1; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py 2; RUN: llc -mtriple=thumbv8.1m.main -mattr=+mve.fp -verify-machineinstrs -o - %s | FileCheck %s 3 4define arm_aapcs_vfpcc <4 x i32> @test_vldrwq_gather_base_wb_s32(<4 x i32>* %addr) { 5; CHECK-LABEL: test_vldrwq_gather_base_wb_s32: 6; CHECK: @ %bb.0: @ %entry 7; CHECK-NEXT: vldrw.u32 q1, [r0] 8; CHECK-NEXT: vldrw.u32 q0, [q1, #80]! 9; CHECK-NEXT: vstrw.32 q1, [r0] 10; CHECK-NEXT: bx lr 11entry: 12 %0 = load <4 x i32>, <4 x i32>* %addr, align 8 13 %1 = tail call { <4 x i32>, <4 x i32> } @llvm.arm.mve.vldr.gather.base.wb.v4i32.v4i32(<4 x i32> %0, i32 80) 14 %2 = extractvalue { <4 x i32>, <4 x i32> } %1, 1 15 store <4 x i32> %2, <4 x i32>* %addr, align 8 16 %3 = extractvalue { <4 x i32>, <4 x i32> } %1, 0 17 ret <4 x i32> %3 18} 19 20declare { <4 x i32>, <4 x i32> } @llvm.arm.mve.vldr.gather.base.wb.v4i32.v4i32(<4 x i32>, i32) 21 22define arm_aapcs_vfpcc <4 x float> @test_vldrwq_gather_base_wb_f32(<4 x i32>* %addr) { 23; CHECK-LABEL: test_vldrwq_gather_base_wb_f32: 24; CHECK: @ %bb.0: @ %entry 25; CHECK-NEXT: vldrw.u32 q1, [r0] 26; CHECK-NEXT: vldrw.u32 q0, [q1, #64]! 27; CHECK-NEXT: vstrw.32 q1, [r0] 28; CHECK-NEXT: bx lr 29entry: 30 %0 = load <4 x i32>, <4 x i32>* %addr, align 8 31 %1 = tail call { <4 x float>, <4 x i32> } @llvm.arm.mve.vldr.gather.base.wb.v4f32.v4i32(<4 x i32> %0, i32 64) 32 %2 = extractvalue { <4 x float>, <4 x i32> } %1, 1 33 store <4 x i32> %2, <4 x i32>* %addr, align 8 34 %3 = extractvalue { <4 x float>, <4 x i32> } %1, 0 35 ret <4 x float> %3 36} 37 38declare { <4 x float>, <4 x i32> } @llvm.arm.mve.vldr.gather.base.wb.v4f32.v4i32(<4 x i32>, i32) 39 40define arm_aapcs_vfpcc <2 x i64> @test_vldrdq_gather_base_wb_z_u64(<2 x i64>* %addr, i16 zeroext %p) { 41; CHECK-LABEL: test_vldrdq_gather_base_wb_z_u64: 42; CHECK: @ %bb.0: @ %entry 43; CHECK-NEXT: vmsr p0, r1 44; CHECK-NEXT: vldrw.u32 q1, [r0] 45; CHECK-NEXT: vpst 46; CHECK-NEXT: vldrdt.u64 q0, [q1, #656]! 47; CHECK-NEXT: vstrw.32 q1, [r0] 48; CHECK-NEXT: bx lr 49entry: 50 %0 = load <2 x i64>, <2 x i64>* %addr, align 8 51 %1 = zext i16 %p to i32 52 %2 = tail call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 %1) 53 %3 = tail call { <2 x i64>, <2 x i64> } @llvm.arm.mve.vldr.gather.base.wb.predicated.v2i64.v2i64.v4i1(<2 x i64> %0, i32 656, <4 x i1> %2) 54 %4 = extractvalue { <2 x i64>, <2 x i64> } %3, 1 55 store <2 x i64> %4, <2 x i64>* %addr, align 8 56 %5 = extractvalue { <2 x i64>, <2 x i64> } %3, 0 57 ret <2 x i64> %5 58} 59 60declare <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32) 61 62declare { <2 x i64>, <2 x i64> } @llvm.arm.mve.vldr.gather.base.wb.predicated.v2i64.v2i64.v4i1(<2 x i64>, i32, <4 x i1>) 63