BuildSystem: conversion from jam-based to make-based system.

KNOWN ISSUES:
    1. OSX + Xcode builds do not support parallel builds; attempting to
       use them may cause unbounded number of jobs. However, disabling
       via configure --disable-xcode avoids the issue.
    2. OSX ppc binary produces binary which has audio-scan problems.
    3. OSX ppc64 binary has not been runtime tested.

ADDED:
    00-BuildUserGuide.txt
    contrib/*/module.*
    doc/
    make/

    libhb/module.*
    test/module.*
    macos/module.*
    gtk/module.*

DROPPED:
    BUILD
    BuildContribDarwin.sh
    DownloadMacOsXContribBinaries.sh
    Jamfile
    Jamrules
    MacOsXContribBinariesVersion.txt
    Makefile
    Makefile.config
    jam
    libhb/Jamefile
    libhb/Makefile
    test/BUILDSHARED
    test/Makefile
    contrib/Jamfile
    contrib/Makefile
    contrib/patch-ffmpeg.patch
    contrib/patch-x264-idr.patch
    contrib/patch-x264-linux.patch

RENAMED:
    contrib/*.patch -> contrib/MODULE/[AP]??-*.patch
    macosx/HandBrake.plist -> macosx/Info.plist

MODIFIED: libhb/decavcodec.c
    Patched to use cleaner include "libavcodec/audioconvert".
    Second part to support this cleanup is ffmpeg A02-audioconvert.patch .

MODIFIED: libhb/hb.c
MODIFIED: libhb/hb.h
MODIFIED: libhb/muxmkv.c
MODIFIED: libhb/muxmp4.c
MODIFIED: libhb/update.c
    Patched to use "project.h" for project metadata.
    Renamed HB_BUILD -> HB_PROJECT_BUILD.
    Renamed HB_VERSION -> HB_PROJECT_VERSION.

MODIFIED: test/test.c:
    Patched HandBrakeCLI to support I/O on Mac OS X ZFS filesystems.
    Reference r1803 as similar patch for HandBrake.app .
    Unpatched behavior is crash/buserror when using ZFS.

MODIFIED: macosx/Growl.framework/
    Upgraded 0.7.6 (i386,ppc) -> 1.1.2 (i386,x86_64,ppc,ppc64).
    New architectures facilitate x86_64 and ppc64 builds.

MODIFIED: macosx/HandBrake.xcodeproj/
    Bumped compatibilty mode from 2.4 -> 3.1 .
    Dumped old configurations Deployment, Developer.
    Added configurations standard, sebug as replacements.
    Added standard.i386, standard.x86_64, standard.ppc, standard.ppc64 .
    for explicit architecture build.
    All configurations settings cleaned up and normalized.
    Build output directories adjusted to support new build system.


git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@2180 b64f7644-9d1e-0410-96f1-a4d463321fa5

1 files changed, 56 insertions, 0 deletions

diff --git a/contrib/ffmpeg/A01-mpegleak.patch b/contrib/ffmpeg/A01-mpegleak.patch
new file mode 100644
index 000000000..36179ba57
--- /dev/null
+++ b/contrib/ffmpeg/A01-mpegleak.patch
@@ -0,0 +1,56 @@
+diff -Naur ffmpeg.orig/libavcodec/h264.c ffmpeg/libavcodec/h264.c
+--- ffmpeg.orig/libavcodec/h264.c	2008-12-01 11:24:19.000000000 -0500
++++ ffmpeg/libavcodec/h264.c	2009-02-21 08:23:58.000000000 -0500
+@@ -3298,7 +3298,7 @@
+          * stream. Need to discard one frame. Prevents overrun of the
+          * short_ref and long_ref buffers.
+          */
+-        av_log(h->s.avctx, AV_LOG_ERROR,
++        av_log(h->s.avctx, AV_LOG_DEBUG,
+                "number of reference frames exceeds max (probably "
+                "corrupt input), discarding one\n");
+ 
+@@ -7556,7 +7556,7 @@
+ 
+     if(!(s->flags2 & CODEC_FLAG2_CHUNKS) && !s->current_picture_ptr){
+         if (avctx->skip_frame >= AVDISCARD_NONREF || s->hurry_up) return 0;
+-        av_log(avctx, AV_LOG_ERROR, "no frame!\n");
++        av_log(avctx, AV_LOG_DEBUG, "no frame!\n");
+         return -1;
+     }
+ 
+diff -Naur ffmpeg.orig/libavcodec/mpegvideo.c ffmpeg/libavcodec/mpegvideo.c
+--- ffmpeg.orig/libavcodec/mpegvideo.c	2008-10-01 20:27:09.000000000 -0400
++++ ffmpeg/libavcodec/mpegvideo.c	2009-02-21 08:23:58.000000000 -0500
+@@ -779,19 +779,18 @@
+     }
+ 
+     av_log(s->avctx, AV_LOG_FATAL, "Internal error, picture buffer overflow\n");
+-    /* We could return -1, but the codec would crash trying to draw into a
+-     * non-existing frame anyway. This is safer than waiting for a random crash.
+-     * Also the return of this is never useful, an encoder must only allocate
+-     * as much as allowed in the specification. This has no relationship to how
+-     * much libavcodec could allocate (and MAX_PICTURE_COUNT is always large
+-     * enough for such valid streams).
+-     * Plus, a decoder has to check stream validity and remove frames if too
+-     * many reference frames are around. Waiting for "OOM" is not correct at
+-     * all. Similarly, missing reference frames have to be replaced by
+-     * interpolated/MC frames, anything else is a bug in the codec ...
+-     */
+-    abort();
+-    return -1;
++    /* XXX there seems to be a leak caused by h264 in mpeg transport
++     * streams: Over-the-air streams have a lot of errors. A picture
++     * may be marked as referenced but the actual references get lost
++     * so it never gets released. We take care of that here by releasing
++     * the oldest we have & reusing its slot. */
++    int oldest=0;
++    for(i=0; i<MAX_PICTURE_COUNT; i++){
++        if (s->picture[i].coded_picture_number < s->picture[oldest].coded_picture_number)
++            oldest = i;
++    }
++    s->avctx->release_buffer(s->avctx, (AVFrame*)&s->picture[oldest]);
++    return oldest;
+ }
+ 
+ static void update_noise_reduction(MpegEncContext *s){