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
96
97
98
99
100
101
102
103
|
; NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
; RUN: llc -mtriple=riscv32 -mattr=+f,+d -stop-after=finalize-isel < %s \
; RUN: -target-abi=ilp32d | FileCheck -check-prefixes=RV32IF %s
; RUN: llc -mtriple=riscv64 -mattr=+f,+d -stop-after=finalize-isel < %s \
; RUN: -target-abi=lp64d | FileCheck -check-prefixes=RV64IF %s
; Make sure an implicit FRM dependency is added to instructions with dynamic
; rounding.
define float @fadd_s(float %a, float %b) nounwind {
; RV32IF-LABEL: name: fadd_s
; RV32IF: bb.0 (%ir-block.0):
; RV32IF-NEXT: liveins: $f10_f, $f11_f
; RV32IF-NEXT: {{ $}}
; RV32IF-NEXT: [[COPY:%[0-9]+]]:fpr32 = COPY $f11_f
; RV32IF-NEXT: [[COPY1:%[0-9]+]]:fpr32 = COPY $f10_f
; RV32IF-NEXT: %2:fpr32 = nofpexcept FADD_S [[COPY1]], [[COPY]], 7, implicit $frm
; RV32IF-NEXT: $f10_f = COPY %2
; RV32IF-NEXT: PseudoRET implicit $f10_f
; RV64IF-LABEL: name: fadd_s
; RV64IF: bb.0 (%ir-block.0):
; RV64IF-NEXT: liveins: $f10_f, $f11_f
; RV64IF-NEXT: {{ $}}
; RV64IF-NEXT: [[COPY:%[0-9]+]]:fpr32 = COPY $f11_f
; RV64IF-NEXT: [[COPY1:%[0-9]+]]:fpr32 = COPY $f10_f
; RV64IF-NEXT: %2:fpr32 = nofpexcept FADD_S [[COPY1]], [[COPY]], 7, implicit $frm
; RV64IF-NEXT: $f10_f = COPY %2
; RV64IF-NEXT: PseudoRET implicit $f10_f
%1 = fadd float %a, %b
ret float %1
}
declare float @llvm.fma.f32(float, float, float)
define float @fmadd_s(float %a, float %b, float %c) nounwind {
; RV32IF-LABEL: name: fmadd_s
; RV32IF: bb.0 (%ir-block.0):
; RV32IF-NEXT: liveins: $f10_f, $f11_f, $f12_f
; RV32IF-NEXT: {{ $}}
; RV32IF-NEXT: [[COPY:%[0-9]+]]:fpr32 = COPY $f12_f
; RV32IF-NEXT: [[COPY1:%[0-9]+]]:fpr32 = COPY $f11_f
; RV32IF-NEXT: [[COPY2:%[0-9]+]]:fpr32 = COPY $f10_f
; RV32IF-NEXT: %3:fpr32 = nofpexcept FMADD_S [[COPY2]], [[COPY1]], [[COPY]], 7, implicit $frm
; RV32IF-NEXT: $f10_f = COPY %3
; RV32IF-NEXT: PseudoRET implicit $f10_f
; RV64IF-LABEL: name: fmadd_s
; RV64IF: bb.0 (%ir-block.0):
; RV64IF-NEXT: liveins: $f10_f, $f11_f, $f12_f
; RV64IF-NEXT: {{ $}}
; RV64IF-NEXT: [[COPY:%[0-9]+]]:fpr32 = COPY $f12_f
; RV64IF-NEXT: [[COPY1:%[0-9]+]]:fpr32 = COPY $f11_f
; RV64IF-NEXT: [[COPY2:%[0-9]+]]:fpr32 = COPY $f10_f
; RV64IF-NEXT: %3:fpr32 = nofpexcept FMADD_S [[COPY2]], [[COPY1]], [[COPY]], 7, implicit $frm
; RV64IF-NEXT: $f10_f = COPY %3
; RV64IF-NEXT: PseudoRET implicit $f10_f
%1 = call float @llvm.fma.f32(float %a, float %b, float %c)
ret float %1
}
; This uses rtz instead of dyn rounding mode so shouldn't have an FRM dependncy.
define i32 @fcvt_w_s(float %a) nounwind {
; RV32IF-LABEL: name: fcvt_w_s
; RV32IF: bb.0 (%ir-block.0):
; RV32IF-NEXT: liveins: $f10_f
; RV32IF-NEXT: {{ $}}
; RV32IF-NEXT: [[COPY:%[0-9]+]]:fpr32 = COPY $f10_f
; RV32IF-NEXT: %1:gpr = nofpexcept FCVT_W_S [[COPY]], 1
; RV32IF-NEXT: $x10 = COPY %1
; RV32IF-NEXT: PseudoRET implicit $x10
; RV64IF-LABEL: name: fcvt_w_s
; RV64IF: bb.0 (%ir-block.0):
; RV64IF-NEXT: liveins: $f10_f
; RV64IF-NEXT: {{ $}}
; RV64IF-NEXT: [[COPY:%[0-9]+]]:fpr32 = COPY $f10_f
; RV64IF-NEXT: %1:gpr = nofpexcept FCVT_W_S [[COPY]], 1
; RV64IF-NEXT: $x10 = COPY %1
; RV64IF-NEXT: PseudoRET implicit $x10
%1 = fptosi float %a to i32
ret i32 %1
}
; This doesn't use a rounding mode since i32 can be represented exactly as a
; double.
define double @fcvt_d_w(i32 %a) nounwind {
; RV32IF-LABEL: name: fcvt_d_w
; RV32IF: bb.0 (%ir-block.0):
; RV32IF-NEXT: liveins: $x10
; RV32IF-NEXT: {{ $}}
; RV32IF-NEXT: [[COPY:%[0-9]+]]:gpr = COPY $x10
; RV32IF-NEXT: %1:fpr64 = nofpexcept FCVT_D_W [[COPY]]
; RV32IF-NEXT: $f10_d = COPY %1
; RV32IF-NEXT: PseudoRET implicit $f10_d
; RV64IF-LABEL: name: fcvt_d_w
; RV64IF: bb.0 (%ir-block.0):
; RV64IF-NEXT: liveins: $x10
; RV64IF-NEXT: {{ $}}
; RV64IF-NEXT: [[COPY:%[0-9]+]]:gpr = COPY $x10
; RV64IF-NEXT: %1:fpr64 = nofpexcept FCVT_D_W [[COPY]]
; RV64IF-NEXT: $f10_d = COPY %1
; RV64IF-NEXT: PseudoRET implicit $f10_d
%1 = sitofp i32 %a to double
ret double %1
}
|
