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
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
| ; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=SI -check-prefix=SIVI %s
; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=VI -check-prefix=SIVI %s
; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=gfx900 -mattr=-flat-for-global,-fp64-fp16-denormals -verify-machineinstrs -enable-unsafe-fp-math < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GFX9 %s
; GCN-LABEL: {{^}}fptrunc_f32_to_f16:
; GCN: buffer_load_dword v[[A_F32:[0-9]+]]
; GCN: v_cvt_f16_f32_e32 v[[R_F16:[0-9]+]], v[[A_F32]]
; GCN: buffer_store_short v[[R_F16]]
; GCN: s_endpgm
define amdgpu_kernel void @fptrunc_f32_to_f16(
half addrspace(1)* %r,
float addrspace(1)* %a) {
entry:
%a.val = load float, float addrspace(1)* %a
%r.val = fptrunc float %a.val to half
store half %r.val, half addrspace(1)* %r
ret void
}
; GCN-LABEL: {{^}}fptrunc_f64_to_f16:
; GCN: buffer_load_dwordx2 v{{\[}}[[A_F64_0:[0-9]+]]:[[A_F64_1:[0-9]+]]{{\]}}
; GCN: v_cvt_f32_f64_e32 v[[A_F32:[0-9]+]], v{{\[}}[[A_F64_0]]:[[A_F64_1]]{{\]}}
; GCN: v_cvt_f16_f32_e32 v[[R_F16:[0-9]+]], v[[A_F32]]
; GCN: buffer_store_short v[[R_F16]]
; GCN: s_endpgm
define amdgpu_kernel void @fptrunc_f64_to_f16(
half addrspace(1)* %r,
double addrspace(1)* %a) {
entry:
%a.val = load double, double addrspace(1)* %a
%r.val = fptrunc double %a.val to half
store half %r.val, half addrspace(1)* %r
ret void
}
; GCN-LABEL: {{^}}fptrunc_v2f32_to_v2f16:
; GCN: buffer_load_dwordx2 v{{\[}}[[A_F32_0:[0-9]+]]:[[A_F32_1:[0-9]+]]{{\]}}
; GCN-DAG: v_cvt_f16_f32_e32 v[[R_F16_0:[0-9]+]], v[[A_F32_0]]
; SI-DAG: v_cvt_f16_f32_e32 v[[R_F16_1:[0-9]+]], v[[A_F32_1]]
; SI-DAG: v_lshlrev_b32_e32 v[[R_F16_HI:[0-9]+]], 16, v[[R_F16_1]]
; SI: v_or_b32_e32 v[[R_V2_F16:[0-9]+]], v[[R_F16_0]], v[[R_F16_HI]]
; VI-DAG: v_cvt_f16_f32_sdwa v[[R_F16_1:[0-9]+]], v[[A_F32_1]] dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:DWORD
; VI: v_or_b32_e32 v[[R_V2_F16:[0-9]+]], v[[R_F16_0]], v[[R_F16_1]]
; GFX9-DAG: v_cvt_f16_f32_e32 v[[R_F16_1:[0-9]+]], v[[A_F32_1]]
; GFX9: v_and_b32_e32 v[[R_F16_LO:[0-9]+]], 0xffff, v[[R_F16_0]]
; GFX9: v_lshl_or_b32 v[[R_V2_F16:[0-9]+]], v[[R_F16_1]], 16, v[[R_F16_LO]]
; GCN: buffer_store_dword v[[R_V2_F16]]
; GCN: s_endpgm
define amdgpu_kernel void @fptrunc_v2f32_to_v2f16(
<2 x half> addrspace(1)* %r,
<2 x float> addrspace(1)* %a) {
entry:
%a.val = load <2 x float>, <2 x float> addrspace(1)* %a
%r.val = fptrunc <2 x float> %a.val to <2 x half>
store <2 x half> %r.val, <2 x half> addrspace(1)* %r
ret void
}
; GCN-LABEL: {{^}}fptrunc_v2f64_to_v2f16:
; GCN: buffer_load_dwordx4 v{{\[}}[[A_F64_0:[0-9]+]]:[[A_F64_3:[0-9]+]]{{\]}}
; GCN-DAG: v_cvt_f32_f64_e32 v[[A_F32_0:[0-9]+]], v{{\[}}[[A_F64_0]]:{{[0-9]+}}{{\]}}
; GCN-DAG: v_cvt_f32_f64_e32 v[[A_F32_1:[0-9]+]], v{{\[}}{{[0-9]+}}:[[A_F64_3]]{{\]}}
; GCN-DAG: v_cvt_f16_f32_e32 v[[R_F16_0:[0-9]+]], v[[A_F32_0]]
;
; SI-DAG: v_cvt_f16_f32_e32 v[[CVTHI:[0-9]+]], v[[A_F32_1]]
; SI-DAG: v_lshlrev_b32_e32 v[[R_F16_HI:[0-9]+]], 16, v[[CVTHI]]
; VI: v_cvt_f16_f32_sdwa v[[R_F16_HI:[0-9]+]], v[[A_F32_1]] dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:DWORD
; SIVI: v_or_b32_e32 v[[R_V2_F16:[0-9]+]], v[[R_F16_0]], v[[R_F16_HI]]
; GFX9-DAG: v_cvt_f16_f32_e32 v[[R_F16_1:[0-9]+]], v[[A_F32_1]]
; GFX9: v_and_b32_e32 v[[R_F16_LO:[0-9]+]], 0xffff, v[[R_F16_0]]
; GFX9: v_lshl_or_b32 v[[R_V2_F16:[0-9]+]], v[[R_F16_1]], 16, v[[R_F16_LO]]
; GCN: buffer_store_dword v[[R_V2_F16]]
define amdgpu_kernel void @fptrunc_v2f64_to_v2f16(
