; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py ; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s define arm_aapcs_vfpcc void @test_fadd(float* noalias nocapture readonly %A, float %B, float* noalias nocapture %C, i32 %n) { ; CHECK-LABEL: test_fadd: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r2, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB0_1: @ %vector.ph ; CHECK-NEXT: vmov r3, s0 ; CHECK-NEXT: .LBB0_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r0], #16 ; CHECK-NEXT: subs r2, #4 ; CHECK-NEXT: vadd.f32 q0, q0, r3 ; CHECK-NEXT: vstrb.8 q0, [r1], #16 ; CHECK-NEXT: bne .LBB0_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp18 = icmp sgt i32 %n, 0 br i1 %cmp18, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert10 = insertelement <4 x float> undef, float %B, i32 0 %broadcast.splat11 = shufflevector <4 x float> %broadcast.splatinsert10, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = fadd fast <4 x float> %wide.load, %broadcast.splat11 %4 = getelementptr inbounds float, float* %C, i32 %index %5 = bitcast float* %4 to <4 x float>* store <4 x float> %3, <4 x float>* %5, align 4 %index.next = add i32 %index, 4 %6 = icmp eq i32 %index.next, %n br i1 %6, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fadd_r(float* noalias nocapture readonly %A, float %B, float* noalias nocapture %C, i32 %n) { ; CHECK-LABEL: test_fadd_r: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r2, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB1_1: @ %vector.ph ; CHECK-NEXT: vmov r3, s0 ; CHECK-NEXT: .LBB1_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r0], #16 ; CHECK-NEXT: subs r2, #4 ; CHECK-NEXT: vadd.f32 q0, q0, r3 ; CHECK-NEXT: vstrb.8 q0, [r1], #16 ; CHECK-NEXT: bne .LBB1_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp18 = icmp sgt i32 %n, 0 br i1 %cmp18, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert10 = insertelement <4 x float> undef, float %B, i32 0 %broadcast.splat11 = shufflevector <4 x float> %broadcast.splatinsert10, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = fadd fast <4 x float> %broadcast.splat11, %wide.load %4 = getelementptr inbounds float, float* %C, i32 %index %5 = bitcast float* %4 to <4 x float>* store <4 x float> %3, <4 x float>* %5, align 4 %index.next = add i32 %index, 4 %6 = icmp eq i32 %index.next, %n br i1 %6, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fmul(float* noalias nocapture readonly %A, float %B, float* noalias nocapture %C, i32 %n) { ; CHECK-LABEL: test_fmul: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r2, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB2_1: @ %vector.ph ; CHECK-NEXT: vmov r3, s0 ; CHECK-NEXT: .LBB2_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r0], #16 ; CHECK-NEXT: subs r2, #4 ; CHECK-NEXT: vmul.f32 q0, q0, r3 ; CHECK-NEXT: vstrb.8 q0, [r1], #16 ; CHECK-NEXT: bne .LBB2_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp18 = icmp sgt i32 %n, 0 br i1 %cmp18, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert10 = insertelement <4 x float> undef, float %B, i32 0 %broadcast.splat11 = shufflevector <4 x float> %broadcast.splatinsert10, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = fmul fast <4 x float> %wide.load, %broadcast.splat11 %4 = getelementptr inbounds float, float* %C, i32 %index %5 = bitcast float* %4 to <4 x float>* store <4 x float> %3, <4 x float>* %5, align 4 %index.next = add i32 %index, 4 %6 = icmp eq i32 %index.next, %n br i1 %6, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fmul_r(float* noalias nocapture readonly %A, float %B, float* noalias nocapture %C, i32 %n) { ; CHECK-LABEL: test_fmul_r: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r2, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB3_1: @ %vector.ph ; CHECK-NEXT: vmov r3, s0 ; CHECK-NEXT: .LBB3_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r0], #16 ; CHECK-NEXT: subs r2, #4 ; CHECK-NEXT: vmul.f32 q0, q0, r3 ; CHECK-NEXT: vstrb.8 q0, [r1], #16 ; CHECK-NEXT: bne .LBB3_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp18 = icmp sgt i32 %n, 0 br i1 %cmp18, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert10 = insertelement <4 x float> undef, float %B, i32 0 %broadcast.splat11 = shufflevector <4 x float> %broadcast.splatinsert10, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = fmul fast <4 x float> %broadcast.splat11, %wide.load %4 = getelementptr inbounds float, float* %C, i32 %index %5 = bitcast float* %4 to <4 x float>* store <4 x float> %3, <4 x float>* %5, align 4 %index.next = add i32 %index, 4 %6 = icmp eq i32 %index.next, %n br i1 %6, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fsub(float* noalias nocapture readonly %A, float %B, float* noalias nocapture %C, i32 %n) { ; CHECK-LABEL: test_fsub: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r2, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB4_1: @ %vector.ph ; CHECK-NEXT: vmov r3, s0 ; CHECK-NEXT: .LBB4_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r0], #16 ; CHECK-NEXT: subs r2, #4 ; CHECK-NEXT: vsub.f32 q0, q0, r3 ; CHECK-NEXT: vstrb.8 q0, [r1], #16 ; CHECK-NEXT: bne .LBB4_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp18 = icmp sgt i32 %n, 0 br i1 %cmp18, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert10 = insertelement <4 x float> undef, float %B, i32 0 %broadcast.splat11 = shufflevector <4 x float> %broadcast.splatinsert10, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = fsub fast <4 x float> %wide.load, %broadcast.splat11 %4 = getelementptr inbounds float, float* %C, i32 %index %5 = bitcast float* %4 to <4 x float>* store <4 x float> %3, <4 x float>* %5, align 4 %index.next = add i32 %index, 4 %6 = icmp eq i32 %index.next, %n br i1 %6, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fsub_r(float* noalias nocapture readonly %A, float %B, float* noalias nocapture %C, i32 %n) { ; CHECK-LABEL: test_fsub_r: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r2, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB5_1: @ %vector.ph ; CHECK-NEXT: vmov r3, s0 ; CHECK-NEXT: vdup.32 q0, r3 ; CHECK-NEXT: .LBB5_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q1, [r0], #16 ; CHECK-NEXT: subs r2, #4 ; CHECK-NEXT: vsub.f32 q1, q0, q1 ; CHECK-NEXT: vstrb.8 q1, [r1], #16 ; CHECK-NEXT: bne .LBB5_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp18 = icmp sgt i32 %n, 0 br i1 %cmp18, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert10 = insertelement <4 x float> undef, float %B, i32 0 %broadcast.splat11 = shufflevector <4 x float> %broadcast.splatinsert10, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = fsub fast <4 x float> %broadcast.splat11, %wide.load %4 = getelementptr inbounds float, float* %C, i32 %index %5 = bitcast float* %4 to <4 x float>* store <4 x float> %3, <4 x float>* %5, align 4 %index.next = add i32 %index, 4 %6 = icmp eq i32 %index.next, %n br i1 %6, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fmas(float* noalias nocapture readonly %A, float* noalias nocapture readonly %B, float %C, float* noalias nocapture %D, i32 %n) { ; CHECK-LABEL: test_fmas: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r3, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB6_1: @ %vector.ph ; CHECK-NEXT: vmov r12, s0 ; CHECK-NEXT: .LBB6_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r0], #16 ; CHECK-NEXT: vldrw.u32 q1, [r1], #16 ; CHECK-NEXT: subs r3, #4 ; CHECK-NEXT: vfmas.f32 q1, q0, r12 ; CHECK-NEXT: vstrb.8 q1, [r2], #16 ; CHECK-NEXT: bne .LBB6_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp110 = icmp sgt i32 %n, 0 br i1 %cmp110, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert13 = insertelement <4 x float> undef, float %C, i32 0 %broadcast.splat14 = shufflevector <4 x float> %broadcast.splatinsert13, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = getelementptr inbounds float, float* %B, i32 %index %4 = bitcast float* %3 to <4 x float>* %wide.load12 = load <4 x float>, <4 x float>* %4, align 4 %5 = fmul fast <4 x float> %wide.load12, %wide.load %6 = fadd fast <4 x float> %5, %broadcast.splat14 %7 = getelementptr inbounds float, float* %D, i32 %index %8 = bitcast float* %7 to <4 x float>* store <4 x float> %6, <4 x float>* %8, align 4 %index.next = add i32 %index, 4 %9 = icmp eq i32 %index.next, %n br i1 %9, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fmas_r(float* noalias nocapture readonly %A, float* noalias nocapture readonly %B, float %C, float* noalias nocapture %D, i32 %n) { ; CHECK-LABEL: test_fmas_r: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r3, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB7_1: @ %vector.ph ; CHECK-NEXT: vmov r12, s0 ; CHECK-NEXT: .LBB7_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r0], #16 ; CHECK-NEXT: vldrw.u32 q1, [r1], #16 ; CHECK-NEXT: subs r3, #4 ; CHECK-NEXT: vfmas.f32 q1, q0, r12 ; CHECK-NEXT: vstrb.8 q1, [r2], #16 ; CHECK-NEXT: bne .LBB7_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp110 = icmp sgt i32 %n, 0 br i1 %cmp110, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert13 = insertelement <4 x float> undef, float %C, i32 0 %broadcast.splat14 = shufflevector <4 x float> %broadcast.splatinsert13, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = getelementptr inbounds float, float* %B, i32 %index %4 = bitcast float* %3 to <4 x float>* %wide.load12 = load <4 x float>, <4 x float>* %4, align 4 %5 = fmul fast <4 x float> %wide.load12, %wide.load %6 = fadd fast <4 x float> %broadcast.splat14, %5 %7 = getelementptr inbounds float, float* %D, i32 %index %8 = bitcast float* %7 to <4 x float>* store <4 x float> %6, <4 x float>* %8, align 4 %index.next = add i32 %index, 4 %9 = icmp eq i32 %index.next, %n br i1 %9, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fma(float* noalias nocapture readonly %A, float* noalias nocapture readonly %B, float %C, float* noalias nocapture %D, i32 %n) { ; CHECK-LABEL: test_fma: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r3, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB8_1: @ %vector.ph ; CHECK-NEXT: vmov r12, s0 ; CHECK-NEXT: .LBB8_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r0], #16 ; CHECK-NEXT: vldrw.u32 q1, [r1], #16 ; CHECK-NEXT: subs r3, #4 ; CHECK-NEXT: vfma.f32 q1, q0, r12 ; CHECK-NEXT: vstrb.8 q1, [r2], #16 ; CHECK-NEXT: bne .LBB8_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp110 = icmp sgt i32 %n, 0 br i1 %cmp110, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert12 = insertelement <4 x float> undef, float %C, i32 0 %broadcast.splat13 = shufflevector <4 x float> %broadcast.splatinsert12, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = fmul fast <4 x float> %wide.load, %broadcast.splat13 %4 = getelementptr inbounds float, float* %B, i32 %index %5 = bitcast float* %4 to <4 x float>* %wide.load14 = load <4 x float>, <4 x float>* %5, align 4 %6 = fadd fast <4 x float> %3, %wide.load14 %7 = getelementptr inbounds float, float* %D, i32 %index %8 = bitcast float* %7 to <4 x float>* store <4 x float> %6, <4 x float>* %8, align 4 %index.next = add i32 %index, 4 %9 = icmp eq i32 %index.next, %n br i1 %9, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fma_r(float* noalias nocapture readonly %A, float* noalias nocapture readonly %B, float %C, float* noalias nocapture %D, i32 %n) { ; CHECK-LABEL: test_fma_r: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r3, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB9_1: @ %vector.ph ; CHECK-NEXT: vmov r12, s0 ; CHECK-NEXT: .LBB9_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r0], #16 ; CHECK-NEXT: vldrw.u32 q1, [r1], #16 ; CHECK-NEXT: subs r3, #4 ; CHECK-NEXT: vfma.f32 q1, q0, r12 ; CHECK-NEXT: vstrb.8 q1, [r2], #16 ; CHECK-NEXT: bne .LBB9_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp110 = icmp sgt i32 %n, 0 br i1 %cmp110, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert12 = insertelement <4 x float> undef, float %C, i32 0 %broadcast.splat13 = shufflevector <4 x float> %broadcast.splatinsert12, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = fmul fast <4 x float> %broadcast.splat13, %wide.load %4 = getelementptr inbounds float, float* %B, i32 %index %5 = bitcast float* %4 to <4 x float>* %wide.load14 = load <4 x float>, <4 x float>* %5, align 4 %6 = fadd fast <4 x float> %3, %wide.load14 %7 = getelementptr inbounds float, float* %D, i32 %index %8 = bitcast float* %7 to <4 x float>* store <4 x float> %6, <4 x float>* %8, align 4 %index.next = add i32 %index, 4 %9 = icmp eq i32 %index.next, %n br i1 %9, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fmss(float* noalias nocapture readonly %A, float* noalias nocapture readonly %B, float %C, float* noalias nocapture %D, i32 %n) { ; CHECK-LABEL: test_fmss: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r3, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB10_1: @ %vector.ph ; CHECK-NEXT: vmov r12, s0 ; CHECK-NEXT: vdup.32 q0, r12 ; CHECK-NEXT: vneg.f32 q0, q0 ; CHECK-NEXT: .LBB10_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q1, [r0], #16 ; CHECK-NEXT: vldrw.u32 q2, [r1], #16 ; CHECK-NEXT: vmov q3, q0 ; CHECK-NEXT: subs r3, #4 ; CHECK-NEXT: vfma.f32 q3, q2, q1 ; CHECK-NEXT: vstrb.8 q3, [r2], #16 ; CHECK-NEXT: bne .LBB10_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp110 = icmp sgt i32 %n, 0 br i1 %cmp110, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert13 = insertelement <4 x float> undef, float %C, i32 0 %broadcast.splat14 = shufflevector <4 x float> %broadcast.splatinsert13, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = getelementptr inbounds float, float* %B, i32 %index %4 = bitcast float* %3 to <4 x float>* %wide.load12 = load <4 x float>, <4 x float>* %4, align 4 %5 = fmul fast <4 x float> %wide.load12, %wide.load %6 = fsub fast <4 x float> %5, %broadcast.splat14 %7 = getelementptr inbounds float, float* %D, i32 %index %8 = bitcast float* %7 to <4 x float>* store <4 x float> %6, <4 x float>* %8, align 4 %index.next = add i32 %index, 4 %9 = icmp eq i32 %index.next, %n br i1 %9, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fmss_r(float* noalias nocapture readonly %A, float* noalias nocapture readonly %B, float %C, float* noalias nocapture %D, i32 %n) { ; CHECK-LABEL: test_fmss_r: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r3, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB11_1: @ %vector.ph ; CHECK-NEXT: vmov r12, s0 ; CHECK-NEXT: vdup.32 q0, r12 ; CHECK-NEXT: .LBB11_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q1, [r0], #16 ; CHECK-NEXT: vldrw.u32 q2, [r1], #16 ; CHECK-NEXT: vmov q3, q0 ; CHECK-NEXT: subs r3, #4 ; CHECK-NEXT: vfms.f32 q3, q2, q1 ; CHECK-NEXT: vstrb.8 q3, [r2], #16 ; CHECK-NEXT: bne .LBB11_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp110 = icmp sgt i32 %n, 0 br i1 %cmp110, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert13 = insertelement <4 x float> undef, float %C, i32 0 %broadcast.splat14 = shufflevector <4 x float> %broadcast.splatinsert13, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = getelementptr inbounds float, float* %B, i32 %index %4 = bitcast float* %3 to <4 x float>* %wide.load12 = load <4 x float>, <4 x float>* %4, align 4 %5 = fmul fast <4 x float> %wide.load12, %wide.load %6 = fsub fast <4 x float> %broadcast.splat14, %5 %7 = getelementptr inbounds float, float* %D, i32 %index %8 = bitcast float* %7 to <4 x float>* store <4 x float> %6, <4 x float>* %8, align 4 %index.next = add i32 %index, 4 %9 = icmp eq i32 %index.next, %n br i1 %9, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fms(float* noalias nocapture readonly %A, float* noalias nocapture readonly %B, float %C, float* noalias nocapture %D, i32 %n) { ; CHECK-LABEL: test_fms: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r3, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB12_1: @ %vector.ph ; CHECK-NEXT: vmov r12, s0 ; CHECK-NEXT: .LBB12_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r1], #16 ; CHECK-NEXT: vldrw.u32 q1, [r0], #16 ; CHECK-NEXT: subs r3, #4 ; CHECK-NEXT: vneg.f32 q0, q0 ; CHECK-NEXT: vfma.f32 q0, q1, r12 ; CHECK-NEXT: vstrb.8 q0, [r2], #16 ; CHECK-NEXT: bne .LBB12_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp110 = icmp sgt i32 %n, 0 br i1 %cmp110, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert12 = insertelement <4 x float> undef, float %C, i32 0 %broadcast.splat13 = shufflevector <4 x float> %broadcast.splatinsert12, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = fmul fast <4 x float> %wide.load, %broadcast.splat13 %4 = getelementptr inbounds float, float* %B, i32 %index %5 = bitcast float* %4 to <4 x float>* %wide.load14 = load <4 x float>, <4 x float>* %5, align 4 %6 = fsub fast <4 x float> %3, %wide.load14 %7 = getelementptr inbounds float, float* %D, i32 %index %8 = bitcast float* %7 to <4 x float>* store <4 x float> %6, <4 x float>* %8, align 4 %index.next = add i32 %index, 4 %9 = icmp eq i32 %index.next, %n br i1 %9, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define arm_aapcs_vfpcc void @test_fms_r(float* noalias nocapture readonly %A, float* noalias nocapture readonly %B, float %C, float* noalias nocapture %D, i32 %n) { ; CHECK-LABEL: test_fms_r: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: cmp r3, #1 ; CHECK-NEXT: it lt ; CHECK-NEXT: bxlt lr ; CHECK-NEXT: .LBB13_1: @ %vector.ph ; CHECK-NEXT: vmov r12, s0 ; CHECK-NEXT: .LBB13_2: @ %vector.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r1], #16 ; CHECK-NEXT: vldrw.u32 q1, [r0], #16 ; CHECK-NEXT: subs r3, #4 ; CHECK-NEXT: vneg.f32 q0, q0 ; CHECK-NEXT: vfma.f32 q0, q1, r12 ; CHECK-NEXT: vstrb.8 q0, [r2], #16 ; CHECK-NEXT: bne .LBB13_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond.cleanup ; CHECK-NEXT: bx lr entry: %0 = and i32 %n, 7 %cmp = icmp eq i32 %0, 0 tail call void @llvm.assume(i1 %cmp) %cmp110 = icmp sgt i32 %n, 0 br i1 %cmp110, label %vector.ph, label %for.cond.cleanup vector.ph: ; preds = %entry %broadcast.splatinsert12 = insertelement <4 x float> undef, float %C, i32 0 %broadcast.splat13 = shufflevector <4 x float> %broadcast.splatinsert12, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %vector.ph %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] %1 = getelementptr inbounds float, float* %A, i32 %index %2 = bitcast float* %1 to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %2, align 4 %3 = fmul fast <4 x float> %broadcast.splat13, %wide.load %4 = getelementptr inbounds float, float* %B, i32 %index %5 = bitcast float* %4 to <4 x float>* %wide.load14 = load <4 x float>, <4 x float>* %5, align 4 %6 = fsub fast <4 x float> %3, %wide.load14 %7 = getelementptr inbounds float, float* %D, i32 %index %8 = bitcast float* %7 to <4 x float>* store <4 x float> %6, <4 x float>* %8, align 4 %index.next = add i32 %index, 4 %9 = icmp eq i32 %index.next, %n br i1 %9, label %for.cond.cleanup, label %vector.body for.cond.cleanup: ; preds = %vector.body, %entry ret void } define dso_local void @test_nested(float* noalias nocapture %pInT1, float* noalias nocapture readonly %pOutT1, float* noalias nocapture readonly %pPRT_in, float* noalias nocapture readnone %pPRT_pDst, i32 %numRows, i32 %numCols, i32 %l, float %in) local_unnamed_addr #0 { ; CHECK-LABEL: test_nested: ; CHECK: @ %bb.0: @ %for.body.us.preheader ; CHECK-NEXT: .save {r4, r5, r6, lr} ; CHECK-NEXT: push {r4, r5, r6, lr} ; CHECK-NEXT: ldrd lr, r12, [sp, #16] ; CHECK-NEXT: lsl.w r3, r12, #2 ; CHECK-NEXT: dls lr, lr ; CHECK-NEXT: .LBB14_1: @ %for.body.us ; CHECK-NEXT: @ =>This Loop Header: Depth=1 ; CHECK-NEXT: @ Child Loop BB14_2 Depth 2 ; CHECK-NEXT: ldr r4, [r1] ; CHECK-NEXT: mov r5, r2 ; CHECK-NEXT: mov r6, r12 ; CHECK-NEXT: vdup.32 q0, r4 ; CHECK-NEXT: mov r4, r0 ; CHECK-NEXT: .LBB14_2: @ %vector.body ; CHECK-NEXT: @ Parent Loop BB14_1 Depth=1 ; CHECK-NEXT: @ => This Inner Loop Header: Depth=2 ; CHECK-NEXT: vldrw.u32 q1, [r5], #16 ; CHECK-NEXT: vldrw.u32 q2, [r4] ; CHECK-NEXT: subs r6, #4 ; CHECK-NEXT: vfms.f32 q2, q1, q0 ; CHECK-NEXT: vstrb.8 q2, [r4], #16 ; CHECK-NEXT: bne .LBB14_2 ; CHECK-NEXT: @ %bb.3: @ %for.cond6.for.end_crit_edge.us ; CHECK-NEXT: @ in Loop: Header=BB14_1 Depth=1 ; CHECK-NEXT: add r0, r3 ; CHECK-NEXT: add r2, r3 ; CHECK-NEXT: adds r1, #4 ; CHECK-NEXT: le lr, .LBB14_1 ; CHECK-NEXT: @ %bb.4: @ %for.end14 ; CHECK-NEXT: pop {r4, r5, r6, pc} for.body.us.preheader: %cmp = icmp sgt i32 %numRows, 0 tail call void @llvm.assume(i1 %cmp) %cmp1 = icmp sgt i32 %numCols, 0 tail call void @llvm.assume(i1 %cmp1) %rem = and i32 %numCols, 7 %cmp2 = icmp eq i32 %rem, 0 tail call void @llvm.assume(i1 %cmp2) %cmp3 = icmp slt i32 %l, %numCols tail call void @llvm.assume(i1 %cmp3) br label %for.body.us for.body.us: ; preds = %for.cond6.for.end_crit_edge.us, %for.body.us.preheader %pInT1.addr.038.us = phi float* [ %scevgep40, %for.cond6.for.end_crit_edge.us ], [ %pInT1, %for.body.us.preheader ] %i.037.us = phi i32 [ %inc13.us, %for.cond6.for.end_crit_edge.us ], [ 0, %for.body.us.preheader ] %pOutT1.addr.036.us = phi float* [ %incdec.ptr.us, %for.cond6.for.end_crit_edge.us ], [ %pOutT1, %for.body.us.preheader ] %pPRT_in.addr.035.us = phi float* [ %scevgep, %for.cond6.for.end_crit_edge.us ], [ %pPRT_in, %for.body.us.preheader ] %scevgep = getelementptr float, float* %pPRT_in.addr.035.us, i32 %numCols %0 = load float, float* %pOutT1.addr.036.us, align 4 %broadcast.splatinsert47 = insertelement <4 x float> undef, float %0, i32 0 %broadcast.splat48 = shufflevector <4 x float> %broadcast.splatinsert47, <4 x float> undef, <4 x i32> zeroinitializer br label %vector.body vector.body: ; preds = %vector.body, %for.body.us %index = phi i32 [ 0, %for.body.us ], [ %index.next, %vector.body ] %next.gep = getelementptr float, float* %pInT1.addr.038.us, i32 %index %next.gep45 = getelementptr float, float* %pPRT_in.addr.035.us, i32 %index %1 = bitcast float* %next.gep to <4 x float>* %wide.load = load <4 x float>, <4 x float>* %1, align 4 %2 = bitcast float* %next.gep45 to <4 x float>* %wide.load46 = load <4 x float>, <4 x float>* %2, align 4 %3 = fmul fast <4 x float> %wide.load46, %broadcast.splat48 %4 = fsub fast <4 x float> %wide.load, %3 store <4 x float> %4, <4 x float>* %1, align 4 %index.next = add i32 %index, 4 %5 = icmp eq i32 %index.next, %numCols br i1 %5, label %for.cond6.for.end_crit_edge.us, label %vector.body for.cond6.for.end_crit_edge.us: ; preds = %vector.body %incdec.ptr.us = getelementptr inbounds float, float* %pOutT1.addr.036.us, i32 1 %scevgep40 = getelementptr float, float* %pInT1.addr.038.us, i32 %numCols %inc13.us = add nuw nsw i32 %i.037.us, 1 %exitcond41 = icmp eq i32 %inc13.us, %numRows br i1 %exitcond41, label %for.end14, label %for.body.us for.end14: ; preds = %for.cond6.for.end_crit_edge.us ret void } %struct.arm_fir_instance_f32 = type { i16, float*, float* } define void @arm_fir_f32_1_4_mve(%struct.arm_fir_instance_f32* nocapture readonly %S, float* nocapture readonly %pSrc, float* %pDst, i32 %blockSize) { ; CHECK-LABEL: arm_fir_f32_1_4_mve: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: .pad #8 ; CHECK-NEXT: sub sp, #8 ; CHECK-NEXT: ldrh.w r9, [r0] ; CHECK-NEXT: mov r11, r1 ; CHECK-NEXT: ldr.w r12, [r0, #4] ; CHECK-NEXT: sub.w r1, r9, #1 ; CHECK-NEXT: cmp r1, #3 ; CHECK-NEXT: bhi .LBB15_6 ; CHECK-NEXT: @ %bb.1: @ %if.then ; CHECK-NEXT: ldr r4, [r0, #8] ; CHECK-NEXT: lsr.w lr, r3, #2 ; CHECK-NEXT: ldrd r7, r6, [r4] ; CHECK-NEXT: ldrd r5, r8, [r4, #8] ; CHECK-NEXT: add.w r4, r12, r1, lsl #2 ; CHECK-NEXT: wls lr, lr, .LBB15_5 ; CHECK-NEXT: @ %bb.2: @ %while.body.lr.ph ; CHECK-NEXT: bic r1, r3, #3 ; CHECK-NEXT: str r1, [sp] @ 4-byte Spill ; CHECK-NEXT: add.w r10, r12, #4 ; CHECK-NEXT: add.w r1, r2, r1, lsl #2 ; CHECK-NEXT: str r1, [sp, #4] @ 4-byte Spill ; CHECK-NEXT: mov r1, r11 ; CHECK-NEXT: .LBB15_3: @ %while.body ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r1], #16 ; CHECK-NEXT: vstrb.8 q0, [r4], #16 ; CHECK-NEXT: vldrw.u32 q0, [r10, #-4] ; CHECK-NEXT: vldrw.u32 q1, [r10], #16 ; CHECK-NEXT: vmul.f32 q0, q0, r7 ; CHECK-NEXT: vldrw.u32 q2, [r10, #-8] ; CHECK-NEXT: vfma.f32 q0, q1, r6 ; CHECK-NEXT: vldrw.u32 q1, [r10, #-12] ; CHECK-NEXT: vfma.f32 q0, q1, r5 ; CHECK-NEXT: vfma.f32 q0, q2, r8 ; CHECK-NEXT: vstrb.8 q0, [r2], #16 ; CHECK-NEXT: le lr, .LBB15_3 ; CHECK-NEXT: @ %bb.4: @ %while.end.loopexit ; CHECK-NEXT: ldr r1, [sp] @ 4-byte Reload ; CHECK-NEXT: ldr r2, [sp, #4] @ 4-byte Reload ; CHECK-NEXT: add.w r12, r12, r1, lsl #2 ; CHECK-NEXT: add.w r11, r11, r1, lsl #2 ; CHECK-NEXT: .LBB15_5: @ %while.end ; CHECK-NEXT: and r1, r3, #3 ; CHECK-NEXT: vldrw.u32 q0, [r11] ; CHECK-NEXT: vctp.32 r1 ; CHECK-NEXT: vpst ; CHECK-NEXT: vstrwt.32 q0, [r4] ; CHECK-NEXT: vldrw.u32 q0, [r12] ; CHECK-NEXT: vldrw.u32 q1, [r12, #4] ; CHECK-NEXT: vmul.f32 q0, q0, r7 ; CHECK-NEXT: vfma.f32 q0, q1, r6 ; CHECK-NEXT: vldrw.u32 q1, [r12, #8] ; CHECK-NEXT: vfma.f32 q0, q1, r5 ; CHECK-NEXT: vldrw.u32 q1, [r12, #12] ; CHECK-NEXT: vfma.f32 q0, q1, r8 ; CHECK-NEXT: vpst ; CHECK-NEXT: vstrwt.32 q0, [r2] ; CHECK-NEXT: ldr.w r12, [r0, #4] ; CHECK-NEXT: .LBB15_6: @ %if.end ; CHECK-NEXT: add.w r0, r12, r3, lsl #2 ; CHECK-NEXT: lsr.w lr, r9, #2 ; CHECK-NEXT: wls lr, lr, .LBB15_10 ; CHECK-NEXT: @ %bb.7: @ %while.body51.preheader ; CHECK-NEXT: bic r2, r9, #3 ; CHECK-NEXT: adds r1, r2, r3 ; CHECK-NEXT: mov r3, r12 ; CHECK-NEXT: add.w r1, r12, r1, lsl #2 ; CHECK-NEXT: .LBB15_8: @ %while.body51 ; CHECK-NEXT: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r0], #16 ; CHECK-NEXT: vstrb.8 q0, [r3], #16 ; CHECK-NEXT: le lr, .LBB15_8 ; CHECK-NEXT: @ %bb.9: @ %while.end55.loopexit ; CHECK-NEXT: add.w r12, r12, r2, lsl #2 ; CHECK-NEXT: mov r0, r1 ; CHECK-NEXT: .LBB15_10: @ %while.end55 ; CHECK-NEXT: ands r1, r9, #3 ; CHECK-NEXT: beq .LBB15_12 ; CHECK-NEXT: @ %bb.11: @ %if.then59 ; CHECK-NEXT: vldrw.u32 q0, [r0] ; CHECK-NEXT: vctp.32 r1 ; CHECK-NEXT: vpst ; CHECK-NEXT: vstrwt.32 q0, [r12] ; CHECK-NEXT: .LBB15_12: @ %if.end61 ; CHECK-NEXT: add sp, #8 ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc} entry: %pState1 = getelementptr inbounds %struct.arm_fir_instance_f32, %struct.arm_fir_instance_f32* %S, i32 0, i32 1 %0 = load float*, float** %pState1, align 4 %pCoeffs2 = getelementptr inbounds %struct.arm_fir_instance_f32, %struct.arm_fir_instance_f32* %S, i32 0, i32 2 %1 = load float*, float** %pCoeffs2, align 4 %numTaps3 = getelementptr inbounds %struct.arm_fir_instance_f32, %struct.arm_fir_instance_f32* %S, i32 0, i32 0 %2 = load i16, i16* %numTaps3, align 4 %conv = zext i16 %2 to i32 %sub = add nsw i32 %conv, -1 %cmp = icmp ult i32 %sub, 4 br i1 %cmp, label %if.then, label %if.end if.then: ; preds = %entry %arrayidx = getelementptr inbounds float, float* %0, i32 %sub %incdec.ptr = getelementptr inbounds float, float* %1, i32 1 %3 = load float, float* %1, align 4 %incdec.ptr6 = getelementptr inbounds float, float* %1, i32 2 %4 = load float, float* %incdec.ptr, align 4 %incdec.ptr7 = getelementptr inbounds float, float* %1, i32 3 %5 = load float, float* %incdec.ptr6, align 4 %6 = load float, float* %incdec.ptr7, align 4 %shr = lshr i32 %blockSize, 2 %cmp9146 = icmp eq i32 %shr, 0 %.pre161 = insertelement <4 x float> undef, float %3, i32 0 %.pre162 = shufflevector <4 x float> %.pre161, <4 x float> undef, <4 x i32> zeroinitializer %.pre163 = insertelement <4 x float> undef, float %4, i32 0 %.pre164 = shufflevector <4 x float> %.pre163, <4 x float> undef, <4 x i32> zeroinitializer %.pre165 = insertelement <4 x float> undef, float %5, i32 0 %.pre166 = shufflevector <4 x float> %.pre165, <4 x float> undef, <4 x i32> zeroinitializer %.pre167 = insertelement <4 x float> undef, float %6, i32 0 %.pre168 = shufflevector <4 x float> %.pre167, <4 x float> undef, <4 x i32> zeroinitializer br i1 %cmp9146, label %while.end, label %while.body.lr.ph while.body.lr.ph: ; preds = %if.then %7 = and i32 %blockSize, -4 %scevgep158 = getelementptr float, float* %pDst, i32 %7 br label %while.body while.body: ; preds = %while.body.lr.ph, %while.body %pStateCur.0151 = phi float* [ %arrayidx, %while.body.lr.ph ], [ %add.ptr, %while.body ] %pSamples.0150 = phi float* [ %0, %while.body.lr.ph ], [ %add.ptr24, %while.body ] %pOutput.0149 = phi float* [ %pDst, %while.body.lr.ph ], [ %add.ptr23, %while.body ] %pTempSrc.0148 = phi float* [ %pSrc, %while.body.lr.ph ], [ %add.ptr11, %while.body ] %blkCnt.0147 = phi i32 [ %shr, %while.body.lr.ph ], [ %dec, %while.body ] %8 = bitcast float* %pTempSrc.0148 to <4 x float>* %9 = load <4 x float>, <4 x float>* %8, align 4 %10 = bitcast float* %pStateCur.0151 to <4 x float>* store <4 x float> %9, <4 x float>* %10, align 4 %add.ptr = getelementptr inbounds float, float* %pStateCur.0151, i32 4 %add.ptr11 = getelementptr inbounds float, float* %pTempSrc.0148, i32 4 %11 = bitcast float* %pSamples.0150 to <4 x float>* %12 = load <4 x float>, <4 x float>* %11, align 4 %13 = fmul fast <4 x float> %12, %.pre162 %arrayidx12 = getelementptr inbounds float, float* %pSamples.0150, i32 1 %14 = bitcast float* %arrayidx12 to <4 x float>* %15 = load <4 x float>, <4 x float>* %14, align 4 %mul = fmul fast <4 x float> %15, %.pre164 %add = fadd fast <4 x float> %mul, %13 %arrayidx13 = getelementptr inbounds float, float* %pSamples.0150, i32 2 %16 = bitcast float* %arrayidx13 to <4 x float>* %17 = load <4 x float>, <4 x float>* %16, align 4 %mul16 = fmul fast <4 x float> %17, %.pre166 %add17 = fadd fast <4 x float> %add, %mul16 %arrayidx18 = getelementptr inbounds float, float* %pSamples.0150, i32 3 %18 = bitcast float* %arrayidx18 to <4 x float>* %19 = load <4 x float>, <4 x float>* %18, align 4 %mul21 = fmul fast <4 x float> %19, %.pre168 %add22 = fadd fast <4 x float> %add17, %mul21 %20 = bitcast float* %pOutput.0149 to <4 x float>* store <4 x float> %add22, <4 x float>* %20, align 4 %add.ptr23 = getelementptr inbounds float, float* %pOutput.0149, i32 4 %add.ptr24 = getelementptr inbounds float, float* %pSamples.0150, i32 4 %dec = add nsw i32 %blkCnt.0147, -1 %cmp9 = icmp eq i32 %dec, 0 br i1 %cmp9, label %while.end.loopexit, label %while.body while.end.loopexit: ; preds = %while.body %scevgep157 = getelementptr float, float* %pSrc, i32 %7 %scevgep159 = getelementptr float, float* %0, i32 %7 br label %while.end while.end: ; preds = %if.then, %while.end.loopexit %pTempSrc.0.lcssa = phi float* [ %scevgep157, %while.end.loopexit ], [ %pSrc, %if.then ] %pOutput.0.lcssa = phi float* [ %scevgep158, %while.end.loopexit ], [ %pDst, %if.then ] %pSamples.0.lcssa = phi float* [ %scevgep159, %while.end.loopexit ], [ %0, %if.then ] %pStateCur.0.lcssa = phi float* [ %add.ptr, %while.end.loopexit ], [ %arrayidx, %if.then ] %and = and i32 %blockSize, 3 %21 = tail call <4 x i1> @llvm.arm.mve.vctp32(i32 %and) %22 = bitcast float* %pTempSrc.0.lcssa to <4 x float>* %23 = load <4 x float>, <4 x float>* %22, align 4 %24 = bitcast float* %pStateCur.0.lcssa to <4 x float>* tail call void @llvm.masked.store.v4f32.p0v4f32(<4 x float> %23, <4 x float>* %24, i32 4, <4 x i1> %21) %25 = bitcast float* %pSamples.0.lcssa to <4 x float>* %26 = load <4 x float>, <4 x float>* %25, align 4 %27 = fmul fast <4 x float> %26, %.pre162 %arrayidx29 = getelementptr inbounds float, float* %pSamples.0.lcssa, i32 1 %28 = bitcast float* %arrayidx29 to <4 x float>* %29 = load <4 x float>, <4 x float>* %28, align 4 %mul32 = fmul fast <4 x float> %29, %.pre164 %add33 = fadd fast <4 x float> %mul32, %27 %arrayidx34 = getelementptr inbounds float, float* %pSamples.0.lcssa, i32 2 %30 = bitcast float* %arrayidx34 to <4 x float>* %31 = load <4 x float>, <4 x float>* %30, align 4 %mul37 = fmul fast <4 x float> %31, %.pre166 %add38 = fadd fast <4 x float> %add33, %mul37 %arrayidx39 = getelementptr inbounds float, float* %pSamples.0.lcssa, i32 3 %32 = bitcast float* %arrayidx39 to <4 x float>* %33 = load <4 x float>, <4 x float>* %32, align 4 %mul42 = fmul fast <4 x float> %33, %.pre168 %add43 = fadd fast <4 x float> %add38, %mul42 %34 = bitcast float* %pOutput.0.lcssa to <4 x float>* tail call void @llvm.masked.store.v4f32.p0v4f32(<4 x float> %add43, <4 x float>* %34, i32 4, <4 x i1> %21) %.pre = load float*, float** %pState1, align 4 br label %if.end if.end: ; preds = %while.end, %entry %35 = phi float* [ %.pre, %while.end ], [ %0, %entry ] %arrayidx45 = getelementptr inbounds float, float* %35, i32 %blockSize %shr47 = lshr i32 %conv, 2 %cmp49141 = icmp eq i32 %shr47, 0 br i1 %cmp49141, label %while.end55, label %while.body51.preheader while.body51.preheader: ; preds = %if.end %36 = and i32 %conv, 65532 %37 = add i32 %36, %blockSize %scevgep = getelementptr float, float* %35, i32 %37 br label %while.body51 while.body51: ; preds = %while.body51.preheader, %while.body51 %pTempSrc.1144 = phi float* [ %add.ptr52, %while.body51 ], [ %arrayidx45, %while.body51.preheader ] %pTempDest.0143 = phi float* [ %add.ptr53, %while.body51 ], [ %35, %while.body51.preheader ] %blkCnt.1142 = phi i32 [ %dec54, %while.body51 ], [ %shr47, %while.body51.preheader ] %38 = bitcast float* %pTempSrc.1144 to <4 x float>* %39 = load <4 x float>, <4 x float>* %38, align 4 %40 = bitcast float* %pTempDest.0143 to <4 x float>* store <4 x float> %39, <4 x float>* %40, align 4 %add.ptr52 = getelementptr inbounds float, float* %pTempSrc.1144, i32 4 %add.ptr53 = getelementptr inbounds float, float* %pTempDest.0143, i32 4 %dec54 = add nsw i32 %blkCnt.1142, -1 %cmp49 = icmp eq i32 %dec54, 0 br i1 %cmp49, label %while.end55.loopexit, label %while.body51 while.end55.loopexit: ; preds = %while.body51 %scevgep156 = getelementptr float, float* %35, i32 %36 br label %while.end55 while.end55: ; preds = %while.end55.loopexit, %if.end %pTempDest.0.lcssa = phi float* [ %35, %if.end ], [ %scevgep156, %while.end55.loopexit ] %pTempSrc.1.lcssa = phi float* [ %arrayidx45, %if.end ], [ %scevgep, %while.end55.loopexit ] %and56 = and i32 %conv, 3 %cmp57 = icmp eq i32 %and56, 0 br i1 %cmp57, label %if.end61, label %if.then59 if.then59: ; preds = %while.end55 %41 = tail call <4 x i1> @llvm.arm.mve.vctp32(i32 %and56) %42 = bitcast float* %pTempSrc.1.lcssa to <4 x float>* %43 = load <4 x float>, <4 x float>* %42, align 4 %44 = bitcast float* %pTempDest.0.lcssa to <4 x float>* tail call void @llvm.masked.store.v4f32.p0v4f32(<4 x float> %43, <4 x float>* %44, i32 4, <4 x i1> %41) br label %if.end61 if.end61: ; preds = %while.end55, %if.then59 ret void } define void @fir(%struct.arm_fir_instance_f32* nocapture readonly %S, float* nocapture readonly %pSrc, float* nocapture %pDst, i32 %blockSize) { ; CHECK-LABEL: fir: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: .pad #4 ; CHECK-NEXT: sub sp, #4 ; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13} ; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13} ; CHECK-NEXT: .pad #32 ; CHECK-NEXT: sub sp, #32 ; CHECK-NEXT: cmp r3, #8 ; CHECK-NEXT: blo.w .LBB16_12 ; CHECK-NEXT: @ %bb.1: @ %entry ; CHECK-NEXT: lsrs.w r12, r3, #2 ; CHECK-NEXT: beq.w .LBB16_12 ; CHECK-NEXT: @ %bb.2: @ %while.body.lr.ph ; CHECK-NEXT: ldrh r6, [r0] ; CHECK-NEXT: movs r4, #1 ; CHECK-NEXT: ldrd r5, r10, [r0, #4] ; CHECK-NEXT: sub.w r3, r6, #8 ; CHECK-NEXT: add.w r0, r3, r3, lsr #29 ; CHECK-NEXT: asrs r7, r0, #3 ; CHECK-NEXT: cmp r7, #1 ; CHECK-NEXT: it gt ; CHECK-NEXT: asrgt r4, r0, #3 ; CHECK-NEXT: add.w r0, r5, r6, lsl #2 ; CHECK-NEXT: sub.w r9, r0, #4 ; CHECK-NEXT: rsbs r0, r6, #0 ; CHECK-NEXT: str r4, [sp, #4] @ 4-byte Spill ; CHECK-NEXT: and r4, r3, #7 ; CHECK-NEXT: str r0, [sp, #16] @ 4-byte Spill ; CHECK-NEXT: add.w r0, r10, #32 ; CHECK-NEXT: str r6, [sp, #20] @ 4-byte Spill ; CHECK-NEXT: str r0, [sp, #8] @ 4-byte Spill ; CHECK-NEXT: str r4, [sp, #12] @ 4-byte Spill ; CHECK-NEXT: b .LBB16_4 ; CHECK-NEXT: .LBB16_3: @ %while.end ; CHECK-NEXT: @ in Loop: Header=BB16_4 Depth=1 ; CHECK-NEXT: ldr r0, [sp, #16] @ 4-byte Reload ; CHECK-NEXT: subs.w r12, r12, #1 ; CHECK-NEXT: vstrb.8 q0, [r2], #16 ; CHECK-NEXT: add.w r0, r5, r0, lsl #2 ; CHECK-NEXT: add.w r5, r0, #16 ; CHECK-NEXT: beq.w .LBB16_12 ; CHECK-NEXT: .LBB16_4: @ %while.body ; CHECK-NEXT: @ =>This Loop Header: Depth=1 ; CHECK-NEXT: @ Child Loop BB16_6 Depth 2 ; CHECK-NEXT: @ Child Loop BB16_10 Depth 2 ; CHECK-NEXT: add.w lr, r10, #8 ; CHECK-NEXT: vldrw.u32 q0, [r1], #16 ; CHECK-NEXT: ldrd r3, r7, [r10] ; CHECK-NEXT: ldm.w lr, {r0, r4, r6, lr} ; CHECK-NEXT: ldrd r11, r8, [r10, #24] ; CHECK-NEXT: vstrb.8 q0, [r9], #16 ; CHECK-NEXT: vldrw.u32 q0, [r5], #32 ; CHECK-NEXT: strd r9, r1, [sp, #24] @ 8-byte Folded Spill ; CHECK-NEXT: vldrw.u32 q1, [r5, #-28] ; CHECK-NEXT: vmul.f32 q0, q0, r3 ; CHECK-NEXT: vldrw.u32 q6, [r5, #-24] ; CHECK-NEXT: vldrw.u32 q4, [r5, #-20] ; CHECK-NEXT: vfma.f32 q0, q1, r7 ; CHECK-NEXT: vldrw.u32 q5, [r5, #-16] ; CHECK-NEXT: vfma.f32 q0, q6, r0 ; CHECK-NEXT: vldrw.u32 q2, [r5, #-12] ; CHECK-NEXT: vfma.f32 q0, q4, r4 ; CHECK-NEXT: vldrw.u32 q3, [r5, #-8] ; CHECK-NEXT: vfma.f32 q0, q5, r6 ; CHECK-NEXT: ldr r0, [sp, #20] @ 4-byte Reload ; CHECK-NEXT: vfma.f32 q0, q2, lr ; CHECK-NEXT: vldrw.u32 q1, [r5, #-4] ; CHECK-NEXT: vfma.f32 q0, q3, r11 ; CHECK-NEXT: cmp r0, #16 ; CHECK-NEXT: vfma.f32 q0, q1, r8 ; CHECK-NEXT: blo .LBB16_7 ; CHECK-NEXT: @ %bb.5: @ %for.body.preheader ; CHECK-NEXT: @ in Loop: Header=BB16_4 Depth=1 ; CHECK-NEXT: ldr r0, [sp, #4] @ 4-byte Reload ; CHECK-NEXT: dls lr, r0 ; CHECK-NEXT: ldr r7, [sp, #8] @ 4-byte Reload ; CHECK-NEXT: .LBB16_6: @ %for.body ; CHECK-NEXT: @ Parent Loop BB16_4 Depth=1 ; CHECK-NEXT: @ => This Inner Loop Header: Depth=2 ; CHECK-NEXT: ldm.w r7, {r0, r3, r4, r6} ; CHECK-NEXT: vldrw.u32 q1, [r5], #32 ; CHECK-NEXT: add.w r11, r7, #16 ; CHECK-NEXT: vldrw.u32 q6, [r5, #-24] ; CHECK-NEXT: vldrw.u32 q4, [r5, #-20] ; CHECK-NEXT: vfma.f32 q0, q1, r0 ; CHECK-NEXT: vldrw.u32 q1, [r5, #-28] ; CHECK-NEXT: ldm.w r11, {r1, r8, r11} ; CHECK-NEXT: vldrw.u32 q5, [r5, #-16] ; CHECK-NEXT: vfma.f32 q0, q1, r3 ; CHECK-NEXT: vldrw.u32 q2, [r5, #-12] ; CHECK-NEXT: vfma.f32 q0, q6, r4 ; CHECK-NEXT: vldrw.u32 q3, [r5, #-8] ; CHECK-NEXT: vfma.f32 q0, q4, r6 ; CHECK-NEXT: ldr.w r9, [r7, #28] ; CHECK-NEXT: vfma.f32 q0, q5, r1 ; CHECK-NEXT: vldrw.u32 q1, [r5, #-4] ; CHECK-NEXT: vfma.f32 q0, q2, r8 ; CHECK-NEXT: adds r7, #32 ; CHECK-NEXT: vfma.f32 q0, q3, r11 ; CHECK-NEXT: vfma.f32 q0, q1, r9 ; CHECK-NEXT: le lr, .LBB16_6 ; CHECK-NEXT: b .LBB16_8 ; CHECK-NEXT: .LBB16_7: @ in Loop: Header=BB16_4 Depth=1 ; CHECK-NEXT: ldr r7, [sp, #8] @ 4-byte Reload ; CHECK-NEXT: .LBB16_8: @ %for.end ; CHECK-NEXT: @ in Loop: Header=BB16_4 Depth=1 ; CHECK-NEXT: ldrd r9, r1, [sp, #24] @ 8-byte Folded Reload ; CHECK-NEXT: ldr r4, [sp, #12] @ 4-byte Reload ; CHECK-NEXT: cmp.w r4, #0 ; CHECK-NEXT: beq .LBB16_3 ; CHECK-NEXT: b .LBB16_9 ; CHECK-NEXT: .LBB16_9: @ %while.body76.preheader ; CHECK-NEXT: @ in Loop: Header=BB16_4 Depth=1 ; CHECK-NEXT: mov r3, r5 ; CHECK-NEXT: mov lr, r4 ; CHECK-NEXT: .LBB16_10: @ %while.body76 ; CHECK-NEXT: @ Parent Loop BB16_4 Depth=1 ; CHECK-NEXT: @ => This Inner Loop Header: Depth=2 ; CHECK-NEXT: ldr r0, [r7], #4 ; CHECK-NEXT: vldrw.u32 q1, [r3], #4 ; CHECK-NEXT: subs.w lr, lr, #1 ; CHECK-NEXT: vfma.f32 q0, q1, r0 ; CHECK-NEXT: bne .LBB16_10 ; CHECK-NEXT: b .LBB16_11 ; CHECK-NEXT: .LBB16_11: @ %while.end.loopexit ; CHECK-NEXT: @ in Loop: Header=BB16_4 Depth=1 ; CHECK-NEXT: add.w r5, r5, r4, lsl #2 ; CHECK-NEXT: b .LBB16_3 ; CHECK-NEXT: .LBB16_12: @ %if.end ; CHECK-NEXT: add sp, #32 ; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13} ; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc} entry: %pState1 = getelementptr inbounds %struct.arm_fir_instance_f32, %struct.arm_fir_instance_f32* %S, i32 0, i32 1 %0 = load float*, float** %pState1, align 4 %pCoeffs2 = getelementptr inbounds %struct.arm_fir_instance_f32, %struct.arm_fir_instance_f32* %S, i32 0, i32 2 %1 = load float*, float** %pCoeffs2, align 4 %numTaps3 = getelementptr inbounds %struct.arm_fir_instance_f32, %struct.arm_fir_instance_f32* %S, i32 0, i32 0 %2 = load i16, i16* %numTaps3, align 4 %conv = zext i16 %2 to i32 %cmp = icmp ugt i32 %blockSize, 7 br i1 %cmp, label %if.then, label %if.end if.then: ; preds = %entry %shr = lshr i32 %blockSize, 2 %cmp5217 = icmp eq i32 %shr, 0 br i1 %cmp5217, label %if.end, label %while.body.lr.ph while.body.lr.ph: ; preds = %if.then %sub = add nsw i32 %conv, -1 %arrayidx = getelementptr inbounds float, float* %0, i32 %sub %incdec.ptr = getelementptr inbounds float, float* %1, i32 1 %incdec.ptr7 = getelementptr inbounds float, float* %1, i32 2 %incdec.ptr8 = getelementptr inbounds float, float* %1, i32 3 %incdec.ptr9 = getelementptr inbounds float, float* %1, i32 4 %incdec.ptr10 = getelementptr inbounds float, float* %1, i32 5 %incdec.ptr11 = getelementptr inbounds float, float* %1, i32 6 %incdec.ptr12 = getelementptr inbounds float, float* %1, i32 7 %sub37 = add nsw i32 %conv, -8 %div = sdiv i32 %sub37, 8 %pCoeffsCur.0199 = getelementptr inbounds float, float* %1, i32 8 %cmp38201 = icmp ugt i16 %2, 15 %and = and i32 %sub37, 7 %cmp74210 = icmp eq i32 %and, 0 %idx.neg = sub nsw i32 0, %conv %3 = icmp sgt i32 %div, 1 %smax = select i1 %3, i32 %div, i32 1 br label %while.body while.body: ; preds = %while.body.lr.ph, %while.end %blkCnt.0222 = phi i32 [ %shr, %while.body.lr.ph ], [ %dec84, %while.end ] %pStateCur.0221 = phi float* [ %arrayidx, %while.body.lr.ph ], [ %add.ptr, %while.end ] %pSamples.0220 = phi float* [ %0, %while.body.lr.ph ], [ %add.ptr83, %while.end ] %pTempSrc.0219 = phi float* [ %pSrc, %while.body.lr.ph ], [ %add.ptr14, %while.end ] %pOutput.0218 = phi float* [ %pDst, %while.body.lr.ph ], [ %add.ptr81, %while.end ] %4 = load float, float* %1, align 4 %5 = load float, float* %incdec.ptr, align 4 %6 = load