I changed the ram and now the build (and other problems) are gone. I compiled from git and tried loading a dpx sequence with the internal engine and it works without any problems. I created a dpx render with "ffmpeg first" and it works. This render loads in CinGG with both "ffmpeg first" and the internal engine. I've fallen behind on the mailing list, though. Are there any patches to test? Tests to run? (no termux or ppc64 patches, please!)
On Thursday, April 7, 2022, Andrea paz via Cin <[email protected]> wrote:
I changed the ram and now the build (and other problems) are gone. I compiled from git and tried loading a dpx sequence with the internal engine and it works without any problems. I created a dpx render with "ffmpeg first" and it works. This render loads in CinGG with both "ffmpeg first" and the internal engine. I've fallen behind on the mailing list, though. Are there any patches to test? Tests to run? (no termux or ppc64 patches, please!)
try ffmpeg5 series? (i developing it slowly on termux, because its rolling distro model forces me to adapt..) https://lists.cinelerra-gg.org/pipermail/cin/2022-April/004861.html most likely 0004 and 0006 from this email, and i lately added conditional compilation around so it should work with both embedded ffmpeg 4.4 and external 4.4/5.0 hopefully attached patches should be enough for making it work on top of git. I am glad replacing ram fixed your machine! --
Cin mailing list [email protected] https://lists.cinelerra-gg.org/mailman/listinfo/cin
Andrea, if you want to add DPX (must specify read only, no writes/ rendering) to the list of native formats in the Manual, that would be good. On Thu, Apr 7, 2022 at 6:09 AM Andrea paz via Cin < [email protected]> wrote:
I changed the ram and now the build (and other problems) are gone. I compiled from git and tried loading a dpx sequence with the internal engine and it works without any problems. I created a dpx render with "ffmpeg first" and it works. This render loads in CinGG with both "ffmpeg first" and the internal engine. I've fallen behind on the mailing list, though. Are there any patches to test? Tests to run? (no termux or ppc64 patches, please!)
@Andrew I used "git am --whitespace-ignore ..." the patches 003, 007 and 0009 give error and I skipped them, the compilation was successful and I found no error messages. I don't know what to test though: am I using ffmpeg 4.4 or 5, with these patches? How do I check it? @Phyllis I have added read support for DPX. I'm wondering, how do you make render presets for the internal engine?
On Thursday, April 7, 2022, Andrea paz via Cin <[email protected]> wrote:
@Andrew I used "git am --whitespace-ignore ..." the patches 003, 007 and 0009 give error and I skipped them, the compilation was successful and I found no error messages. I don't know what to test though: am I using ffmpeg 4.4 or 5, with these patches? How do I check it?
I tried to make ./configure --without-thirdparty works, so in this case you will be using system-wide installation of libv* libs. I heard Arch keep both ffmpeg4 and ffmpeg5 (wisely) for now, so you might look at that libs you have installed and use 'ldd' to see actual ones Cingg linked to. Default still internal ffmpeg 4.4/static linking with intent to NOT breaking this case, so fact it woks as usual in itself quite good news. You probably will notice bunch of new errors from new ffmpeg filters (added in 5.0) at first startup when you successfully compile against system's ffmpeg 5.0.
@Phyllis I have added read support for DPX. I'm wondering, how do you make render presets for the internal engine?
hopefully this patch series should bring dynamic compilation with system's ffmpeg/libav* back into working order, as well fix compilation with system's openEXR 3. --with-git-ffmpeg should also work as long as you manually delete old thirdparty/src/ffmpeg.git* patches = no BluRay encoding out of the box but I hopefully have patches for this too) sorry for mixing in freebsd/termux changes, but whole need for this patch series arise from my desire to compile for those platforms I'll send whole pile of patches as separate email On Thursday, April 7, 2022, Andrew Randrianasulu <[email protected]> wrote:
On Thursday, April 7, 2022, Andrea paz via Cin <[email protected]> wrote:
@Andrew I used "git am --whitespace-ignore ..." the patches 003, 007 and 0009 give error and I skipped them, the compilation was successful and I found no error messages. I don't know what to test though: am I using ffmpeg 4.4 or 5, with these patches? How do I check it?
I tried to make ./configure --without-thirdparty works, so in this case you will be using system-wide installation of libv* libs. I heard Arch keep both ffmpeg4 and ffmpeg5 (wisely) for now, so you might look at that libs you have installed and use 'ldd' to see actual ones Cingg linked to.
Default still internal ffmpeg 4.4/static linking with intent to NOT breaking this case, so fact it woks as usual in itself quite good news. You probably will notice bunch of new errors from new ffmpeg filters (added in 5.0) at first startup when you successfully compile against system's ffmpeg 5.0.
@Phyllis I have added read support for DPX. I'm wondering, how do you make render presets for the internal engine?
I tried to compile with all your patches. I removed patch 0052 according to your instructions. The merge of 3 patches failed: 0010 Hopefully fixed dynamic liba52 0011 Cinelerra/Makefile dynamic ... 0020 Beginning of ffmpeg versioned compiled I skipped these 3 patches and continued the build using --without-thirdparty. The build failed immediately, I attach the cin5.log. PS: In your last email you talk about a "patch archive" but I don't find anything attached.
On Friday, April 8, 2022, Andrea paz <[email protected]> wrote:
I tried to compile with all your patches. I removed patch 0052 according to your instructions. The merge of 3 patches failed:
0010 Hopefully fixed dynamic liba52 0011 Cinelerra/Makefile dynamic ... 0020 Beginning of ffmpeg versioned compiled
can you reset git back to head and copy/paste failures from git am? some patches depend on each other
I skipped these 3 patches and continued the build using --without-thirdparty. The build failed immediately, I attach the cin5.log.
PS: In your last email you talk about a "patch archive" but I don't find anything attached.
hm.... no tar.gz?
