define void @gather(ptr noalias nocapture %A, ptr noalias nocapture readonly %B) {
; CHECK-LABEL: gather:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: li a2, 0
+; CHECK-NEXT: li a2, 1024
; CHECK-NEXT: li a3, 32
; CHECK-NEXT: li a4, 5
-; CHECK-NEXT: li a5, 1024
; CHECK-NEXT: .LBB0_1: # %vector.body
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
; CHECK-NEXT: vsetvli zero, a3, e8, m1, ta, ma
; CHECK-NEXT: vlse8.v v8, (a1), a4
-; CHECK-NEXT: add a6, a0, a2
-; CHECK-NEXT: vle8.v v9, (a6)
+; CHECK-NEXT: vle8.v v9, (a0)
; CHECK-NEXT: vadd.vv v8, v9, v8
-; CHECK-NEXT: vse8.v v8, (a6)
-; CHECK-NEXT: addi a2, a2, 32
+; CHECK-NEXT: vse8.v v8, (a0)
+; CHECK-NEXT: addi a2, a2, -32
+; CHECK-NEXT: addi a0, a0, 32
; CHECK-NEXT: addi a1, a1, 160
-; CHECK-NEXT: bne a2, a5, .LBB0_1
+; CHECK-NEXT: bnez a2, .LBB0_1
; CHECK-NEXT: # %bb.2: # %for.cond.cleanup
; CHECK-NEXT: ret
entry:
define void @gather_masked(ptr noalias nocapture %A, ptr noalias nocapture readonly %B, <32 x i8> %maskedoff) {
; V-LABEL: gather_masked:
; V: # %bb.0: # %entry
-; V-NEXT: li a2, 0
+; V-NEXT: li a2, 1024
; V-NEXT: lui a3, 983765
; V-NEXT: addiw a3, a3, 873
; V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; V-NEXT: vmv.s.x v0, a3
; V-NEXT: li a3, 32
; V-NEXT: li a4, 5
-; V-NEXT: li a5, 1024
; V-NEXT: .LBB1_1: # %vector.body
; V-NEXT: # =>This Inner Loop Header: Depth=1
; V-NEXT: vsetvli zero, a3, e8, m1, ta, mu
; V-NEXT: vmv1r.v v9, v8
; V-NEXT: vlse8.v v9, (a1), a4, v0.t
-; V-NEXT: add a6, a0, a2
-; V-NEXT: vle8.v v10, (a6)
+; V-NEXT: vle8.v v10, (a0)
; V-NEXT: vadd.vv v9, v10, v9
-; V-NEXT: vse8.v v9, (a6)
-; V-NEXT: addi a2, a2, 32
+; V-NEXT: vse8.v v9, (a0)
+; V-NEXT: addi a2, a2, -32
+; V-NEXT: addi a0, a0, 32
; V-NEXT: addi a1, a1, 160
-; V-NEXT: bne a2, a5, .LBB1_1
+; V-NEXT: bnez a2, .LBB1_1
; V-NEXT: # %bb.2: # %for.cond.cleanup
; V-NEXT: ret
;
; ZVE32F-LABEL: gather_masked:
; ZVE32F: # %bb.0: # %entry
-; ZVE32F-NEXT: li a2, 0
+; ZVE32F-NEXT: li a2, 1024
; ZVE32F-NEXT: lui a3, 983765
; ZVE32F-NEXT: addiw a3, a3, 873
; ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; ZVE32F-NEXT: vmv.s.x v0, a3
; ZVE32F-NEXT: li a3, 32
; ZVE32F-NEXT: li a4, 5
-; ZVE32F-NEXT: li a5, 1024
; ZVE32F-NEXT: .LBB1_1: # %vector.body
; ZVE32F-NEXT: # =>This Inner Loop Header: Depth=1
; ZVE32F-NEXT: vsetvli zero, a3, e8, m1, ta, mu
; ZVE32F-NEXT: vmv1r.v v9, v8
; ZVE32F-NEXT: vlse8.v v9, (a1), a4, v0.t
-; ZVE32F-NEXT: add a6, a0, a2
-; ZVE32F-NEXT: vle8.v v10, (a6)
+; ZVE32F-NEXT: vle8.v v10, (a0)
