AMDGPUArgumentUsageInfo.cpp
5.32 KB
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
//===----------------------------------------------------------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
#include "AMDGPU.h"
#include "AMDGPUArgumentUsageInfo.h"
#include "SIRegisterInfo.h"
#include "llvm/Support/NativeFormatting.h"
#include "llvm/Support/raw_ostream.h"
using namespace llvm;
#define DEBUG_TYPE "amdgpu-argument-reg-usage-info"
INITIALIZE_PASS(AMDGPUArgumentUsageInfo, DEBUG_TYPE,
"Argument Register Usage Information Storage", false, true)
void ArgDescriptor::print(raw_ostream &OS,
const TargetRegisterInfo *TRI) const {
if (!isSet()) {
OS << "<not set>\n";
return;
}
if (isRegister())
OS << "Reg " << printReg(getRegister(), TRI);
else
OS << "Stack offset " << getStackOffset();
if (isMasked()) {
OS << " & ";
llvm::write_hex(OS, Mask, llvm::HexPrintStyle::PrefixLower);
}
OS << '\n';
}
char AMDGPUArgumentUsageInfo::ID = 0;
const AMDGPUFunctionArgInfo AMDGPUArgumentUsageInfo::ExternFunctionInfo{};
bool AMDGPUArgumentUsageInfo::doInitialization(Module &M) {
return false;
}
bool AMDGPUArgumentUsageInfo::doFinalization(Module &M) {
ArgInfoMap.clear();
return false;
}
void AMDGPUArgumentUsageInfo::print(raw_ostream &OS, const Module *M) const {
for (const auto &FI : ArgInfoMap) {
OS << "Arguments for " << FI.first->getName() << '\n'
<< " PrivateSegmentBuffer: " << FI.second.PrivateSegmentBuffer
<< " DispatchPtr: " << FI.second.DispatchPtr
<< " QueuePtr: " << FI.second.QueuePtr
<< " KernargSegmentPtr: " << FI.second.KernargSegmentPtr
<< " DispatchID: " << FI.second.DispatchID
<< " FlatScratchInit: " << FI.second.FlatScratchInit
<< " PrivateSegmentSize: " << FI.second.PrivateSegmentSize
<< " WorkGroupIDX: " << FI.second.WorkGroupIDX
<< " WorkGroupIDY: " << FI.second.WorkGroupIDY
<< " WorkGroupIDZ: " << FI.second.WorkGroupIDZ
<< " WorkGroupInfo: " << FI.second.WorkGroupInfo
<< " PrivateSegmentWaveByteOffset: "
<< FI.second.PrivateSegmentWaveByteOffset
<< " ImplicitBufferPtr: " << FI.second.ImplicitBufferPtr
<< " ImplicitArgPtr: " << FI.second.ImplicitArgPtr
<< " WorkItemIDX " << FI.second.WorkItemIDX
<< " WorkItemIDY " << FI.second.WorkItemIDY
<< " WorkItemIDZ " << FI.second.WorkItemIDZ
<< '\n';
}
}
std::pair<const ArgDescriptor *, const TargetRegisterClass *>
AMDGPUFunctionArgInfo::getPreloadedValue(
AMDGPUFunctionArgInfo::PreloadedValue Value) const {
switch (Value) {
case AMDGPUFunctionArgInfo::PRIVATE_SEGMENT_BUFFER: {
return std::make_pair(
PrivateSegmentBuffer ? &PrivateSegmentBuffer : nullptr,
&AMDGPU::SGPR_128RegClass);
}
case AMDGPUFunctionArgInfo::IMPLICIT_BUFFER_PTR:
return std::make_pair(ImplicitBufferPtr ? &ImplicitBufferPtr : nullptr,
&AMDGPU::SGPR_64RegClass);
case AMDGPUFunctionArgInfo::WORKGROUP_ID_X:
return std::make_pair(WorkGroupIDX ? &WorkGroupIDX : nullptr,
&AMDGPU::SGPR_32RegClass);
case AMDGPUFunctionArgInfo::WORKGROUP_ID_Y:
return std::make_pair(WorkGroupIDY ? &WorkGroupIDY : nullptr,
&AMDGPU::SGPR_32RegClass);
case AMDGPUFunctionArgInfo::WORKGROUP_ID_Z:
return std::make_pair(WorkGroupIDZ ? &WorkGroupIDZ : nullptr,
&AMDGPU::SGPR_32RegClass);
case AMDGPUFunctionArgInfo::PRIVATE_SEGMENT_WAVE_BYTE_OFFSET:
return std::make_pair(
PrivateSegmentWaveByteOffset ? &PrivateSegmentWaveByteOffset : nullptr,
&AMDGPU::SGPR_32RegClass);
case AMDGPUFunctionArgInfo::KERNARG_SEGMENT_PTR:
return std::make_pair(KernargSegmentPtr ? &KernargSegmentPtr : nullptr,
&AMDGPU::SGPR_64RegClass);
case AMDGPUFunctionArgInfo::IMPLICIT_ARG_PTR:
return std::make_pair(ImplicitArgPtr ? &ImplicitArgPtr : nullptr,
&AMDGPU::SGPR_64RegClass);
case AMDGPUFunctionArgInfo::DISPATCH_ID:
return std::make_pair(DispatchID ? &DispatchID : nullptr,
&AMDGPU::SGPR_64RegClass);
case AMDGPUFunctionArgInfo::FLAT_SCRATCH_INIT:
return std::make_pair(FlatScratchInit ? &FlatScratchInit : nullptr,
&AMDGPU::SGPR_64RegClass);
case AMDGPUFunctionArgInfo::DISPATCH_PTR:
return std::make_pair(DispatchPtr ? &DispatchPtr : nullptr,
&AMDGPU::SGPR_64RegClass);
case AMDGPUFunctionArgInfo::QUEUE_PTR:
return std::make_pair(QueuePtr ? &QueuePtr : nullptr,
&AMDGPU::SGPR_64RegClass);
case AMDGPUFunctionArgInfo::WORKITEM_ID_X:
return std::make_pair(WorkItemIDX ? &WorkItemIDX : nullptr,
&AMDGPU::VGPR_32RegClass);
case AMDGPUFunctionArgInfo::WORKITEM_ID_Y:
return std::make_pair(WorkItemIDY ? &WorkItemIDY : nullptr,
&AMDGPU::VGPR_32RegClass);
case AMDGPUFunctionArgInfo::WORKITEM_ID_Z:
return std::make_pair(WorkItemIDZ ? &WorkItemIDZ : nullptr,
&AMDGPU::VGPR_32RegClass);
}
llvm_unreachable("unexpected preloaded value type");
}