[llvm] [RISCV][CostModel] cost of vector cttz/ctlz under ZVBB (PR #115800)
Luke Lau via llvm-commits
llvm-commits at lists.llvm.org
Mon Nov 18 23:13:08 PST 2024
================
@@ -157,6 +157,260 @@ define void @bitreverse() {
ret void
}
+define void @ctlz() {
+; NOZVBB-LABEL: 'ctlz'
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %1 = call <2 x i8> @llvm.ctlz.v2i8(<2 x i8> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %2 = call <4 x i8> @llvm.ctlz.v4i8(<4 x i8> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %3 = call <8 x i8> @llvm.ctlz.v8i8(<8 x i8> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %4 = call <16 x i8> @llvm.ctlz.v16i8(<16 x i8> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %5 = call <vscale x 1 x i8> @llvm.ctlz.nxv1i8(<vscale x 1 x i8> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %6 = call <vscale x 2 x i8> @llvm.ctlz.nxv2i8(<vscale x 2 x i8> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %7 = call <vscale x 4 x i8> @llvm.ctlz.nxv4i8(<vscale x 4 x i8> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %8 = call <vscale x 8 x i8> @llvm.ctlz.nxv8i8(<vscale x 8 x i8> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %9 = call <vscale x 16 x i8> @llvm.ctlz.nxv16i8(<vscale x 16 x i8> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Invalid cost for instruction: %10 = call <vscale x 32 x i8> @llvm.ctlz.nxv32i8(<vscale x 32 x i8> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Invalid cost for instruction: %11 = call <vscale x 64 x i8> @llvm.ctlz.nxv64i8(<vscale x 64 x i8> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %12 = call <2 x i16> @llvm.ctlz.v2i16(<2 x i16> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %13 = call <4 x i16> @llvm.ctlz.v4i16(<4 x i16> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %14 = call <8 x i16> @llvm.ctlz.v8i16(<8 x i16> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %15 = call <16 x i16> @llvm.ctlz.v16i16(<16 x i16> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %16 = call <vscale x 1 x i16> @llvm.ctlz.nxv1i16(<vscale x 1 x i16> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %17 = call <vscale x 2 x i16> @llvm.ctlz.nxv2i16(<vscale x 2 x i16> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %18 = call <vscale x 4 x i16> @llvm.ctlz.nxv4i16(<vscale x 4 x i16> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %19 = call <vscale x 8 x i16> @llvm.ctlz.nxv8i16(<vscale x 8 x i16> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %20 = call <vscale x 16 x i16> @llvm.ctlz.nxv16i16(<vscale x 16 x i16> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Invalid cost for instruction: %21 = call <vscale x 32 x i16> @llvm.ctlz.nxv32i16(<vscale x 32 x i16> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %22 = call <2 x i32> @llvm.ctlz.v2i32(<2 x i32> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %23 = call <4 x i32> @llvm.ctlz.v4i32(<4 x i32> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %24 = call <8 x i32> @llvm.ctlz.v8i32(<8 x i32> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %25 = call <16 x i32> @llvm.ctlz.v16i32(<16 x i32> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %26 = call <vscale x 1 x i32> @llvm.ctlz.nxv1i32(<vscale x 1 x i32> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %27 = call <vscale x 2 x i32> @llvm.ctlz.nxv2i32(<vscale x 2 x i32> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %28 = call <vscale x 4 x i32> @llvm.ctlz.nxv4i32(<vscale x 4 x i32> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %29 = call <vscale x 8 x i32> @llvm.ctlz.nxv8i32(<vscale x 8 x i32> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %30 = call <vscale x 16 x i32> @llvm.ctlz.nxv16i32(<vscale x 16 x i32> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %31 = call <2 x i64> @llvm.ctlz.v2i64(<2 x i64> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %32 = call <4 x i64> @llvm.ctlz.v4i64(<4 x i64> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %33 = call <8 x i64> @llvm.ctlz.v8i64(<8 x i64> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %34 = call <16 x i64> @llvm.ctlz.v16i64(<16 x i64> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %35 = call <vscale x 1 x i64> @llvm.ctlz.nxv1i64(<vscale x 1 x i64> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %36 = call <vscale x 2 x i64> @llvm.ctlz.nxv2i64(<vscale x 2 x i64> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %37 = call <vscale x 4 x i64> @llvm.ctlz.nxv4i64(<vscale x 4 x i64> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %38 = call <vscale x 8 x i64> @llvm.ctlz.nxv8i64(<vscale x 8 x i64> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %39 = call <vscale x 16 x i64> @llvm.ctlz.nxv16i64(<vscale x 16 x i64> undef, i1 false)