float, float* %incdec.ptr7, align 4 %7 = load float, float* %incdec.ptr8, align 4 %8 = load float, float* %incdec.ptr9, align 4 %9 = load float, float* %incdec.ptr10, align 4 %10 = load float, float* %incdec.ptr11, align 4 %11 = load float, float* %incdec.ptr12, align 4 %12 = bitcast float* %pTempSrc.0219 to <4 x float>* %13 = load <4 x float>, <4 x float>* %12, align 4 %14 = bitcast float* %pStateCur.0221 to <4 x float>* store <4 x float> %13, <4 x float>* %14, align 4 %add.ptr = getelementptr inbounds float, float* %pStateCur.0221, i32 4 %add.ptr14 = getelementptr inbounds float, float* %pTempSrc.0219, i32 4 %15 = bitcast float* %pSamples.0220 to <4 x float>* %16 = load <4 x float>, <4 x float>* %15, align 4 %.splatinsert = insertelement <4 x float> undef, float %4, i32 0 %.splat = shufflevector <4 x float> %.splatinsert, <4 x float> undef, <4 x i32> zeroinitializer %17 = fmul fast <4 x float> %16, %.splat %arrayidx15 = getelementptr inbounds float, float* %pSamples.0220, i32 1 %18 = bitcast float* %arrayidx15 to <4 x float>* %19 = load <4 x float>, <4 x float>* %18, align 4 %.splatinsert16 = insertelement <4 x float> undef, float %5, i32 0 %.splat17 = shufflevector <4 x float> %.splatinsert16, <4 x float> undef, <4 x i32> zeroinitializer %20 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %19, <4 x float> %.splat17, <4 x float> %17) %arrayidx18 = getelementptr inbounds float, float* %pSamples.0220, i32 2 %21 = bitcast float* %arrayidx18 to <4 x float>* %22 = load <4 x float>, <4 x float>* %21, align 4 %.splatinsert19 = insertelement <4 x float> undef, float %6, i32 0 %.splat20 = shufflevector <4 x float> %.splatinsert19, <4 x float> undef, <4 x i32> zeroinitializer %23 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %22, <4 x float> %.splat20, <4 x float> %20) %arrayidx21 = getelementptr inbounds float, float* %pSamples.0220, i32 3 %24 = bitcast float* %arrayidx21 to <4 x float>* %25 = load <4 x float>, <4 x float>* %24, align 4 %.splatinsert22 = insertelement <4 x float> undef, float %7, i32 0 %.splat23 = shufflevector <4 x float> %.splatinsert22, <4 x float> undef, <4 x i32> zeroinitializer %26 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %25, <4 x float> %.splat23, <4 x float> %23) %arrayidx24 = getelementptr inbounds float, float* %pSamples.0220, i32 4 %27 = bitcast float* %arrayidx24 to <4 x float>* %28 = load <4 x float>, <4 x float>* %27, align 4 %.splatinsert25 = insertelement <4 x float> undef, float %8, i32 0 %.splat26 = shufflevector <4 x float> %.splatinsert25, <4 x float> undef, <4 x i32> zeroinitializer %29 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %28, <4 x float> %.splat26, <4 x float> %26) %arrayidx27 = getelementptr inbounds float, float* %pSamples.0220, i32 5 %30 = bitcast float* %arrayidx27 to <4 x float>* %31 = load <4 x float>, <4 x float>* %30, align 4 %.splatinsert28 = insertelement <4 x float> undef, float %9, i32 0 %.splat29 = shufflevector <4 x float> %.splatinsert28, <4 x float> undef, <4 x i32> zeroinitializer %32 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %31, <4 x float> %.splat29, <4 x float> %29) %arrayidx30 = getelementptr inbounds float, float* %pSamples.0220, i32 6 %33 = bitcast float* %arrayidx30 to <4 x float>* %34 = load <4 x float>, <4 x float>* %33, align 4 %.splatinsert31 = insertelement <4 x float> undef, float %10, i32 0 %.splat32 = shufflevector <4 x float> %.splatinsert31, <4 x float> undef, <4 x i32> zeroinitializer %35 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %34, <4 x float> %.splat32, <4 x float> %32) %arrayidx33 = getelementptr inbounds float, float* %pSamples.0220, i32 7 %36 = bitcast float* %arrayidx33 to <4 x float>* %37 = load <4 x float>, <4 x float>* %36, align 4 %.splatinsert34 = insertelement <4 x float> undef, float %11, i32 0 %.splat35 = shufflevector <4 x float> %.splatinsert34, <4 x float> undef, <4 x i32> zeroinitializer %38 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %37, <4 x float> %.splat35, <4 x float> %35) %pSamples.1200 = getelementptr inbounds float, float* %pSamples.0220, i32 8 br i1 %cmp38201, label %for.body, label %for.end for.body: ; preds = %while.body, %for.body %pSamples.1207 = phi float* [ %pSamples.1, %for.body ], [ %pSamples.1200, %while.body ] %pCoeffsCur.0206 = phi float* [ %pCoeffsCur.0, %for.body ], [ %pCoeffsCur.0199, %while.body ] %.pn205 = phi float* [ %pCoeffsCur.0206, %for.body ], [ %1, %while.body ] %i.0204 = phi i32 [ %inc, %for.body ], [ 0, %while.body ] %vecAcc0.0203 = phi <4 x float> [ %70, %for.body ], [ %38, %while.body ] %pSamples.0.pn202 = phi float* [ %pSamples.1207, %for.body ], [ %pSamples.0220, %while.body ] %incdec.ptr40 = getelementptr inbounds float, float* %.pn205, i32 9 %39 = load float, float* %pCoeffsCur.0206, align 4 %incdec.ptr41 = getelementptr inbounds float, float* %.pn205, i32 10 %40 = load float, float* %incdec.ptr40, align 4 %incdec.ptr42 = getelementptr inbounds float, float* %.pn205, i32 11 %41 = load float, float* %incdec.ptr41, align 4 %incdec.ptr43 = getelementptr inbounds float, float* %.pn205, i32 12 %42 = load float, float* %incdec.ptr42, align 4 %incdec.ptr44 = getelementptr inbounds float, float* %.pn205, i32 13 %43 = load float, float* %incdec.ptr43, align 4 %incdec.ptr45 = getelementptr inbounds float, float* %.pn205, i32 14 %44 = load float, float* %incdec.ptr44, align 4 %incdec.ptr46 = getelementptr inbounds float, float* %.pn205, i32 15 %45 = load float, float* %incdec.ptr45, align 4 %46 = load float, float* %incdec.ptr46, align 4 %47 = bitcast float* %pSamples.1207 to <4 x float>* %48 = load <4 x float>, <4 x float>* %47, align 4 %.splatinsert48 = insertelement <4 x float> undef, float %39, i32 0 %.splat49 = shufflevector <4 x float> %.splatinsert48, <4 x float> undef, <4 x i32> zeroinitializer %49 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %48, <4 x float> %.splat49, <4 x float> %vecAcc0.0203) %arrayidx50 = getelementptr inbounds float, float* %pSamples.0.pn202, i32 9 %50 = bitcast float* %arrayidx50 to <4 x float>* %51 = load <4 x float>, <4 x float>* %50, align 4 %.splatinsert51 = insertelement <4 x float> undef, float %40, i32 0 %.splat52 = shufflevector <4 x float> %.splatinsert51, <4 x float> undef, <4 x i32> zeroinitializer %52 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %51, <4 x float> %.splat52, <4 x float> %49) %arrayidx53 = getelementptr inbounds float, float* %pSamples.0.pn202, i32 10 %53 = bitcast float* %arrayidx53 to <4 x float>* %54 = load <4 x float>, <4 x float>* %53, align 4 %.splatinsert54 = insertelement <4 x float> undef, float %41, i32 0 %.splat55 = shufflevector <4 x float> %.splatinsert54, <4 x float> undef, <4 x i32> zeroinitializer %55 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %54, <4 x float> %.splat55, <4 x float> %52) %arrayidx56 = getelementptr inbounds float, float* %pSamples.0.pn202, i32 11 %56 = bitcast float* %arrayidx56 to <4 x float>* %57 = load <4 x float>, <4 x float>* %56, align 4 %.splatinsert57 = insertelement <4 x float> undef, float %42, i32 0 %.splat58 = shufflevector <4 x float> %.splatinsert57, <4 x float> undef, <4 x i32> zeroinitializer %58 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %57, <4 x float> %.splat58, <4 x float> %55) %arrayidx59 = getelementptr inbounds float, float* %pSamples.0.pn202, i32 12 %59 = bitcast float* %arrayidx59 to <4 x float>* %60 = load <4 x float>, <4 x float>* %59, align 4 %.splatinsert60 = insertelement <4 x float> undef, float %43, i32 0 %.splat61 = shufflevector <4 x float> %.splatinsert60, <4 x float> undef, <4 x i32> zeroinitializer %61 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %60, <4 x float> %.splat61, <4 x float> %58) %arrayidx62 = getelementptr inbounds float, float* %pSamples.0.pn202, i32 13 %62 = bitcast float* %arrayidx62 to <4 x float>* %63 = load <4 x float>, <4 x float>* %62, align 4 %.splatinsert63 = insertelement <4 x float> undef, float %44, i32 0 %.splat64 = shufflevector <4 x float> %.splatinsert63, <4 x float> undef, <4 x i32> zeroinitializer %64 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %63, <4 x float> %.splat64, <4 x float> %61) %arrayidx65 = getelementptr inbounds float, float* %pSamples.0.pn202, i32 14 %65 = bitcast float* %arrayidx65 to <4 x float>* %66 = load <4 x float>, <4 x float>* %65, align 4 %.splatinsert66 = insertelement <4 x float> undef, float %45, i32 0 %.splat67 = shufflevector <4 x float> %.splatinsert66, <4 x float> undef, <4 x i32> zeroinitializer %67 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %66, <4 x float> %.splat67, <4 x float> %64) %arrayidx68 = getelementptr inbounds float, float* %pSamples.0.pn202, i32 15 %68 = bitcast float* %arrayidx68 to <4 x float>* %69 = load <4 x float>, <4 x float>* %68, align 4 %.splatinsert69 = insertelement <4 x float> undef, float %46, i32 0 %.splat70 = shufflevector <4 x float> %.splatinsert69, <4 x float> undef, <4 x i32> zeroinitializer %70 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %69, <4 x float> %.splat70, <4 x float> %67) %inc = add nuw nsw i32 %i.0204, 1 %pCoeffsCur.0 = getelementptr inbounds float, float* %pCoeffsCur.0206, i32 8 %pSamples.1 = getelementptr inbounds float, float* %pSamples.1207, i32 8 %exitcond = icmp eq i32 %inc, %smax br i1 %exitcond, label %for.end, label %for.body for.end: ; preds = %for.body, %while.body %vecAcc0.0.lcssa = phi <4 x float> [ %38, %while.body ], [ %70, %for.body ] %pCoeffsCur.0.lcssa = phi float* [ %pCoeffsCur.0199, %while.body ], [ %pCoeffsCur.0, %for.body ] %pSamples.1.lcssa = phi float* [ %pSamples.1200, %while.body ], [ %pSamples.1, %for.body ] br i1 %cmp74210, label %while.end, label %while.body76 while.body76: ; preds = %for.end, %while.body76 %pCoeffsCur.1214 = phi float* [ %incdec.ptr77, %while.body76 ], [ %pCoeffsCur.0.lcssa, %for.end ] %vecAcc0.1213 = phi <4 x float> [ %74, %while.body76 ], [ %vecAcc0.0.lcssa, %for.end ] %numCnt.0212 = phi i32 [ %dec, %while.body76 ], [ %and, %for.end ] %pSamples.2211 = phi float* [ %incdec.ptr80, %while.body76 ], [ %pSamples.1.lcssa, %for.end ] %incdec.ptr77 = getelementptr inbounds float, float* %pCoeffsCur.1214, i32 1 %71 = load float, float* %pCoeffsCur.1214, align 4 %72 = bitcast float* %pSamples.2211 to <4 x float>* %73 = load <4 x float>, <4 x float>* %72, align 4 %.splatinsert78 = insertelement <4 x float> undef, float %71, i32 