[FFmpeg-devel] Added require fallback for libmfx in the case that pkg-config cannot find libmfx

[Michael Niedermayer]

On Fri, May 05, 2017 at 02:17:40PM +0200, wm4 wrote:
> On Fri, 5 May 2017 14:12:56 +0200
> Michael Niedermayer <michael at niedermayer.cc> wrote:
> 
> > On Fri, May 05, 2017 at 09:42:51AM +0200, Clément Bœsch wrote:
> > > On Fri, May 05, 2017 at 12:11:27AM -0700, Aaron Levinson wrote:  
> > > > On 4/19/2017 10:43 AM, Aaron Levinson wrote:  
> > > > > On 4/14/2017 6:51 PM, Aaron Levinson wrote:  
> > > > >> From e0c73c054add0137901d0bf7a7893e42e7e566c8 Mon Sep 17 00:00:00 2001
> > > > >> From: Aaron Levinson <alevinsn at aracnet.com>
> > > > >> Date: Fri, 14 Apr 2017 18:38:37 -0700
> > > > >> Subject: [PATCH] Added require fallback for libmfx in the case that
> > > > >>  pkg-config cannot find libmfx
> > > > >>
> > > > >> Purpose: Added require fallback for libmfx in the case that pkg-config
> > > > >> cannot find libmfx.  On Linux, most people likely get libmfx via
> > > > >> https://github.com/lu-zero/mfx_dispatch , but on Windows, the most
> > > > >> well-known way to get libmfx is via the Intel Media SDK, which
> > > > >> provides a static build of libmfx.lib and also provides the source
> > > > >> code for building libmfx yourself.  If built this way, there are no
> > > > >> pkg-config files to be found.  The changes utilize a similar approach
> > > > >> to that already done for libx264 in configure.
> > > > >>
> > > > >> Comments:
> > > > >>
> > > > >> -- configure: Altered enabled libmfx step to use use_pkg_config()
> > > > >>    instead of require_pkg_config(), and, if use_pkg_config() fails, it
> > > > >>    falls back to require().  Note that the reason that require() is
> > > > >>    passed -llibmfx as the last argument, instead of -lmfx, is the file
> > > > >>    name for the library produced from the Intel Media SDK starts with
> > > > >>    "libmfx".  Apparently, the filename for the library produced via
> > > > >>    https://github.com/lu-zero/mfx_dispatch starts with "mfx".
> > > > >> ---
> > > > >>  configure | 3 ++-
> > > > >>  1 file changed, 2 insertions(+), 1 deletion(-)
> > > > >>
> > > > >> diff --git a/configure b/configure
> > > > >> index 3bea057..b20a0b4 100755
> > > > >> --- a/configure
> > > > >> +++ b/configure
> > > > >> @@ -5819,7 +5819,8 @@ enabled libgsm            && { for gsm_hdr in
> > > > >> "gsm.h" "gsm/gsm.h"; do
> > > > >>                                 done || die "ERROR: libgsm not found"; }
> > > > >>  enabled libilbc           && require libilbc ilbc.h
> > > > >> WebRtcIlbcfix_InitDecode -lilbc
> > > > >>  enabled libkvazaar        && require_pkg_config "kvazaar >= 0.8.1"
> > > > >> kvazaar.h kvz_api_get
> > > > >> -enabled libmfx            && require_pkg_config libmfx
> > > > >> "mfx/mfxvideo.h" MFXInit
> > > > >> +enabled libmfx            && { use_pkg_config libmfx "mfx/mfxvideo.h"
> > > > >> MFXInit ||
> > > > >> +                               { require libmfx "mfx/mfxvideo.h"
> > > > >> MFXInit -llibmfx && warn "using libmfx without pkg-config"; } }
> > > > >>  enabled libmodplug        && require_pkg_config libmodplug
> > > > >> libmodplug/modplug.h ModPlug_Load
> > > > >>  enabled libmp3lame        && require "libmp3lame >= 3.98.3"
> > > > >> lame/lame.h lame_set_VBR_quality -lmp3lame
> > > > >>  enabled libnut            && require libnut libnut.h nut_demuxer_init
> > > > >> -lnut
> > > > >>  
> > > > > 
> > > > > Pinging this patch submission again.
> > > > > 
> > > > > Thanks,
> > > > > Aaron Levinson  
> > > > 
> > > > And again.  This patch is pretty straightforward, and considering that this approach was deemed suitable for libx264  
> > > 
> > > It wasn't, Carl Eugen insisted in doing this  
> > 
> > 
> > > against everyone opinion, but
> > > we tolerated his whim because he didn't want to use or install a 80kB
> > > package installed on virtually every system that supports x264.  
> > 
> > i remember it somewhat different but
> > Do you belive this text belongs on a technical open source mailing list
> > ?
> 
> Do you believe we should just let CE continually sabotage development
> and let him openly insult us and others?

I belive all developers have good intends and many put huge amounts of
time into improving FFmpeg. But they themselfs see each other sometimes
as enemies ...

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Does the universe only have a finite lifespan? No, its going to go on
forever, its just that you wont like living in it. -- Hiranya Peiri
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: Digital signature
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20170505/f8ae06c3/attachment.sig>