; ZVE32F-NEXT: vadd.vv v9, v10, v9
-; ZVE32F-NEXT: vse8.v v9, (a6)
-; ZVE32F-NEXT: addi a2, a2, 32
+; ZVE32F-NEXT: vse8.v v9, (a0)
+; ZVE32F-NEXT: addi a2, a2, -32
+; ZVE32F-NEXT: addi a0, a0, 32
; ZVE32F-NEXT: addi a1, a1, 160
-; ZVE32F-NEXT: bne a2, a5, .LBB1_1
+; ZVE32F-NEXT: bnez a2, .LBB1_1
; ZVE32F-NEXT: # %bb.2: # %for.cond.cleanup
; ZVE32F-NEXT: ret
entry:
define void @gather_negative_stride(ptr noalias nocapture %A, ptr noalias nocapture readonly %B) {
; CHECK-LABEL: gather_negative_stride:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: li a2, 0
; CHECK-NEXT: addi a1, a1, 155
+; CHECK-NEXT: li a2, 1024
; CHECK-NEXT: li a3, 32
; CHECK-NEXT: li a4, -5
-; CHECK-NEXT: li a5, 1024
; CHECK-NEXT: .LBB2_1: # %vector.body
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
; CHECK-NEXT: vsetvli zero, a3, e8, m1, ta, ma
; CHECK-NEXT: vlse8.v v8, (a1), a4
-; CHECK-NEXT: add a6, a0, a2
-; CHECK-NEXT: vle8.v v9, (a6)
+; CHECK-NEXT: vle8.v v9, (a0)
; CHECK-NEXT: vadd.vv v8, v9, v8
-; CHECK-NEXT: vse8.v v8, (a6)
-; CHECK-NEXT: addi a2, a2, 32
+; CHECK-NEXT: vse8.v v8, (a0)
+; CHECK-NEXT: addi a2, a2, -32
+; CHECK-NEXT: addi a0, a0, 32
; CHECK-NEXT: addi a1, a1, 160
-; CHECK-NEXT: bne a2, a5, .LBB2_1
+; CHECK-NEXT: bnez a2, .LBB2_1
; CHECK-NEXT: # %bb.2: # %for.cond.cleanup
; CHECK-NEXT: ret
entry:
define void @gather_zero_stride(ptr noalias nocapture %A, ptr noalias nocapture readonly %B) {
; CHECK-LABEL: gather_zero_stride:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: li a2, 0
+; CHECK-NEXT: li a2, 1024
; CHECK-NEXT: li a3, 32
-; CHECK-NEXT: li a4, 1024
; CHECK-NEXT: .LBB3_1: # %vector.body
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT: lbu a5, 0(a1)
-; CHECK-NEXT: add a6, a0, a2
+; CHECK-NEXT: lbu a4, 0(a1)
; CHECK-NEXT: vsetvli zero, a3, e8, m1, ta, ma
-; CHECK-NEXT: vle8.v v8, (a6)
-; CHECK-NEXT: vadd.vx v8, v8, a5
-; CHECK-NEXT: vse8.v v8, (a6)
-; CHECK-NEXT: addi a2, a2, 32
+; CHECK-NEXT: vle8.v v8, (a0)
+; CHECK-NEXT: vadd.vx v8, v8, a4
+; CHECK-NEXT: vse8.v v8, (a0)
+; CHECK-NEXT: addi a2, a2, -32
+; CHECK-NEXT: addi a0, a0, 32
; CHECK-NEXT: addi a1, a1, 160
-; CHECK-NEXT: bne a2, a4, .LBB3_1
+; CHECK-NEXT: bnez a2, .LBB3_1
; CHECK-NEXT: # %bb.2: # %for.cond.cleanup
; CHECK-NEXT: ret
entry:
define void @gather_zero_stride_unfold(ptr noalias nocapture %A, ptr noalias nocapture readonly %B) {
; V-LABEL: gather_zero_stride_unfold:
; V: # %bb.0: # %entry
-; V-NEXT: li a2, 0
+; V-NEXT: li a2, 1024
; V-NEXT: li a3, 32
-; V-NEXT: li a4, 1024
