Segfault on RHEL 8

[Mike_Rochefort]

Messing about on a personal workstation running RHEL 8, it does not seem like Arnold likes the distribution. Just running kick --version results in an instant segfault. Running an strace on the process shows kick writing the correct output at the very end, but breaks before this can be displayed to stdout for some reason. Running kick in a RHEL 7 container operates just fine. Checking all binaries and shared objects with ldd reports no missing objects with the exception of libanimomath_1_0.so and libbifrostapi_1_0.so on extensions/bifrost/1.5.0/libbifrost_procedural_0_2.so. Any potential pointers on figuring out what's going on?

$ echo $LD_LIBRARY_PATH
/usr/autodesk/maya2020/lib:/usr/autodesk/maya2020/plug-ins/xgen/lib:/home/myuser/arnold/maya2020/bin
$ ./kick --version ## MtoA 4.0.1.1
Segmentation fault (core dumped)

Cheers,

Mike

strace.txt

[Stephen.Blair]

Does the processor support SSE 4.1 ?

[Mike_Rochefort]

Yes, lscpu is reporting sse4_1 and sse4_2. This is a Broadwell chip, well within the hardware requirements and I haven't had an issue for several years.

For software, RHEL 8 is built on glibc-2.28 and libstdc++-8.3.1 (so.6.0.25, supporting up to GLIBCXX_3.4.25).

This can be replicated by using the ubi8 image from Red Hat and installing libX11, libxcb, and libXau.