[llvm] r313026 - [X86][AVX2] Add integer broadcast scheduling tests
Simon Pilgrim via llvm-commits
llvm-commits at lists.llvm.org
Tue Sep 12 05:59:21 PDT 2017
Author: rksimon
Date: Tue Sep 12 05:59:20 2017
New Revision: 313026
URL: http://llvm.org/viewvc/llvm-project?rev=313026&view=rev
Log:
[X86][AVX2] Add integer broadcast scheduling tests
Modified:
llvm/trunk/test/CodeGen/X86/avx2-schedule.ll
Modified: llvm/trunk/test/CodeGen/X86/avx2-schedule.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/avx2-schedule.ll?rev=313026&r1=313025&r2=313026&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/avx2-schedule.ll (original)
+++ llvm/trunk/test/CodeGen/X86/avx2-schedule.ll Tue Sep 12 05:59:20 2017
@@ -4,6 +4,36 @@
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=skylake | FileCheck %s --check-prefix=CHECK --check-prefix=SKYLAKE
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=znver1 | FileCheck %s --check-prefix=CHECK --check-prefix=ZNVER1
+define <8 x i32> @test_broadcasti128(<8 x i32> %a0, <4 x i32> *%a1) {
+; GENERIC-LABEL: test_broadcasti128:
+; GENERIC: # BB#0:
+; GENERIC-NEXT: vbroadcasti128 {{.*#+}} ymm1 = mem[0,1,0,1] sched: [4:0.50]
+; GENERIC-NEXT: vpaddd %ymm0, %ymm1, %ymm0 # sched: [3:1.00]
+; GENERIC-NEXT: retq # sched: [1:1.00]
+;
+; HASWELL-LABEL: test_broadcasti128:
+; HASWELL: # BB#0:
+; HASWELL-NEXT: vbroadcasti128 {{.*#+}} ymm1 = mem[0,1,0,1] sched: [1:0.50]
+; HASWELL-NEXT: vpaddd %ymm0, %ymm1, %ymm0 # sched: [1:0.50]
+; HASWELL-NEXT: retq # sched: [2:1.00]
+;
+; SKYLAKE-LABEL: test_broadcasti128:
+; SKYLAKE: # BB#0:
+; SKYLAKE-NEXT: vbroadcasti128 {{.*#+}} ymm1 = mem[0,1,0,1] sched: [1:0.50]
+; SKYLAKE-NEXT: vpaddd %ymm0, %ymm1, %ymm0 # sched: [1:0.50]
+; SKYLAKE-NEXT: retq # sched: [2:1.00]
+;
+; ZNVER1-LABEL: test_broadcasti128:
+; ZNVER1: # BB#0:
+; ZNVER1-NEXT: vbroadcasti128 {{.*#+}} ymm1 = mem[0,1,0,1] sched: [8:0.50]
+; ZNVER1-NEXT: vpaddd %ymm0, %ymm1, %ymm0 # sched: [1:0.25]
+; ZNVER1-NEXT: retq # sched: [1:0.50]
+ %1 = load <4 x i32>, <4 x i32> *%a1, align 16
+ %2 = shufflevector <4 x i32> %1, <4 x i32> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
+ %3 = add <8 x i32> %2, %a0
+ ret <8 x i32> %3
+}
+
define <4 x double> @test_broadcastsd_ymm(<2 x double> %a0) {
; GENERIC-LABEL: test_broadcastsd_ymm:
; GENERIC: # BB#0:
@@ -472,6 +502,286 @@ define <4 x i64> @test_pandn(<4 x i64> %
ret <4 x i64> %6