The application of the patch archive occurs without errors this time. However I report the terminal messages (sorry for italian): $ git am --ignore-whitespace /home/paz/patch/*.patch Applicazione in corso: ffmpeg 4.4 bluray lpcm encoder Applicazione in corso: Fix ffmpeg 4.4 bluray lpcm patch for 5.1 ? Applicazione in corso: Add simple m2ts lpcm profile Applicazione in corso: Improve truehd decoder/encoder from ffmpeg.git Applicazione in corso: EXPERIMENT: update bdwrite for lpcm/hevc Applicazione in corso: EXPERIMENTAL: add variable chapter interval to bdwrite Applicazione in corso: EXPERIMENTAL: add bd_profile array to bdcreate.C (so in future we can diff between encoders) Applicazione in corso: EXPERIMENTAL: bdcreate: add m2ts profiles, part 1 Applicazione in corso: Fix bdcreate.C layout Applicazione in corso: EXPERIMENTAL: attempt to get value from popuptextbox .git/rebase-apply/patch:44: new blank line at EOF. + warning: 1 riga aggiunge errori di spazi bianchi. Applicazione in corso: Rename m2ts_pcm.m2ts to bluray_lpcm.m2ts Applicazione in corso: truehd experimental bluray profile Applicazione in corso: EXPERIMENTAL: add truehd audio profile for bdcreate.C Applicazione in corso: EXPERIMENTAL: add tsmuxer checkbox Applicazione in corso: EXPERIMENTAL: fix layout in bdcreate after tsmuxer checkbox Applicazione in corso: NOT YET Applicazione in corso: EXPERIMENTAL: shell script run tsmuxer conditionally on bd.meta, bd.meta conditionally written on gui var Applicazione in corso: Attempt at fixing bdwrite (stream_type/coding_type confusion) Applicazione in corso: EXPERIMENTAL: attempt at fixing bdwrite on Android after my changes, still crash Applicazione in corso: EXPERIMENTAL: real fix for bdwrite? Applicazione in corso: EXPERIMENTAL: try to add lpcm to tsmuxer meta file Applicazione in corso: EXPERIMENTAL: really fix audio names fir tsmuxer meta? Applicazione in corso: mem and resource leaks in indexfile/indextate (cppcheck) Applicazione in corso: cppcheck in cursor.C Applicazione in corso: cppcheck in resourcepixmap.C Applicazione in corso: Add objrem target for thirdparty/Makefile, removes only .o objs Applicazione in corso: Freebsd 13 conditional include in exportedl.C Applicazione in corso: fileexr forward port (openexr 3) Applicazione in corso: filegif addition of nearly all quntize.c from giflib dist, needed on freebsd/external giflib > 5.1.9 .git/rebase-apply/patch:257: trailing whitespace. sizeof(QuantizedColorType *) * warning: 1 riga aggiunge errori di spazi bianchi. Applicazione in corso: unsigned long cast for freebsd in bctrace.C Applicazione in corso: dirent64 and readdir64 aliases for freebsd in guicast/filesystem.h Applicazione in corso: realtime prio only on linux and termux for now Applicazione in corso: freebsd includes in plugins/titler Applicazione in corso: freebsd defines in guicast/bcresources.C Applicazione in corso: freebsd in indexfile.C Applicazione in corso: freebsd in file.C Applicazione in corso: freebsd in bdcreate/bdwrite Applicazione in corso: freebsd in dvdcreate.C Applicazione in corso: pkg-config based dynamic openexr link fix for freebsdv13 Applicazione in corso: disable frei0r and libvmaf for freebsd 13/dynamic ffmpeg Applicazione in corso: update blds/bsd.bld for freebsd 13 Applicazione in corso: Attempt at adding freebsd detect in plugins/Makefile Applicazione in corso: Attempt at freebsd/netbsd detection in cinelerra/Makefile Applicazione in corso: experimental: switch python in guicast to python3 unconditionally Applicazione in corso: Add android-only hacked Dynamic makefile libs .git/rebase-apply/patch:20: trailing whitespace. LIBS += -lgif -l:libtheora.a -l:libvorbis.a -ltiff warning: 1 riga aggiunge errori di spazi bianchi. Applicazione in corso: Make cin compile with ffmpeg 5 (unconditionally) .git/rebase-apply/patch:132: trailing whitespace. //ret = avcodec_parameters_from_context(codec_ctx, warning: 1 riga aggiunge errori di spazi bianchi. Applicazione in corso: Dynamic termux configure Applicazione in corso: Hopefully fixed dynamic liba52 Applicazione in corso: Cinelerra/Makefile dynamic section fix for non-android, i hope Applicazione in corso: Hopefully fix dyn liba52 on termux Applicazione in corso: Dynamic termux + libzmpeg Applicazione in corso: Really fix build without dpx Applicazione in corso: Patches for attempting to make netbsd/i386 compile .git/rebase-apply/patch:49: trailing whitespace. .git/rebase-apply/patch:54: trailing whitespace. .git/rebase-apply/patch:95: trailing whitespace. .git/rebase-apply/patch:99: trailing whitespace. .git/rebase-apply/patch:102: trailing whitespace. +LDFLAGS2 = $(LDFLAGS) warning: 101 errori di spazi bianchi soppressi warning: 106 righe aggiungono errori di spazi bianchi. Applicazione in corso: video4linux2 and DVB support currently depend on libzmpeg Applicazione in corso: Try to abort earlier if xinerama/xfixes support not here Applicazione in corso: Beginning of ffmpeg versioned compilation Applicazione in corso: Ffmpeg versioning in bdwrite.C Applicazione in corso: More libavcodec versionkng in ffmpeg Applicazione in corso: Even more ffmpeg versioning in ffmpeg.C Applicazione in corso: Nearly all versioning in ffmpeg.C Applicazione in corso: Finished versioning in ffmpeg.C Applicazione in corso: Really fix ffmpeg.C for ffmpeg 4.4 Applicazione in corso: Slightly more conditions in probe order Applicazione in corso: ffmpeg.git conditional fixes Applicazione in corso: Limit git clone depth to 1 (faster download, less space) Applicazione in corso: Add c++11 std override for filedox (fixes compile on gcc 5.5) Building with --without-thirdparty leads to error, I attach the cin5.log. I tried the build "with thirdparty" and everything is fine. Can you try to write me the ./configure that you would use? Maybe I'm wrong with mine (which is the following: ./configure --with-single-user --without-thirdparty --with-booby) Maybe this configure leads to a static build and a dynamic one is needed instead? Or should I also use --prefix=/usr/?
