From mboxrd@z Thu Jan  1 00:00:00 1970
From: Michael Tremer <michael.tremer@ipfire.org>
To: development@lists.ipfire.org
Subject: Re: Optimization for 'monit'
Date: Wed, 05 Jun 2019 09:10:21 +0100
Message-ID: <B13509A3-1CF2-4AD4-A5BC-B5CF95183D73@ipfire.org>
In-Reply-To: <3bcf533e-7c7b-8d6d-b1af-56ede79a6161@ipfire.org>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="===============6650907069470877634=="
List-Id: <development.lists.ipfire.org>

--===============6650907069470877634==
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable

Morning :)

> On 5 Jun 2019, at 08:52, Matthias Fischer <matthias.fischer(a)ipfire.org> w=
rote:
>=20
> Hi,
>=20
> could someone please take a look at the '--enable-optimized'-option in
> 'monit'-configure:
>=20
> The corresponding code from 'configure' reads:
>=20
> ...
> # Check whether --enable-optimized was given.
> if test "${enable_optimized+set}" =3D set; then :
>  enableval=3D$enable_optimized;
>        CFLAGS=3D`echo $CFLAGS|sed 's/-O.//g'`
>        if test "x$enableval" =3D "xyes" ; then
>                CFLAGS=3D`echo $CFLAGS|sed 's/-g[^ ]*//g'`
>                CFLAGS=3D`echo $CFLAGS|sed 's/-O.//g'`
>                CFLAGS=3D"$CFLAGS -O3 -DNDEBUG"
>                OPTIMIZED=3D1
>        else
>                OPTIMIZED=3D0
>        fi
> ...
>=20
> Since I'm not so good in 'CFGLAG'-options ( ;-) ), I'd like to know what
> this option REALLY does and if we should use it - prior to pushing it to
> GIT.

So this removes all -O options and replaces them with -O3.

That is the highest level of code optimisation in the compiler. It will spend=
 a lot of extra time to find out how to make the code faster. That might all =
sound desirable, but in reality it does not really help much. The code will g=
row in size and on computers with smaller caches (i.e. Atom series) the code =
will perform significantly slower.

We have this enabled for a couple things where the processor spends a lot of =
time (glib used to be compiled with -O3). Boost uses it by default, because t=
hey only care about performance on large systems.

Leave it disabled, please. We use -O2 everywhere.

> Second option would be '--without-ipv6' since I think we do not really
> need this in v2.23.

Not sure if is necessary but we have IPv6 compiled into many things.

> Its running here on Core 131 - compiled with these two options and
> without seen problems - but I just want to be sure.
>=20
> What do you think?

See above :)

-Michael

>=20
> Best,
> Matthias


--===============6650907069470877634==--