HI_SOFTIRQ and TASKLET_SOFTIRQ. They were introduced
> > in the commit 3c53776e29f8 ("Mark HI and TASKLET softirq synchronous")
> > which says if we don't mask them, it will cause excessive latencies in
> > some cases.
>
> As we are ripping this out, I'll ignore this patch.

Sure, please ignore this heuristic patch. Paolo and I have already
tested that revert patch in the production environment before and
verified its usefulness, please take that one if you could.

Thanks,
Jason