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
104
105
106
|
; RUN: llc -mtriple=amdgcn < %s | FileCheck -check-prefix=GCN %s
; RUN: llc -mtriple=amdgcn -mcpu=tonga < %s | FileCheck -check-prefix=GCN %s
; GCN-LABEL: {{^}}test_fmed3:
; GCN: v_med3_f32 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}
define amdgpu_kernel void @test_fmed3(ptr addrspace(1) %out, float %src0, float %src1, float %src2) #1 {
%med3 = call float @llvm.amdgcn.fmed3.f32(float %src0, float %src1, float %src2)
store float %med3, ptr addrspace(1) %out
ret void
}
; GCN-LABEL: {{^}}test_fmed3_srcmods:
; GCN: v_med3_f32 v{{[0-9]+}}, -s{{[0-9]+}}, |v{{[0-9]+}}|, -|v{{[0-9]+}}|
define amdgpu_kernel void @test_fmed3_srcmods(ptr addrspace(1) %out, float %src0, float %src1, float %src2) #1 {
%src0.fneg = fsub float -0.0, %src0
%src1.fabs = call float @llvm.fabs.f32(float %src1)
%src2.fabs = call float @llvm.fabs.f32(float %src2)
%src2.fneg.fabs = fsub float -0.0, %src2.fabs
%med3 = call float @llvm.amdgcn.fmed3.f32(float %src0.fneg, float %src1.fabs, float %src2.fneg.fabs)
store float %med3, ptr addrspace(1) %out
ret void
}
; GCN-LABEL: {{^}}test_fneg_fmed3:
; GCN: v_med3_f32 v{{[0-9]+}}, -s{{[0-9]+}}, -v{{[0-9]+}}, -v{{[0-9]+}}
define amdgpu_kernel void @test_fneg_fmed3(ptr addrspace(1) %out, float %src0, float %src1, float %src2) #1 {
%med3 = call float @llvm.amdgcn.fmed3.f32(float %src0, float %src1, float %src2)
%neg.med3 = fsub float -0.0, %med3
store float %neg.med3, ptr addrspace(1) %out
ret void
}
; GCN-LABEL: {{^}}test_fneg_fmed3_multi_use:
; GCN: v_med3_f32 [[MED3:v[0-9]+]], -s{{[0-9]+}}, -v{{[0-9]+}}, -v{{[0-9]+}}
; GCN: v_mul_f32_e32 v{{[0-9]+}}, -4.0, [[MED3]]
define amdgpu_kernel void @test_fneg_fmed3_multi_use(ptr addrspace(1) %out, float %src0, float %src1, float %src2) #1 {
%med3 = call float @llvm.amdgcn.fmed3.f32(float %src0, float %src1, float %src2)
%neg.med3 = fsub float -0.0, %med3
%med3.user = fmul float %med3, 4.0
store volatile float %med3.user, ptr addrspace(1) %out
store volatile float %neg.med3, ptr addrspace(1) %out
ret void
}
; GCN-LABEL: {{^}}test_fabs_fmed3:
; GCN: v_med3_f32 [[MED3:v[0-9]+]], s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}
; GCN: v_and_b32_e32 v{{[0-9]+}}, 0x7fffffff, [[MED3]]
define amdgpu_kernel void @test_fabs_fmed3(ptr addrspace(1) %out, float %src0, float %src1, float %src2) #1 {
%med3 = call float @llvm.amdgcn.fmed3.f32(float %src0, float %src1, float %src2)
%fabs.med3 = call float @llvm.fabs.f32(float %med3)
store float %fabs.med3, ptr addrspace(1) %out
ret void
}
; GCN-LABEL: {{^}}test_fneg_fmed3_rr_0:
; GCN: v_bfrev_b32_e32 [[NEG0:v[0-9]+]], 1
; GCN: v_med3_f32 v{{[0-9]+}}, -s{{[0-9]+}}, -v{{[0-9]+}}, [[NEG0]]
define amdgpu_kernel void @test_fneg_fmed3_rr_0(ptr addrspace(1) %out, float %src0, float %src1) #1 {
%med3 = call float @llvm.amdgcn.fmed3.f32(float %src0, float %src1, float 0.0)
%neg.med3 = fsub float -0.0, %med3
store float %neg.med3, ptr addrspace(1) %out
ret void
}
; FIXME: Worse off from folding this
; GCN-LABEL: {{^}}test_fneg_fmed3_rr_0_foldable_user:
; GCN: v_bfrev_b32_e32 [[NEG0:v[0-9]+]], 1
; GCN: v_med3_f32 [[MED3:v[0-9]+]], -s{{[0-9]+}}, -v{{[0-9]+}}, [[NEG0]]
; GCN: v_mul_f32_e32 v{{[0-9]+}}, s{{[0-9]+}}, [[MED3]]
define amdgpu_kernel void @test_fneg_fmed3_rr_0_foldable_user(ptr addrspace(1) %out, float %src0, float %src1, float %mul.arg) #1 {
%med3 = call float @llvm.amdgcn.fmed3.f32(float %src0, float %src1, float 0.0)
%neg.med3 = fsub float -0.0, %med3
%mul = fmul float %neg.med3, %mul.arg
store float %mul, ptr addrspace(1) %out
ret void
}
; GCN-LABEL: {{^}}test_fneg_fmed3_r_inv2pi_0:
; GCN-DAG: v_bfrev_b32_e32 [[NEG0:v[0-9]+]], 1
; GCN-DAG: v_mov_b32_e32 [[NEG_INV:v[0-9]+]], 0xbe22f983
; GCN: v_med3_f32 v{{[0-9]+}}, -s{{[0-9]+}}, [[NEG_INV]], [[NEG0]]
define amdgpu_kernel void @test_fneg_fmed3_r_inv2pi_0(ptr addrspace(1) %out, float %src0) #1 {
%med3 = call float @llvm.amdgcn.fmed3.f32(float %src0, float 0x3FC45F3060000000, float 0.0)
%neg.med3 = fsub float -0.0, %med3
store float %neg.med3, ptr addrspace(1) %out
ret void
}
; GCN-LABEL: {{^}}test_fneg_fmed3_r_inv2pi_0_foldable_user:
; GCN-DAG: v_bfrev_b32_e32 [[NEG0:v[0-9]+]], 1
; GCN-DAG: v_mov_b32_e32 [[NEG_INV:v[0-9]+]], 0xbe22f983
; GCN: v_med3_f32 [[MED3:v[0-9]+]], -s{{[0-9]+}}, [[NEG_INV]], [[NEG0]]
; GCN: v_mul_f32_e32 v{{[0-9]+}}, s{{[0-9]+}}, [[MED3]]
define amdgpu_kernel void @test_fneg_fmed3_r_inv2pi_0_foldable_user(ptr addrspace(1) %out, float %src0, float %mul.arg) #1 {
%med3 = call float @llvm.amdgcn.fmed3.f32(float %src0, float 0x3FC45F3060000000, float 0.0)
%neg.med3 = fsub float -0.0, %med3
%mul = fmul float %neg.med3, %mul.arg
store float %mul, ptr addrspace(1) %out
ret void
}
declare float @llvm.amdgcn.fmed3.f32(float, float, float) #0
declare float @llvm.fabs.f32(float) #0
attributes #0 = { nounwind readnone }
attributes #1 = { nounwind }
|