On Friday, April 8, 2022, Andrea paz <[email protected]> wrote:
The application of the patch archive occurs without errors this time. However I report the terminal messages (sorry for italian):
$ git am --ignore-whitespace /home/paz/patch/*.patch Applicazione in corso: ffmpeg 4.4 bluray lpcm encoder Applicazione in corso: Fix ffmpeg 4.4 bluray lpcm patch for 5.1 ? Applicazione in corso: Add simple m2ts lpcm profile Applicazione in corso: Improve truehd decoder/encoder from ffmpeg.git Applicazione in corso: EXPERIMENT: update bdwrite for lpcm/hevc Applicazione in corso: EXPERIMENTAL: add variable chapter interval to bdwrite Applicazione in corso: EXPERIMENTAL: add bd_profile array to bdcreate.C (so in future we can diff between encoders) Applicazione in corso: EXPERIMENTAL: bdcreate: add m2ts profiles, part 1 Applicazione in corso: Fix bdcreate.C layout Applicazione in corso: EXPERIMENTAL: attempt to get value from popuptextbox .git/rebase-apply/patch:44: new blank line at EOF. + warning: 1 riga aggiunge errori di spazi bianchi. Applicazione in corso: Rename m2ts_pcm.m2ts to bluray_lpcm.m2ts Applicazione in corso: truehd experimental bluray profile Applicazione in corso: EXPERIMENTAL: add truehd audio profile for bdcreate.C Applicazione in corso: EXPERIMENTAL: add tsmuxer checkbox Applicazione in corso: EXPERIMENTAL: fix layout in bdcreate after tsmuxer checkbox Applicazione in corso: NOT YET Applicazione in corso: EXPERIMENTAL: shell script run tsmuxer conditionally on bd.meta, bd.meta conditionally written on gui var Applicazione in corso: Attempt at fixing bdwrite (stream_type/coding_type confusion) Applicazione in corso: EXPERIMENTAL: attempt at fixing bdwrite on Android after my changes, still crash Applicazione in corso: EXPERIMENTAL: real fix for bdwrite? Applicazione in corso: EXPERIMENTAL: try to add lpcm to tsmuxer meta file Applicazione in corso: EXPERIMENTAL: really fix audio names fir tsmuxer meta? Applicazione in corso: mem and resource leaks in indexfile/indextate (cppcheck) Applicazione in corso: cppcheck in cursor.C Applicazione in corso: cppcheck in resourcepixmap.C Applicazione in corso: Add objrem target for thirdparty/Makefile, removes only .o objs Applicazione in corso: Freebsd 13 conditional include in exportedl.C Applicazione in corso: fileexr forward port (openexr 3) Applicazione in corso: filegif addition of nearly all quntize.c from giflib dist, needed on freebsd/external giflib > 5.1.9 .git/rebase-apply/patch:257: trailing whitespace. sizeof(QuantizedColorType *) * warning: 1 riga aggiunge errori di spazi bianchi. Applicazione in corso: unsigned long cast for freebsd in bctrace.C Applicazione in corso: dirent64 and readdir64 aliases for freebsd in guicast/filesystem.h Applicazione in corso: realtime prio only on linux and termux for now Applicazione in corso: freebsd includes in plugins/titler Applicazione in corso: freebsd defines in guicast/bcresources.C Applicazione in corso: freebsd in indexfile.C Applicazione in corso: freebsd in file.C Applicazione in corso: freebsd in bdcreate/bdwrite Applicazione in corso: freebsd in dvdcreate.C Applicazione in corso: pkg-config based dynamic openexr link fix for freebsdv13 Applicazione in corso: disable frei0r and libvmaf for freebsd 13/dynamic ffmpeg Applicazione in corso: update blds/bsd.bld for freebsd 13 Applicazione in corso: Attempt at adding freebsd detect in plugins/Makefile Applicazione in corso: Attempt at freebsd/netbsd detection in cinelerra/Makefile Applicazione in corso: experimental: switch python in guicast to python3 unconditionally Applicazione in corso: Add android-only hacked Dynamic makefile libs .git/rebase-apply/patch:20: trailing whitespace. LIBS += -lgif -l:libtheora.a -l:libvorbis.a -ltiff warning: 1 riga aggiunge errori di spazi bianchi. Applicazione in corso: Make cin compile with ffmpeg 5 (unconditionally) .git/rebase-apply/patch:132: trailing whitespace. //ret = avcodec_parameters_from_context(codec_ctx, warning: 1 riga aggiunge errori di spazi bianchi. Applicazione in corso: Dynamic termux configure Applicazione in corso: Hopefully fixed dynamic liba52 Applicazione in corso: Cinelerra/Makefile dynamic section fix for non-android, i hope Applicazione in corso: Hopefully fix dyn liba52 on termux Applicazione in corso: Dynamic termux + libzmpeg Applicazione in corso: Really fix build without dpx Applicazione in corso: Patches for attempting to make netbsd/i386 compile .git/rebase-apply/patch:49: trailing whitespace.
.git/rebase-apply/patch:54: trailing whitespace.
.git/rebase-apply/patch:95: trailing whitespace.
.git/rebase-apply/patch:99: trailing whitespace.
.git/rebase-apply/patch:102: trailing whitespace. +LDFLAGS2 = $(LDFLAGS) warning: 101 errori di spazi bianchi soppressi warning: 106 righe aggiungono errori di spazi bianchi. Applicazione in corso: video4linux2 and DVB support currently depend on libzmpeg Applicazione in corso: Try to abort earlier if xinerama/xfixes support not here Applicazione in corso: Beginning of ffmpeg versioned compilation Applicazione in corso: Ffmpeg versioning in bdwrite.C Applicazione in corso: More libavcodec versionkng in ffmpeg Applicazione in corso: Even more ffmpeg versioning in ffmpeg.C Applicazione in corso: Nearly all versioning in ffmpeg.C Applicazione in corso: Finished versioning in ffmpeg.C Applicazione in corso: Really fix ffmpeg.C for ffmpeg 4.4 Applicazione in corso: Slightly more conditions in probe order Applicazione in corso: ffmpeg.git conditional fixes Applicazione in corso: Limit git clone depth to 1 (faster download, less space) Applicazione in corso: Add c++11 std override for filedox (fixes compile on gcc 5.5)
Building with --without-thirdparty leads to error, I attach the cin5.log. I tried the build "with thirdparty" and everything is fine. Can you try to write me the ./configure that you would use?
#!/bin/sh
./configure --with-single-user --without-thirdparty \ --without-dv \ --without-firewire --without-vaapi \ --without-vdpau --without-lv2 --without-openexr --without-ogg \ --without-gl but this is on termux, I think more libs here dynamic-only. you can also look at bsd.bld: ./configure --with-single-user \ --disable-static-build --without-lv2 --disable-lame --disable-twolame \ --with-oss --without-alsa --without-firewire --without-dv --without-dvb \ --without-video4linux2 --without-xxf86vm --without-ladspa-build \ --without-libzmpeg --without-commercial --without-thirdparty \ --without-vaapi --without-vdpau --without-shuttle --without-shuttle-usb \ --without-x10tv --with-jobs=4 --without-wintv --disable-static-build hopefully will force dynamic build even if bith shared and static versions of libs oresent on system? I'll look into your build log...
Maybe I'm wrong with mine (which is the following: ./configure --with-single-user --without-thirdparty --with-booby) Maybe this configure leads to a static build and a dynamic one is needed instead? Or should I also use --prefix=/usr/?
