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: <20181106134630.29591-1-smuchun@gmail.com> <CAJZ5v0ifsQqkoFw2tossFKJAUkqDLdR9NpN3eZTn3t0xUz6K-g@mail.gmail.com>
 <CAPSr9jE8EbWue31LP18eF9+KysZHNRqHNrzif8H4NkCzHHtueQ@mail.gmail.com> <CAJZ5v0iZo95ssiu3JcGJMgP+QrKLy0wu9vshk212FE3aY_kwoQ@mail.gmail.com>
In-Reply-To: <CAJZ5v0iZo95ssiu3JcGJMgP+QrKLy0wu9vshk212FE3aY_kwoQ@mail.gmail.com>
From:   Muchun Song <smuchun@gmail.com>
Date:   Tue, 6 Nov 2018 23:12:51 +0800
Message-ID: <CAPSr9jG0X9x7=krPf1OS_N2LndWcfHkREgiaQwAgY=sBQgXijg@mail.gmail.com>
Subject: Re: [PATCH] driver core: Add branch prediction hints in really_probe()
To:     rafael@kernel.org
Cc:     gregkh@linuxfoundation.org, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi Rafael,

>
> On Tue, Nov 6, 2018 at 3:43 PM Muchun Song <smuchun@gmail.com> wrote:
> >
> > Hi Rafael,
> >
> > If we want the driver core to test driver remove functions, we can
> > enable CONFIG_DEBUG_TEST_DRIVER_REMOVE. This option is
> > just for testing it. So, in most cases, the option is disabled and the if
> > condition is false. So I think we can add an unlikely() to it.
>
> Yes, it can be added there, but does it really need to be added?
>
> If the conditions are false all the time, the branch predictor in the
> processor should be able to deal with it just fine.
>
> And if they are false already at build time, the compiler should just
> optimize them away.

Thank you for your explanation. I really didn't take into account the
situation you said. Yeah, the compiler can optimize them away or
the processor can deal with it.

So,  we don't need to add unlikely() to it. The patch does't make sense.
Thanks.