GitLab

Merge branch 'rocblas_api_opt' of github.com:ROCmSoftwarePlatform/AMDMIGraphX into print_matmul_perf_flops

Support nonstandard shapes like slice, broadcast and transpose for the unsqueeze op

* add submodule test
* remove for loop
* simplify reshape test

Update copyright for a python file that was modified when added to our code 

There is now a MIGRAPHX_DISABLE_POINTWISE_FUSION to disable it

Support slice, broadcast and transpose shapes for the squeeze op.