| module { | |
| tt.func public @triton__0d1de(%arg0: !tt.ptr<f32, 1> {tt.divisibility = 16 : i32}, %arg1: i64 {tt.divisibility = 16 : i32, tt.max_divisibility = 16 : i32}) attributes {noinline = false} { | |
| %c512_i64 = arith.constant 512 : i64 | |
| %cst = arith.constant dense<0.000000e+00> : tensor<512xf32> | |
| %0 = tt.get_program_id x : i32 | |
| %1 = arith.extsi %0 : i32 to i64 | |
| %2 = arith.muli %1, %c512_i64 : i64 | |
| %3 = tt.make_range {end = 512 : i32, start = 0 : i32} : tensor<512xi32> | |
| %4 = arith.extsi %3 : tensor<512xi32> to tensor<512xi64> | |
| %5 = tt.splat %2 : (i64) -> tensor<512xi64> | |
| %6 = arith.addi %5, %4 : tensor<512xi64> | |
| %7 = tt.splat %arg0 : (!tt.ptr<f32, 1>) -> tensor<512x!tt.ptr<f32, 1>> | |
| %8 = tt.addptr %7, %6 : tensor<512x!tt.ptr<f32, 1>>, tensor<512xi64> | |
| tt.store %8, %cst {cache = 1 : i32, evict = 1 : i32} : tensor<512xf32> | |
| tt.return | |
| } | |
| } | |