File: llvm.amdgcn.trig.preop.ll

package info (click to toggle)
llvm-toolchain-16 1%3A16.0.6-15~deb12u1
  • links: PTS, VCS
  • area: main
  • in suites: bookworm
  • size: 1,634,792 kB
  • sloc: cpp: 6,179,261; ansic: 1,216,205; asm: 741,319; python: 196,614; objc: 75,325; f90: 49,640; lisp: 32,396; pascal: 12,286; sh: 9,394; perl: 7,442; ml: 5,494; awk: 3,523; makefile: 2,723; javascript: 1,206; xml: 886; fortran: 581; cs: 573
file content (138 lines) | stat: -rw-r--r-- 5,281 bytes parent folder | download | duplicates (2)
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
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=hawaii < %s | FileCheck -check-prefixes=GCN,CI %s
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=fiji < %s | FileCheck -check-prefixes=GCN,VI %s
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 < %s | FileCheck -check-prefixes=GCN,GFX9 %s
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 < %s | FileCheck -check-prefixes=GFX10PLUS,GFX10 %s
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 < %s | FileCheck -check-prefixes=GFX10PLUS,GFX11 %s

define double @v_trig_preop_f64(double %a, i32 %b) {
; GCN-LABEL: v_trig_preop_f64:
; GCN:       ; %bb.0:
; GCN-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GCN-NEXT:    v_trig_preop_f64 v[0:1], v[0:1], v2
; GCN-NEXT:    s_setpc_b64 s[30:31]
;
; GFX10PLUS-LABEL: v_trig_preop_f64:
; GFX10PLUS:       ; %bb.0:
; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX10PLUS-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10PLUS-NEXT:    v_trig_preop_f64 v[0:1], v[0:1], v2
; GFX10PLUS-NEXT:    s_setpc_b64 s[30:31]
  %result = call double @llvm.amdgcn.trig.preop.f64(double %a, i32 %b)
  ret double %result
}

define double @v_trig_preop_f64_imm(double %a, i32 %b) {
; GCN-LABEL: v_trig_preop_f64_imm:
; GCN:       ; %bb.0:
; GCN-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GCN-NEXT:    v_trig_preop_f64 v[0:1], v[0:1], 7
; GCN-NEXT:    s_setpc_b64 s[30:31]
;
; GFX10PLUS-LABEL: v_trig_preop_f64_imm:
; GFX10PLUS:       ; %bb.0:
; GFX10PLUS-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX10PLUS-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10PLUS-NEXT:    v_trig_preop_f64 v[0:1], v[0:1], 7
; GFX10PLUS-NEXT:    s_setpc_b64 s[30:31]
  %result = call double @llvm.amdgcn.trig.preop.f64(double %a, i32 7)
  ret double %result
}

define amdgpu_kernel void @s_trig_preop_f64(double %a, i32 %b) {
; CI-LABEL: s_trig_preop_f64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dword s2, s[4:5], 0x2
; CI-NEXT:    s_load_dwordx2 s[0:1], s[4:5], 0x0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_trig_preop_f64 v[0:1], s[0:1], v0
; CI-NEXT:    flat_store_dwordx2 v[0:1], v[0:1]
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    s_endpgm
;
; VI-LABEL: s_trig_preop_f64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dword s2, s[4:5], 0x8
; VI-NEXT:    s_load_dwordx2 s[0:1], s[4:5], 0x0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_trig_preop_f64 v[0:1], s[0:1], v0
; VI-NEXT:    flat_store_dwordx2 v[0:1], v[0:1]
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: s_trig_preop_f64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dword s2, s[4:5], 0x8
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[4:5], 0x0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v0, s2
; GFX9-NEXT:    v_trig_preop_f64 v[0:1], s[0:1], v0
; GFX9-NEXT:    flat_store_dwordx2 v[0:1], v[0:1]
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: s_trig_preop_f64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_clause 0x1
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[4:5], 0x0
; GFX10-NEXT:    s_load_dword s2, s[4:5], 0x8
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_trig_preop_f64 v[0:1], s[0:1], s2
; GFX10-NEXT:    flat_store_dwordx2 v[0:1], v[0:1]
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: s_trig_preop_f64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_clause 0x1
; GFX11-NEXT:    s_load_b64 s[2:3], s[0:1], 0x0
; GFX11-NEXT:    s_load_b32 s0, s[0:1], 0x8
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_trig_preop_f64 v[0:1], s[2:3], s0
; GFX11-NEXT:    flat_store_b64 v[0:1], v[0:1] dlc
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %result = call double @llvm.amdgcn.trig.preop.f64(double %a, i32 %b)
  store volatile double %result, ptr undef
  ret void
}

define amdgpu_kernel void @s_trig_preop_f64_imm(double %a, i32 %b) {
; GCN-LABEL: s_trig_preop_f64_imm:
; GCN:       ; %bb.0:
; GCN-NEXT:    s_load_dwordx2 s[0:1], s[4:5], 0x0
; GCN-NEXT:    s_waitcnt lgkmcnt(0)
; GCN-NEXT:    v_trig_preop_f64 v[0:1], s[0:1], 7
; GCN-NEXT:    flat_store_dwordx2 v[0:1], v[0:1]
; GCN-NEXT:    s_waitcnt vmcnt(0)
; GCN-NEXT:    s_endpgm
;
; GFX10-LABEL: s_trig_preop_f64_imm:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[4:5], 0x0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_trig_preop_f64 v[0:1], s[0:1], 7
; GFX10-NEXT:    flat_store_dwordx2 v[0:1], v[0:1]
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: s_trig_preop_f64_imm:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[0:1], 0x0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_trig_preop_f64 v[0:1], s[0:1], 7
; GFX11-NEXT:    flat_store_b64 v[0:1], v[0:1] dlc
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %result = call double @llvm.amdgcn.trig.preop.f64(double %a, i32 7)
  store volatile double %result, ptr undef
  ret void
}

declare double @llvm.amdgcn.trig.preop.f64(double, i32) #0

attributes #0 = { nounwind readnone speculatable }