https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122281
--- Comment #22 from Benjamin Schulz <schulz.benjamin at googlemail dot com> --- Created attachment 62845 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=62845&action=edit mathedemonstrations-compute-sanitizer-log.txt this is the compute sanitizer log from mathdemonstrations.cpp of https://github.com/bschulz81/AcceleratedLinearAlgebra/tree/main full of suff like ========= Program hit CUDA_ERROR_INVALID_CONTEXT (error 201) due to "invalid device context" on CUDA API call to cuCtxGetDevice. ========= Program hit CUDA_ERROR_NOT_FOUND (error 500) due to "named symbol not found" on CUDA API call to cuModuleGetGlobal_v2. ========= Saved host backtrace up to driver entry point at error but it really seems to fail at stuff like this: With the advanced algorithms on GPU ========= Invalid __global__ read of size 8 bytes When compiled with clang, running ompute-sanitizer --tool memcheck ./mathdemonstrations > /home/benni/mathdemonstrations-compute-sanitizer-log.txt shows no errors from mathdemonstrations whatsoever, just the regular program output appears in the log