0 %.splat79 = shufflevector <4 x float> %.splatinsert78, <4 x float> undef, <4 x i32> zeroinitializer %74 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %73, <4 x float> %.splat79, <4 x float> %vecAcc0.1213) %incdec.ptr80 = getelementptr inbounds float, float* %pSamples.2211, i32 1 %dec = add nsw i32 %numCnt.0212, -1 %cmp74 = icmp sgt i32 %numCnt.0212, 1 br i1 %cmp74, label %while.body76, label %while.end.loopexit while.end.loopexit: ; preds = %while.body76 %scevgep = getelementptr float, float* %pSamples.1.lcssa, i32 %and br label %while.end while.end: ; preds = %while.end.loopexit, %for.end %pSamples.2.lcssa = phi float* [ %pSamples.1.lcssa, %for.end ], [ %scevgep, %while.end.loopexit ] %vecAcc0.1.lcssa = phi <4 x float> [ %vecAcc0.0.lcssa, %for.end ], [ %74, %while.end.loopexit ] %75 = bitcast float* %pOutput.0218 to <4 x float>* store <4 x float> %vecAcc0.1.lcssa, <4 x float>* %75, align 4 %add.ptr81 = getelementptr inbounds float, float* %pOutput.0218, i32 4 %add.ptr82 = getelementptr inbounds float, float* %pSamples.2.lcssa, i32 4 %add.ptr83 = getelementptr inbounds float, float* %add.ptr82, i32 %idx.neg %dec84 = add nsw i32 %blkCnt.0222, -1 %cmp5 = icmp eq i32 %dec84, 0 br i1 %cmp5, label %if.end, label %while.body if.end: ; preds = %while.end, %if.then, %entry ret void } %struct.arm_biquad_cascade_stereo_df2T_instance_f32 = type { i8, float*, float* } define arm_aapcs_vfpcc void @arm_biquad_cascade_stereo_df2T_f32(%struct.arm_biquad_cascade_stereo_df2T_instance_f32* nocapture readonly %0, float* %1, float* %2, i32 %3) { ; CHECK-LABEL: arm_biquad_cascade_stereo_df2T_f32: ; CHECK: @ %bb.0: ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, lr} ; CHECK-NEXT: .vsave {d8, d9, d10, d11} ; CHECK-NEXT: vpush {d8, d9, d10, d11} ; CHECK-NEXT: .pad #24 ; CHECK-NEXT: sub sp, #24 ; CHECK-NEXT: mov r8, r3 ; CHECK-NEXT: ldrb.w r12, [r0] ; CHECK-NEXT: ldrd r3, r0, [r0, #4] ; CHECK-NEXT: movs r4, #0 ; CHECK-NEXT: cmp.w r8, #0 ; CHECK-NEXT: strd r4, r4, [sp, #16] ; CHECK-NEXT: beq .LBB17_5 ; CHECK-NEXT: @ %bb.1: ; CHECK-NEXT: movs r5, #2 ; CHECK-NEXT: viwdup.u32 q0, r4, r5, #1 ; CHECK-NEXT: mov r4, sp ; CHECK-NEXT: .LBB17_2: @ =>This Loop Header: Depth=1 ; CHECK-NEXT: @ Child Loop BB17_3 Depth 2 ; CHECK-NEXT: ldrd r5, r7, [r0] ; CHECK-NEXT: vldrw.u32 q1, [r3] ; CHECK-NEXT: vldr s8, [r0, #8] ; CHECK-NEXT: ldr r6, [r0, #12] ; CHECK-NEXT: vstrw.32 q1, [r4] ; CHECK-NEXT: vdup.32 q1, r7 ; CHECK-NEXT: vldr s12, [r0, #16] ; CHECK-NEXT: vmov.f32 s6, s8 ; CHECK-NEXT: dls lr, r8 ; CHECK-NEXT: vmov.f32 s7, s8 ; CHECK-NEXT: vdup.32 q2, r6 ; CHECK-NEXT: vmov.f32 s10, s12 ; CHECK-NEXT: mov r7, r2 ; CHECK-NEXT: vmov.f32 s11, s12 ; CHECK-NEXT: .LBB17_3: @ Parent Loop BB17_2 Depth=1 ; CHECK-NEXT: @ => This Inner Loop Header: Depth=2 ; CHECK-NEXT: vldrw.u32 q4, [r1, q0, uxtw #2] ; CHECK-NEXT: vldrw.u32 q5, [r4, q0, uxtw #2] ; CHECK-NEXT: vldrw.u32 q3, [sp, #8] ; CHECK-NEXT: adds r1, #8 ; CHECK-NEXT: vfma.f32 q5, q4, r5 ; CHECK-NEXT: vfma.f32 q3, q5, q2 ; CHECK-NEXT: vstmia r7, {s20, s21} ; CHECK-NEXT: adds r7, #8 ; CHECK-NEXT: vfma.f32 q3, q4, q1 ; CHECK-NEXT: vstrw.32 q3, [r4] ; CHECK-NEXT: le lr, .LBB17_3 ; CHECK-NEXT: @ %bb.4: @ in Loop: Header=BB17_2 Depth=1 ; CHECK-NEXT: subs.w r12, r12, #1 ; CHECK-NEXT: add.w r0, r0, #20 ; CHECK-NEXT: vstrb.8 q3, [r3], #16 ; CHECK-NEXT: mov r1, r2 ; CHECK-NEXT: bne .LBB17_2 ; CHECK-NEXT: b .LBB17_7 ; CHECK-NEXT: .LBB17_5: @ %.preheader ; CHECK-NEXT: dls lr, r12 ; CHECK-NEXT: mov r0, sp ; CHECK-NEXT: .LBB17_6: @ =>This Inner Loop Header: Depth=1 ; CHECK-NEXT: vldrw.u32 q0, [r3], #16 ; CHECK-NEXT: vstrw.32 q0, [r0] ; CHECK-NEXT: le lr, .LBB17_6 ; CHECK-NEXT: .LBB17_7: ; CHECK-NEXT: add sp, #24 ; CHECK-NEXT: vpop {d8, d9, d10, d11} ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, pc} %5 = alloca [6 x float], align 4 %6 = getelementptr inbounds %struct.arm_biquad_cascade_stereo_df2T_instance_f32, %struct.arm_biquad_cascade_stereo_df2T_instance_f32* %0, i32 0, i32 1 %7 = load float*, float** %6, align 4 %8 = getelementptr inbounds %struct.arm_biquad_cascade_stereo_df2T_instance_f32, %struct.arm_biquad_cascade_stereo_df2T_instance_f32* %0, i32 0, i32 2 %9 = load float*, float** %8, align 4 %10 = getelementptr inbounds %struct.arm_biquad_cascade_stereo_df2T_instance_f32, %struct.arm_biquad_cascade_stereo_df2T_instance_f32* %0, i32 0, i32 0 %11 = load i8, i8* %10, align 4 %12 = zext i8 %11 to i32 %13 = bitcast [6 x float]* %5 to i8* call void @llvm.lifetime.start.p0i8(i64 24, i8* nonnull %13) #5 %14 = tail call { <4 x i32>, i32 } @llvm.arm.mve.viwdup.v4i32(i32 0, i32 2, i32 1) %15 = extractvalue { <4 x i32>, i32 } %14, 0 %16 = getelementptr inbounds [6 x float], [6 x float]* %5, i32 0, i32 4 store float 0.000000e+00, float* %16, align 4 %17 = getelementptr inbounds [6 x float], [6 x float]* %5, i32 0, i32 5 store float 0.000000e+00, float* %17, align 4 %18 = bitcast [6 x float]* %5 to <4 x float>* %19 = icmp eq i32 %3, 0 %20 = bitcast [6 x float]* %5 to i32* %21 = getelementptr inbounds [6 x float], [6 x float]* %5, i32 0, i32 2 %22 = bitcast float* %21 to <4 x float>* br i1 %19, label %23, label %31 23: ; preds = %4, %23 %24 = phi i32 [ %29, %23 ], [ %12, %4 ] %25 = phi float* [ %28, %23 ], [ %7, %4 ] %26 = bitcast float* %25 to <4 x float>* %27 = load <4 x float>, <4 x float>* %26, align 8 store <4 x float> %27, <4 x float>* %18, align 4 %28 = getelementptr inbounds float, float* %25, i32 4 %29 = add i32 %24, -1 %30 = icmp eq i32 %29, 0 br i1 %30, label %82, label %23 31: ; preds = %4, %77 %32 = phi i32 [ %80, %77 ], [ %12, %4 ] %33 = phi float* [ %78, %77 ], [ %9, %4 ] %34 = phi float* [ %79, %77 ], [ %7, %4 ] %35 = phi float* [ %2, %77 ], [ %1, %4 ] %36 = getelementptr inbounds float, float* %33, i32 1 %37 = load float, float* %33, align 4 %38 = getelementptr inbounds float, float* %33, i32 2 %39 = load float, float* %36, align 4 %40 = getelementptr inbounds float, float* %33, i32 3 %41 = load float, float* %38, align 4 %42 = getelementptr inbounds float, float* %33, i32 4 %43 = load float, float* %40, align 4 %44 = load float, float* %42, align 4 %45 = insertelement <4 x float> undef, float %43, i32 0 %46 = shufflevector <4 x float> %45, <4 x float> undef, <4 x i32> %47 = insertelement <4 x float> %46, float %44, i32 2 %48 = insertelement <4 x float> %47, float %44, i32 3 %49 = insertelement <4 x float> undef, float %39, i32 0 %50 = shufflevector <4 x float> %49, <4 x float> undef, <4 x i32> %51 = insertelement <4 x float> %50, float %41, i32 2 %52 = insertelement <4 x float> %51, float %41, i32 3 %53 = bitcast float* %34 to <4 x float>* %54 = load <4 x float>, <4 x float>* %53, align 8 store <4 x float> %54, <4 x float>* %18, align 4 %55 = insertelement <4 x float> undef, float %37, i32 0 %56 = shufflevector <4 x float> %55, <4 x float> undef, <4 x i32> zeroinitializer br label %57 57: ; preds = %31, %57 %58 = phi float* [ %35, %31 ], [ %74, %57 ] %59 = phi float* [ %2, %31 ], [ %70, %57 ] %60 = phi i32 [ %3, %31 ], [ %75, %57 ] %61 = call <4 x i32> @llvm.arm.mve.vldr.gather.offset.v4i32.p0i32.v4i32(i32* nonnull %20, <4 x i32> %15, i32 32, i32 2, i32 1) %62 = bitcast <4 x i32> %61 to <4 x float> %63 = bitcast float* %58 to i32* %64 = call <4 x i32> @llvm.arm.mve.vldr.gather.offset.v4i32.p0i32.v4i32(i32* %63, <4 x i32> %15, i32 32, i32 2, i32 1) %65 = bitcast <4 x i32> %64 to <4 x float> %66 = call fast <4 x float> @llvm.fma.v4f32(<4 x float> %65, <4 x float> %56, <4 x float> %62) %67 = extractelement <4 x float> %66, i32 0 %68 = getelementptr inbounds float, float* %59, i32 1 store float %67, float* %59, align 4 %69 = extractelement <4 x float> %66, i32 1 %70 = getelementptr inbounds float, float* %59, i32 2 store float %69, float* %68, align 4 %71 = load <4 x float>, <4 x float>* %22, align 4 %72 = call fast <4 x float> @llvm.fma.v4f32(<4 x float> %66, <4 x float> %48, <4 x float> %71) %73 = call fast <4 x float> @llvm.fma.v4f32(<4 x float> %65, <4 x float> %52, <4 x float> %72) store <4 x float> %73, <4 x float>* %18, align 4 %74 = getelementptr inbounds float, float* %58, i32 2 %75 = add i32 %60, -1 %76 = icmp eq i32 %75, 0 br i1 %76, label %77, label %57 77: ; preds = %57 %78 = getelementptr inbounds float, float* %33, i32 5 store <4 x float> %73, <4 x float>* %53, align 4 %79 = getelementptr inbounds float, float* %34, i32 4 %80 = add i32 %32, -1 %81 = icmp eq i32 %80, 0 br i1 %81, label %82, label %31 82: ; preds = %77, %23 call void @llvm.lifetime.end.p0i8(i64 24, i8* nonnull %13) #5 ret void } define arm_aapcs_vfpcc void @fms(float* nocapture readonly %pSrc1, float* nocapture readonly %pSrc2, float* nocapture readonly %pSrc3, float* nocapture %pDst, i32 %N, i32 %M) { ; CHECK-LABEL: fms: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: .save {r4, r5, r7, lr} ; CHECK-NEXT: push {r4, r5, r7, lr} ; CHECK-NEXT: ldr.w lr, [sp, #16] ; CHECK-NEXT: movs r4, #0 ; CHECK-NEXT: cmp.w r4, lr, lsr #2 ; CHECK-NEXT: beq .LBB18_5 ; CHECK-NEXT: @ %bb.1: @ %do.body.preheader ; CHECK-NEXT: ldr.w r12, [sp, #20] ; CHECK-NEXT: lsr.w r5, lr, #2 ; CHECK-NEXT: .LBB18_2: @ %do.body ; CHECK-NEXT: @ =>This Loop Header: Depth=1 ; CHECK-NEXT: @ Child Loop BB18_3 Depth 2 ; CHECK-NEXT: ldr r4, [r2] ; CHECK-NEXT: dls lr, r5 ; CHECK-NEXT: vdup.32 q0, r4 ; CHECK-NEXT: .LBB18_3: @ %while.body ; CHECK-NEXT: @ Parent Loop BB18_2 Depth=1 ; CHECK-NEXT: @ => This Inner Loop Header: Depth=2 ; CHECK-NEXT: vldrw.u32 q1, [r1], #16 ; CHECK-NEXT: vldrw.u32 q2, [r0], #16 ; CHECK-NEXT: vfms.f32 q2, q1, q0 ; CHECK-NEXT: vstrb.8 q2, [r3], #16 ; CHECK-NEXT: le lr, .LBB18_3 ; CHECK-NEXT: @ %bb.4: @ %while.end ; CHECK-NEXT: @ in Loop: Header=BB18_2 Depth=1 ; CHECK-NEXT: subs.w r12, r12, #1 ; CHECK-NEXT: add.w r2, r2, #4 ; CHECK-NEXT: bne .LBB18_2 ; CHECK-NEXT: .LBB18_5: @ %do.end ; CHECK-NEXT: pop {r4, r5, r7, pc} entry: %shr = lshr i32 %N, 2 %cmp15 = icmp eq i32 %shr, 0 br i1 %cmp15, label %do.end, label %do.body do.body: ; preds = %entry, %while.end %pDst.addr.0 = phi float* [ %add.ptr2, %while.end ], [ %pDst, %entry ] %M.addr.0 = phi i32 [ %dec3, %while.end ], [ %M, %entry ] %pSrc3.addr.0 = phi float* [ %incdec.ptr, %while.end ], [ %pSrc3, %entry ] %pSrc2.addr.0 = phi float* [ %add.ptr1, %while.end ], [ %pSrc2, %entry ] %pSrc1.addr.0 = phi float* [ %add.ptr, %while.end ], [ %pSrc1, %entry ] %0 = load float, float* %pSrc3.addr.0, align 4 %.splatinsert = insertelement <4 