donbowman
commented
4 years ago it appears this happens with "composite and transform" (e.g. insert this as a composition).
Open donbowman opened 4 years ago
donbowman
commented
4 years ago it appears this happens with "composite and transform" (e.g. insert this as a composition).
sandsmark
commented
4 years ago there's no assert on line 200 in filter_movit_convert.cpp in master (and hasn't been since at least early 2019, that's as far back as I bothered checking if the line numbers have changed).
do you have any local changes?
donbowman
commented
4 years ago https://github.com/mltframework/mlt/blob/master/src/modules/opengl/filter_movit_convert.cpp#L204
is the line number. it has an assert.
i have no local changes, i've tried this both with pre-packaged build and local build from git master.
sandsmark
commented
4 years ago ok, now I have no idea what I was looking at when I didn't find it...
do you get any warning messages printed before it happens? the issue seems to be that there's no _movit effect property attached to the frame, fwiw., so something doesn't attach that property when there's still images, I guess (not sure if mlt or kdenlive is supposed to be doing it)
donbowman
commented
4 years ago it crashes instantly. No warning, just instant death due to assert.
sandsmark
commented
4 years ago I don't understand the code (it looks a bit weird), but it seems like for some reason a convert filter is inserted, and that assumes there's another movit effect in front of it. This patch should in theory make it work, if you want to test:
diff --git a/src/modules/opengl/filter_movit_convert.cpp b/src/modules/opengl/filter_movit_convert.cpp
index 3d06d618..6c7089b7 100644
--- a/src/modules/opengl/filter_movit_convert.cpp
+++ b/src/modules/opengl/filter_movit_convert.cpp
@@ -201,7 +201,11 @@ static void build_fingerprint( mlt_service service, mlt_frame frame, std::string
}
Effect* effect = GlslManager::get_effect( service, frame );
- assert( effect );
+ if ( !effect ) {
+ fingerprint->append( "NO EFFECTS" );
+ return;
+ }
+
mlt_service input_a = GlslManager::get_effect_input( service, frame );
fingerprint->push_back( '(' );
build_fingerprint( input_a, frame, fingerprint );
@@ -242,7 +246,7 @@ static void build_fingerprint( mlt_service service, mlt_frame frame, std::string
static Effect* build_movit_chain( mlt_service service, mlt_frame frame, GlslChain *chain )
{
- if ( service == (mlt_service) -1 ) {
+ if ( service == (mlt_service) -1 || GlslManager::get_effect( service, frame ) == nullptr ) {
mlt_producer producer = mlt_producer_cut_parent( mlt_frame_get_original_producer( frame ) );
MltInput* input = GlslManager::get_input( producer, frame );
GlslManager::set_input( producer, frame, NULL );
faridosc
commented
4 years ago Hmmm, still crashes and has artifacts.: gdb-patch.txt
`Thread 1211 (Thread 0x7ffe77780640 (LWP 113997)):
#0 0x00000000400aff31 in ()
#1 0x00007fff3cb59710 in ()
#2 0x00007ffe13745c10 in ()
#3 0x00007ffe7777f050 in ()
#4 0x0000000000000337 in ()
#5 0x00007fffe7a1c3f6 in () at /usr/lib/libnvidia-glcore.so.455.23.04
#6 0x00007fffe7a26ffa in () at /usr/lib/libnvidia-glcore.so.455.23.04
#7 0x00007fffe7b9aa9b in () at /usr/lib/libnvidia-glcore.so.455.23.04
#8 0x00007fffe7b676de in () at /usr/lib/libnvidia-glcore.so.455.23.04
#9 0x00007fffe7b69016 in () at /usr/lib/libnvidia-glcore.so.455.23.04
#10 0x00007fffe7b81514 in () at /usr/lib/libnvidia-glcore.so.455.23.04
#11 0x00007fffcbf8c3df in movit::FlatInput::set_gl_state(unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, unsigned int*) (this=0x617000aa3b00, glsl_program_num=<optimized out>, prefix=..., sampler_num=0x7ffe7777f5d0) at flat_input.cpp:149