; V-NEXT: .LBB4_1: # %vector.body
; V-NEXT: # =>This Inner Loop Header: Depth=1
; V-NEXT: vsetvli zero, a3, e8, m1, ta, ma
; V-NEXT: vlse8.v v8, (a1), zero
-; V-NEXT: add a5, a0, a2
-; V-NEXT: vle8.v v9, (a5)
+; V-NEXT: vle8.v v9, (a0)
; V-NEXT: vdivu.vv v8, v8, v9
-; V-NEXT: vse8.v v8, (a5)
-; V-NEXT: addi a2, a2, 32
+; V-NEXT: vse8.v v8, (a0)
+; V-NEXT: addi a2, a2, -32
+; V-NEXT: addi a0, a0, 32
; V-NEXT: addi a1, a1, 160
-; V-NEXT: bne a2, a4, .LBB4_1
+; V-NEXT: bnez a2, .LBB4_1
; V-NEXT: # %bb.2: # %for.cond.cleanup
; V-NEXT: ret
;
; ZVE32F-LABEL: gather_zero_stride_unfold:
; ZVE32F: # %bb.0: # %entry
-; ZVE32F-NEXT: li a2, 0
+; ZVE32F-NEXT: li a2, 1024
; ZVE32F-NEXT: li a3, 32
-; ZVE32F-NEXT: li a4, 1024
; ZVE32F-NEXT: .LBB4_1: # %vector.body
; ZVE32F-NEXT: # =>This Inner Loop Header: Depth=1
; ZVE32F-NEXT: vsetvli zero, a3, e8, m1, ta, ma
; ZVE32F-NEXT: vlse8.v v8, (a1), zero
-; ZVE32F-NEXT: add a5, a0, a2
-; ZVE32F-NEXT: vle8.v v9, (a5)
+; ZVE32F-NEXT: vle8.v v9, (a0)
; ZVE32F-NEXT: vdivu.vv v8, v8, v9
-; ZVE32F-NEXT: vse8.v v8, (a5)
-; ZVE32F-NEXT: addi a2, a2, 32
+; ZVE32F-NEXT: vse8.v v8, (a0)
+; ZVE32F-NEXT: addi a2, a2, -32
+; ZVE32F-NEXT: addi a0, a0, 32
; ZVE32F-NEXT: addi a1, a1, 160
-; ZVE32F-NEXT: bne a2, a4, .LBB4_1
+; ZVE32F-NEXT: bnez a2, .LBB4_1
; ZVE32F-NEXT: # %bb.2: # %for.cond.cleanup
; ZVE32F-NEXT: ret
;
; NOT-OPTIMIZED-LABEL: gather_zero_stride_unfold:
; NOT-OPTIMIZED: # %bb.0: # %entry
-; NOT-OPTIMIZED-NEXT: li a2, 0
+; NOT-OPTIMIZED-NEXT: li a2, 1024
; NOT-OPTIMIZED-NEXT: li a3, 32
-; NOT-OPTIMIZED-NEXT: li a4, 1024
; NOT-OPTIMIZED-NEXT: .LBB4_1: # %vector.body
; NOT-OPTIMIZED-NEXT: # =>This Inner Loop Header: Depth=1
-; NOT-OPTIMIZED-NEXT: lbu a5, 0(a1)
+; NOT-OPTIMIZED-NEXT: lbu a4, 0(a1)
; NOT-OPTIMIZED-NEXT: vsetvli zero, a3, e8, m1, ta, ma
-; NOT-OPTIMIZED-NEXT: add a6, a0, a2
-; NOT-OPTIMIZED-NEXT: vle8.v v8, (a6)
-; NOT-OPTIMIZED-NEXT: vmv.v.x v9, a5
+; NOT-OPTIMIZED-NEXT: vle8.v v8, (a0)
+; NOT-OPTIMIZED-NEXT: vmv.v.x v9, a4
; NOT-OPTIMIZED-NEXT: vdivu.vv v8, v9, v8
-; NOT-OPTIMIZED-NEXT: vse8.v v8, (a6)
-; NOT-OPTIMIZED-NEXT: addi a2, a2, 32
+; NOT-OPTIMIZED-NEXT: vse8.v v8, (a0)
+; NOT-OPTIMIZED-NEXT: addi a2, a2, -32
+; NOT-OPTIMIZED-NEXT: addi a0, a0, 32
; NOT-OPTIMIZED-NEXT: addi a1, a1, 160
-; NOT-OPTIMIZED-NEXT: bne a2, a4, .LBB4_1
+; NOT-OPTIMIZED-NEXT: bnez a2, .LBB4_1
; NOT-OPTIMIZED-NEXT: # %bb.2: # %for.cond.cleanup
; NOT-OPTIMIZED-NEXT: ret
entry:
define void @scatter(ptr noalias nocapture %A, ptr noalias nocapture readonly %B) {