}
+define <16 x i8> @test_pbroadcastb(<16 x i8> %a0, <16 x i8> *%a1) {
+; GENERIC-LABEL: test_pbroadcastb:
+; GENERIC: # BB#0:
+; GENERIC-NEXT: vpbroadcastb %xmm0, %xmm0 # sched: [1:1.00]
+; GENERIC-NEXT: vpbroadcastb (%rdi), %xmm1 # sched: [4:0.50]
+; GENERIC-NEXT: vpaddb %xmm1, %xmm0, %xmm0 # sched: [1:0.50]
+; GENERIC-NEXT: retq # sched: [1:1.00]
+;
+; HASWELL-LABEL: test_pbroadcastb:
+; HASWELL: # BB#0:
+; HASWELL-NEXT: vpbroadcastb %xmm0, %xmm0 # sched: [3:1.00]
+; HASWELL-NEXT: vpbroadcastb (%rdi), %xmm1 # sched: [4:1.00]
+; HASWELL-NEXT: vpaddb %xmm1, %xmm0, %xmm0 # sched: [1:0.50]
+; HASWELL-NEXT: retq # sched: [2:1.00]
+;
+; SKYLAKE-LABEL: test_pbroadcastb:
+; SKYLAKE: # BB#0:
+; SKYLAKE-NEXT: vpbroadcastb %xmm0, %xmm0 # sched: [3:1.00]
+; SKYLAKE-NEXT: vpbroadcastb (%rdi), %xmm1 # sched: [4:1.00]
+; SKYLAKE-NEXT: vpaddb %xmm1, %xmm0, %xmm0 # sched: [1:0.50]
+; SKYLAKE-NEXT: retq # sched: [2:1.00]
+;
+; ZNVER1-LABEL: test_pbroadcastb:
+; ZNVER1: # BB#0:
+; ZNVER1-NEXT: vpbroadcastb (%rdi), %xmm1 # sched: [8:1.00]
+; ZNVER1-NEXT: vpbroadcastb %xmm0, %xmm0 # sched: [1:0.25]
+; ZNVER1-NEXT: vpaddb %xmm1, %xmm0, %xmm0 # sched: [1:0.25]
+; ZNVER1-NEXT: retq # sched: [1:0.50]
+ %1 = shufflevector <16 x i8> %a0, <16 x i8> undef, <16 x i32> zeroinitializer
+ %2 = load <16 x i8>, <16 x i8> *%a1, align 16
+ %3 = shufflevector <16 x i8> %2, <16 x i8> undef, <16 x i32> zeroinitializer
+ %4 = add <16 x i8> %1, %3
+ ret <16 x i8> %4
+}
+
+define <32 x i8> @test_pbroadcastb_ymm(<32 x i8> %a0, <32 x i8> *%a1) {
+; GENERIC-LABEL: test_pbroadcastb_ymm:
+; GENERIC: # BB#0:
+; GENERIC-NEXT: vpbroadcastb %xmm0, %ymm0 # sched: [1:1.00]
+; GENERIC-NEXT: vpbroadcastb (%rdi), %ymm1 # sched: [4:0.50]
+; GENERIC-NEXT: vpaddb %ymm1, %ymm0, %ymm0 # sched: [3:1.00]
+; GENERIC-NEXT: retq # sched: [1:1.00]
+;
+; HASWELL-LABEL: test_pbroadcastb_ymm:
+; HASWELL: # BB#0:
+; HASWELL-NEXT: vpbroadcastb %xmm0, %ymm0 # sched: [3:1.00]
+; HASWELL-NEXT: vpbroadcastb (%rdi), %ymm1 # sched: [4:1.00]
+; HASWELL-NEXT: vpaddb %ymm1, %ymm0, %ymm0 # sched: [1:0.50]
+; HASWELL-NEXT: retq # sched: [2:1.00]
+;
+; SKYLAKE-LABEL: test_pbroadcastb_ymm:
+; SKYLAKE: # BB#0:
+; SKYLAKE-NEXT: vpbroadcastb %xmm0, %ymm0 # sched: [3:1.00]
+; SKYLAKE-NEXT: vpbroadcastb (%rdi), %ymm1 # sched: [4:1.00]
+; SKYLAKE-NEXT: vpaddb %ymm1, %ymm0, %ymm0 # sched: [1:0.50]
+; SKYLAKE-NEXT: retq # sched: [2:1.00]