internal_format = <optimized out>
format = 6407
__PRETTY_FUNCTION__ = "virtual void movit::FlatInput::set_gl_state(GLuint, const string&, unsigned int*)"
#12 0x00007fffcbf7a717 in movit::EffectChain::execute_phase(movit::Phase*, std::map<movit::Phase*, unsigned int, std::less<movit::Phase*>, std::allocator<std::pair<movit::Phase* const, unsigned int> > > const&, std::vector<movit::EffectChain::DestinationTexture, std::allocator<movit::EffectChain::DestinationTexture> > const&, std::set<movit::Phase*, std::less<movit::Phase*>, std::allocator<movit::Phase*> >*) (this=this@entry=0x612000e14cc0, phase=phase@entry=0x616004128c80, output_textures=std::map with 0 elements, destinations=std::vector of length 0, capacity 0, generated_mipmaps=generated_mipmaps@entry=0x7ffe7777f6f0) at /usr/include/c++/10.1.0/bits/stl_tree.h:350
node = 0x60d000992b80
old_sampler_num = <optimized out>
i = 2
__PRETTY_FUNCTION__ = "void movit::EffectChain::execute_phase(movit::Phase*, const std::map<movit::Phase*, unsigned int>&, const std::vector<movit::EffectChain::DestinationTexture>&, std::set<movit::Phase*>*)"
instance_program_num = 54
fbo = 0
sampler_num = 1
#13 0x00007fffcbf7b266 in movit::EffectChain::render(unsigned int, std::vector<movit::EffectChain::DestinationTexture, std::allocator<movit::EffectChain::DestinationTexture> > const&, unsigned int, unsigned int, unsigned int, unsigned int) (this=0x612000e14cc0, dest_fbo=13, destinations=std::vector of length 0, capacity 0, x=0, y=0, width=1920, height=1080) at effect_chain.cpp:2109
phase = 0x616004128c80
last_phase = <optimized out>
needs_srgb = <optimized out>
phase_destinations = std::vector of length 0, capacity 0
phase_num = 0
__PRETTY_FUNCTION__ = "void movit::EffectChain::render(GLuint, const std::vector<movit::EffectChain::DestinationTexture>&, unsigned int, unsigned int, unsigned int, unsigned int)"
final_srgb = <optimized out>
current_srgb = false
generated_mipmaps = std::set with 0 elements
output_textures = std::map with 0 elements
ref_counts = std::map with 0 elements
num_phases = 1
#14 0x00007fffcbf7c075 in movit::EffectChain::render_to_fbo(unsigned int, unsigned int, unsigned int) (this=this@entry=0x612000e14cc0, dest_fbo=13, width=<optimized out>, width@entry=1920, height=<optimized out>, height@entry=1080) at /usr/include/c++/10.1.0/bits/stl_vector.h:97
x = <optimized out>
y = <optimized out>
#15 0x00007fffd113e0bb in GlslManager::render_frame_texture(movit::EffectChain*, mlt_frame_s*, int, int, unsigned char**) (this=0x6080000ee9a0, chain=chain@entry=0x612000e14cc0, frame=frame@entry=0x608000df6920, width=width@entry=1920, height=height@entry=1080, image=image@entry=0x7ffe7777fb58) at filter_glsl_manager.cpp:434
texture = 0x60300441d1e0
fbo = 13
sync = <optimized out>
#16 0x00007fffd11418bf in movit_render (image=0x7ffe7777fb58, height=1080, width=1920, output_format=mlt_image_glsl_texture, format=0x61400198d64c, frame=0x608000df6920, chain=0x612000e14cc0) at filter_movit_convert.cpp:458
glsl = <optimized out>
error = <optimized out>
leaf_service = 0x60c00592fb40
chain = 0x60d000ec0910
properties = 0x608000df6920
glsl = <optimized out>
error = 0
width = 1920
height = 1080
#17 movit_render (image=0x7ffe7777fb58, height=1080, width=1920, output_format=mlt_image_glsl_texture, format=0x61400198d64c, frame=0x608000df6920, chain=0x612000e14cc0) at filter_movit_convert.cpp:448
leaf_service = 0x60c00592fb40
chain = 0x60d000ec0910
properties = 0x608000df6920
glsl = <optimized out>
error = 0
width = 1920
height = 1080
