Rambo f64 emulation

Author: Hardcode84

dpbench/benchmarks/rambo/rambo_numba_mlir_k.py

@@ -7,7 +7,7 @@
 import numba_mlir.kernel as nb
 
 
-@nb.kernel
+@nb.kernel(gpu_fp64_truncate="auto")
 def _rambo(C1, F1, Q1, nout, output):
     i = nb.get_global_id(0)
     for j in range(nout):

dpbench/benchmarks/rambo/rambo_numba_mlir_n.py

@@ -6,7 +6,7 @@
 import numpy as np
 
 
-@nb.njit
+@nb.njit(parallel=True, fastmath=True, gpu_fp64_truncate="auto")
 def rambo(nevts, nout, C1, F1, Q1, output):
     C = 2.0 * C1 - 1.0
     S = np.sqrt(1 - np.square(C))

dpbench/benchmarks/rambo/rambo_numba_mlir_p.py

@@ -7,7 +7,7 @@
 import numpy as np
 
 
-@nb.njit
+@nb.njit(parallel=True, fastmath=True, gpu_fp64_truncate="auto")
 def rambo(nevts, nout, C1, F1, Q1, output):
     for i in numba.prange(nevts):
         for j in numba.prange(nout):