On Friday, April 8, 2022, Andrew Randrianasulu <[email protected]> wrote:
On Friday, April 8, 2022, Andrea paz <[email protected]> wrote:
The application of the patch archive occurs without errors this time. However I report the terminal messages (sorry for italian):
$ git am --ignore-whitespace /home/paz/patch/*.patch Applicazione in corso: ffmpeg 4.4 bluray lpcm encoder Applicazione in corso: Fix ffmpeg 4.4 bluray lpcm patch for 5.1 ? Applicazione in corso: Add simple m2ts lpcm profile Applicazione in corso: Improve truehd decoder/encoder from ffmpeg.git Applicazione in corso: EXPERIMENT: update bdwrite for lpcm/hevc Applicazione in corso: EXPERIMENTAL: add variable chapter interval to bdwrite Applicazione in corso: EXPERIMENTAL: add bd_profile array to bdcreate.C (so in future we can diff between encoders) Applicazione in corso: EXPERIMENTAL: bdcreate: add m2ts profiles, part 1 Applicazione in corso: Fix bdcreate.C layout Applicazione in corso: EXPERIMENTAL: attempt to get value from popuptextbox .git/rebase-apply/patch:44: new blank line at EOF. + warning: 1 riga aggiunge errori di spazi bianchi. Applicazione in corso: Rename m2ts_pcm.m2ts to bluray_lpcm.m2ts Applicazione in corso: truehd experimental bluray profile Applicazione in corso: EXPERIMENTAL: add truehd audio profile for bdcreate.C Applicazione in corso: EXPERIMENTAL: add tsmuxer checkbox Applicazione in corso: EXPERIMENTAL: fix layout in bdcreate after tsmuxer checkbox Applicazione in corso: NOT YET Applicazione in corso: EXPERIMENTAL: shell script run tsmuxer conditionally on bd.meta, bd.meta conditionally written on gui var Applicazione in corso: Attempt at fixing bdwrite (stream_type/coding_type confusion) Applicazione in corso: EXPERIMENTAL: attempt at fixing bdwrite on Android after my changes, still crash Applicazione in corso: EXPERIMENTAL: real fix for bdwrite? Applicazione in corso: EXPERIMENTAL: try to add lpcm to tsmuxer meta file Applicazione in corso: EXPERIMENTAL: really fix audio names fir tsmuxer meta? Applicazione in corso: mem and resource leaks in indexfile/indextate (cppcheck) Applicazione in corso: cppcheck in cursor.C Applicazione in corso: cppcheck in resourcepixmap.C Applicazione in corso: Add objrem target for thirdparty/Makefile, removes only .o objs Applicazione in corso: Freebsd 13 conditional include in exportedl.C Applicazione in corso: fileexr forward port (openexr 3) Applicazione in corso: filegif addition of nearly all quntize.c from giflib dist, needed on freebsd/external giflib > 5.1.9 .git/rebase-apply/patch:257: trailing whitespace. sizeof(QuantizedColorType *) * warning: 1 riga aggiunge errori di spazi bianchi. Applicazione in corso: unsigned long cast for freebsd in bctrace.C Applicazione in corso: dirent64 and readdir64 aliases for freebsd in guicast/filesystem.h Applicazione in corso: realtime prio only on linux and termux for now Applicazione in corso: freebsd includes in plugins/titler Applicazione in corso: freebsd defines in guicast/bcresources.C Applicazione in corso: freebsd in indexfile.C Applicazione in corso: freebsd in file.C Applicazione in corso: freebsd in bdcreate/bdwrite Applicazione in corso: freebsd in dvdcreate.C Applicazione in corso: pkg-config based dynamic openexr link fix for freebsdv13 Applicazione in corso: disable frei0r and libvmaf for freebsd 13/dynamic ffmpeg Applicazione in corso: update blds/bsd.bld for freebsd 13 Applicazione in corso: Attempt at adding freebsd detect in plugins/Makefile Applicazione in corso: Attempt at freebsd/netbsd detection in cinelerra/Makefile Applicazione in corso: experimental: switch python in guicast to python3 unconditionally Applicazione in corso: Add android-only hacked Dynamic makefile libs .git/rebase-apply/patch:20: trailing whitespace. LIBS += -lgif -l:libtheora.a -l:libvorbis.a -ltiff warning: 1 riga aggiunge errori di spazi bianchi. Applicazione in corso: Make cin compile with ffmpeg 5 (unconditionally) .git/rebase-apply/patch:132: trailing whitespace. //ret = avcodec_parameters_from_contex t(codec_ctx, warning: 1 riga aggiunge errori di spazi bianchi. Applicazione in corso: Dynamic termux configure Applicazione in corso: Hopefully fixed dynamic liba52 Applicazione in corso: Cinelerra/Makefile dynamic section fix for non-android, i hope Applicazione in corso: Hopefully fix dyn liba52 on termux Applicazione in corso: Dynamic termux + libzmpeg Applicazione in corso: Really fix build without dpx Applicazione in corso: Patches for attempting to make netbsd/i386 compile .git/rebase-apply/patch:49: trailing whitespace.
.git/rebase-apply/patch:54: trailing whitespace.
.git/rebase-apply/patch:95: trailing whitespace.
.git/rebase-apply/patch:99: trailing whitespace.
.git/rebase-apply/patch:102: trailing whitespace. +LDFLAGS2 = $(LDFLAGS) warning: 101 errori di spazi bianchi soppressi warning: 106 righe aggiungono errori di spazi bianchi. Applicazione in corso: video4linux2 and DVB support currently depend on libzmpeg Applicazione in corso: Try to abort earlier if xinerama/xfixes support not here Applicazione in corso: Beginning of ffmpeg versioned compilation Applicazione in corso: Ffmpeg versioning in bdwrite.C Applicazione in corso: More libavcodec versionkng in ffmpeg Applicazione in corso: Even more ffmpeg versioning in ffmpeg.C Applicazione in corso: Nearly all versioning in ffmpeg.C Applicazione in corso: Finished versioning in ffmpeg.C Applicazione in corso: Really fix ffmpeg.C for ffmpeg 4.4 Applicazione in corso: Slightly more conditions in probe order Applicazione in corso: ffmpeg.git conditional fixes Applicazione in corso: Limit git clone depth to 1 (faster download, less space) Applicazione in corso: Add c++11 std override for filedox (fixes compile on gcc 5.5)
Building with --without-thirdparty leads to error, I attach the cin5.log. I tried the build "with thirdparty" and everything is fine. Can you try to write me the ./configure that you would use?
#!/bin/sh
./configure --with-single-user --without-thirdparty \ --without-dv \ --without-firewire --without-vaapi \ --without-vdpau --without-lv2 --without-openexr --without-ogg \ --without-gl
but this is on termux, I think more libs here dynamic-only.
you can also look at bsd.bld:
./configure --with-single-user \ --disable-static-build --without-lv2 --disable-lame --disable-twolame \ --with-oss --without-alsa --without-firewire --without-dv --without-dvb \ --without-video4linux2 --without-xxf86vm --without-ladspa-build \ --without-libzmpeg --without-commercial --without-thirdparty \ --without-vaapi --without-vdpau --without-shuttle --without-shuttle-usb \ --without-x10tv --with-jobs=4 --without-wintv
--disable-static-build hopefully will force dynamic build even if bith shared and static versions of libs oresent on system?
