creates a SIGSEGV and even corrupts a part of my screen.
Full output is: Setting pipeline to PAUSED ... Pipeline is PREROLLING ... nvbuf_utils: dmabuf_fd -1 mapped entry NOT found nvbuf_utils: Can not get HW buffer from FD... Exiting... Caught SIGSEGV #0 0x0000007f984b9e28 in __GI___poll (fds=0x55be758fe0, nfds=548017041976, timeout=<optimized out>) at ../sysdeps/unix/sysv/linux/poll.c:41 #1 0x0000007f985c6e08 in () at /usr/lib/aarch64-linux-gnu/libglib-2.0.so.0 #2 0x00000055be483b30 in () Spinning. Please run 'gdb gst-launch-1.0 31455' to continue debugging, Ctrl-C to quit, or Ctrl-\ to dump core.
Am I doing anything wrong?
It works fine without the nvvideoconvert, but I`ll need that block for a later extension of the pipeline.
But the SIGSEGV should not happen, even when I pipe in a “random” bitstream, should it?
Meanwhile I also managed to get this pipeline to work: gst-launch-1.0 filesrc location=sample_720p.mjpeg ! nvjpegdec ! nvvideoconvert ! nvegltransform ! nveglglessink
while this variant SIGSEGVs again (and makes an unrelated process crash): gst-launch-1.0 filesrc location=sample_720p.mjpeg ! nvjpegdec ! nvvideoconvert ! 'video/x-raw' ! xvimagesink