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
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -global-isel -mtriple=amdgcn-amd-amdpal -mcpu=bonaire -verify-machineinstrs < %s | FileCheck -check-prefix=GFX7 %s
; RUN: llc -global-isel -mtriple=amdgcn-amd-amdpal -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=GFX8 %s
define i16 @v_trunc_i32_to_i16(i32 %src) {
; GFX7-LABEL: v_trunc_i32_to_i16:
; GFX7: ; %bb.0:
; GFX7-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX7-NEXT: s_setpc_b64 s[30:31]
;
; GFX8-LABEL: v_trunc_i32_to_i16:
; GFX8: ; %bb.0:
; GFX8-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX8-NEXT: s_setpc_b64 s[30:31]
%trunc = trunc i32 %src to i16
ret i16 %trunc
}
define amdgpu_ps i16 @s_trunc_i32_to_i16(i32 inreg %src) {
; GFX7-LABEL: s_trunc_i32_to_i16:
; GFX7: ; %bb.0:
; GFX7-NEXT: ; return to shader part epilog
;
; GFX8-LABEL: s_trunc_i32_to_i16:
; GFX8: ; %bb.0:
; GFX8-NEXT: ; return to shader part epilog
%trunc = trunc i32 %src to i16
ret i16 %trunc
}
define i16 @v_trunc_i64_to_i16(i64 %src) {
; GFX7-LABEL: v_trunc_i64_to_i16:
; GFX7: ; %bb.0:
; GFX7-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX7-NEXT: s_setpc_b64 s[30:31]
;
; GFX8-LABEL: v_trunc_i64_to_i16:
; GFX8: ; %bb.0:
; GFX8-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX8-NEXT: s_setpc_b64 s[30:31]
%trunc = trunc i64 %src to i16
ret i16 %trunc
}
define amdgpu_ps i16 @s_trunc_i64_to_i16(i64 inreg %src) {
; GFX7-LABEL: s_trunc_i64_to_i16:
; GFX7: ; %bb.0:
; GFX7-NEXT: ; return to shader part epilog
;
; GFX8-LABEL: s_trunc_i64_to_i16:
; GFX8: ; %bb.0:
; GFX8-NEXT: ; return to shader part epilog
%trunc = trunc i64 %src to i16
ret i16 %trunc
}
define amdgpu_ps i16 @s_trunc_i128_to_i16(i128 inreg %src) {
; GFX7-LABEL: s_trunc_i128_to_i16:
; GFX7: ; %bb.0:
; GFX7-NEXT: ; return to shader part epilog
;
; GFX8-LABEL: s_trunc_i128_to_i16:
; GFX8: ; %bb.0:
; GFX8-NEXT: ; return to shader part epilog
%trunc = trunc i128 %src to i16
ret i16 %trunc
}
define i16 @v_trunc_i128_to_i16(i128 %src) {
; GFX7-LABEL: v_trunc_i128_to_i16:
; GFX7: ; %bb.0:
; GFX7-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX7-NEXT: s_setpc_b64 s[30:31]
;
; GFX8-LABEL: v_trunc_i128_to_i16:
; GFX8: ; %bb.0:
; GFX8-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX8-NEXT: s_setpc_b64 s[30:31]
%trunc = trunc i128 %src to i16
ret i16 %trunc
}
define i32 @v_trunc_v2i32_to_v2i16(<2 x i32> %src) {
; GFX7-LABEL: v_trunc_v2i32_to_v2i16:
; GFX7: ; %bb.0:
; GFX7-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX7-NEXT: v_lshlrev_b32_e32 v1, 16, v1
; GFX7-NEXT: v_and_b32_e32 v0, 0xffff, v0
; GFX7-NEXT: v_or_b32_e32 v0, v1, v0
; GFX7-NEXT: s_setpc_b64 s[30:31]
;
; GFX8-LABEL: v_trunc_v2i32_to_v2i16:
; GFX8: ; %bb.0:
; GFX8-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX8-NEXT: v_mov_b32_sdwa v0, v1 dst_sel:WORD_1 dst_unused:UNUSED_PRESERVE src0_sel:WORD_0
; GFX8-NEXT: s_setpc_b64 s[30:31]
%trunc = trunc <2 x i32> %src to <2 x i16>
%cast = bitcast <2 x i16> %trunc to i32
ret i32 %cast
}
