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
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=amdgcn -mcpu=gfx1200 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX12 %s
define amdgpu_cs float @test_cvt_f32_bf8_byte0(i32 %a) {
; GFX12-LABEL: test_cvt_f32_bf8_byte0:
; GFX12: ; %bb.0:
; GFX12-NEXT: v_cvt_f32_bf8_dpp v0, v0 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf bound_ctrl:1
; GFX12-NEXT: ; return to shader part epilog
%tmp0 = call i32 @llvm.amdgcn.mov.dpp.i32(i32 %a, i32 228, i32 15, i32 15, i1 1)
%ret = tail call float @llvm.amdgcn.cvt.f32.bf8(i32 %tmp0, i32 0)
ret float %ret
}
define amdgpu_cs float @test_cvt_f32_bf8_byte1(i32 %a) {
; GFX12-LABEL: test_cvt_f32_bf8_byte1:
; GFX12: ; %bb.0:
; GFX12-NEXT: v_cvt_f32_bf8_e64_dpp v0, v0 byte_sel:1 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf bound_ctrl:1
; GFX12-NEXT: ; return to shader part epilog
%tmp0 = call i32 @llvm.amdgcn.mov.dpp.i32(i32 %a, i32 228, i32 15, i32 15, i1 1)
%ret = tail call float @llvm.amdgcn.cvt.f32.bf8(i32 %tmp0, i32 1)
ret float %ret
}
define amdgpu_cs float @test_cvt_f32_bf8_byte2(i32 %a) {
; GFX12-LABEL: test_cvt_f32_bf8_byte2:
; GFX12: ; %bb.0:
; GFX12-NEXT: v_cvt_f32_bf8_e64_dpp v0, v0 byte_sel:2 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf bound_ctrl:1
; GFX12-NEXT: ; return to shader part epilog
%tmp0 = call i32 @llvm.amdgcn.mov.dpp.i32(i32 %a, i32 228, i32 15, i32 15, i1 1)
%ret = tail call float @llvm.amdgcn.cvt.f32.bf8(i32 %tmp0, i32 2)
ret float %ret
}
define amdgpu_cs float @test_cvt_f32_fp8_byte3(i32 %a) {
; GFX12-LABEL: test_cvt_f32_fp8_byte3:
; GFX12: ; %bb.0:
; GFX12-NEXT: v_cvt_f32_fp8_e64_dpp v0, v0 byte_sel:3 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf bound_ctrl:1
; GFX12-NEXT: ; return to shader part epilog
%tmp0 = call i32 @llvm.amdgcn.mov.dpp.i32(i32 %a, i32 228, i32 15, i32 15, i1 1)
%ret = tail call float @llvm.amdgcn.cvt.f32.fp8(i32 %tmp0, i32 3)
ret float %ret
}
define amdgpu_cs void @test_cvt_pk_bf8_f32_word0(i32 %a, float %y, i32 %old, ptr addrspace(1) %out) {
; GFX12-LABEL: test_cvt_pk_bf8_f32_word0:
; GFX12: ; %bb.0:
; GFX12-NEXT: v_cvt_pk_bf8_f32_e64_dpp v2, v0, v1 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf bound_ctrl:1
; GFX12-NEXT: global_store_b32 v[3:4], v2, off
; GFX12-NEXT: s_nop 0
; GFX12-NEXT: s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX12-NEXT: s_endpgm
%tmp0 = call i32 @llvm.amdgcn.mov.dpp.i32(i32 %a, i32 228, i32 15, i32 15, i1 1)
%tmp1 = bitcast i32 %tmp0 to float
%ret = tail call i32 @llvm.amdgcn.cvt.pk.bf8.f32(float %tmp1, float %y, i32 %old, i1 false)
store i32 %ret, ptr addrspace(1) %out
ret void
}
define amdgpu_cs void @test_cvt_pk_fp8_f32_word1(i32 %a, float %y, i32 %old, ptr addrspace(1) %out) {
; GFX12-LABEL: test_cvt_pk_fp8_f32_word1:
; GFX12: ; %bb.0:
; GFX12-NEXT: v_mov_b32_dpp v0, v0 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf bound_ctrl:1
; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1)
; GFX12-NEXT: v_cvt_pk_fp8_f32 v2, v0, v1 op_sel:[0,0,1]