x float> undef, float %0, i32 0 %.splat = shufflevector <4 x float> %.splatinsert, <4 x float> undef, <4 x i32> zeroinitializer br label %while.body while.body: ; preds = %do.body, %while.body %pSrc1.addr.119 = phi float* [ %pSrc1.addr.0, %do.body ], [ %add.ptr, %while.body ] %pSrc2.addr.118 = phi float* [ %pSrc2.addr.0, %do.body ], [ %add.ptr1, %while.body ] %blkCnt.017 = phi i32 [ %shr, %do.body ], [ %dec, %while.body ] %pDst.addr.116 = phi float* [ %pDst.addr.0, %do.body ], [ %add.ptr2, %while.body ] %1 = bitcast float* %pSrc1.addr.119 to <4 x float>* %2 = load <4 x float>, <4 x float>* %1, align 4 %3 = bitcast float* %pSrc2.addr.118 to <4 x float>* %4 = load <4 x float>, <4 x float>* %3, align 4 %5 = fneg fast <4 x float> %4 %6 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %.splat, <4 x float> %5, <4 x float> %2) %7 = bitcast float* %pDst.addr.116 to <4 x float>* store <4 x float> %6, <4 x float>* %7, align 4 %add.ptr = getelementptr inbounds float, float* %pSrc1.addr.119, i32 4 %add.ptr1 = getelementptr inbounds float, float* %pSrc2.addr.118, i32 4 %add.ptr2 = getelementptr inbounds float, float* %pDst.addr.116, i32 4 %dec = add nsw i32 %blkCnt.017, -1 %cmp = icmp eq i32 %dec, 0 br i1 %cmp, label %while.end, label %while.body while.end: ; preds = %while.body %incdec.ptr = getelementptr inbounds float, float* %pSrc3.addr.0, i32 1 %dec3 = add i32 %M.addr.0, -1 %cmp4 = icmp eq i32 %dec3, 0 br i1 %cmp4, label %do.end, label %do.body do.end: ; preds = %while.end, %entry ret void } %struct.arm_biquad_casd_df1_inst_f32 = type { i32, float*, float* } define arm_aapcs_vfpcc void @arm_biquad_cascade_df1_f32(%struct.arm_biquad_casd_df1_inst_f32* nocapture readonly %S, float* nocapture readonly %pSrc, float* nocapture %pDst, i32 %blockSize) { ; CHECK-LABEL: arm_biquad_cascade_df1_f32: ; CHECK: @ %bb.0: @ %entry ; CHECK-NEXT: .save {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: push.w {r4, r5, r6, r7, r8, r9, r10, r11, lr} ; CHECK-NEXT: .pad #4 ; CHECK-NEXT: sub sp, #4 ; CHECK-NEXT: .vsave {d8, d9, d10, d11, d12, d13, d14, d15} ; CHECK-NEXT: vpush {d8, d9, d10, d11, d12, d13, d14, d15} ; CHECK-NEXT: .pad #88 ; CHECK-NEXT: sub sp, #88 ; CHECK-NEXT: ldrd r12, r10, [r0] ; CHECK-NEXT: @ implicit-def: $s2 ; CHECK-NEXT: and r7, r3, #3 ; CHECK-NEXT: ldr.w r11, [r0, #8] ; CHECK-NEXT: lsrs r0, r3, #2 ; CHECK-NEXT: str r0, [sp, #60] @ 4-byte Spill ; CHECK-NEXT: str r7, [sp, #12] @ 4-byte Spill ; CHECK-NEXT: str r2, [sp, #56] @ 4-byte Spill ; CHECK-NEXT: b .LBB19_3 ; CHECK-NEXT: .LBB19_1: @ in Loop: Header=BB19_3 Depth=1 ; CHECK-NEXT: vmov.f32 s14, s7 ; CHECK-NEXT: ldr r2, [sp, #56] @ 4-byte Reload ; CHECK-NEXT: vmov.f32 s0, s10 ; CHECK-NEXT: vmov.f32 s7, s6 ; CHECK-NEXT: .LBB19_2: @ %if.end69 ; CHECK-NEXT: @ in Loop: Header=BB19_3 Depth=1 ; CHECK-NEXT: vstr s8, [r10] ; CHECK-NEXT: subs.w r12, r12, #1 ; CHECK-NEXT: vstr s0, [r10, #4] ; CHECK-NEXT: add.w r11, r11, #128 ; CHECK-NEXT: vstr s14, [r10, #8] ; CHECK-NEXT: mov r1, r2 ; CHECK-NEXT: vstr s7, [r10, #12] ; CHECK-NEXT: add.w r10, r10, #16 ; CHECK-NEXT: beq.w .LBB19_13 ; CHECK-NEXT: .LBB19_3: @ %do.body ; CHECK-NEXT: @ =>This Loop Header: Depth=1 ; CHECK-NEXT: @ Child Loop BB19_5 Depth 2 ; CHECK-NEXT: vldr s7, [r10, #8] ; CHECK-NEXT: mov r5, r2 ; CHECK-NEXT: ldr r0, [sp, #60] @ 4-byte Reload ; CHECK-NEXT: vldr s8, [r10] ; CHECK-NEXT: vldr s10, [r10, #4] ; CHECK-NEXT: vldr s6, [r10, #12] ; CHECK-NEXT: wls lr, r0, .LBB19_6 ; CHECK-NEXT: @ %bb.4: @ %while.body.lr.ph ; CHECK-NEXT: @ in Loop: Header=BB19_3 Depth=1 ; CHECK-NEXT: ldrd r5, lr, [sp, #56] @ 8-byte Folded Reload ; CHECK-NEXT: .LBB19_5: @ %while.body ; CHECK-NEXT: @ Parent Loop BB19_3 Depth=1 ; CHECK-NEXT: @ => This Inner Loop Header: Depth=2 ; CHECK-NEXT: vmov r4, s8 ; CHECK-NEXT: vldr s8, [r1, #12] ; CHECK-NEXT: vldrw.u32 q0, [r11, #112] ; CHECK-NEXT: vmov r0, s10 ; CHECK-NEXT: vldr s10, [r1, #8] ; CHECK-NEXT: vmov r7, s7 ; CHECK-NEXT: vmov r9, s6 ; CHECK-NEXT: vldrw.u32 q1, [r11] ; CHECK-NEXT: vstrw.32 q0, [sp, #64] @ 16-byte Spill ; CHECK-NEXT: vmov r8, s8 ; CHECK-NEXT: vldrw.u32 q0, [r11, #16] ; CHECK-NEXT: ldr r6, [r1, #4] ; CHECK-NEXT: vldrw.u32 q7, [r11, #32] ; CHECK-NEXT: vmul.f32 q1, q1, r8 ; CHECK-NEXT: vmov r3, s10 ; CHECK-NEXT: vldrw.u32 q3, [r11, #48] ; CHECK-NEXT: vfma.f32 q1, q0, r3 ; CHECK-NEXT: ldr r3, [r1] ; CHECK-NEXT: vfma.f32 q1, q7, r6 ; CHECK-NEXT: vldrw.u32 q6, [r11, #64] ; CHECK-NEXT: vfma.f32 q1, q3, r3 ; CHECK-NEXT: vldrw.u32 q5, [r11, #80] ; CHECK-NEXT: vfma.f32 q1, q6, r4 ; CHECK-NEXT: vldrw.u32 q4, [r11, #96] ; CHECK-NEXT: vfma.f32 q1, q5, r0 ; CHECK-NEXT: vldrw.u32 q0, [sp, #64] @ 16-byte Reload ; CHECK-NEXT: vfma.f32 q1, q4, r7 ; CHECK-NEXT: adds r1, #16 ; CHECK-NEXT: vfma.f32 q1, q0, r9 ; CHECK-NEXT: vmov.f32 s2, s8 ; CHECK-NEXT: vstrb.8 q1, [r5], #16 ; CHECK-NEXT: le lr, .LBB19_5 ; CHECK-NEXT: .LBB19_6: @ %while.end ; CHECK-NEXT: @ in Loop: Header=BB19_3 Depth=1 ; CHECK-NEXT: ldr r7, [sp, #12] @ 4-byte Reload ; CHECK-NEXT: cmp r7, #0 ; CHECK-NEXT: beq .LBB19_1 ; CHECK-NEXT: @ %bb.7: @ %if.then ; CHECK-NEXT: @ in Loop: Header=BB19_3 Depth=1 ; CHECK-NEXT: vldr s24, [r1] ; CHECK-NEXT: vmov r0, s8 ; CHECK-NEXT: vldr s0, [r1, #4] ; CHECK-NEXT: vldrw.u32 q3, [r11] ; CHECK-NEXT: vldr s3, [r1, #12] ; CHECK-NEXT: vldrw.u32 q4, [r11, #32] ; CHECK-NEXT: vldr s1, [r1, #8] ; CHECK-NEXT: vmov r1, s10 ; CHECK-NEXT: vldrw.u32 q2, [r11, #96] ; CHECK-NEXT: vmov r6, s3 ; CHECK-NEXT: vmul.f32 q3, q3, r6 ; CHECK-NEXT: vmov r6, s1 ; CHECK-NEXT: vstrw.32 q2, [sp, #32] @ 16-byte Spill ; CHECK-NEXT: vldrw.u32 q2, [r11, #112] ; CHECK-NEXT: vldrw.u32 q5, [r11, #48] ; CHECK-NEXT: vmov r4, s0 ; CHECK-NEXT: vstrw.32 q2, [sp, #64] @ 16-byte Spill ; CHECK-NEXT: vldrw.u32 q2, [r11, #80] ; CHECK-NEXT: vldrw.u32 q7, [r11, #64] ; CHECK-NEXT: vmov r3, s24 ; CHECK-NEXT: vstrw.32 q2, [sp, #16] @ 16-byte Spill ; CHECK-NEXT: vldrw.u32 q2, [r11, #16] ; CHECK-NEXT: vmov r2, s7 ; CHECK-NEXT: cmp r7, #1 ; CHECK-NEXT: vfma.f32 q3, q2, r6 ; CHECK-NEXT: vldrw.u32 q2, [sp, #16] @ 16-byte Reload ; CHECK-NEXT: vfma.f32 q3, q4, r4 ; CHECK-NEXT: vmov lr, s6 ; CHECK-NEXT: vfma.f32 q3, q5, r3 ; CHECK-NEXT: vfma.f32 q3, q7, r0 ; CHECK-NEXT: vfma.f32 q3, q2, r1 ; CHECK-NEXT: vldrw.u32 q2, [sp, #32] @ 16-byte Reload ; CHECK-NEXT: vfma.f32 q3, q2, r2 ; CHECK-NEXT: vldrw.u32 q2, [sp, #64] @ 16-byte Reload ; CHECK-NEXT: vfma.f32 q3, q2, lr ; CHECK-NEXT: bne .LBB19_9 ; CHECK-NEXT: @ %bb.8: @ %if.then58 ; CHECK-NEXT: @ in Loop: Header=BB19_3 Depth=1 ; CHECK-NEXT: vstr s12, [r5] ; CHECK-NEXT: vmov.f32 s8, s24 ; CHECK-NEXT: vmov.f32 s0, s2 ; CHECK-NEXT: vmov.f32 s14, s12 ; CHECK-NEXT: b .LBB19_11 ; CHECK-NEXT: .LBB19_9: @ %if.else ; CHECK-NEXT: @ in Loop: Header=BB19_3 Depth=1 ; CHECK-NEXT: cmp r7, #2 ; CHECK-NEXT: vstmia r5, {s12, s13} ; CHECK-NEXT: bne .LBB19_12 ; CHECK-NEXT: @ %bb.10: @ in Loop: Header=BB19_3 Depth=1 ; CHECK-NEXT: vmov.f32 s8, s0 ; CHECK-NEXT: vmov.f32 s14, s13 ; CHECK-NEXT: vmov.f32 s0, s24 ; CHECK-NEXT: vmov.f32 s7, s12 ; CHECK-NEXT: .LBB19_11: @ %if.end69 ; CHECK-NEXT: @ in Loop: Header=BB19_3 Depth=1 ; CHECK-NEXT: vmov.f32 s2, s3 ; CHECK-NEXT: ldr r2, [sp, #56] @ 4-byte Reload ; CHECK-NEXT: b .LBB19_2 ; CHECK-NEXT: .LBB19_12: @ %if.else64 ; CHECK-NEXT: @ in Loop: Header=BB19_3 Depth=1 ; CHECK-NEXT: vmov.f32 s7, s13 ; CHECK-NEXT: ldr r2, [sp, #56] @ 4-byte Reload ; CHECK-NEXT: vmov.f32 s2, s3 ; CHECK-NEXT: vstr s14, [r5, #8] ; CHECK-NEXT: vmov.f32 s8, s1 ; CHECK-NEXT: b .LBB19_2 ; CHECK-NEXT: .LBB19_13: @ %do.end ; CHECK-NEXT: add sp, #88 ; CHECK-NEXT: vpop {d8, d9, d10, d11, d12, d13, d14, d15} ; CHECK-NEXT: add sp, #4 ; CHECK-NEXT: pop.w {r4, r5, r6, r7, r8, r9, r10, r11, pc} entry: %pState1 = getelementptr inbounds %struct.arm_biquad_casd_df1_inst_f32, %struct.arm_biquad_casd_df1_inst_f32* %S, i32 0, i32 1 %0 = load float*, float** %pState1, align 4 %pCoeffs2 = getelementptr inbounds %struct.arm_biquad_casd_df1_inst_f32, %struct.arm_biquad_casd_df1_inst_f32* %S, i32 0, i32 2 %1 = load float*, float** %pCoeffs2, align 4 %numStages = getelementptr inbounds %struct.arm_biquad_casd_df1_inst_f32, %struct.arm_biquad_casd_df1_inst_f32* %S, i32 0, i32 0 %2 = load i32, i32* %numStages, align 4 %shr = lshr i32 %blockSize, 2 %cmp201 = icmp eq i32 %shr, 0 %and = and i32 %blockSize, 3 %tobool = icmp eq i32 %and, 0 %cmp57 = icmp eq i32 %and, 1 %cmp60 = icmp eq i32 %and, 2 br label %do.body do.body: ; preds = %if.end69, %entry %pState.0 = phi float* [ %0, %entry ], [ %incdec.ptr73, %if.end69 ] %pCoeffs.0 = phi float* [ %1, %entry ], [ %add.ptr74, %if.end69 ] %pIn.0 = phi float* [ %pSrc, %entry ], [ %pDst, %if.end69 ] %X3.0 = phi float [ undef, %entry ], [ %X3.2, %if.end69 ] %stage.0 = phi i32 [ %2, %entry ], [ %dec75, %if.end69 ] %3 = load float, float* %pState.0, align 4 %arrayidx3 = getelementptr inbounds float, float* %pState.0, i32 1 %4 = load float, float* %arrayidx3, align 4 %arrayidx4 = getelementptr inbounds float, float* %pState.0, i32 2 %5 = load float, float* %arrayidx4, align 4 %arrayidx5 = getelementptr inbounds float, float* %pState.0, i32 3 %6 = load float, float* %arrayidx5, align 4 br i1 %cmp201, label %while.end, label %while.body.lr.ph while.body.lr.ph: ; preds = %do.body %7 = bitcast float* %pCoeffs.0 to <4 x float>* %arrayidx9 = getelementptr inbounds float, float* %pCoeffs.0, i32 4 %8 = bitcast float* %arrayidx9 to <4 x float>* %arrayidx12 = getelementptr inbounds float, float* %pCoeffs.0, i32 8 %9 = bitcast float* %arrayidx12 to <4 x float>* %arrayidx15 = getelementptr inbounds float, float* %pCoeffs.0, i32 12 %10 = bitcast float* %arrayidx15 to <4 x float>* %arrayidx18 = getelementptr inbounds float, float* %pCoeffs.0, i32 16 %11 = bitcast float* %arrayidx18 to <4 x float>* %arrayidx21 = getelementptr inbounds float, float* %pCoeffs.0, i32 20 %12 = bitcast float* %arrayidx21 to <4 x float>* %arrayidx24 = getelementptr inbounds float, float* %pCoeffs.0, i32 24 %13 = bitcast float* %arrayidx24 to <4 x float>* %arrayidx27 = getelementptr inbounds float, float* %pCoeffs.0, i32 28 %14 = bitcast float* %arrayidx27 to <4 x float>* br label %while.body while.body: ; preds = %while.body.lr.ph, %while.body %sample.0208 = phi i32 [ %shr, %while.body.lr.ph ], [ %dec, %while.body ] %pIn.1207 = phi float* [ %pIn.0, %while.body.lr.ph ], [ %incdec.ptr8, %while.body ] %pOut.1206 = phi float* [ %pDst, %while.body.lr.ph ], [ %add.ptr, %while.body ] %Yn2.0205 = phi float [ %6, %while.body.lr.ph ], [ %37, %while.body ] %Yn1.0204 = phi float [ %5, %while.body.lr.ph ], [ %36, %while.body ] %Xn2.0203 = phi float [ %4, %while.body.lr.ph ], [ %17, %while.body ] %Xn1.0202 = phi float [ %3, %while.body.lr.ph ], [ %18, %while.body ] %incdec.ptr = getelementptr inbounds float, float* %pIn.1207, i32 1 %15 = load float, float* %pIn.1207, align 4 %incdec.ptr6 = getelementptr inbounds float, float* %pIn.1207, i32 2 %16 = load float, float* %incdec.ptr, align 4 %incdec.ptr7 = getelementptr inbounds float, float* %pIn.1207, i32 3 %17 = load float, float* %incdec.ptr6, align 4 %incdec.ptr8 = getelementptr inbounds float, float* %pIn.1207, i32 4 %18 = load float, float* %incdec.ptr7, align 4 %19 = load <4 x float>, <4 x float>* %7, align 4 %.splatinsert = insertelement <4 x float> undef, float %18, i32 0 %.splat = shufflevector <4 x float> %.splatinsert, <4 x float> undef, <4 x i32> zeroinitializer %20 = fmul fast <4 x float> %.splat, %19 %21 = load <4 x float>, <4 x float>* %8, align 4 %.splatinsert10 = insertelement <4 x float> undef, float %17, i32 0 %.splat11 = shufflevector <4 x float> %.splatinsert10, <4 x float> undef, <4 x i32> zeroinitializer %22 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %21, <4 x float> %.splat11, <4 x float> %20) %23 = load <4 x float>, <4 x float>* %9, align 4 %.splatinsert13 = insertelement <4 x float> undef, float %16, i32 0 %.splat14 = shufflevector <4 x float> %.splatinsert13, <4 x float> undef, <4 x i32> zeroinitializer %24 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %23, <4 x float> %.splat14, <4 x float> %22) %25 = load <4 x float>, <4 x float>* %10, align 4 %.splatinsert16 = insertelement <4 x float> undef, float %15, i32 0 %.splat17 = shufflevector <4 x float> %.splatinsert16, <4 x float> undef, <4 x i32> zeroinitializer %26 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %25, <4 x float> %.splat17, <4 x float> %24) %27 = load <4 x float>, <4 x float>* %11, align 4 %.splatinsert19 = insertelement <4 x float> undef, float %Xn1.0202, i32 0 %.splat20 = shufflevector <4 x float> %.splatinsert19, <4 x float> undef, <4 x i32> zeroinitializer %28 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %27, <4 x float> %.splat20, <4 x float> %26) %29 = load <4 x float>, <4 x float>* %12, align 4 %.splatinsert22 = insertelement <4 x float> undef, float %Xn2.0203, i32 0 %.splat23 = shufflevector <4 x float> %.splatinsert22, <4 x float> undef, <4 x i32> zeroinitializer %30 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %29, <4 x float> %.splat23, <4 x float> %28) %31 = load <4 x float>, <4 x float>* %13, align 4 %.splatinsert25 = insertelement <4 x float> undef, float %Yn1.0204, i32 0 %.splat26 = shufflevector <4 x float> %.splatinsert25, <4 x float> undef, <4 x i32> zeroinitializer %32 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %31, <4 x float> %.splat26, <4 x float> %30) %33 = load <4 x float>, <4 x float>* %14, align 4 %.splatinsert28 = insertelement <4 x float> undef, float %Yn2.0205, i32 0 %.splat29 = shufflevector <4 x float> %.splatinsert28, <4 x float> undef, <4 x i32> zeroinitializer %34 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %33, <4 x float> %.splat29, <4 x float> %32) %35 = bitcast float* %pOut.1206 to <4 x float>* store <4 x float> %34, <4 x float>* %35, align 4 %add.ptr = getelementptr inbounds float, float* %pOut.1206, i32 4 %36 = extractelement <4 x float> %34, i32 3 %37 = extractelement <4 x float> %34, i32 2 %dec = add nsw i32 %sample.0208, -1 %cmp = icmp eq i32 %dec, 0 br i1 %cmp, label %while.end, label %while.body while.end: ; preds = %while.body, %do.body %Xn1.0.lcssa = phi float [ %3, %do.body ], [ %18, %while.body ] %Xn2.0.lcssa = phi float [ %4, %do.body ], [ %17, %while.body ] %Yn1.0.lcssa = phi float [ %5, %do.body ], [ %36, %while.body ] %Yn2.0.lcssa = phi float [ %6, %do.body ], [ %37, %while.body ] %pOut.1.lcssa = phi float* [ %pDst, %do.body ], [ %add.ptr, %while.body ] %pIn.1.lcssa = phi float* [ %pIn.0, %do.body ], [ %incdec.ptr8, %while.body ] %X3.1.lcssa = phi float [ %X3.0, %do.body ], [ %18, %while.body ] br i1 %tobool, label %if.end69, label %if.then if.then: ; preds = %while.end %incdec.ptr30 = getelementptr inbounds float, float* %pIn.1.lcssa, i32 1 %38 = load float, float* %pIn.1.lcssa, align 4 %incdec.ptr31 = getelementptr inbounds float, float* %pIn.1.lcssa, i32 2 %39 = load float, float* %incdec.ptr30, align 4 %incdec.ptr32 = getelementptr inbounds float, float* %pIn.1.lcssa, i32 3 %40 = load float, float* %incdec.ptr31, align 4 %41 = load float, float* %incdec.ptr32, align 4 %42 = bitcast float* %pCoeffs.0 to <4 x float>* %43 = load <4 x float>, <4 x float>* %42, align 4 %.splatinsert34 = insertelement <4 x float> undef, float %41, i32 0 %.splat35 = shufflevector <4 x float> %.splatinsert34, <4 x float> undef, <4 x i32> zeroinitializer %44 = fmul fast <4 x float> %.splat35, %43 %arrayidx36 = getelementptr inbounds float, float* %pCoeffs.0, i32 4 %45 = bitcast float* %arrayidx36 to <4 x float>* %46 = load <4 x float>, <4 x float>* %45, align 4 %.splatinsert37 = insertelement <4 x float> undef, float %40, i32 0 %.splat38 = shufflevector <4 x float> %.splatinsert37, <4 x float> undef, <4 x i32> zeroinitializer %47 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %46, <4 x float> %.splat38, <4 x float> %44) %arrayidx39 = getelementptr inbounds float, float* %pCoeffs.0, i32 8 %48 = bitcast float* %arrayidx39 to <4 x float>* %49 = load <4 x float>, <4 x float>* %48, align 4 %.splatinsert40 = insertelement <4 x float> undef, float %39, i32 0 %.splat41 = shufflevector <4 x float> %.splatinsert40, <4 x float> undef, <4 x i32> zeroinitializer %50 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %49, <4 x float> %.splat41, <4 x float> %47) %arrayidx42 = getelementptr inbounds float, float* %pCoeffs.0, i32 12 %51 = bitcast float* %arrayidx42 to <4 x float>* %52 = load <4 x float>, <4 x float>* %51, align 4 %.splatinsert43 = insertelement <4 x float> undef, float %38, i32 0 %.splat44 = shufflevector <4 x float> %.splatinsert43, <4 x float> undef, <4 x i32> zeroinitializer %53 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %52, <4 x float> %.splat44, <4 x float> %50) %arrayidx45 = getelementptr inbounds float, float* %pCoeffs.0, i32 16 %54 = bitcast float* %arrayidx45 to <4 x float>* %55 = load <4 x float>, <4 x float>* %54, align 4 %.splatinsert46 = insertelement <4 x float> undef, float %Xn1.0.lcssa, i32 0 %.splat47 = shufflevector <4 x float> %.splatinsert46, <4 x float> undef, <4 x i32> zeroinitializer %56 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %55, <4 x float> %.splat47, <4 x float> %53) %arrayidx48 = getelementptr inbounds float, float* %pCoeffs.0, i32 20 %57 = bitcast float* %arrayidx48 to <4 x float>* %58 = load <4 x float>, <4 x float>* %57, align 4 %.splatinsert49 = insertelement <4 x float> undef, float %Xn2.0.lcssa, i32 0 %.splat50 = shufflevector <4 x float> %.splatinsert49, <4 x float> undef, <4 x i32> zeroinitializer %59 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %58, <4 x float> %.splat50, <4 x float> %56) %arrayidx51 = getelementptr inbounds float, float* %pCoeffs.0, i32 24 %60 = bitcast float* %arrayidx51 to <4 x float>* %61 = load <4 x float>, <4 x float>* %60, align 4 %.splatinsert52 = insertelement <4 x float> undef, float %Yn1.0.lcssa, i32 0 %.splat53 = shufflevector <4 x float> %.splatinsert52, <4 x float> undef, <4 x i32> zeroinitializer %62 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %61, <4 x float> %.splat53, <4 x float> %59) %arrayidx54 = getelementptr inbounds float, float* %pCoeffs.0, i32 28 %63 = bitcast float* %arrayidx54 to <4 x float>* %64 = load <4 x float>, <4 x float>* %63, align 4 %.splatinsert55 = insertelement <4 x float> undef, float %Yn2.0.lcssa, i32 0 %.splat56 = shufflevector <4 x float> %.splatinsert55, <4 x float> undef, <4 x i32> zeroinitializer %65 = tail call fast <4 x float> @llvm.fma.v4f32(<4 x float> %64, <4 x float> %.splat56, <4 x float> %62) %66 = extractelement <4 x float> %65, i32 0 br i1 %cmp57, label %if.then58, label %if.else if.then58: ; preds = %if.then store float %66, float* %pOut.1.lcssa, align 4 br label %if.end69 if.else: ; preds = %if.then %incdec.ptr62 = getelementptr inbounds float, float* %pOut.1.lcssa, i32 1 store float %66, float* %pOut.1.lcssa, align 4 %67 = extractelement <4 x float> %65, i32 1 store float %67, float* %incdec.ptr62, align 4 br i1 %cmp60, label %if.end69, label %if.else64 if.else64: ; preds = %if.else %incdec.ptr63 = getelementptr inbounds float, float* %pOut.1.lcssa, i32 2 %68 = extractelement <4 x float> %65, i32 2 store float %68, float* %incdec.ptr63, align 4 br label %if.end69 if.end69: ; preds = %if.else, %while.end, %if.then58, %if.else64 %Xn1.1 = phi float [ %38, %if.then58 ], [ %40, %if.else64 ], [ %Xn1.0.lcssa, %while.end ], [ %39, %if.else ] %Xn2.1 = phi float [ %X3.1.lcssa, %if.then58 ], [ %39, %if.else64 ], [ %Xn2.0.lcssa, %while.end ], [ %38, %if.else ] %Yn1.1 = phi float [ %66, %if.then58 ], [ %68, %if.else64 ], [ %Yn1.0.lcssa, %while.end ], [ %67, %if.else ] %Yn2.1 = phi float [ %Yn1.0.lcssa, %if.then58 ], [ %67, %if.else64 ], [ %Yn2.0.lcssa, %while.end ], [ %66, %if.else ] %X3.2 = phi float [ %41, %if.then58 ], [ %41, %if.else64 ], [ %X3.1.lcssa, %while.end ], [ %41, %if.else ] store float %Xn1.1, float* %pState.0, align 4 store float %Xn2.1, float* %arrayidx3, align 4 store float %Yn1.1, float* %arrayidx4, align 4 %incdec.ptr73 = getelementptr inbounds float, float* %pState.0, i32 4 store float %Yn2.1, float* %arrayidx5, align 4 %add.ptr74 = getelementptr inbounds float, float* %pCoeffs.0, i32 32 %dec75 = add i32 %stage.0, -1 %cmp76 = icmp eq i32 %dec75, 0 br i1 %cmp76, label %do.end, label %do.body do.end: ; preds = %if.end69 ret void } declare void @llvm.lifetime.end.p0i8(i64 immarg, i8* nocapture) declare void @llvm.lifetime.start.p0i8(i64 immarg, i8* nocapture) declare { <4 x i32>, i32 } @llvm.arm.mve.viwdup.v4i32(i32, i32, i32) declare <4 x i32> @llvm.arm.mve.vldr.gather.offset.v4i32.p0i32.v4i32(i32*, <4 x i32>, i32, i32, i32) declare void @llvm.assume(i1) declare <4 x i1> @llvm.arm.mve.vctp32(i32) declare <4 x float> @llvm.fma.v4f32(<4 x float>, <4 x float>, <4 x float>) declare void @llvm.masked.store.v4f32.p0v4f32(<4 x float>, <4 x float>*, i32 immarg, <4 x i1>)