define amdgpu_ps i32 @s_trunc_v2i32_to_v2i16(<2 x i32> inreg %src) {
; GFX7-LABEL: s_trunc_v2i32_to_v2i16:
; GFX7: ; %bb.0:
; GFX7-NEXT: s_lshl_b32 s1, s1, 16
; GFX7-NEXT: s_and_b32 s0, s0, 0xffff
; GFX7-NEXT: s_or_b32 s0, s1, s0
; GFX7-NEXT: ; return to shader part epilog
;
; GFX8-LABEL: s_trunc_v2i32_to_v2i16:
; GFX8: ; %bb.0:
; GFX8-NEXT: s_lshl_b32 s1, s1, 16
; GFX8-NEXT: s_and_b32 s0, s0, 0xffff
; GFX8-NEXT: s_or_b32 s0, s1, s0
; GFX8-NEXT: ; return to shader part epilog
%trunc = trunc <2 x i32> %src to <2 x i16>
%cast = bitcast <2 x i16> %trunc to i32
ret i32 %cast
}
; ; FIXME: G_INSERT mishandled
; define <2 x i32> @v_trunc_v3i32_to_v3i16(<3 x i32> %src) {
; %trunc = trunc <3 x i32> %src to <3 x i16>
; %ext = shufflevector <3 x i16> %trunc, <3 x i16> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; %cast = bitcast <4 x i16> %ext to <2 x i32>
; ret <2 x i32> %cast
; }
; ; FIXME: G_INSERT mishandled
; define amdgpu_ps <2 x i32> @s_trunc_v3i32_to_v3i16(<3 x i32> inreg %src) {
; %trunc = trunc <3 x i32> %src to <3 x i16>
; %ext = shufflevector <3 x i16> %trunc, <3 x i16> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
; %cast = bitcast <4 x i16> %ext to <2 x i32>
; ret <2 x i32> %cast
; }
define <2 x i32> @v_trunc_v4i32_to_v4i16(<4 x i32> %src) {
; GFX7-LABEL: v_trunc_v4i32_to_v4i16:
; GFX7: ; %bb.0:
; GFX7-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX7-NEXT: v_mov_b32_e32 v4, 0xffff
; GFX7-NEXT: v_lshlrev_b32_e32 v1, 16, v1
; GFX7-NEXT: v_and_b32_e32 v0, v0, v4
; GFX7-NEXT: v_or_b32_e32 v0, v1, v0
; GFX7-NEXT: v_lshlrev_b32_e32 v1, 16, v3
; GFX7-NEXT: v_and_b32_e32 v2, v2, v4
; GFX7-NEXT: v_or_b32_e32 v1, v1, v2
; GFX7-NEXT: s_setpc_b64 s[30:31]
;
; GFX8-LABEL: v_trunc_v4i32_to_v4i16:
; GFX8: ; %bb.0:
; GFX8-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX8-NEXT: v_mov_b32_sdwa v2, v3 dst_sel:WORD_1 dst_unused:UNUSED_PRESERVE src0_sel:WORD_0
; GFX8-NEXT: v_mov_b32_sdwa v0, v1 dst_sel:WORD_1 dst_unused:UNUSED_PRESERVE src0_sel:WORD_0
; GFX8-NEXT: v_mov_b32_e32 v1, v2
; GFX8-NEXT: s_setpc_b64 s[30:31]
%trunc = trunc <4 x i32> %src to <4 x i16>
%cast = bitcast <4 x i16> %trunc to <2 x i32>
ret <2 x i32> %cast
}
define amdgpu_ps <2 x i32> @s_trunc_v4i32_to_v4i16(<4 x i32> inreg %src) {
; GFX7-LABEL: s_trunc_v4i32_to_v4i16:
; GFX7: ; %bb.0:
; GFX7-NEXT: s_mov_b32 s4, 0xffff
; GFX7-NEXT: s_lshl_b32 s1, s1, 16
; GFX7-NEXT: s_and_b32 s0, s0, s4
; GFX7-NEXT: s_or_b32 s0, s1, s0
; GFX7-NEXT: s_lshl_b32 s1, s3, 16
; GFX7-NEXT: s_and_b32 s2, s2, s4
; GFX7-NEXT: s_or_b32 s1, s1, s2
; GFX7-NEXT: ; return to shader part epilog
;
; GFX8-LABEL: s_trunc_v4i32_to_v4i16:
; GFX8: ; %bb.0:
; GFX8-NEXT: s_mov_b32 s4, 0xffff
; GFX8-NEXT: s_lshl_b32 s1, s1, 16
; GFX8-NEXT: s_and_b32 s0, s0, s4
; GFX8-NEXT: s_or_b32 s0, s1, s0
; GFX8-NEXT: s_lshl_b32 s1, s3, 16
; GFX8-NEXT: s_and_b32 s2, s2, s4
; GFX8-NEXT: s_or_b32 s1, s1, s2
; GFX8-NEXT: ; return to shader part epilog
%trunc = trunc <4 x i32> %src to <4 x i16>
%cast = bitcast <4 x i16> %trunc to <2 x i32>
ret <2 x i32> %cast
}
|