+;
+; ZNVER1-LABEL: test_pbroadcastb_ymm:
+; ZNVER1: # BB#0:
+; ZNVER1-NEXT: vpbroadcastb (%rdi), %ymm1 # sched: [8:2.00]
+; ZNVER1-NEXT: vpbroadcastb %xmm0, %ymm0 # sched: [2:0.25]
+; ZNVER1-NEXT: vpaddb %ymm1, %ymm0, %ymm0 # sched: [1:0.25]
+; ZNVER1-NEXT: retq # sched: [1:0.50]
+ %1 = shufflevector <32 x i8> %a0, <32 x i8> undef, <32 x i32> zeroinitializer
+ %2 = load <32 x i8>, <32 x i8> *%a1, align 32
+ %3 = shufflevector <32 x i8> %2, <32 x i8> undef, <32 x i32> zeroinitializer
+ %4 = add <32 x i8> %1, %3
+ ret <32 x i8> %4
+}
+
+define <4 x i32> @test_pbroadcastd(<4 x i32> %a0, <4 x i32> *%a1) {
+; GENERIC-LABEL: test_pbroadcastd:
+; GENERIC: # BB#0:
+; GENERIC-NEXT: vpbroadcastd %xmm0, %xmm0 # sched: [1:1.00]
+; GENERIC-NEXT: vpbroadcastd (%rdi), %xmm1 # sched: [4:0.50]
+; GENERIC-NEXT: vpaddd %xmm1, %xmm0, %xmm0 # sched: [1:0.50]
+; GENERIC-NEXT: retq # sched: [1:1.00]
+;
+; HASWELL-LABEL: test_pbroadcastd:
+; HASWELL: # BB#0:
+; HASWELL-NEXT: vpbroadcastd %xmm0, %xmm0 # sched: [1:1.00]
+; HASWELL-NEXT: vpbroadcastd (%rdi), %xmm1 # sched: [1:0.50]
+; HASWELL-NEXT: vpaddd %xmm1, %xmm0, %xmm0 # sched: [1:0.50]
+; HASWELL-NEXT: retq # sched: [2:1.00]
+;
+; SKYLAKE-LABEL: test_pbroadcastd:
+; SKYLAKE: # BB#0:
+; SKYLAKE-NEXT: vpbroadcastd %xmm0, %xmm0 # sched: [1:1.00]
+; SKYLAKE-NEXT: vpbroadcastd (%rdi), %xmm1 # sched: [1:0.50]
+; SKYLAKE-NEXT: vpaddd %xmm1, %xmm0, %xmm0 # sched: [1:0.50]
+; SKYLAKE-NEXT: retq # sched: [2:1.00]
+;
+; ZNVER1-LABEL: test_pbroadcastd:
+; ZNVER1: # BB#0:
+; ZNVER1-NEXT: vpbroadcastd (%rdi), %xmm1 # sched: [8:0.50]
+; ZNVER1-NEXT: vpbroadcastd %xmm0, %xmm0 # sched: [1:0.25]
+; ZNVER1-NEXT: vpaddd %xmm1, %xmm0, %xmm0 # sched: [1:0.25]
+; ZNVER1-NEXT: retq # sched: [1:0.50]
+ %1 = shufflevector <4 x i32> %a0, <4 x i32> undef, <4 x i32> zeroinitializer
+ %2 = load <4 x i32>, <4 x i32> *%a1, align 16
+ %3 = shufflevector <4 x i32> %2, <4 x i32> undef, <4 x i32> zeroinitializer
+ %4 = add <4 x i32> %1, %3
+ ret <4 x i32> %4
+}
+
+define <8 x i32> @test_pbroadcastd_ymm(<8 x i32> %a0, <8 x i32> *%a1) {
+; GENERIC-LABEL: test_pbroadcastd_ymm:
+; GENERIC: # BB#0:
+; GENERIC-NEXT: vpbroadcastd %xmm0, %ymm0 # sched: [1:1.00]
+; GENERIC-NEXT: vpbroadcastd (%rdi), %ymm1 # sched: [4:0.50]
+; GENERIC-NEXT: vpaddd %ymm1, %ymm0, %ymm0 # sched: [3:1.00]
+; GENERIC-NEXT: retq # sched: [1:1.00]
+;
