nir: Add AMD cooperative matrix intrinsics.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24683>

src/compiler/nir/nir_divergence_analysis.c

@@ -610,6 +610,7 @@ visit_intrinsic(nir_shader *shader, nir_intrinsic_instr *instr)
    case nir_intrinsic_rq_load:
    case nir_intrinsic_load_ray_triangle_vertex_positions:
    case nir_intrinsic_cmat_extract:
+   case nir_intrinsic_cmat_muladd_amd:
       is_divergent = true;
       break;
 

src/compiler/nir/nir_intrinsics.py

@@ -1671,6 +1671,9 @@ system_value("barycentric_optimize_amd", dest_comp=1, bit_sizes=[1])
 intrinsic("strict_wqm_coord_amd", src_comp=[0], dest_comp=0, bit_sizes=[32], indices=[BASE],
           flags=[CAN_ELIMINATE])
 
+intrinsic("cmat_muladd_amd", src_comp=[16, 16, 0], dest_comp=0, bit_sizes=src2,
+          indices=[SATURATE, CMAT_SIGNED_MASK], flags=[CAN_ELIMINATE])
+
 # V3D-specific instrinc for tile buffer color reads.
 #
 # The hardware requires that we read the samples and components of a pixel