Workgroup Reversal

[shivadbhavsar]

Add functionality to apply workgroup reversals to increase cache hits.

[shivadbhavsar]

Initial work resulted in no perf difference. Rocprofiler results on trimmed unet:

1. using MIGRAPHX_MLIR_USE_SPECIFIC_OPS=attention - Cache hits are mostly the same with and without reversal (with some being considerably lower)
2. using `MIGRAPHX_MLIR_USE_SPECIFIC_OPS=convolution,dot,fused,attention' - Cache hits are noticeably higher for some kernels with reversals, but overall perf is consistently worse with reversals

[shivadbhavsar]

Next Steps: Understand cache hits with even smaller graphs

1. Performed test with mul -> dot -> add program which is compiled as mul -> dot_add where mlir_dot_add is reverse indexed when reversal is applied. There is no change in cache hits when reversal is applied