+; HASWELL-LABEL: test_pbroadcastd_ymm:
+; HASWELL: # BB#0:
+; HASWELL-NEXT: vpbroadcastd %xmm0, %ymm0 # sched: [3:1.00]
+; HASWELL-NEXT: vpbroadcastd (%rdi), %ymm1 # sched: [1:0.50]
+; HASWELL-NEXT: vpaddd %ymm1, %ymm0, %ymm0 # sched: [1:0.50]
+; HASWELL-NEXT: retq # sched: [2:1.00]
+;
+; SKYLAKE-LABEL: test_pbroadcastd_ymm:
+; SKYLAKE: # BB#0:
+; SKYLAKE-NEXT: vpbroadcastd %xmm0, %ymm0 # sched: [3:1.00]
+; SKYLAKE-NEXT: vpbroadcastd (%rdi), %ymm1 # sched: [1:0.50]
+; SKYLAKE-NEXT: vpaddd %ymm1, %ymm0, %ymm0 # sched: [1:0.50]
+; SKYLAKE-NEXT: retq # sched: [2:1.00]
+;
+; ZNVER1-LABEL: test_pbroadcastd_ymm:
+; ZNVER1: # BB#0:
+; ZNVER1-NEXT: vpbroadcastd (%rdi), %ymm1 # sched: [8:0.50]
+; ZNVER1-NEXT: vpbroadcastd %xmm0, %ymm0 # sched: [2:0.25]
+; ZNVER1-NEXT: vpaddd %ymm1, %ymm0, %ymm0 # sched: [1:0.25]
+; ZNVER1-NEXT: retq # sched: [1:0.50]
+ %1 = shufflevector <8 x i32> %a0, <8 x i32> undef, <8 x i32> zeroinitializer
+ %2 = load <8 x i32>, <8 x i32> *%a1, align 32
+ %3 = shufflevector <8 x i32> %2, <8 x i32> undef, <8 x i32> zeroinitializer
+ %4 = add <8 x i32> %1, %3
+ ret <8 x i32> %4
+}
+
+define <2 x i64> @test_pbroadcastq(<2 x i64> %a0, <2 x i64> *%a1) {
+; GENERIC-LABEL: test_pbroadcastq:
+; GENERIC: # BB#0:
+; GENERIC-NEXT: vpbroadcastq %xmm0, %xmm0 # sched: [1:1.00]
+; GENERIC-NEXT: vpbroadcastq (%rdi), %xmm1 # sched: [4:0.50]
+; GENERIC-NEXT: vpaddq %xmm1, %xmm0, %xmm0 # sched: [1:0.50]
+; GENERIC-NEXT: retq # sched: [1:1.00]
+;
+; HASWELL-LABEL: test_pbroadcastq:
+; HASWELL: # BB#0:
+; HASWELL-NEXT: vpbroadcastq %xmm0, %xmm0 # sched: [1:1.00]
+; HASWELL-NEXT: vpbroadcastq (%rdi), %xmm1 # sched: [1:0.50]
+; HASWELL-NEXT: vpaddq %xmm1, %xmm0, %xmm0 # sched: [1:0.50]
+; HASWELL-NEXT: retq # sched: [2:1.00]
+;
+; SKYLAKE-LABEL: test_pbroadcastq:
+; SKYLAKE: # BB#0:
+; SKYLAKE-NEXT: vpbroadcastq %xmm0, %xmm0 # sched: [1:1.00]
+; SKYLAKE-NEXT: vpbroadcastq (%rdi), %xmm1 # sched: [1:0.50]
+; SKYLAKE-NEXT: vpaddq %xmm1, %xmm0, %xmm0 # sched: [1:0.50]
+; SKYLAKE-NEXT: retq # sched: [2:1.00]
+;
+; ZNVER1-LABEL: test_pbroadcastq:
+; ZNVER1: # BB#0:
+; ZNVER1-NEXT: vpbroadcastq (%rdi), %xmm1 # sched: [8:0.50]
+; ZNVER1-NEXT: vpbroadcastq %xmm0, %xmm0 # sched: [1:0.25]
+; ZNVER1-NEXT: vpaddq %xmm1, %xmm0, %xmm0 # sched: [1:0.25]
+; ZNVER1-NEXT: retq # sched: [1:0.50]
+ %1 = shufflevector <2 x i64> %a0, <2 x i64> undef, <2 x i32> zeroinitializer