; CHECK-LABEL: scatter:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: li a2, 0
+; CHECK-NEXT: li a2, 1024
; CHECK-NEXT: li a3, 32
; CHECK-NEXT: li a4, 5
-; CHECK-NEXT: li a5, 1024
; CHECK-NEXT: .LBB5_1: # %vector.body
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT: add a6, a1, a2
; CHECK-NEXT: vsetvli zero, a3, e8, m1, ta, ma
-; CHECK-NEXT: vle8.v v8, (a6)
+; CHECK-NEXT: vle8.v v8, (a1)
; CHECK-NEXT: vlse8.v v9, (a0), a4
; CHECK-NEXT: vadd.vv v8, v9, v8
; CHECK-NEXT: vsse8.v v8, (a0), a4
-; CHECK-NEXT: addi a2, a2, 32
+; CHECK-NEXT: addi a2, a2, -32
+; CHECK-NEXT: addi a1, a1, 32
; CHECK-NEXT: addi a0, a0, 160
-; CHECK-NEXT: bne a2, a5, .LBB5_1
+; CHECK-NEXT: bnez a2, .LBB5_1
; CHECK-NEXT: # %bb.2: # %for.cond.cleanup
; CHECK-NEXT: ret
entry:
define void @scatter_masked(ptr noalias nocapture %A, ptr noalias nocapture readonly %B, <32 x i8> %maskedoff) {
; V-LABEL: scatter_masked:
; V: # %bb.0: # %entry
-; V-NEXT: li a2, 0
+; V-NEXT: li a2, 1024
; V-NEXT: li a3, 32
; V-NEXT: lui a4, 983765
; V-NEXT: addiw a4, a4, 873
; V-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; V-NEXT: vmv.s.x v0, a4
; V-NEXT: li a4, 5
-; V-NEXT: li a5, 1024
; V-NEXT: .LBB6_1: # %vector.body
; V-NEXT: # =>This Inner Loop Header: Depth=1
-; V-NEXT: add a6, a1, a2
; V-NEXT: vsetvli zero, a3, e8, m1, ta, mu
-; V-NEXT: vle8.v v9, (a6)
+; V-NEXT: vle8.v v9, (a1)
; V-NEXT: vmv1r.v v10, v8
; V-NEXT: vlse8.v v10, (a0), a4, v0.t
; V-NEXT: vadd.vv v9, v10, v9
; V-NEXT: vsse8.v v9, (a0), a4, v0.t
-; V-NEXT: addi a2, a2, 32
+; V-NEXT: addi a2, a2, -32
+; V-NEXT: addi a1, a1, 32
; V-NEXT: addi a0, a0, 160
-; V-NEXT: bne a2, a5, .LBB6_1
+; V-NEXT: bnez a2, .LBB6_1
; V-NEXT: # %bb.2: # %for.cond.cleanup
; V-NEXT: ret
;
; ZVE32F-LABEL: scatter_masked:
; ZVE32F: # %bb.0: # %entry
-; ZVE32F-NEXT: li a2, 0
+; ZVE32F-NEXT: li a2, 1024
; ZVE32F-NEXT: li a3, 32
; ZVE32F-NEXT: lui a4, 983765
; ZVE32F-NEXT: addiw a4, a4, 873
; ZVE32F-NEXT: vsetivli zero, 1, e32, m1, ta, ma
; ZVE32F-NEXT: vmv.s.x v0, a4
; ZVE32F-NEXT: li a4, 5
-; ZVE32F-NEXT: li a5, 1024
; ZVE32F-NEXT: .LBB6_1: # %vector.body
; ZVE32F-NEXT: # =>This Inner Loop Header: Depth=1
-; ZVE32F-NEXT: add a6, a1, a2
; ZVE32F-NEXT: vsetvli zero, a3, e8, m1, ta, mu
-; ZVE32F-NEXT: vle8.v v9, (a6)
+; ZVE32F-NEXT: vle8.v v9, (a1)
; ZVE32F-NEXT: vmv1r.v v10, v8
; ZVE32F-NEXT: vlse8.v v10, (a0), a4, v0.t
; ZVE32F-NEXT: vadd.vv v9, v10, v9
; ZVE32F-NEXT: vsse8.v v9, (a0), a4, v0.t
-; ZVE32F-NEXT: addi a2, a2, 32
+; ZVE32F-NEXT: addi a2, a2, -32