I'll look into your build log...
if only I can find it (log). resend please?
Maybe I'm wrong with mine (which is the following: ./configure --with-single-user --without-thirdparty --with-booby) Maybe this configure leads to a static build and a dynamic one is needed instead? Or should I also use --prefix=/usr/?
ah, a52.h header... can you send me config.log too? (to see if configure picked up liba52/a52.h header) On Friday, April 8, 2022, Andrea paz <[email protected]> wrote:
Sorry.
so, it detected header but cppflags/cflags still not updated. I found gl_ABSOLUTE_HEADER macro from gnulib/m4 but it only adds '-D PATH_TO_HEADER=///path_to_header' gcc flags and PATH_TO_HEADER define to config.h I thought confuge not just find libs but alters cflags/cppflags/ldflags - but apparently our does not quite work this way.. theres was AC_LIB_LINKFLAGS https://www.gnu.org/software/gnulib/manual/html_node/Searching-for-Libraries... but it also assumes " These macros assume that when a library is installed in some_directory/lib, its include files are installed in some_directory/include." in our case this is wrong.. so I am stuck at adding one more " -I" directive to configure.ac... On Friday, April 8, 2022, Andrea paz <[email protected]> wrote:
can you send me config.log too? (to see if configure picked up liba52/a52.h header)
I tried https://www.gnu.org/software/autoconf-archive/ax_check_library.html like this $ git diff configure.ac diff --git a/cinelerra-5.1/configure.ac b/cinelerra-5.1/configure.ac index fb65862d..3ce9b14e 100644 --- a/cinelerra-5.1/configure.ac +++ b/cinelerra-5.1/configure.ac @@ -28,7 +28,7 @@ CFG_CFLAGS+=" -D__STDC_CONSTANT_MACROS" CFG_CFLAGS+=" -D__STDC_LIMIT_MACROS" CFG_CFLAGS+=" -DPNG_SKIP_SETJMP_CHECK=1" CFG_CFLAGS+=" -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64" -CFG_CFLAGS+=" -I/usr/include -I/usr/local/include - I/data/data/com.termux/files/usr/include/a52dec" +CFG_CFLAGS+=" -I/usr/include -I/usr/local/include" CFG_CFLAGS+=" -I/usr/include/freetype2 -I/usr/local/include/freetype2" CFG_CFLAGS+=" -I/data/data/com.termux/files/usr/include/freetype2" CFG_CFLAGS+=" -I/usr/include/uuid -I/usr/local/include/uuid" @@ -645,6 +645,8 @@ CHECK_LIB([libwebp], [webp], [WebPGetEncoderVersion]) CHECK_HEADERS([libwebp], [libwebp headers], [webp/encode.h]) CHECK_LIB([a52dec], [a52], [a52_init]) CHECK_HEADERS([a52dec], [a52 headers], [stdint.h a52dec/a52.h]) +AX_CHECK_LIBRARY([a52dec], [a52dec/a52.h], [a52]) +echo $a52dec_CPPFLAGS CHECK_LIB([encore], [encore], [encore]) CHECK_HEADERS([encore], [encore headers], [encore.h]) CHECK_LIB([giflib], [gif], [DGifOpen]) @@ -1114,6 +1116,7 @@ echo " using: thirdparty build = $WANT_CIN_3RDPARTY" echo " using: single-user = $WANT_CINBIN_BUILD" echo " using: static-build = $WANT_STATIC_BUILD" echo " using: ladspa-build = $WANT_LADSPA_BUILD" +echo " using CPPFLAGS = $CPPFLAGS" echo "" AC_SUBST(WANT_CIN_3RDPARTY) @@ -1124,6 +1127,7 @@ AC_SUBST(WANT_X265_HIDEPTH) AC_SUBST(CFG_CFLAGS) AC_SUBST(CFG_CXXFLAGS) +AC_SUBST(CPPFLAGS) AC_CONFIG_FILES([Makefile]) AC_OUTPUT No bracket in bottom line (press RETURN) it prints: checking a52 headers... yes checking for stdio.h... (cached) yes checking for stdlib.h... yes checking for string.h... yes checking for inttypes.h... yes checking for stdint.h... (cached) yes checking for strings.h... yes checking for sys/stat.h... yes checking for sys/types.h... yes checking for unistd.h... yes checking for a52dec/a52.h... yes checking for main in -la52... yes checking for encore.h... no but in top-level Makefile corresponding variables are empty: .... WANT_XV = yes WANT_XXF86VM = yes a52dec_CPPFLAGS = a52dec_LDFLAGS = abs_builddir = /data/data/com.termux/files/home/cingg/cinelerra/cinelerra-5.1 abs_srcdir = /data/data/com.termux/files/home/cingg/cinelerra/cinelerra-5.1 abs_top_builddir = /data/data/com.termux/files/home/cingg/cinelerra/cinelerra-5.1 abs_top_srcdir = /data/data/com.termux/files/home/cingg/cinelerra/cinelerra-5.1 ac_ct_CC = gcc ac_ct_CXX = g++ .... I am at lost On Saturday, April 9, 2022, Andrew Randrianasulu <[email protected]> wrote:
so, it detected header but cppflags/cflags still not updated.
I found gl_ABSOLUTE_HEADER macro from gnulib/m4 but it only adds '-D PATH_TO_HEADER=///path_to_header' gcc flags and PATH_TO_HEADER define to config.h
I thought confuge not just find libs but alters cflags/cppflags/ldflags - but apparently our does not quite work this way..
theres was AC_LIB_LINKFLAGS https://www.gnu.org/software/gnulib/manual/html_node/ Searching-for-Libraries.html
but it also assumes " These macros assume that when a library is installed in some_directory/lib, its include files are installed in some_directory/include."
in our case this is wrong..
so I am stuck at adding one more " -I" directive to configure.ac...
