From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jim Thompson <jim@netgate.com>
Subject: Re: [PATCH] eal/common: better likely() and unlikely()
Date: Mon, 20 Nov 2017 11:21:47 -0600
Message-ID: <E303DF32-0977-4341-A632-B39D7A02C252@netgate.com>
References: <1511129764-23123-1-git-send-email-Aleksey.Baulin@gmail.com>
 <D3302FD6-A529-497E-A470-5EBC1729EDF5@intel.com>
Mime-Version: 1.0 (Mac OS X Mail 11.1 \(3445.4.7\))
Content-Type: text/plain;
	charset=utf-8
Content-Transfer-Encoding: quoted-printable
Cc: Aleksey Baulin <aleksey.baulin@gmail.com>,
 Thomas Monjalon <thomas@monjalon.net>, "dev@dpdk.org" <dev@dpdk.org>
To: "Wiles, Keith" <keith.wiles@intel.com>
Return-path: <dev-bounces@dpdk.org>
Received: from mail-ot0-f193.google.com (mail-ot0-f193.google.com
 [74.125.82.193]) by dpdk.org (Postfix) with ESMTP id 65A4223A
 for <dev@dpdk.org>; Mon, 20 Nov 2017 18:21:49 +0100 (CET)
Received: by mail-ot0-f193.google.com with SMTP id o23so8262086otd.1
 for <dev@dpdk.org>; Mon, 20 Nov 2017 09:21:49 -0800 (PST)
In-Reply-To: <D3302FD6-A529-497E-A470-5EBC1729EDF5@intel.com>
List-Id: DPDK patches and discussions <dev.dpdk.org>
List-Unsubscribe: <http://dpdk.org/ml/options/dev>,
 <mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://dpdk.org/ml/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <http://dpdk.org/ml/listinfo/dev>,
 <mailto:dev-request@dpdk.org?subject=subscribe>
Errors-To: dev-bounces@dpdk.org
Sender: "dev" <dev-bounces@dpdk.org>


> On Nov 20, 2017, at 7:36 AM, Wiles, Keith <keith.wiles@intel.com> =
wrote:
>=20
>> On Nov 19, 2017, at 4:16 PM, Aleksey Baulin =
<aleksey.baulin@gmail.com> wrote:
>>=20
>> #ifndef unlikely
>> -#define unlikely(x)  __builtin_expect((x),0)
>> +#define unlikely(x)	__builtin_expect(!!(x), 0)
>=20
> I have not looked at the generated code, but does this add some extra =
instruction now to do the !!(x) ?
>=20
>> #endif /* unlikely */
>>=20
>> #endif /* _RTE_BRANCH_PREDICTION_H_ */
>> --=20
>> 2.7.4
>>=20
>=20
> Regards,
> Keith
>=20

With no =E2=80=98-O=E2=80=99, you get an extra cmpl instruction with the =
double negated unlikely() .vs the one without the =E2=80=98!!=E2=80=99.
The same assembly is generated with =E2=80=98-O=E2=80=99 as a compiler =
switch.

Tested on:
[jim@blackbox-1 ~]$ uname -a
Linux blackbox-1.netgate.com 3.10.0-514.26.2.el7.x86_64 #1 SMP Tue Jul 4 =
15:04:05 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux
[jim@blackbox-1 ~]$ gcc --version
gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-11)
Copyright (C) 2015 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is =
NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR =
PURPOSE.

Jim=