+ %2 = load <2 x i64>, <2 x i64> *%a1, align 16
+ %3 = shufflevector <2 x i64> %2, <2 x i64> undef, <2 x i32> zeroinitializer
+ %4 = add <2 x i64> %1, %3
+ ret <2 x i64> %4
+}
+
+define <4 x i64> @test_pbroadcastq_ymm(<4 x i64> %a0, <4 x i64> *%a1) {
+; GENERIC-LABEL: test_pbroadcastq_ymm:
+; GENERIC: # BB#0:
+; GENERIC-NEXT: vpbroadcastq %xmm0, %ymm0 # sched: [1:1.00]
+; GENERIC-NEXT: vpbroadcastq (%rdi), %ymm1 # sched: [4:0.50]
+; GENERIC-NEXT: vpaddq %ymm1, %ymm0, %ymm0 # sched: [3:1.00]
+; GENERIC-NEXT: retq # sched: [1:1.00]
+;
+; HASWELL-LABEL: test_pbroadcastq_ymm:
+; HASWELL: # BB#0:
+; HASWELL-NEXT: vpbroadcastq %xmm0, %ymm0 # sched: [3:1.00]
+; HASWELL-NEXT: vpbroadcastq (%rdi), %ymm1 # sched: [1:0.50]
+; HASWELL-NEXT: vpaddq %ymm1, %ymm0, %ymm0 # sched: [1:0.50]
+; HASWELL-NEXT: retq # sched: [2:1.00]
+;
+; SKYLAKE-LABEL: test_pbroadcastq_ymm:
+; SKYLAKE: # BB#0:
+; SKYLAKE-NEXT: vpbroadcastq %xmm0, %ymm0 # sched: [3:1.00]
+; SKYLAKE-NEXT: vpbroadcastq (%rdi), %ymm1 # sched: [1:0.50]
+; SKYLAKE-NEXT: vpaddq %ymm1, %ymm0, %ymm0 # sched: [1:0.50]
+; SKYLAKE-NEXT: retq # sched: [2:1.00]
+;
+; ZNVER1-LABEL: test_pbroadcastq_ymm:
+; ZNVER1: # BB#0:
+; ZNVER1-NEXT: vpbroadcastq (%rdi), %ymm1 # sched: [8:0.50]
+; ZNVER1-NEXT: vpbroadcastq %xmm0, %ymm0 # sched: [2:0.25]
+; ZNVER1-NEXT: vpaddq %ymm1, %ymm0, %ymm0 # sched: [1:0.25]
+; ZNVER1-NEXT: retq # sched: [1:0.50]
+ %1 = shufflevector <4 x i64> %a0, <4 x i64> undef, <4 x i32> zeroinitializer
+ %2 = load <4 x i64>, <4 x i64> *%a1, align 32
+ %3 = shufflevector <4 x i64> %2, <4 x i64> undef, <4 x i32> zeroinitializer
+ %4 = add <4 x i64> %1, %3
+ ret <4 x i64> %4
+}
+
+define <8 x i16> @test_pbroadcastw(<8 x i16> %a0, <8 x i16> *%a1) {
+; GENERIC-LABEL: test_pbroadcastw:
+; GENERIC: # BB#0:
+; GENERIC-NEXT: vpbroadcastw %xmm0, %xmm0 # sched: [1:1.00]
+; GENERIC-NEXT: vpbroadcastw (%rdi), %xmm1 # sched: [4:0.50]
+; GENERIC-NEXT: vpaddw %xmm1, %xmm0, %xmm0 # sched: [1:0.50]
+; GENERIC-NEXT: retq # sched: [1:1.00]
+;
+; HASWELL-LABEL: test_pbroadcastw:
+; HASWELL: # BB#0:
+; HASWELL-NEXT: vpbroadcastw %xmm0, %xmm0 # sched: [3:1.00]
+; HASWELL-NEXT: vpbroadcastw (%rdi), %xmm1 # sched: [4:1.00]
+; HASWELL-NEXT: vpaddw %xmm1, %xmm0, %xmm0 # sched: [1:0.50]
+; HASWELL-NEXT: retq # sched: [2:1.00]
+;
+; SKYLAKE-LABEL: test_pbroadcastw:
+; SKYLAKE: # BB#0:
+; SKYLAKE-NEXT: vpbroadcastw %xmm0, %xmm0 # sched: [3:1.00]