<2 x half> addrspace(1)* %r,
<2 x double> addrspace(1)* %a) {
entry:
%a.val = load <2 x double>, <2 x double> addrspace(1)* %a
%r.val = fptrunc <2 x double> %a.val to <2 x half>
store <2 x half> %r.val, <2 x half> addrspace(1)* %r
ret void
}
; GCN-LABEL: {{^}}fneg_fptrunc_f32_to_f16:
; GCN: buffer_load_dword v[[A_F32:[0-9]+]]
; GCN: v_cvt_f16_f32_e64 v[[R_F16:[0-9]+]], -v[[A_F32]]
; GCN: buffer_store_short v[[R_F16]]
; GCN: s_endpgm
define amdgpu_kernel void @fneg_fptrunc_f32_to_f16(
half addrspace(1)* %r,
float addrspace(1)* %a) {
entry:
%a.val = load float, float addrspace(1)* %a
%a.fneg = fsub float -0.0, %a.val
%r.val = fptrunc float %a.fneg to half
store half %r.val, half addrspace(1)* %r
ret void
}
; GCN-LABEL: {{^}}fabs_fptrunc_f32_to_f16:
; GCN: buffer_load_dword v[[A_F32:[0-9]+]]
; GCN: v_cvt_f16_f32_e64 v[[R_F16:[0-9]+]], |v[[A_F32]]|
; GCN: buffer_store_short v[[R_F16]]
; GCN: s_endpgm
define amdgpu_kernel void @fabs_fptrunc_f32_to_f16(
half addrspace(1)* %r,
float addrspace(1)* %a) {
entry:
%a.val = load float, float addrspace(1)* %a
%a.fabs = call float @llvm.fabs.f32(float %a.val)
%r.val = fptrunc float %a.fabs to half
store half %r.val, half addrspace(1)* %r
ret void
}
; GCN-LABEL: {{^}}fneg_fabs_fptrunc_f32_to_f16:
; GCN: buffer_load_dword v[[A_F32:[0-9]+]]
; GCN: v_cvt_f16_f32_e64 v[[R_F16:[0-9]+]], -|v[[A_F32]]|
; GCN: buffer_store_short v[[R_F16]]
; GCN: s_endpgm
define amdgpu_kernel void @fneg_fabs_fptrunc_f32_to_f16(
half addrspace(1)* %r,
float addrspace(1)* %a) #0 {
entry:
%a.val = load float, float addrspace(1)* %a
%a.fabs = call float @llvm.fabs.f32(float %a.val)
%a.fneg.fabs = fsub float -0.0, %a.fabs
%r.val = fptrunc float %a.fneg.fabs to half
store half %r.val, half addrspace(1)* %r
ret void
}
; GCN-LABEL: {{^}}fptrunc_f32_to_f16_zext_i32:
; GCN: buffer_load_dword v[[A_F32:[0-9]+]]
; GCN: v_cvt_f16_f32_e32 v[[R_F16:[0-9]+]], v[[A_F32]]
; GCN-NOT: v[[R_F16]]
; GCN: buffer_store_dword v[[R_F16]]
define amdgpu_kernel void @fptrunc_f32_to_f16_zext_i32(
i32 addrspace(1)* %r,
float addrspace(1)* %a) #0 {
entry:
%a.val = load float, float addrspace(1)* %a
%r.val = fptrunc float %a.val to half
%r.i16 = bitcast half %r.val to i16
%zext = zext i16 %r.i16 to i32
store i32 %zext, i32 addrspace(1)* %r
ret void
}
; GCN-LABEL: {{^}}fptrunc_fabs_f32_to_f16_zext_i32:
; GCN: buffer_load_dword v[[A_F32:[0-9]+]]
; GCN: v_cvt_f16_f32_e64 v[[R_F16:[0-9]+]], |v[[A_F32]]|
; GCN-NOT: v[[R_F16]]
; GCN: buffer_store_dword v[[R_F16]]
define amdgpu_kernel void @fptrunc_fabs_f32_to_f16_zext_i32(
i32 addrspace(1)* %r,
float addrspace(1)* %a) #0 {
entry:
%a.val = load float, float addrspace(1)* %a
%a.fabs = call float @llvm.fabs.f32(float %a.val)
%r.val = fptrunc float %a.fabs to half
%r.i16 = bitcast half %r.val to i16
%zext = zext i16 %r.i16 to i32
store i32 %zext, i32 addrspace(1)* %r
ret void
}
; GCN-LABEL: {{^}}fptrunc_f32_to_f16_sext_i32:
; GCN: buffer_load_dword v[[A_F32:[0-9]+]]
; GCN: v_cvt_f16_f32_e32 v[[R_F16:[0-9]+]], v[[A_F32]]
; GCN: v_bfe_i32 v[[R_F16_SEXT:[0-9]+]], v[[R_F16]], 0, 16
; GCN: buffer_store_dword v[[R_F16_SEXT]]
define amdgpu_kernel void @fptrunc_f32_to_f16_sext_i32(
i32 addrspace(1)* %r,
float addrspace(1)* %a) #0 {
entry:
%a.val = load float, float addrspace(1)* %a
%r.val = fptrunc float %a.val to half
%r.i16 = bitcast half %r.val to i16
%zext = sext i16 %r.i16 to i32
store i32 %zext, i32 addrspace(1)* %r
ret void
}
declare float @llvm.fabs.f32(float) #1
attributes #0 = { nounwind }
attributes #1 = { nounwind readnone }
|