On Friday, April 8, 2022, Andrea paz <[email protected]> wrote:
can you send me config.log too? (to see if configure picked up liba52/a52.h header)
there also migh be some roadblock down the road: https://lists.cinelerra-cv.org/pipermail/cinelerra/2017q3/006867.html but for now you either can try to run configure with CFLAGS=-I/path/to/a52dec or add your system's include path to configure.ac (like it done at beginning of this file for freetype) unfortunately liba52 does not provide .pc (pkg-config) file, so.. may be we can augment all our includes with ifdef so it will be #include "a52.h" if we have both thirdparty and a52dec building, and #include "a52dec/a52.h" if we try to use system copy... also https://www.cinelerra-gg.org/bugtracker/view.php?id=359&nbn=2 (closed, but see comment.. i wonder if compile failure was due to missing include or something else..) On Saturday, April 9, 2022, Andrew Randrianasulu <[email protected]> wrote:
I tried
https://www.gnu.org/software/autoconf-archive/ax_check_library.html
like this
$ git diff configure.ac diff --git a/cinelerra-5.1/configure.ac b/cinelerra-5.1/configure.ac index fb65862d..3ce9b14e 100644 --- a/cinelerra-5.1/configure.ac +++ b/cinelerra-5.1/configure.ac @@ -28,7 +28,7 @@ CFG_CFLAGS+=" -D__STDC_CONSTANT_MACROS" CFG_CFLAGS+=" -D__STDC_LIMIT_MACROS" CFG_CFLAGS+=" -DPNG_SKIP_SETJMP_CHECK=1" CFG_CFLAGS+=" -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64"
-CFG_CFLAGS+=" -I/usr/include -I/usr/local/include - I/data/data/com.termux/files/usr/include/a52dec" +CFG_CFLAGS+=" -I/usr/include -I/usr/local/include"
CFG_CFLAGS+=" -I/usr/include/freetype2 -I/usr/local/include/freetype2" CFG_CFLAGS+=" -I/data/data/com.termux/files/usr/include/freetype2" CFG_CFLAGS+=" -I/usr/include/uuid -I/usr/local/include/uuid" @@ -645,6 +645,8 @@ CHECK_LIB([libwebp], [webp], [WebPGetEncoderVersion]) CHECK_HEADERS([libwebp], [libwebp headers], [webp/encode.h]) CHECK_LIB([a52dec], [a52], [a52_init]) CHECK_HEADERS([a52dec], [a52 headers], [stdint.h a52dec/a52.h])
+AX_CHECK_LIBRARY([a52dec], [a52dec/a52.h], [a52]) +echo $a52dec_CPPFLAGS
CHECK_LIB([encore], [encore], [encore]) CHECK_HEADERS([encore], [encore headers], [encore.h]) CHECK_LIB([giflib], [gif], [DGifOpen]) @@ -1114,6 +1116,7 @@ echo " using: thirdparty build = $WANT_CIN_3RDPARTY" echo " using: single-user = $WANT_CINBIN_BUILD" echo " using: static-build = $WANT_STATIC_BUILD" echo " using: ladspa-build = $WANT_LADSPA_BUILD"
+echo " using CPPFLAGS = $CPPFLAGS" echo ""
AC_SUBST(WANT_CIN_3RDPARTY) @@ -1124,6 +1127,7 @@ AC_SUBST(WANT_X265_HIDEPTH) AC_SUBST(CFG_CFLAGS) AC_SUBST(CFG_CXXFLAGS) +AC_SUBST(CPPFLAGS) AC_CONFIG_FILES([Makefile]) AC_OUTPUT No bracket in bottom line (press RETURN)
it prints:
checking a52 headers... yes checking for stdio.h... (cached) yes checking for stdlib.h... yes checking for string.h... yes checking for inttypes.h... yes checking for stdint.h... (cached) yes checking for strings.h... yes checking for sys/stat.h... yes checking for sys/types.h... yes checking for unistd.h... yes checking for a52dec/a52.h... yes checking for main in -la52... yes checking for encore.h... no
but in top-level Makefile corresponding variables are empty: .... WANT_XV = yes WANT_XXF86VM = yes a52dec_CPPFLAGS = a52dec_LDFLAGS = abs_builddir = /data/data/com.termux/files/home/cingg/cinelerra/ cinelerra-5.1 abs_srcdir = /data/data/com.termux/files/home/cingg/cinelerra/ cinelerra-5.1 abs_top_builddir = /data/data/com.termux/files/home/cingg/cinelerra/cinelerra-5.1 abs_top_srcdir = /data/data/com.termux/files/home/cingg/cinelerra/ cinelerra-5.1 ac_ct_CC = gcc ac_ct_CXX = g++ ....
I am at lost
On Saturday, April 9, 2022, Andrew Randrianasulu <[email protected]> wrote:
so, it detected header but cppflags/cflags still not updated.
I found gl_ABSOLUTE_HEADER macro from gnulib/m4 but it only adds '-D PATH_TO_HEADER=///path_to_header' gcc flags and PATH_TO_HEADER define to config.h
I thought confuge not just find libs but alters cflags/cppflags/ldflags - but apparently our does not quite work this way..
theres was AC_LIB_LINKFLAGS https://www.gnu.org/software/gnulib/manual/html_node/Searchi ng-for-Libraries.html
but it also assumes " These macros assume that when a library is installed in some_directory/lib, its include files are installed in some_directory/include."
in our case this is wrong..
so I am stuck at adding one more " -I" directive to configure.ac...
On Friday, April 8, 2022, Andrea paz <[email protected]> wrote:
can you send me config.log too? (to see if configure picked up liba52/a52.h header)
Sorry, I don't understand what I need to do. In configure.ac I find the following line: 31 CFG_CFLAGS+=" -I/usr/include -I/usr/local/include -I/data/data/com.termux/files/usr/include/a52dec" How should I edit it? In my system a52dec is located in: /usr/include/a52dec (a52.h; a52_internal.h; attributes.h; audio_out.h; mm_accel.h) /usr/bin/a52dec I tried to do "CFLAGS=-I/usr/include/a52dec ./configure ...." and then compile. Still error, but later than the other times. I think it's a new error. I attach cin5.log and config.log. PS: contrary to yesterday, a patch gives me error: Applicazione in corso: Slightly more conditions in probe order error: patch non riuscita: cinelerra-5.1/cinelerra/file.C:390 error: cinelerra-5.1/cinelerra/file.C: la patch non si applica correttamente Patch non riuscita a 0063 Slightly more conditions in probe order suggerimento: Usa 'git am --show-current-patch=diff' per visualizzare la patch non riuscita Una volta risolto questo problema, esegui "git am --continue". Se preferisci saltare questa patch, esegui invece "git am --skip". Per ripristinare il branch originario e terminare il patching, esegui "git am --abort". And then: $ git am --show-current-patch=diff --- cinelerra-5.1/cinelerra/file.C | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/cinelerra-5.1/cinelerra/file.C b/cinelerra-5.1/cinelerra/file.C index 946cca1e..2a133005 100644 --- a/cinelerra-5.1/cinelerra/file.C +++ b/cinelerra-5.1/cinelerra/file.C @@ -390,7 +390,9 @@ const char *File::default_probes[] = { #ifdef HAVE_GIFLIB "GIF", #endif +#ifdef HAVE_LIBDPX "DPX", +#endif #ifdef HAVE_OPENEXR "EXR", #endif @@ -402,7 +404,9 @@ const char *File::default_probes[] = { "OGG", "Vorbis", #endif +#ifdef HAVE_LIBZMPEG "MPEG", +#endif "EDL", "FFMPEG_Late", }; -- 2.35.1
On Saturday, April 9, 2022, Andrea paz <[email protected]> wrote:
Sorry, I don't understand what I need to do. In configure.ac I find the following line:
31 CFG_CFLAGS+=" -I/usr/include -I/usr/local/include -I/data/data/com.termux/files/usr/include/a52dec"
How should I edit it?
put /usr/include/a52dec right after '-I...' (and delete termux-specific path)
In my system a52dec is located in:
/usr/include/a52dec (a52.h; a52_internal.h; attributes.h; audio_out.h; mm_accel.h)
/usr/bin/a52dec
I tried to do "CFLAGS=-I/usr/include/a52dec ./configure ...." and then compile.
Still error, but later than the other times. I think it's a new error. I attach cin5.log and config.log.
thanks, looking....
PS: contrary to yesterday, a patch gives me error:
Applicazione in corso: Slightly more conditions in probe order error: patch non riuscita: cinelerra-5.1/cinelerra/file.C:390 error: cinelerra-5.1/cinelerra/file.C: la patch non si applica correttamente Patch non riuscita a 0063 Slightly more conditions in probe order suggerimento: Usa 'git am --show-current-patch=diff' per visualizzare la patch non riuscita Una volta risolto questo problema, esegui "git am --continue". Se preferisci saltare questa patch, esegui invece "git am --skip". Per ripristinare il branch originario e terminare il patching, esegui "git am --abort".
And then:
$ git am --show-current-patch=diff
just this part was merjed, so you can skip..
--- cinelerra-5.1/cinelerra/file.C | 4 ++++ 1 file changed, 4 insertions(+)
diff --git a/cinelerra-5.1/cinelerra/file.C b/cinelerra-5.1/cinelerra/ file.C index 946cca1e..2a133005 100644 --- a/cinelerra-5.1/cinelerra/file.C +++ b/cinelerra-5.1/cinelerra/file.C @@ -390,7 +390,9 @@ const char *File::default_probes[] = { #ifdef HAVE_GIFLIB "GIF", #endif +#ifdef HAVE_LIBDPX "DPX", +#endif #ifdef HAVE_OPENEXR "EXR", #endif @@ -402,7 +404,9 @@ const char *File::default_probes[] = { "OGG", "Vorbis", #endif +#ifdef HAVE_LIBZMPEG "MPEG", +#endif "EDL", "FFMPEG_Late", }; -- 2.35.1
On Saturday, April 9, 2022, Andrea paz <[email protected]> wrote:
Sorry, I don't understand what I need to do. In configure.ac I find the following line:
31 CFG_CFLAGS+=" -I/usr/include -I/usr/local/include -I/data/data/com.termux/files/usr/include/a52dec"
How should I edit it? In my system a52dec is located in:
/usr/include/a52dec (a52.h; a52_internal.h; attributes.h; audio_out.h; mm_accel.h)
/usr/bin/a52dec
I tried to do "CFLAGS=-I/usr/include/a52dec ./configure ...." and then compile.
Still error, but later than the other times. I think it's a new error. I attach cin5.log and config.log.
ah, openexr includes... either disable openexr or add more cflags before configure ... CFLAGS=$(pkg-config --cflags OpenEXR) i tend to create small sh script so i can modify those switches and variables without re-typing a lot.. ps: for me libogg/libvorbis linking fail on termux, it may work better on Linux/gnu ld systems... but if you meet linking errs at very end you can add --without-ogg switch too. sorry for rough edges!
PS: contrary to yesterday, a patch gives me error:
Applicazione in corso: Slightly more conditions in probe order error: patch non riuscita: cinelerra-5.1/cinelerra/file.C:390 error: cinelerra-5.1/cinelerra/file.C: la patch non si applica correttamente Patch non riuscita a 0063 Slightly more conditions in probe order suggerimento: Usa 'git am --show-current-patch=diff' per visualizzare la patch non riuscita Una volta risolto questo problema, esegui "git am --continue". Se preferisci saltare questa patch, esegui invece "git am --skip". Per ripristinare il branch originario e terminare il patching, esegui "git am --abort".
And then:
$ git am --show-current-patch=diff --- cinelerra-5.1/cinelerra/file.C | 4 ++++ 1 file changed, 4 insertions(+)
diff --git a/cinelerra-5.1/cinelerra/file.C b/cinelerra-5.1/cinelerra/ file.C index 946cca1e..2a133005 100644 --- a/cinelerra-5.1/cinelerra/file.C +++ b/cinelerra-5.1/cinelerra/file.C @@ -390,7 +390,9 @@ const char *File::default_probes[] = { #ifdef HAVE_GIFLIB "GIF", #endif +#ifdef HAVE_LIBDPX "DPX", +#endif #ifdef HAVE_OPENEXR "EXR", #endif @@ -402,7 +404,9 @@ const char *File::default_probes[] = { "OGG", "Vorbis", #endif +#ifdef HAVE_LIBZMPEG "MPEG", +#endif "EDL", "FFMPEG_Late", }; -- 2.35.1
try this on top of pile... for me it detects openEXR2 (recently added to termux on my request) and compiles fine... On Saturday, April 9, 2022, Andrew Randrianasulu <[email protected]> wrote:
On Saturday, April 9, 2022, Andrea paz <[email protected]> wrote:
Sorry, I don't understand what I need to do. In configure.ac I find the following line:
31 CFG_CFLAGS+=" -I/usr/include -I/usr/local/include -I/data/data/com.termux/files/usr/include/a52dec"
How should I edit it? In my system a52dec is located in:
/usr/include/a52dec (a52.h; a52_internal.h; attributes.h; audio_out.h; mm_accel.h)
/usr/bin/a52dec
I tried to do "CFLAGS=-I/usr/include/a52dec ./configure ...." and then compile.
Still error, but later than the other times. I think it's a new error. I attach cin5.log and config.log.
ah, openexr includes...
either disable openexr or add more cflags before configure ...
CFLAGS=$(pkg-config --cflags OpenEXR)
i tend to create small sh script so i can modify those switches and variables without re-typing a lot..
ps: for me libogg/libvorbis linking fail on termux, it may work better on Linux/gnu ld systems... but if you meet linking errs at very end you can add --without-ogg switch too.
sorry for rough edges!
PS: contrary to yesterday, a patch gives me error:
Applicazione in corso: Slightly more conditions in probe order error: patch non riuscita: cinelerra-5.1/cinelerra/file.C:390 error: cinelerra-5.1/cinelerra/file.C: la patch non si applica correttamente Patch non riuscita a 0063 Slightly more conditions in probe order suggerimento: Usa 'git am --show-current-patch=diff' per visualizzare la patch non riuscita Una volta risolto questo problema, esegui "git am --continue". Se preferisci saltare questa patch, esegui invece "git am --skip". Per ripristinare il branch originario e terminare il patching, esegui "git am --abort".
And then:
$ git am --show-current-patch=diff --- cinelerra-5.1/cinelerra/file.C | 4 ++++ 1 file changed, 4 insertions(+)
diff --git a/cinelerra-5.1/cinelerra/file.C b/cinelerra-5.1/cinelerra/file.C index 946cca1e..2a133005 100644 --- a/cinelerra-5.1/cinelerra/file.C +++ b/cinelerra-5.1/cinelerra/file.C @@ -390,7 +390,9 @@ const char *File::default_probes[] = { #ifdef HAVE_GIFLIB "GIF", #endif +#ifdef HAVE_LIBDPX "DPX", +#endif #ifdef HAVE_OPENEXR "EXR", #endif @@ -402,7 +404,9 @@ const char *File::default_probes[] = { "OGG", "Vorbis", #endif +#ifdef HAVE_LIBZMPEG "MPEG", +#endif "EDL", "FFMPEG_Late", }; -- 2.35.1
this one seems to err again on fileexr (you run ./autogen.sh before configure but after applying last patch?) and filedpx - try --without-libdpx... In file included from filedpx.C:25: filedpx.h:29:10: fatal error: DPX.h: File o directory non esistente 29 | #include "DPX.h" | ^~~~~~~ compilation terminated. make[2]: *** [Makefile:585: x86_64/filedpx.o] Errore 1 make[2]: *** Attesa per i processi non terminati.... fileexr.C:35:10: fatal error: ImfChannelList.h: File o directory non es 35 | #include "ImfChannelList.h" | ^~~~~~~~~~~~~~~~~~ compilation terminated. make[2]: *** [Makefile:585: x86_64/fileexr.o] Errore 1 On Saturday, April 9, 2022, Andrea paz <[email protected]> wrote:
try this on top of pile... One more error at a later stage. Now it looks like bdwrite and lv2.
Sorry, I didn't even realize to apply the patches after autogen.sh... Still error; I attach cin5.log and config.log NOTE: Reading the config.log I noticed 2 entries that probably have nothing to do with the problems encountered, but are 2 critical things in my system: 1- in config.log you refer to ROCM; my video card on Arch has many problems with rocm. 2- Arch recently introduced experimental support for LTO in gcc. Fedora and others have had it for years, but Arch may have problems and instability.
On Sunday, April 10, 2022, Andrea paz <[email protected]> wrote:
Sorry, I didn't even realize to apply the patches after autogen.sh...
sorry, I mean you need to run autogen.sh after patching configure.ac from log: checking for ImfOpenInputFile in -lIlmImf... no checking for openexr available... no can you post output of 'pkg-config --libs OpenEXR'? also be sure you have OpenEXR headers installed we obviously better to abort configure early in case we want openexr but it not found on system for dynamic build.. need to figure out some argument for those macro functions (I read documentation on autotools, just slow to experiment)
Still error; I attach cin5.log and config.log
NOTE: Reading the config.log I noticed 2 entries that probably have nothing to do with the problems encountered, but are 2 critical things in my system:
1- in config.log you refer to ROCM; my video card on Arch has many problems with rocm.
I am sorry about this.. I once looked into their codebase.. never moved past compiling custom llvm... guess it mostly needed for Opencl > 1.1 and other forms of gpu compute... 2- Arch recently introduced experimental support for LTO in gcc.
Fedora and others have had it for years, but Arch may have problems and instability.
can you post output of 'pkg-config --libs OpenEXR'?
also be sure you have OpenEXR headers installed $ pkg-config --libs OpenEXR -pthread -lOpenEXR-3_1 -lOpenEXRUtil-3_1 -lOpenEXRCore-3_1 -lIex-3_1 -lIlmThread-3_1 -lImath-3_1
On Arch there is only the "openexr" package, which provides: usr/include/OpenEXR/ImfHeader.h and usr/bin/exrheader I don't know if they fit.
On Sunday, April 10, 2022, Andrea paz <[email protected]> wrote:
can you post output of 'pkg-config --libs OpenEXR'?
also be sure you have OpenEXR headers installed $ pkg-config --libs OpenEXR -pthread -lOpenEXR-3_1 -lOpenEXRUtil-3_1 -lOpenEXRCore-3_1 -lIex-3_1 -lIlmThread-3_1 -lImath-3_1
On Arch there is only the "openexr" package, which provides:
usr/include/OpenEXR/ImfHeader.h and usr/bin/exrheader
I don't know if they fit.
there is also https://archlinux.org/packages/extra/x86_64/imath/ as dependency, but I guess this mean we need more work for supporting openexr3 AND openexr2.... for now try to disable it manually
this patch hopefully should abort configure earlier if non-thirdparty build requested openexr, but openexr was not found on system... On Sunday, April 10, 2022, Andrew Randrianasulu <[email protected]> wrote:
On Sunday, April 10, 2022, Andrea paz <[email protected]> wrote:
can you post output of 'pkg-config --libs OpenEXR'?
also be sure you have OpenEXR headers installed $ pkg-config --libs OpenEXR -pthread -lOpenEXR-3_1 -lOpenEXRUtil-3_1 -lOpenEXRCore-3_1 -lIex-3_1 -lIlmThread-3_1 -lImath-3_1
On Arch there is only the "openexr" package, which provides:
usr/include/OpenEXR/ImfHeader.h and usr/bin/exrheader
I don't know if they fit.
there is also https://archlinux.org/packages/extra/x86_64/imath/
as dependency, but I guess this mean we need more work for supporting openexr3 AND openexr2....
for now try to disable it manually
pluginlv2config.h:35:10: fatal error: lilv/lilv.h: File o directory non 35 | #include <lilv/lilv.h> install lv2 (not just plugins but main package with this header) or add --without-lv2 ... On Monday, April 11, 2022, Andrea paz <[email protected]> wrote:
Always build error. I enclose cin5.log and config.log
attached patch should warn user at configure time about missed lilv.h header if non-thirdparty but otherwise default build attempted... On Monday, April 11, 2022, Andrew Randrianasulu <[email protected]> wrote:
pluginlv2config.h:35:10: fatal error: lilv/lilv.h: File o directory non 35 | #include <lilv/lilv.h>
install lv2 (not just plugins but main package with this header) or add --without-lv2 ...
On Monday, April 11, 2022, Andrea paz <[email protected]> wrote:
Always build error. I enclose cin5.log and config.log
participants (3)
-
Andrea paz -
Andrew Randrianasulu -
Phyllis Smith