#18 convert_image(mlt_frame, uint8_t**, mlt_image_format*, mlt_image_format) (frame=0x608000df6920, image=0x7ffe7777fb58, format=0x61400198d64c, output_format=mlt_image_glsl_texture) at filter_movit_convert.cpp:609
leaf_service = 0x60c00592fb40
chain = 0x60d000ec0910
properties = 0x608000df6920
glsl = <optimized out>
error = 0
width = 1920
height = 1080
#19 0x00007ffff70fc554 in mlt_frame_get_image (self=self@entry=0x608000df6920, buffer=buffer@entry=0x7ffe7777fb58, format=format@entry=0x61400198d64c, width=width@entry=0x7ffe7777fb44, height=height@entry=0x7ffe7777fb48, writable=writable@entry=0) at mlt_frame.c:626
properties = 0x608000df6920
get_image = 0x7fffd11473c0 <get_image(mlt_frame, uint8_t**, mlt_image_format*, int*, int*, int)>
requested_format = <optimized out>
error = <optimized out>
#20 0x00007ffff711537e in producer_get_image (self=self@entry=0x608000df31a0, buffer=buffer@entry=0x7ffe7777fb58, format=format@entry=0x61400198d64c, width=width@entry=0x7ffe7777fb44, height=height@entry=0x7ffe7777fb48, writable=writable@entry=0) at mlt_tractor.c:371
data = 0x0
size = 0
properties = 0x608000df31a0
frame = 0x608000df6920
frame_properties = 0x608000df6920
i = <optimized out>
#21 0x00007ffff70fc4f2 in mlt_frame_get_image (self=self@entry=0x608000df31a0, buffer=buffer@entry=0x7ffe7777fb58, format=format@entry=0x61400198d64c, width=width@entry=0x7ffe7777fb44, height=height@entry=0x7ffe7777fb48, writable=writable@entry=0) at mlt_frame.c:620
properties = 0x608000df31a0
get_image = 0x7ffff7115200 <producer_get_image>
requested_format = <optimized out>
error = 0
#22 0x00007ffff711255c in consumer_read_ahead_thread (arg=<optimized out>) at mlt_consumer.c:911
_mlt_log_timings_begin = 1600987042436677
_mlt_log_timings_end = <optimized out>
buffer = 1
time_current = <optimized out>
self = <optimized out>
priv = 0x61400198d640
properties = <optimized out>
width = 1920
height = 1080
video_off = <optimized out>
preview_off = <optimized out>
preview_format = <optimized out>
samples = 2002
audio = 0x625001d80910
audio_off = <optimized out>
frame = 0x608000df31a0
image = 0x60c00592fb40 "@\373\222\005\300`"
ante = {tv_sec = 1600987042, tv_usec = 430851}
count = <optimized out>
skipped = <optimized out>
time_process = <optimized out>
skip_next = <optimized out>
pos = <optimized out>
start_pos = 394
last_pos = <optimized out>
frame_duration = <optimized out>
drop_max = <optimized out>
#23 0x000055555655a1c1 in RenderThread::run() (this=0x604000af4f10) at /home/farid/kdenlive/src/monitor/glwidget.cpp:1448
#24 0x00007ffff21d2e8f in () at /usr/lib/libQt5Core.so.5
#25 0x00007ffff1db73e9 in start_thread () at /usr/lib/libpthread.so.0
#26 0x00007ffff1cc0293 in clone () at /usr/lib/libc.so.6`patch try number 2 (it deleted the input instance, but didn't clear it):
diff --git a/src/modules/opengl/filter_movit_convert.cpp b/src/modules/opengl/filter_movit_convert.cpp
index 3d06d618..10405787 100644
--- a/src/modules/opengl/filter_movit_convert.cpp
+++ b/src/modules/opengl/filter_movit_convert.cpp
@@ -594,6 +594,7 @@ static int convert_image( mlt_frame frame, uint8_t **image, mlt_image_format *fo
*image = GlslManager::get_input_pixel_pointer( producer, frame );
*format = input->get_format();
delete input;
+ GlslManager::set_input( producer, frame, nullptr );
GlslManager::get_instance()->unlock_service( frame );
return convert_on_cpu( frame, image, format, output_format );
}there's more code there I don't trust, but it's probably correct..
for fun, valgrind gives up when trying to run mlt: More than 1000 different errors detected. I'm not reporting any more. (lots of unitialized reads)
faridosc
commented
4 years ago Still crashing: gdb-patch2.txt
I left your other patch applied as well as this one, did I do it right?
