1 ; RUN: llc -mtriple=aarch64-none-linux-gnu -max-bytes-for-alignment=8 --align-loops=32 < %s -o -| FileCheck %s --check-prefixes=CHECK,CHECK-EXPLICIT
2 ; RUN: llc -mtriple=aarch64-none-linux-gnu --align-loops=32 < %s -o -| FileCheck %s --check-prefixes=CHECK,CHECK-IMPLICIT
3 ; RUN: llc -mtriple=aarch64-none-linux-gnu --align-loops=32 < %s -o - --filetype=obj | llvm-objdump --arch=aarch64 -d -| FileCheck %s --check-prefixes=CHECK-OBJ,CHECK-OBJ-IMPLICIT
4 ; RUN: llc -mtriple=aarch64-none-linux-gnu -max-bytes-for-alignment=8 --align-loops=32 < %s -o - --filetype=obj | llvm-objdump --arch=aarch64 -d -| FileCheck %s --check-prefixes=CHECK-OBJ,CHECK-OBJ-EXPLICIT
6 ; This test is checking that the correct operands to the .p2align are emitted correctly, and that the resulting obj
7 ; is padded as expected. The key interest in the CHECK-OBJ-* sections is the size of the padding region (the nops),
8 ; and not the exact instructions either side of them (But the last instruction of the EXPLICIT and IMPLICIT checks
9 ; should be the same, at different locations)
10 define i32 @a(i32 %x, i32* nocapture readonly %y, i32* nocapture readonly %z) {
12 ; CHECK-EXPLICIT: .p2align 5, , 8
13 ; CHECK-IMPLICIT: .p2align 5
14 ; CHECK-NEXT: .LBB0_5: // %vector.body
15 ; CHECK-EXPLICIT: .p2align 5, , 8
16 ; CHECK-IMPLICIT: .p2align 5
17 ; CHECK-NEXT: .LBB0_8: // %for.body
18 ; CHECK-OBJ;Disassembly of section .text:
19 ; CHECK-OBJ: 88: 8b0a002a add
20 ; CHECK-OBJ-IMPLICIT-NEXT: 8c: d503201f nop
21 ; CHECK-OBJ-IMPLICIT-NEXT: 90: d503201f nop
22 ; CHECK-OBJ-IMPLICIT-NEXT: 94: d503201f nop
23 ; CHECK-OBJ-IMPLICIT-NEXT: 98: d503201f nop
24 ; CHECK-OBJ-IMPLICIT-NEXT: 9c: d503201f nop
25 ; CHECK-OBJ-IMPLICIT-NEXT: a0: b840454b ldr
26 ; CHECK-OBJ-EXPLICIT-NEXT: 8c: b840454b ldr
28 %cmp10 = icmp sgt i32 %x, 0
29 br i1 %cmp10, label %for.body.preheader, label %for.cond.cleanup
31 for.body.preheader: ; preds = %entry
32 %wide.trip.count = zext i32 %x to i64
33 %min.iters.check = icmp ult i32 %x, 8
34 br i1 %min.iters.check, label %for.body.preheader17, label %vector.ph
36 vector.ph: ; preds = %for.body.preheader
37 %n.vec = and i64 %wide.trip.count, 4294967288
40 vector.body: ; preds = %vector.body, %vector.ph
41 %index = phi i64 [ 0, %vector.ph ], [ %index.next, %vector.body ]
42 %vec.phi = phi <4 x i32> [ zeroinitializer, %vector.ph ], [ %10, %vector.body ]
43 %vec.phi13 = phi <4 x i32> [ zeroinitializer, %vector.ph ], [ %11, %vector.body ]
44 %0 = getelementptr inbounds i32, i32* %y, i64 %index
45 %1 = bitcast i32* %0 to <4 x i32>*
46 %wide.load = load <4 x i32>, <4 x i32>* %1, align 4
47 %2 = getelementptr inbounds i32, i32* %0, i64 4
48 %3 = bitcast i32* %2 to <4 x i32>*
49 %wide.load14 = load <4 x i32>, <4 x i32>* %3, align 4
50 %4 = getelementptr inbounds i32, i32* %z, i64 %index
51 %5 = bitcast i32* %4 to <4 x i32>*
52 %wide.load15 = load <4 x i32>, <4 x i32>* %5, align 4
53 %6 = getelementptr inbounds i32, i32* %4, i64 4
54 %7 = bitcast i32* %6 to <4 x i32>*
55 %wide.load16 = load <4 x i32>, <4 x i32>* %7, align 4
56 %8 = add <4 x i32> %wide.load, %vec.phi
57 %9 = add <4 x i32> %wide.load14, %vec.phi13
58 %10 = add <4 x i32> %8, %wide.load15
59 %11 = add <4 x i32> %9, %wide.load16
60 %index.next = add nuw i64 %index, 8
61 %12 = icmp eq i64 %index.next, %n.vec
62 br i1 %12, label %middle.block, label %vector.body
64 middle.block: ; preds = %vector.body
65 %bin.rdx = add <4 x i32> %11, %10
66 %13 = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> %bin.rdx)
67 %cmp.n = icmp eq i64 %n.vec, %wide.trip.count
68 br i1 %cmp.n, label %for.cond.cleanup, label %for.body.preheader17
70 for.body.preheader17: ; preds = %for.body.preheader, %middle.block
71 %indvars.iv.ph = phi i64 [ 0, %for.body.preheader ], [ %n.vec, %middle.block ]
72 %b.011.ph = phi i32 [ 0, %for.body.preheader ], [ %13, %middle.block ]
75 for.cond.cleanup: ; preds = %for.body, %middle.block, %entry
76 %b.0.lcssa = phi i32 [ 0, %entry ], [ %13, %middle.block ], [ %add3, %for.body ]
79 for.body: ; preds = %for.body.preheader17, %for.body
80 %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ %indvars.iv.ph, %for.body.preheader17 ]
81 %b.011 = phi i32 [ %add3, %for.body ], [ %b.011.ph, %for.body.preheader17 ]
82 %arrayidx = getelementptr inbounds i32, i32* %y, i64 %indvars.iv
83 %14 = load i32, i32* %arrayidx, align 4
84 %arrayidx2 = getelementptr inbounds i32, i32* %z, i64 %indvars.iv
85 %15 = load i32, i32* %arrayidx2, align 4
86 %add = add i32 %14, %b.011
87 %add3 = add i32 %add, %15
88 %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
89 %exitcond.not = icmp eq i64 %indvars.iv.next, %wide.trip.count
90 br i1 %exitcond.not, label %for.cond.cleanup, label %for.body
93 declare i32 @llvm.vector.reduce.add.v4i32(<4 x i32>)