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
|
# NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 4
# RUN: llc %s -mtriple=riscv32 -mattr=+zdinx -start-before=prologepilog -o - | FileCheck %s
# We want to make sure eliminateFrameIndex doesn't fold sp+2044 as an offset in
# a GPR pair spill/reload instruction. When we split the pair spill, we would be
# unable to add 4 to the immediate without overflowing simm12.
--- |
define void @foo() {
; CHECK-LABEL: foo:
; CHECK: # %bb.0:
; CHECK-NEXT: addi sp, sp, -2048
; CHECK-NEXT: addi sp, sp, -16
; CHECK-NEXT: .cfi_def_cfa_offset 2064
; CHECK-NEXT: lui t0, 1
; CHECK-NEXT: add t0, sp, t0
; CHECK-NEXT: sw a0, -2040(t0) # 4-byte Folded Spill
; CHECK-NEXT: sw a1, -2036(t0) # 4-byte Folded Spill
; CHECK-NEXT: lui a0, 1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: sw a2, -2048(a0) # 4-byte Folded Spill
; CHECK-NEXT: sw a3, -2044(a0) # 4-byte Folded Spill
; CHECK-NEXT: sw a4, 2040(sp) # 4-byte Folded Spill
; CHECK-NEXT: sw a5, 2044(sp) # 4-byte Folded Spill
; CHECK-NEXT: sw a6, 2032(sp) # 4-byte Folded Spill
; CHECK-NEXT: sw a7, 2036(sp) # 4-byte Folded Spill
; CHECK-NEXT: lui a0, 1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: lw a1, -2036(a0) # 4-byte Folded Reload
; CHECK-NEXT: lw a0, -2040(a0) # 4-byte Folded Reload
; CHECK-NEXT: lui a0, 1
; CHECK-NEXT: add a0, sp, a0
; CHECK-NEXT: lw a2, -2048(a0) # 4-byte Folded Reload
; CHECK-NEXT: lw a3, -2044(a0) # 4-byte Folded Reload
; CHECK-NEXT: lw a4, 2040(sp) # 4-byte Folded Reload
; CHECK-NEXT: lw a5, 2044(sp) # 4-byte Folded Reload
; CHECK-NEXT: lw a6, 2032(sp) # 4-byte Folded Reload
; CHECK-NEXT: lw a7, 2036(sp) # 4-byte Folded Reload
; CHECK-NEXT: addi sp, sp, 2032
; CHECK-NEXT: addi sp, sp, 32
; CHECK-NEXT: .cfi_def_cfa_offset 0
; CHECK-NEXT: ret
ret void
}
...
---
name: foo
tracksRegLiveness: true
tracksDebugUserValues: true
frameInfo:
maxAlignment: 4
stack:
- { id: 0, type: spill-slot, size: 8, alignment: 4 }
- { id: 1, type: spill-slot, size: 8, alignment: 4 }
- { id: 2, type: spill-slot, size: 8, alignment: 4 }
- { id: 3, type: spill-slot, size: 8, alignment: 4 }
- { id: 4, type: spill-slot, size: 2024, alignment: 4 }
machineFunctionInfo:
varArgsFrameIndex: 0
varArgsSaveSize: 0
body: |
bb.0:
liveins: $x10_x11, $x12_x13, $x14_x15, $x16_x17
PseudoRV32ZdinxSD killed renamable $x10_x11, %stack.0, 0 :: (store (s64) into %stack.0, align 4)
PseudoRV32ZdinxSD killed renamable $x12_x13, %stack.1, 0 :: (store (s64) into %stack.1, align 4)
PseudoRV32ZdinxSD killed renamable $x14_x15, %stack.2, 0 :: (store (s64) into %stack.2, align 4)
PseudoRV32ZdinxSD killed renamable $x16_x17, %stack.3, 0 :: (store (s64) into %stack.3, align 4)
renamable $x10_x11 = PseudoRV32ZdinxLD %stack.0, 0 :: (load (s64) from %stack.0, align 4)
renamable $x12_x13 = PseudoRV32ZdinxLD %stack.1, 0 :: (load (s64) from %stack.1, align 4)
renamable $x14_x15 = PseudoRV32ZdinxLD %stack.2, 0 :: (load (s64) from %stack.2, align 4)
renamable $x16_x17 = PseudoRV32ZdinxLD %stack.3, 0 :: (load (s64) from %stack.3, align 4)
PseudoRET
...
|