+; ZVE32F-NEXT: addi a1, a1, 32
; ZVE32F-NEXT: addi a0, a0, 160
-; ZVE32F-NEXT: bne a2, a5, .LBB6_1
+; ZVE32F-NEXT: bnez a2, .LBB6_1
; ZVE32F-NEXT: # %bb.2: # %for.cond.cleanup
; ZVE32F-NEXT: ret
entry:
define void @sink_splat_fma_scalable(ptr noalias nocapture %a, ptr noalias nocapture readonly %b, float %x) {
; CHECK-LABEL: sink_splat_fma_scalable:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: srli a3, a2, 2
-; CHECK-NEXT: li a4, 1024
-; CHECK-NEXT: bgeu a4, a3, .LBB34_2
+; CHECK-NEXT: csrr a3, vlenb
+; CHECK-NEXT: srli a4, a3, 2
+; CHECK-NEXT: li a2, 1024
+; CHECK-NEXT: bgeu a2, a4, .LBB34_2
; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: li a4, 0
+; CHECK-NEXT: li a2, 0
; CHECK-NEXT: j .LBB34_5
; CHECK-NEXT: .LBB34_2: # %vector.ph
-; CHECK-NEXT: li a6, 0
-; CHECK-NEXT: addiw a4, a3, -1
-; CHECK-NEXT: andi a5, a4, 1024
-; CHECK-NEXT: xori a4, a5, 1024
-; CHECK-NEXT: vsetvli a7, zero, e32, m1, ta, ma
-; CHECK-NEXT: mv a7, a4
+; CHECK-NEXT: addiw a2, a4, -1
+; CHECK-NEXT: andi a5, a2, 1024
+; CHECK-NEXT: xori a2, a5, 1024
+; CHECK-NEXT: vsetvli a6, zero, e32, m1, ta, ma
+; CHECK-NEXT: mv a6, a0
+; CHECK-NEXT: mv a7, a1
+; CHECK-NEXT: mv t0, a2
; CHECK-NEXT: .LBB34_3: # %vector.body
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT: add t0, a0, a6
-; CHECK-NEXT: vl1re32.v v8, (t0)
-; CHECK-NEXT: add t1, a1, a6
-; CHECK-NEXT: vl1re32.v v9, (t1)
+; CHECK-NEXT: vl1re32.v v8, (a6)
+; CHECK-NEXT: vl1re32.v v9, (a7)
; CHECK-NEXT: vfmacc.vf v9, fa0, v8
-; CHECK-NEXT: vs1r.v v9, (t0)
-; CHECK-NEXT: sub a7, a7, a3
-; CHECK-NEXT: add a6, a6, a2
-; CHECK-NEXT: bnez a7, .LBB34_3
+; CHECK-NEXT: vs1r.v v9, (a6)
+; CHECK-NEXT: sub t0, t0, a4
+; CHECK-NEXT: add a7, a7, a3
+; CHECK-NEXT: add a6, a6, a3
+; CHECK-NEXT: bnez t0, .LBB34_3
; CHECK-NEXT: # %bb.4: # %middle.block
; CHECK-NEXT: beqz a5, .LBB34_7
; CHECK-NEXT: .LBB34_5: # %for.body.preheader
-; CHECK-NEXT: addi a2, a4, -1024
-; CHECK-NEXT: slli a4, a4, 2
-; CHECK-NEXT: add a1, a1, a4
-; CHECK-NEXT: add a0, a0, a4
+; CHECK-NEXT: addi a3, a2, -1024
+; CHECK-NEXT: slli a2, a2, 2
+; CHECK-NEXT: add a1, a1, a2
+; CHECK-NEXT: add a0, a0, a2
; CHECK-NEXT: .LBB34_6: # %for.body
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
; CHECK-NEXT: flw ft0, 0(a0)
; CHECK-NEXT: flw ft1, 0(a1)
; CHECK-NEXT: fmadd.s ft0, ft0, fa0, ft1
; CHECK-NEXT: fsw ft0, 0(a0)
-; CHECK-NEXT: addi a2, a2, 1
+; CHECK-NEXT: addi a3, a3, 1
; CHECK-NEXT: addi a1, a1, 4
; CHECK-NEXT: addi a0, a0, 4
-; CHECK-NEXT: bnez a2, .LBB34_6