+; NOZVBB-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
+;
+; ZVBB-LABEL: 'ctlz'
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %1 = call <2 x i8> @llvm.ctlz.v2i8(<2 x i8> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %2 = call <4 x i8> @llvm.ctlz.v4i8(<4 x i8> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %3 = call <8 x i8> @llvm.ctlz.v8i8(<8 x i8> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %4 = call <16 x i8> @llvm.ctlz.v16i8(<16 x i8> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %5 = call <vscale x 1 x i8> @llvm.ctlz.nxv1i8(<vscale x 1 x i8> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %6 = call <vscale x 2 x i8> @llvm.ctlz.nxv2i8(<vscale x 2 x i8> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %7 = call <vscale x 4 x i8> @llvm.ctlz.nxv4i8(<vscale x 4 x i8> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %8 = call <vscale x 8 x i8> @llvm.ctlz.nxv8i8(<vscale x 8 x i8> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %9 = call <vscale x 16 x i8> @llvm.ctlz.nxv16i8(<vscale x 16 x i8> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %10 = call <vscale x 32 x i8> @llvm.ctlz.nxv32i8(<vscale x 32 x i8> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %11 = call <vscale x 64 x i8> @llvm.ctlz.nxv64i8(<vscale x 64 x i8> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %12 = call <2 x i16> @llvm.ctlz.v2i16(<2 x i16> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %13 = call <4 x i16> @llvm.ctlz.v4i16(<4 x i16> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %14 = call <8 x i16> @llvm.ctlz.v8i16(<8 x i16> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %15 = call <16 x i16> @llvm.ctlz.v16i16(<16 x i16> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %16 = call <vscale x 1 x i16> @llvm.ctlz.nxv1i16(<vscale x 1 x i16> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %17 = call <vscale x 2 x i16> @llvm.ctlz.nxv2i16(<vscale x 2 x i16> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %18 = call <vscale x 4 x i16> @llvm.ctlz.nxv4i16(<vscale x 4 x i16> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %19 = call <vscale x 8 x i16> @llvm.ctlz.nxv8i16(<vscale x 8 x i16> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %20 = call <vscale x 16 x i16> @llvm.ctlz.nxv16i16(<vscale x 16 x i16> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %21 = call <vscale x 32 x i16> @llvm.ctlz.nxv32i16(<vscale x 32 x i16> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %22 = call <2 x i32> @llvm.ctlz.v2i32(<2 x i32> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %23 = call <4 x i32> @llvm.ctlz.v4i32(<4 x i32> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %24 = call <8 x i32> @llvm.ctlz.v8i32(<8 x i32> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %25 = call <16 x i32> @llvm.ctlz.v16i32(<16 x i32> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %26 = call <vscale x 1 x i32> @llvm.ctlz.nxv1i32(<vscale x 1 x i32> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %27 = call <vscale x 2 x i32> @llvm.ctlz.nxv2i32(<vscale x 2 x i32> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %28 = call <vscale x 4 x i32> @llvm.ctlz.nxv4i32(<vscale x 4 x i32> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %29 = call <vscale x 8 x i32> @llvm.ctlz.nxv8i32(<vscale x 8 x i32> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %30 = call <vscale x 16 x i32> @llvm.ctlz.nxv16i32(<vscale x 16 x i32> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %31 = call <2 x i64> @llvm.ctlz.v2i64(<2 x i64> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %32 = call <4 x i64> @llvm.ctlz.v4i64(<4 x i64> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %33 = call <8 x i64> @llvm.ctlz.v8i64(<8 x i64> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %34 = call <16 x i64> @llvm.ctlz.v16i64(<16 x i64> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %35 = call <vscale x 1 x i64> @llvm.ctlz.nxv1i64(<vscale x 1 x i64> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %36 = call <vscale x 2 x i64> @llvm.ctlz.nxv2i64(<vscale x 2 x i64> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %37 = call <vscale x 4 x i64> @llvm.ctlz.nxv4i64(<vscale x 4 x i64> undef, i1 false)
+; ZVBB-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %38 = call <vscale x 8 x i64> @llvm.ctlz.nxv8i64(<vscale x 8 x i64> undef, i1 false)
----------------
lukel97 wrote:
It looks like we still need to scale the cost by LMUL, for both the zvbb and non-zvbb costs. This might be worth doing in a follow up PR
https://github.com/llvm/llvm-project/pull/115800
More information about the llvm-commits
mailing list