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;
Date:   Mon, 27 Nov 2023 11:27:43 +0100
From:   Christian Brauner <brauner@kernel.org>
To:     Linus Torvalds <torvalds@linux-foundation.org>
Cc:     kernel test robot <oliver.sang@intel.com>, oe-lkp@lists.linux.dev,
        lkp@intel.com, linux-kernel@vger.kernel.org,
        Jann Horn <jannh@google.com>, linux-doc@vger.kernel.org,
        linuxppc-dev@lists.ozlabs.org, intel-gfx@lists.freedesktop.org,
        linux-fsdevel@vger.kernel.org, gfs2@lists.linux.dev,
        bpf@vger.kernel.org, ying.huang@intel.com, feng.tang@intel.com,
        fengwei.yin@intel.com
Subject: Re: [linus:master] [file] 0ede61d858: will-it-scale.per_thread_ops
 -2.9% regression
Message-ID: <20231127-kirschen-dissens-b511900fa85a@brauner>
References: <202311201406.2022ca3f-oliver.sang@intel.com>
 <CAHk-=wjMKONPsXAJ=yJuPBEAx6HdYRkYE8TdYVBvpm3=x_EnCw@mail.gmail.com>
 <CAHk-=wiCJtLbFWNURB34b9a_R_unaH3CiMRXfkR0-iihB_z68A@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <CAHk-=wiCJtLbFWNURB34b9a_R_unaH3CiMRXfkR0-iihB_z68A@mail.gmail.com>
Precedence: bulk

> So that nobody else would waste any time on this, attached is a new
> attempt. This time actually tested *after* the changes.

So I've picked up your patch (vfs.misc). It's clever alright so thanks
for the comments in there otherwise I would've stared at this for far
too long.

It's a little unpleasant because of the cast-orama going on before we
check the file pointer but I don't see that it's in any way wrong. And
given how focussed people are with __fget_* performance I think it might
even be the right thing to do.

But the cleverness means we have the same logic slightly differently
twice. Not too bad ofc but not too nice either especially because that
rcu lookup is pretty complicated already.

A few days ago I did just write a long explanatory off-list email to
someone who had questions about this and who is fairly experienced so
we're not making it easy on people. But performance or simplicity; one
can't necessarily always have both.