Thanks :)
Thread 995 (LWP 19444):
#0 0x00000000400aff31 in ()
#1 0x00007fff152cba90 in ()
#2 0x00007ffe93dde610 in ()
#3 0x00007fff432a7050 in ()
#4 0x00000000000002da in ()
#5 0x00007fffe7a1c3f6 in () at /usr/lib/libnvidia-glcore.so.455.23.04
#6 0x00007fffe7a26ffa in () at /usr/lib/libnvidia-glcore.so.455.23.04
#7 0x00007fffe7b9aa9b in () at /usr/lib/libnvidia-glcore.so.455.23.04
#8 0x00007fffe7b676de in () at /usr/lib/libnvidia-glcore.so.455.23.04
#9 0x00007fffe7b69016 in () at /usr/lib/libnvidia-glcore.so.455.23.04
#10 0x00007fffe7b81514 in () at /usr/lib/libnvidia-glcore.so.455.23.04
#11 0x00007fffcbf8c3df in movit::FlatInput::set_gl_state(unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, unsigned int*) (this=0x617002ac4b00, glsl_program_num=<optimized out>, prefix=..., sampler_num=0x7fff432a75d0) at flat_input.cpp:149
internal_format = <optimized out>
format = 6407
__PRETTY_FUNCTION__ = "virtual void movit::FlatInput::set_gl_state(GLuint, const string&, unsigned int*)"
#12 0x00007fffcbf7a717 in movit::EffectChain::execute_phase(movit::Phase*, std::map<movit::Phase*, unsigned int, std::less<movit::Phase*>, std::allocator<std::pair<movit::Phase* const, unsigned int> > > const&, std::vector<movit::EffectChain::DestinationTexture, std::allocator<movit::EffectChain::DestinationTexture> > const&, std::set<movit::Phase*, std::less<movit::Phase*>, std::allocator<movit::Phase*> >*) (this=this@entry=0x612000119740, phase=phase@entry=0x616003abff80, output_textures=std::map with 0 elements, destinations=std::vector of length 0, capacity 0, generated_mipmaps=generated_mipmaps@entry=0x7fff432a76f0) at /usr/include/c++/10.1.0/bits/stl_tree.h:350
node = 0x60d000930f70
old_sampler_num = <optimized out>
i = 5
__PRETTY_FUNCTION__ = "void movit::EffectChain::execute_phase(movit::Phase*, const std::map<movit::Phase*, unsigned int>&, const std::vector<movit::EffectChain::DestinationTexture>&, std::set<movit::Phase*>*)"
instance_program_num = 77
fbo = 0
sampler_num = 4
#13 0x00007fffcbf7b266 in movit::EffectChain::render(unsigned int, std::vector<movit::EffectChain::DestinationTexture, std::allocator<movit::EffectChain::DestinationTexture> > const&, unsigned int, unsigned int, unsigned int, unsigned int) (this=0x612000119740, dest_fbo=13, destinations=std::vector of length 0, capacity 0, x=0, y=0, width=1920, height=1080) at effect_chain.cpp:2109
phase = 0x616003abff80
last_phase = <optimized out>
needs_srgb = <optimized out>
phase_destinations = std::vector of length 0, capacity 0
phase_num = 0
__PRETTY_FUNCTION__ = "void movit::EffectChain::render(GLuint, const std::vector<movit::EffectChain::DestinationTexture>&, unsigned int, unsigned int, unsigned int, unsigned int)"
final_srgb = <optimized out>
current_srgb = false
generated_mipmaps = std::set with 0 elements
output_textures = std::map with 0 elements
ref_counts = std::map with 0 elements
num_phases = 1
#14 0x00007fffcbf7c075 in movit::EffectChain::render_to_fbo(unsigned int, unsigned int, unsigned int) (this=this@entry=0x612000119740, dest_fbo=13, width=<optimized out>, width@entry=1920, height=<optimized out>, height@entry=1080) at /usr/include/c++/10.1.0/bits/stl_vector.h:97
x = <optimized out>
y = <optimized out>
#15 0x00007fffd113e0bb in GlslManager::render_frame_texture(movit::EffectChain*, mlt_frame_s*, int, int, unsigned char**) (this=0x6080000eeaa0, chain=chain@entry=0x612000119740, frame=frame@entry=0x608000a94d20, width=width@entry=1920, height=height@entry=1080, image=image@entry=0x7fff432a7b58) at filter_glsl_manager.cpp:434
texture = 0x603001ae6ed0
fbo = 13
sync = <optimized out>
