Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 3 Jan 2021 11:20:08 +0100
From:      Matthias Andree <mandree@FreeBSD.org>
To:        Gerald Pfeifer <gerald@pfeifer.com>
Cc:        ports-committers@freebsd.org, svn-ports-all@freebsd.org, svn-ports-head@freebsd.org
Subject:   Re: svn commit: r521278 - head/graphics/rawtherapee
Message-ID:  <ada0f8ab-fe68-0daf-29c2-a8be69a4e4c2@FreeBSD.org>
In-Reply-To: <455a37b-e2cd-5c23-f118-27a31ac12a@pfeifer.com>
References:  <201912291234.xBTCYYJK087997@repo.freebsd.org> <455a37b-e2cd-5c23-f118-27a31ac12a@pfeifer.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Am 02.01.21 um 21:24 schrieb Gerald Pfeifer:
> On Sun, 29 Dec 2019, Matthias Andree wrote:
>> +# FreeBSD 11.2/12.0's base clang 6.0.1 emits worse code with
>> +# considerably higher processing times, and on i386 additional calls to
>> +# __atomic_*(), so we shall stick to GCC 9.x for now
>> +USE_GCC=	9+
> 
> We generally try to avoid referring to specific versions of GCC.
> 
> Would you mind considering something like the below? (It will 
> default to GCC 9 right now except for powerpcspe.)

> -# So we shall stick to GCC 9.x for now
> -USE_GCC=	9+
> +# So we shall stick to modern GCC for now
> +USE_GCC=	yes

Gerald,

I will sneak something similar in without revision bump (this will
follow with the ilmbase/openexr updates to 2.5.4 a bit later), but I
will add warning notes - the code is (un)practically somewhat too
sensitive to the compiler details, so in doubt we may need to pin the
GCC version later, and hope we don't incur failures on GCC 10 or 11 later...

Every now and then I've tried to move to our base compilers, without
much success. :-/

Regards
Matthias



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?ada0f8ab-fe68-0daf-29c2-a8be69a4e4c2>