+; SKYLAKE-NEXT: vpbroadcastw (%rdi), %xmm1 # sched: [4:1.00]
+; SKYLAKE-NEXT: vpaddw %xmm1, %xmm0, %xmm0 # sched: [1:0.50]
+; SKYLAKE-NEXT: retq # sched: [2:1.00]
+;
+; ZNVER1-LABEL: test_pbroadcastw:
+; ZNVER1: # BB#0:
+; ZNVER1-NEXT: vpbroadcastw (%rdi), %xmm1 # sched: [8:1.00]
+; ZNVER1-NEXT: vpbroadcastw %xmm0, %xmm0 # sched: [1:0.25]
+; ZNVER1-NEXT: vpaddw %xmm1, %xmm0, %xmm0 # sched: [1:0.25]
+; ZNVER1-NEXT: retq # sched: [1:0.50]
+ %1 = shufflevector <8 x i16> %a0, <8 x i16> undef, <8 x i32> zeroinitializer
+ %2 = load <8 x i16>, <8 x i16> *%a1, align 16
+ %3 = shufflevector <8 x i16> %2, <8 x i16> undef, <8 x i32> zeroinitializer
+ %4 = add <8 x i16> %1, %3
+ ret <8 x i16> %4
+}
+
+define <16 x i16> @test_pbroadcastw_ymm(<16 x i16> %a0, <16 x i16> *%a1) {
+; GENERIC-LABEL: test_pbroadcastw_ymm:
+; GENERIC: # BB#0:
+; GENERIC-NEXT: vpbroadcastw %xmm0, %ymm0 # sched: [1:1.00]
+; GENERIC-NEXT: vpbroadcastw (%rdi), %ymm1 # sched: [4:0.50]
+; GENERIC-NEXT: vpaddw %ymm1, %ymm0, %ymm0 # sched: [3:1.00]
+; GENERIC-NEXT: retq # sched: [1:1.00]
+;
+; HASWELL-LABEL: test_pbroadcastw_ymm:
+; HASWELL: # BB#0:
+; HASWELL-NEXT: vpbroadcastw %xmm0, %ymm0 # sched: [3:1.00]
+; HASWELL-NEXT: vpbroadcastw (%rdi), %ymm1 # sched: [4:1.00]
+; HASWELL-NEXT: vpaddw %ymm1, %ymm0, %ymm0 # sched: [1:0.50]
+; HASWELL-NEXT: retq # sched: [2:1.00]
+;
+; SKYLAKE-LABEL: test_pbroadcastw_ymm:
+; SKYLAKE: # BB#0:
+; SKYLAKE-NEXT: vpbroadcastw %xmm0, %ymm0 # sched: [3:1.00]
+; SKYLAKE-NEXT: vpbroadcastw (%rdi), %ymm1 # sched: [4:1.00]
+; SKYLAKE-NEXT: vpaddw %ymm1, %ymm0, %ymm0 # sched: [1:0.50]
+; SKYLAKE-NEXT: retq # sched: [2:1.00]
+;
+; ZNVER1-LABEL: test_pbroadcastw_ymm:
+; ZNVER1: # BB#0:
+; ZNVER1-NEXT: vpbroadcastw (%rdi), %ymm1 # sched: [8:2.00]
+; ZNVER1-NEXT: vpbroadcastw %xmm0, %ymm0 # sched: [2:0.25]
+; ZNVER1-NEXT: vpaddw %ymm1, %ymm0, %ymm0 # sched: [1:0.25]
+; ZNVER1-NEXT: retq # sched: [1:0.50]
+ %1 = shufflevector <16 x i16> %a0, <16 x i16> undef, <16 x i32> zeroinitializer
+ %2 = load <16 x i16>, <16 x i16> *%a1, align 32
+ %3 = shufflevector <16 x i16> %2, <16 x i16> undef, <16 x i32> zeroinitializer
+ %4 = add <16 x i16> %1, %3
+ ret <16 x i16> %4
+}
+
define <4 x i64> @test_perm2i128(<4 x i64> %a0, <4 x i64> %a1, <4 x i64> *%a2) {
; GENERIC-LABEL: test_perm2i128:
; GENERIC: # BB#0:
More information about the llvm-commits
mailing list