#16 0x00007fffd11418bf in movit_render (image=0x7fff432a7b58, height=1080, width=1920, output_format=mlt_image_glsl_texture, format=0x614001b00e4c, frame=0x608000a94d20, chain=0x612000119740) at filter_movit_convert.cpp:458
glsl = <optimized out>
error = <optimized out>
leaf_service = 0x60c0066af840
chain = 0x60d001a393f0
properties = 0x608000a94d20
glsl = <optimized out>
error = 0
width = 1920
height = 1080
#17 movit_render (image=0x7fff432a7b58, height=1080, width=1920, output_format=mlt_image_glsl_texture, format=0x614001b00e4c, frame=0x608000a94d20, chain=0x612000119740) at filter_movit_convert.cpp:448
leaf_service = 0x60c0066af840
chain = 0x60d001a393f0
properties = 0x608000a94d20
glsl = <optimized out>
error = 0
width = 1920
height = 1080
#18 convert_image(mlt_frame, uint8_t**, mlt_image_format*, mlt_image_format) (frame=0x608000a94d20, image=0x7fff432a7b58, format=0x614001b00e4c, output_format=mlt_image_glsl_texture) at filter_movit_convert.cpp:609
leaf_service = 0x60c0066af840
chain = 0x60d001a393f0
properties = 0x608000a94d20
glsl = <optimized out>
error = 0
width = 1920
height = 1080
#19 0x00007ffff70fc554 in mlt_frame_get_image (self=self@entry=0x608000a94d20, buffer=buffer@entry=0x7fff432a7b58, format=format@entry=0x614001b00e4c, width=width@entry=0x7fff432a7b44, height=height@entry=0x7fff432a7b48, writable=writable@entry=0) at mlt_frame.c:626
properties = 0x608000a94d20
get_image = 0x7fffd11473c0 <get_image(mlt_frame, uint8_t**, mlt_image_format*, int*, int*, int)>
requested_format = <optimized out>
error = <optimized out>
#20 0x00007ffff711537e in producer_get_image (self=self@entry=0x6080004fb7a0, buffer=buffer@entry=0x7fff432a7b58, format=format@entry=0x614001b00e4c, width=width@entry=0x7fff432a7b44, height=height@entry=0x7fff432a7b48, writable=writable@entry=0) at mlt_tractor.c:371
data = 0x0
size = 0
properties = 0x6080004fb7a0
frame = 0x608000a94d20
frame_properties = 0x608000a94d20
i = <optimized out>
#21 0x00007ffff70fc4f2 in mlt_frame_get_image (self=self@entry=0x6080004fb7a0, buffer=buffer@entry=0x7fff432a7b58, format=format@entry=0x614001b00e4c, width=width@entry=0x7fff432a7b44, height=height@entry=0x7fff432a7b48, writable=writable@entry=0) at mlt_frame.c:620
properties = 0x6080004fb7a0
get_image = 0x7ffff7115200 <producer_get_image>
requested_format = <optimized out>
error = 0
#22 0x00007ffff711255c in consumer_read_ahead_thread (arg=<optimized out>) at mlt_consumer.c:911
_mlt_log_timings_begin = 1601058839765672
_mlt_log_timings_end = <optimized out>
buffer = 1
time_current = <optimized out>
self = <optimized out>
priv = 0x614001b00e40
properties = <optimized out>
width = 1920
height = 1080
video_off = <optimized out>
preview_off = <optimized out>
preview_format = <optimized out>
samples = 2002
audio = 0x625000d45910
audio_off = <optimized out>
frame = 0x6080004fb7a0
image = 0x60c0066af840 "@\370j\006\300`"
ante = {tv_sec = 1601058839, tv_usec = 760081}
count = <optimized out>
skipped = <optimized out>
time_process = <optimized out>
skip_next = <optimized out>
pos = <optimized out>
start_pos = 1309
last_pos = <optimized out>
frame_duration = <optimized out>
drop_max = <optimized out>
#23 0x000055555655d911 in RenderThread::run() (this=0x604001996010) at /home/farid/kdenlive/src/monitor/glwidget.cpp:1448
#24 0x00007ffff21d2e8f in () at /usr/lib/libQt5Core.so.5
#25 0x00007ffff1db73e9 in start_thread () at /usr/lib/libpthread.so.0
#26 0x00007ffff1cc0293 in clone () at /usr/lib/libc.so.6yup
donbowman
commented
4 years ago any update? anything i can do to help?
sandsmark
commented
4 years ago tldr: I'm a bit stuck, things point to mlt not handling RGB input correctly, valgrind is useless, but trying with asan and ubsan might point out something.
