[OPTIMIZER] Improved layout simplification pass so it handles swizzled layouts better (#789)

Note: uncommented `test_gemm`, since backend has an issue with swizzling. This will get uncommented in a subsequent PR.

python/triton/compiler.py

@@ -861,6 +861,9 @@ def optimize_tritongpu_ir(mod, num_stages):
     pm.add_cse_pass()
     pm.add_coalesce_pass()
     pm.add_triton_gpu_combine_pass()
+    pm.add_triton_gpu_swizzle_pass()
+    pm.add_triton_gpu_combine_pass()
+    pm.add_cse_pass()
     pm.run(mod)
     return mod