1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 2
; RUN: llc -aarch64-streaming-hazard-size=0 < %s | FileCheck %s
target triple = "aarch64"
; This function would normally scavenge a stackslot from the callee-save
; area, which would lead to spilling 's0' to that stackslot before the
; smstop and filling it with 'addvl + <offset>' after the smstop because
; the frame-pointer is not available.
; This would not be valid, since the vector-length has changed so 'addvl'
; cannot be used. This is testing that the stackslot-scavenging is disabled
; when there are any streaming-mode-changing call-sequences in the
; function.
define void @test_no_stackslot_scavenging(float %f) #0 {
; CHECK-LABEL: test_no_stackslot_scavenging:
; CHECK: // %bb.0:
; CHECK-NEXT: stp d15, d14, [sp, #-96]! // 16-byte Folded Spill
; CHECK-NEXT: cntd x9
; CHECK-NEXT: stp d13, d12, [sp, #16] // 16-byte Folded Spill
; CHECK-NEXT: stp d11, d10, [sp, #32] // 16-byte Folded Spill
; CHECK-NEXT: stp d9, d8, [sp, #48] // 16-byte Folded Spill
; CHECK-NEXT: stp x29, x30, [sp, #64] // 16-byte Folded Spill
; CHECK-NEXT: stp x9, x24, [sp, #80] // 16-byte Folded Spill
; CHECK-NEXT: sub sp, sp, #16
; CHECK-NEXT: addvl sp, sp, #-1
; CHECK-NEXT: str s0, [sp, #12] // 4-byte Folded Spill
; CHECK-NEXT: //APP
; CHECK-NEXT: //NO_APP
; CHECK-NEXT: smstop sm
; CHECK-NEXT: ldr s0, [sp, #12] // 4-byte Folded Reload
; CHECK-NEXT: bl use_f
; CHECK-NEXT: smstart sm
; CHECK-NEXT: addvl sp, sp, #1
; CHECK-NEXT: add sp, sp, #16
; CHECK-NEXT: ldp x29, x30, [sp, #64] // 16-byte Folded Reload
; CHECK-NEXT: ldr x24, [sp, #88] // 8-byte Folded Reload
; CHECK-NEXT: ldp d9, d8, [sp, #48] // 16-byte Folded Reload
; CHECK-NEXT: ldp d11, d10, [sp, #32] // 16-byte Folded Reload
; CHECK-NEXT: ldp d13, d12, [sp, #16] // 16-byte Folded Reload
; CHECK-NEXT: ldp d15, d14, [sp], #96 // 16-byte Folded Reload
; CHECK-NEXT: ret
%ptr = alloca <vscale x 16 x i8>
call void asm sideeffect "", "~{x24}"() nounwind
call void @use_f(float %f)
ret void
}
define void @test_no_stackslot_scavenging_with_fp(float %f, i64 %n) #0 "frame-pointer"="all" {
; CHECK-LABEL: test_no_stackslot_scavenging_with_fp:
; CHECK: // %bb.0:
; CHECK-NEXT: stp d15, d14, [sp, #-128]! // 16-byte Folded Spill
; CHECK-NEXT: cntd x9
; CHECK-NEXT: stp d13, d12, [sp, #16] // 16-byte Folded Spill
; CHECK-NEXT: stp d11, d10, [sp, #32] // 16-byte Folded Spill
; CHECK-NEXT: stp d9, d8, [sp, #48] // 16-byte Folded Spill
; CHECK-NEXT: stp x29, x30, [sp, #64] // 16-byte Folded Spill
; CHECK-NEXT: add x29, sp, #64
; CHECK-NEXT: str x9, [sp, #80] // 8-byte Folded Spill
; CHECK-NEXT: stp x28, x25, [sp, #96] // 16-byte Folded Spill
; CHECK-NEXT: stp x24, x19, [sp, #112] // 16-byte Folded Spill
; CHECK-NEXT: addvl sp, sp, #-1
; CHECK-NEXT: lsl x9, x0, #3
; CHECK-NEXT: mov x8, sp
; CHECK-NEXT: mov x19, sp
; CHECK-NEXT: str s0, [x29, #28] // 4-byte Folded Spill
; CHECK-NEXT: add x9, x9, #15
; CHECK-NEXT: and x9, x9, #0xfffffffffffffff0
; CHECK-NEXT: sub x8, x8, x9
; CHECK-NEXT: mov sp, x8
; CHECK-NEXT: //APP
; CHECK-NEXT: //NO_APP
; CHECK-NEXT: smstop sm
; CHECK-NEXT: ldr s0, [x29, #28] // 4-byte Folded Reload
; CHECK-NEXT: bl use_f
; CHECK-NEXT: smstart sm
; CHECK-NEXT: sub sp, x29, #64
; CHECK-NEXT: ldp x24, x19, [sp, #112] // 16-byte Folded Reload
; CHECK-NEXT: ldp x28, x25, [sp, #96] // 16-byte Folded Reload
; CHECK-NEXT: ldp x29, x30, [sp, #64] // 16-byte Folded Reload
; CHECK-NEXT: ldp d9, d8, [sp, #48] // 16-byte Folded Reload
; CHECK-NEXT: ldp d11, d10, [sp, #32] // 16-byte Folded Reload
; CHECK-NEXT: ldp d13, d12, [sp, #16] // 16-byte Folded Reload
; CHECK-NEXT: ldp d15, d14, [sp], #128 // 16-byte Folded Reload
; CHECK-NEXT: ret
%ptr2 = alloca i64, i64 %n, align 8
%ptr = alloca <vscale x 16 x i8>
call void asm sideeffect "", "~{x24},~{x25}"() nounwind
call void @use_f(float %f)
ret void
}
declare void @use_f(float)
declare void @use_f_and_ptr(float, ptr)
attributes #0 = { nounwind "target-features"="+sve,+sme" "aarch64_pstate_sm_enabled" }
|
