Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20200124093047.008739095@linuxfoundation.org> <20200124093204.391643194@linuxfoundation.org>
 <20200126171505.GD19082@duo.ucw.cz>
In-Reply-To: <20200126171505.GD19082@duo.ucw.cz>
From:   Eric Dumazet <edumazet@google.com>
Date:   Sun, 26 Jan 2020 11:03:34 -0800
Message-ID: <CANn89iJLuoQ=R3k_Fz7ZFdhrcsvhH5WxDKbLLhEN-jkzpzz6+w@mail.gmail.com>
Subject: Re: [PATCH 4.19 599/639] net: avoid possible false sharing in sk_leave_memory_pressure()
To:     Pavel Machek <pavel@denx.de>
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        LKML <linux-kernel@vger.kernel.org>, stable@vger.kernel.org,
        Jakub Kicinski <jakub.kicinski@netronome.com>,
        Sasha Levin <sashal@kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Sun, Jan 26, 2020 at 9:15 AM Pavel Machek <pavel@denx.de> wrote:
>
> On Fri 2020-01-24 10:32:49, Greg Kroah-Hartman wrote:
> > From: Eric Dumazet <edumazet@google.com>
> >
> > [ Upstream commit 503978aca46124cd714703e180b9c8292ba50ba7 ]
> >
> > As mentioned in https://github.com/google/ktsan/wiki/READ_ONCE-and-WRITE_ONCE#it-may-improve-performance
> > a C compiler can legally transform :
> >
> > if (memory_pressure && *memory_pressure)
> >         *memory_pressure = 0;
> >
> > to :
> >
> > if (memory_pressure)
> >         *memory_pressure = 0;
>
> Well, C compiler can do a lot of stuff, and we rely on C compiler
> being "sane" -- that is gcc.
>
> Even if compiler did the transformation, that will only result in
> slightly slower performance, right?
>
> Is there any evidence this is problem in practice? Should this be in
> stable?

There is evidence of this problem in practice yes.

Should it be in stable I do not know.
Should stable kernels only be compiled by old compilers, I do not know.