| // RUN: iree-dialects-opt -linalg-interp-transforms %s | FileCheck %s |
| |
| func @fun_to_benchmark(%arg0: tensor<128x128xf32>, %arg1: tensor<128x128xf32>, %arg2: tensor<128x128xf32>) -> |
| tensor<128x128xf32> attributes {passthrough = ["noinline", ["target-cpu", "skylake-avx512"], ["prefer-vector-width", "512"]]} { |
| // With scalarization we expect vectorization to still work albeit with a leading |
| // `1` dimension. |
| // CHECK: vector.contract {{.*}} : vector<1x32xf32>, vector<32x16xf32> into vector<1x16xf32> |
| %0 = linalg.matmul ins(%arg0, %arg1 : tensor<128x128xf32>, tensor<128x128xf32>) |
| outs(%arg2 : tensor<128x128xf32>) -> tensor<128x128xf32> |
| return %0 : tensor<128x128xf32> |
| } |
| |
| pdl.pattern @isa_linalg.matmul : benefit(1) { |
| %0 = operands |
| %1 = types |
| %2 = operation "linalg.matmul"(%0 : !pdl.range<value>) -> (%1 : !pdl.range<type>) |
| rewrite %2 with "iree_linalg_transform.apply" |
| } |
| |
| iree_linalg_transform.sequence { |
| %0 = match @isa_linalg.matmul |
| %tiled_linalg_op, %loops:3 = tile %0 {interchange = [1, 0, 2], sizes = [6, 16, 32]} |
| %1 = peel_loop %loops#0 |
| // This test checks the proper handling of the scalarize dims attribute. |
| // The first dimension does not divide but we can always scalarize a `?` into `1` |
| // and enable vectorization of a lower-rank op this way. |
| %tiled_linalg_op_0 = scalarize %tiled_linalg_op |
| vectorize {vectorize_padding = false} |
| } |