; GFX12-NEXT: global_store_b32 v[3:4], v2, off
; GFX12-NEXT: s_nop 0
; GFX12-NEXT: s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX12-NEXT: s_endpgm
%tmp0 = call i32 @llvm.amdgcn.mov.dpp.i32(i32 %a, i32 228, i32 15, i32 15, i1 1)
%tmp1 = bitcast i32 %tmp0 to float
%ret = tail call i32 @llvm.amdgcn.cvt.pk.fp8.f32(float %tmp1, float %y, i32 %old, i1 true)
store i32 %ret, ptr addrspace(1) %out
ret void
}
define amdgpu_cs void @test_cvt_sr_bf8_f32_byte0(i32 %a, i32 %r, i32 %old, ptr addrspace(1) %out) {
; GFX12-LABEL: test_cvt_sr_bf8_f32_byte0:
; GFX12: ; %bb.0:
; GFX12-NEXT: v_cvt_sr_bf8_f32_e64_dpp v2, v0, v1 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf bound_ctrl:1
; GFX12-NEXT: global_store_b32 v[3:4], v2, off
; GFX12-NEXT: s_nop 0
; GFX12-NEXT: s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX12-NEXT: s_endpgm
%tmp0 = call i32 @llvm.amdgcn.mov.dpp.i32(i32 %a, i32 228, i32 15, i32 15, i1 1)
%tmp1 = bitcast i32 %tmp0 to float
%ret = tail call i32 @llvm.amdgcn.cvt.sr.bf8.f32(float %tmp1, i32 %r, i32 %old, i32 0)
store i32 %ret, ptr addrspace(1) %out
ret void
}
define amdgpu_cs void @test_cvt_sr_fp8_f32_byte1(i32 %a, i32 %r, i32 %old, ptr addrspace(1) %out) {
; GFX12-LABEL: test_cvt_sr_fp8_f32_byte1:
; GFX12: ; %bb.0:
; GFX12-NEXT: v_cvt_sr_fp8_f32_e64_dpp v2, v0, v1 byte_sel:1 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf bound_ctrl:1
; GFX12-NEXT: global_store_b32 v[3:4], v2, off
; GFX12-NEXT: s_nop 0
; GFX12-NEXT: s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX12-NEXT: s_endpgm
%tmp0 = call i32 @llvm.amdgcn.mov.dpp.i32(i32 %a, i32 228, i32 15, i32 15, i1 1)
%tmp1 = bitcast i32 %tmp0 to float
%ret = tail call i32 @llvm.amdgcn.cvt.sr.fp8.f32(float %tmp1, i32 %r, i32 %old, i32 1)
store i32 %ret, ptr addrspace(1) %out
ret void
}
define amdgpu_cs void @test_cvt_sr_fp8_f32_byte2(i32 %a, i32 %r, i32 %old, ptr addrspace(1) %out) {
; GFX12-LABEL: test_cvt_sr_fp8_f32_byte2:
; GFX12: ; %bb.0:
; GFX12-NEXT: v_cvt_sr_fp8_f32_e64_dpp v2, v0, v1 byte_sel:2 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf bound_ctrl:1
; GFX12-NEXT: global_store_b32 v[3:4], v2, off
; GFX12-NEXT: s_nop 0
; GFX12-NEXT: s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX12-NEXT: s_endpgm
%tmp0 = call i32 @llvm.amdgcn.mov.dpp.i32(i32 %a, i32 228, i32 15, i32 15, i1 1)
%tmp1 = bitcast i32 %tmp0 to float
%ret = tail call i32 @llvm.amdgcn.cvt.sr.fp8.f32(float %tmp1, i32 %r, i32 %old, i32 2)
store i32 %ret, ptr addrspace(1) %out
ret void
}
declare float @llvm.amdgcn.cvt.f32.bf8(i32, i32)
declare float @llvm.amdgcn.cvt.f32.fp8(i32, i32)
declare i32 @llvm.amdgcn.cvt.pk.bf8.f32(float, float, i32, i1)
declare i32 @llvm.amdgcn.cvt.pk.fp8.f32(float, float, i32, i1)
declare i32 @llvm.amdgcn.cvt.sr.bf8.f32(float, i32, i32, i32)
declare i32 @llvm.amdgcn.cvt.sr.fp8.f32(float, i32, i32, i32)
declare i32 @llvm.amdgcn.mov.dpp.i32(i32, i32, i32, i32, i1) #1
declare i32 @llvm.amdgcn.mov.dpp8.i32(i32, i32) #1
attributes #0 = { nounwind convergent }
attributes #1 = { nounwind readnone convergent }
|
