Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 24 Oct 2023 22:23:45 +0200
From:      fuz@freebsd.org
To:        Lorenzo Salvadore <developer@lorenzosalvadore.it>
Cc:        ports@freebsd.org
Subject:   Re: We need to do something about build times
Message-ID:  <ZTgn0SZuFXsUDnU5@fuz.su>
In-Reply-To: <mr-ZfzDI6LnKXUiMwCbZaHnhR6PWuHdXBIFrJRrXWU1wXiykghYHlSGtB0FicDHypxbLxcG81Sk12NI65HbxXtOwqF7TnkrIHebxsZJqFcI=@lorenzosalvadore.it>
References:  <ZTgXDSmpAq6lpT3f@fuz.su> <mr-ZfzDI6LnKXUiMwCbZaHnhR6PWuHdXBIFrJRrXWU1wXiykghYHlSGtB0FicDHypxbLxcG81Sk12NI65HbxXtOwqF7TnkrIHebxsZJqFcI=@lorenzosalvadore.it>

next in thread | previous in thread | raw e-mail | index | archive | help

--1OTt4BIAZnAKakTV
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Hi Lorenzo,

Am Tue, Oct 24, 2023 at 08:11:29PM +0000 schrieb Lorenzo Salvadore:
> Disabling LTO_BOOTSTRAP option by default has
> already been done for the devel ports on the i386, amd64
> and aarch64 architectures (so for all tier 1 platforms):
> https://cgit.freebsd.org/ports/commit/?id=3Ddaf12c76bd531335f58e7602afd3b=
c8da8fc5c33
> The same is about to be done for the production ports
> too as part of a larger update:
> https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D273397

This is great news and I hope sincerely that this change will be
applied to all ports.

> However, please note that LTO_BOOTSTRAP is an option:
> if the build times are too long for you, you can disable
> the option. The option has been enabled by default to
> distribute optimized packages, not because it is recommend
> to build the port with that option enabled. Actually, I even
> wrote a warning about it in pkg-help.

As mentioned earlier, it is impossible for me to know if non-default
build options have any effect downstream, so for acceptance testing
(i.e. testing before I commit an update), I don't think I should use
any non-default options.

Perhaps we can reach a consensus on whether LTO_BOOTSTRAP being the
default really makes sense.  The same applies to LTO use in other
ports.  Sure it's nice for performance, but for most software, there
is barely any difference and build times suffer significantly.

Maybe LTO could be enabled on a per-port basis where it has been
established that the port significantly benefits?

> I think that for most people disabling any bootstrap is
> perfectly fine, still enabling it by default should produce
> more stable official packages (I have not checked: GCC ports are
> very complex to maintain and I have lots of higher priorities
> issues to solve). Again, users are free to disable
> the option if they want faster builds and do not need any bootstrap.

Thank you for your hard work on this port.

> Cheers,
>=20
> Lorenzo Salvadore

Yours,
Robert Clausecker

--=20
()  ascii ribbon campaign - for an 8-bit clean world=20
/\  - against html email  - against proprietary attachments

--1OTt4BIAZnAKakTV
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----
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=Cq4v
-----END PGP SIGNATURE-----

--1OTt4BIAZnAKakTV--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?ZTgn0SZuFXsUDnU5>