I'm a bit stuck, I don't know the mlt code well enough (you can also see some more progress here: https://github.com/mltframework/mlt/issues/608).
The issue seems to be something related to when mlt tries to use RGB input. And since the crash always seem to happen in when movit tries to upload the texture I'm fairly sure it's because MLT sends in some invalid memory (e. g.either an old buffer it has freed, it sets the type of data wrong so Movit thinks the buffer it gets passed is bigger than it is, or something similar).
So the crash isn't really in movit at all, movit is just telling the graphics driver to upload the buffer it has gotten from MLT with glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, width, height, format, type, pixel_data);.
And as I indicated above, there's so much use of uninitialized memory in MLT that valgrind gives up (More than 1000 different errors detected. I'm not reporting any more.), which I hope are false positives, but it makes it useless to use valgrind to see if MLT is trying to use uninitialized or freed memory.
One possible next step would be to use -fsanitize=address -fsanitize=undefined when building mlt, but that doesn't always catch everything (valgrind is a bit better there). In addition it might trigger too early on the same issues that valgrind detects, but it isn't as good as valgrind wrt. continuing execution with invalid memory usage.
faridosc
commented
4 years ago Yikes. Godspeed! :)
faridosc
commented
4 years ago Hi @sesse sorry to bother you with this, I am trying to compile everything from clean and building movit with enable debug flag is not working anymore (I am on arch so using latest release):
configure: WARNING: unrecognized options: --enable-debug
make: *** No rule to make target '/usr/src/gtest/src/gtest-all.cc', needed by 'gtest-all.o'. Stop.I did install gtest but still get the same message... Any clue?
Thanks <3
Btw,are you still maintaining movit?
sesse
commented
4 years ago On Tue, Oct 06, 2020 at 05:56:10AM -0700, faridosc wrote:
Hi @sesse sorry to bother you with this, I am trying to compile everything from clean and building movit with enable debug flag is not working anymore (I am on arch so using latest release):
Hi,
configure: WARNING: unrecognized options: --enable-debug
I don't think --enable-debug has ever done anything in Movit's configure.
make: *** No rule to make target '/usr/src/gtest/src/gtest-all.cc', needed by 'gtest-all.o'. Stop.
I did install gtest but still get the same message... Any clue?
I believe Arch Linux has a gtest package with a precompiled .so, which is not recommended by Google Test upstream since 2010 (see http://groups.google.com/group/googletestframework/browse_thread/thread/668eff1cebf5309d). Movit does not support this; you'll need to point GTEST_DIR to wherever one can find gtest-all.cc, so it can compile it itself.
Btw,are you still maintaining movit?
I still respond to bugs and such if needed, but I don't add a whole not of new filters and such these days, no.
/ Steinar / -- Homepage: https://www.sesse.net/
donbowman
commented
4 years ago ./configure --enable-debug --enable-gpl3 --enable-gpl
worked for me, followed by make -j128
files have -g:
cc -I../../ -DARCH_X86_64 -Wall -DPIC -DUSE_MMX -mmmx -DUSE_SSE -msse -DUSE_SSE2 -msse2 -g -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -fPIC -pthread -DARCH_X86 -c -o cpu_accel.o cpu_accel.cin configure, --enable-debug causes removal of -O2 -pipe -fno-tree-pre -fno-tree-dominator-opts
sandsmark
commented
3 years ago configure: WARNING: unrecognized options: --enable-debug make: *** No rule to make target '/usr/src/gtest/src/gtest-all.cc', needed by 'gtest-all.o'. Stop.
gtest-all.cc is at /usr/src/googletest/src/gtest-all.cc in archlinux (according to https://www.archlinux.org/packages/community/x86_64/gtest/).
TheCompGeek
commented
3 years ago Kdenlive 20.12.0, instant death upon encountering a Composite and Transform transition with GPU rendering enabled.
(reproduction steps and materiall in the attachment along with system setup, libraries, etc).
Whenever GPU acceleration is enabled in kdenlive master + libmlt master, if I do composite transform of an image onto a video, it crashes with this assertion (kdenlive crashes on load). I can't generate the mlt script, kdenlive crashes immediately.
libmlt 6.21+git202005312040-1~ubuntu20.04.1 kdenlive 4:20.07+git202006011026~ubuntu20.04.1
this happens all the time, regardless of source material.
It also happens w/ just melt command line.
bug.zip