+; CHECK-NEXT: bnez a3, .LBB34_6
; CHECK-NEXT: .LBB34_7: # %for.cond.cleanup
; CHECK-NEXT: ret
entry:
define void @sink_splat_fma_commute_scalable(ptr noalias nocapture %a, ptr noalias nocapture readonly %b, float %x) {
; CHECK-LABEL: sink_splat_fma_commute_scalable:
; CHECK: # %bb.0: # %entry
-; CHECK-NEXT: csrr a2, vlenb
-; CHECK-NEXT: srli a3, a2, 2
-; CHECK-NEXT: li a4, 1024
-; CHECK-NEXT: bgeu a4, a3, .LBB35_2
+; CHECK-NEXT: csrr a3, vlenb
+; CHECK-NEXT: srli a4, a3, 2
+; CHECK-NEXT: li a2, 1024
+; CHECK-NEXT: bgeu a2, a4, .LBB35_2
; CHECK-NEXT: # %bb.1:
-; CHECK-NEXT: li a4, 0
+; CHECK-NEXT: li a2, 0
; CHECK-NEXT: j .LBB35_5
; CHECK-NEXT: .LBB35_2: # %vector.ph
-; CHECK-NEXT: li a6, 0
-; CHECK-NEXT: addiw a4, a3, -1
-; CHECK-NEXT: andi a5, a4, 1024
-; CHECK-NEXT: xori a4, a5, 1024
-; CHECK-NEXT: vsetvli a7, zero, e32, m1, ta, ma
-; CHECK-NEXT: mv a7, a4
+; CHECK-NEXT: addiw a2, a4, -1
+; CHECK-NEXT: andi a5, a2, 1024
+; CHECK-NEXT: xori a2, a5, 1024
+; CHECK-NEXT: vsetvli a6, zero, e32, m1, ta, ma
+; CHECK-NEXT: mv a6, a0
+; CHECK-NEXT: mv a7, a1
+; CHECK-NEXT: mv t0, a2
; CHECK-NEXT: .LBB35_3: # %vector.body
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT: add t0, a0, a6
-; CHECK-NEXT: vl1re32.v v8, (t0)
-; CHECK-NEXT: add t1, a1, a6
-; CHECK-NEXT: vl1re32.v v9, (t1)
+; CHECK-NEXT: vl1re32.v v8, (a6)
+; CHECK-NEXT: vl1re32.v v9, (a7)
; CHECK-NEXT: vfmacc.vf v9, fa0, v8
-; CHECK-NEXT: vs1r.v v9, (t0)
-; CHECK-NEXT: sub a7, a7, a3
-; CHECK-NEXT: add a6, a6, a2
-; CHECK-NEXT: bnez a7, .LBB35_3
+; CHECK-NEXT: vs1r.v v9, (a6)
+; CHECK-NEXT: sub t0, t0, a4
+; CHECK-NEXT: add a7, a7, a3
+; CHECK-NEXT: add a6, a6, a3
+; CHECK-NEXT: bnez t0, .LBB35_3
; CHECK-NEXT: # %bb.4: # %middle.block
; CHECK-NEXT: beqz a5, .LBB35_7
; CHECK-NEXT: .LBB35_5: # %for.body.preheader
-; CHECK-NEXT: addi a2, a4, -1024
-; CHECK-NEXT: slli a4, a4, 2
-; CHECK-NEXT: add a1, a1, a4
-; CHECK-NEXT: add a0, a0, a4
+; CHECK-NEXT: addi a3, a2, -1024
+; CHECK-NEXT: slli a2, a2, 2
+; CHECK-NEXT: add a1, a1, a2
+; CHECK-NEXT: add a0, a0, a2
; CHECK-NEXT: .LBB35_6: # %for.body
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
; CHECK-NEXT: flw ft0, 0(a0)
; CHECK-NEXT: flw ft1, 0(a1)
; CHECK-NEXT: fmadd.s ft0, fa0, ft0, ft1
; CHECK-NEXT: fsw ft0, 0(a0)
-; CHECK-NEXT: addi a2, a2, 1
+; CHECK-NEXT: addi a3, a3, 1
; CHECK-NEXT: addi a1, a1, 4
; CHECK-NEXT: addi a0, a0, 4
-; CHECK-NEXT: bnez a2, .LBB35_6
+; CHECK-NEXT: bnez a3, .LBB35_6
; CHECK-NEXT: .LBB35_7: # %for.cond.cleanup
; CHECK-NEXT: ret
entry: