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
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
|
//===- llvm/lib/Target/X86/X86CallLowering.cpp - Call lowering ------------===//
//
// 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
//
//===----------------------------------------------------------------------===//
//
/// \file
/// This file implements the lowering of LLVM calls to machine code calls for
/// GlobalISel.
//
//===----------------------------------------------------------------------===//
#include "X86CallLowering.h"
#include "X86CallingConv.h"
#include "X86ISelLowering.h"
#include "X86InstrInfo.h"
#include "X86RegisterInfo.h"
#include "X86Subtarget.h"
#include "llvm/ADT/ArrayRef.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/CodeGen/Analysis.h"
#include "llvm/CodeGen/CallingConvLower.h"
#include "llvm/CodeGen/GlobalISel/MachineIRBuilder.h"
#include "llvm/CodeGen/GlobalISel/Utils.h"
#include "llvm/CodeGen/LowLevelType.h"
#include "llvm/CodeGen/MachineBasicBlock.h"
#include "llvm/CodeGen/MachineFrameInfo.h"
#include "llvm/CodeGen/MachineFunction.h"
#include "llvm/CodeGen/MachineInstrBuilder.h"
#include "llvm/CodeGen/MachineMemOperand.h"
#include "llvm/CodeGen/MachineOperand.h"
#include "llvm/CodeGen/MachineRegisterInfo.h"
#include "llvm/CodeGen/TargetInstrInfo.h"
#include "llvm/CodeGen/TargetSubtargetInfo.h"
#include "llvm/CodeGen/ValueTypes.h"
#include "llvm/IR/Attributes.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/Value.h"
#include "llvm/MC/MCRegisterInfo.h"
#include "llvm/Support/LowLevelTypeImpl.h"
#include "llvm/Support/MachineValueType.h"
#include <cassert>
#include <cstdint>
using namespace llvm;
X86CallLowering::X86CallLowering(const X86TargetLowering &TLI)
: CallLowering(&TLI) {}
namespace {
struct X86OutgoingValueAssigner : public CallLowering::OutgoingValueAssigner {
private:
uint64_t StackSize = 0;
unsigned NumXMMRegs = 0;
public:
uint64_t getStackSize() { return StackSize; }
unsigned getNumXmmRegs() { return NumXMMRegs; }
X86OutgoingValueAssigner(CCAssignFn *AssignFn_)
: CallLowering::OutgoingValueAssigner(AssignFn_) {}
bool assignArg(unsigned ValNo, EVT OrigVT, MVT ValVT, MVT LocVT,
CCValAssign::LocInfo LocInfo,
const CallLowering::ArgInfo &Info, ISD::ArgFlagsTy Flags,
CCState &State) override {
bool Res = AssignFn(ValNo, ValVT, LocVT, LocInfo, Flags, State);
StackSize = State.getNextStackOffset();
static const MCPhysReg XMMArgRegs[] = {X86::XMM0, X86::XMM1, X86::XMM2,
X86::XMM3, X86::XMM4, X86::XMM5,
X86::XMM6, X86::XMM7};
if (!Info.IsFixed)
NumXMMRegs = State.getFirstUnallocated(XMMArgRegs);
return Res;
}
};
struct X86OutgoingValueHandler : public CallLowering::OutgoingValueHandler {
X86OutgoingValueHandler(MachineIRBuilder &MIRBuilder,
MachineRegisterInfo &MRI, MachineInstrBuilder &MIB)
: OutgoingValueHandler(MIRBuilder, MRI), MIB(MIB),
DL(MIRBuilder.getMF().getDataLayout()),
STI(MIRBuilder.getMF().getSubtarget<X86Subtarget>()) {}
Register getStackAddress(uint64_t Size, int64_t Offset,
MachinePointerInfo &MPO,
ISD::ArgFlagsTy Flags) override {
LLT p0 = LLT::pointer(0, DL.getPointerSizeInBits(0));
LLT SType = LLT::scalar(DL.getPointerSizeInBits(0));
auto SPReg =
MIRBuilder.buildCopy(p0, STI.getRegisterInfo()->getStackRegister());
auto OffsetReg = MIRBuilder.buildConstant(SType, Offset);
auto AddrReg = MIRBuilder.buildPtrAdd(p0, SPReg, OffsetReg);
MPO = MachinePointerInfo::getStack(MIRBuilder.getMF(), Offset);
return AddrReg.getReg(0);
}
void assignValueToReg(Register ValVReg, Register PhysReg,
CCValAssign VA) override {
MIB.addUse(PhysReg, RegState::Implicit);
Register ExtReg = extendRegister(ValVReg, VA);
MIRBuilder.buildCopy(PhysReg, ExtReg);
}
void assignValueToAddress(Register ValVReg, Register Addr, LLT MemTy,
MachinePointerInfo &MPO, CCValAssign &VA) override {
MachineFunction &MF = MIRBuilder.getMF();
Register ExtReg = extendRegister(ValVReg, VA);
auto *MMO = MF.getMachineMemOperand(MPO, MachineMemOperand::MOStore, MemTy,
inferAlignFromPtrInfo(MF, MPO));
MIRBuilder.buildStore(ExtReg, Addr, *MMO);
}
protected:
MachineInstrBuilder &MIB;
const DataLayout &DL;
const X86Subtarget &STI;
};
} // end anonymous namespace
bool X86CallLowering::lowerReturn(MachineIRBuilder &MIRBuilder,
const Value *Val, ArrayRef<Register> VRegs,
FunctionLoweringInfo &FLI) const {
assert(((Val && !VRegs.empty()) || (!Val && VRegs.empty())) &&
"Return value without a vreg");
auto MIB = MIRBuilder.buildInstrNoInsert(X86::RET).addImm(0);
if (!VRegs.empty()) {
MachineFunction &MF = MIRBuilder.getMF();
const Function &F = MF.getFunction();
MachineRegisterInfo &MRI = MF.getRegInfo();
const DataLayout &DL = MF.getDataLayout();
ArgInfo OrigRetInfo(VRegs, Val->getType(), 0);
setArgFlags(OrigRetInfo, AttributeList::ReturnIndex, DL, F);
SmallVector<ArgInfo, 4> SplitRetInfos;
splitToValueTypes(OrigRetInfo, SplitRetInfos, DL, F.getCallingConv());
X86OutgoingValueAssigner Assigner(RetCC_X86);
X86OutgoingValueHandler Handler(MIRBuilder, MRI, MIB);
if (!determineAndHandleAssignments(Handler, Assigner, SplitRetInfos,
MIRBuilder, F.getCallingConv(),
F.isVarArg()))
return false;
}
MIRBuilder.insertInstr(MIB);
return true;
}
namespace {
struct X86IncomingValueHandler : public CallLowering::IncomingValueHandler {
X86IncomingValueHandler(MachineIRBuilder &MIRBuilder,
MachineRegisterInfo &MRI)
: IncomingValueHandler(MIRBuilder, MRI),
DL(MIRBuilder.getMF().getDataLayout()) {}
Register getStackAddress(uint64_t Size, int64_t Offset,
MachinePointerInfo &MPO,
ISD::ArgFlagsTy Flags) override {
auto &MFI = MIRBuilder.getMF().getFrameInfo();
// Byval is assumed to be writable memory, but other stack passed arguments
// are not.
const bool IsImmutable = !Flags.isByVal();
int FI = MFI.CreateFixedObject(Size, Offset, IsImmutable);
MPO = MachinePointerInfo::getFixedStack(MIRBuilder.getMF(), FI);
return MIRBuilder
.buildFrameIndex(LLT::pointer(0, DL.getPointerSizeInBits(0)), FI)
.getReg(0);
}
void assignValueToAddress(Register ValVReg, Register Addr, LLT MemTy,
MachinePointerInfo &MPO, CCValAssign &VA) override {
MachineFunction &MF = MIRBuilder.getMF();
auto *MMO = MF.getMachineMemOperand(
MPO, MachineMemOperand::MOLoad | MachineMemOperand::MOInvariant, MemTy,
inferAlignFromPtrInfo(MF, MPO));
MIRBuilder.buildLoad(ValVReg, Addr, *MMO);
}
void assignValueToReg(Register ValVReg, Register PhysReg,
CCValAssign VA) override {
markPhysRegUsed(PhysReg);
IncomingValueHandler::assignValueToReg(ValVReg, PhysReg, VA);
}
/// How the physical register gets marked varies between formal
/// parameters (it's a basic-block live-in), and a call instruction
/// (it's an implicit-def of the BL).
virtual void markPhysRegUsed(unsigned PhysReg) = 0;
protected:
const DataLayout &DL;
};
struct FormalArgHandler : public X86IncomingValueHandler {
FormalArgHandler(MachineIRBuilder &MIRBuilder, MachineRegisterInfo &MRI)
: X86IncomingValueHandler(MIRBuilder, MRI) {}
void markPhysRegUsed(unsigned PhysReg) override {
MIRBuilder.getMRI()->addLiveIn(PhysReg);
MIRBuilder.getMBB().addLiveIn(PhysReg);
}
};
struct CallReturnHandler : public X86IncomingValueHandler {
CallReturnHandler(MachineIRBuilder &MIRBuilder, MachineRegisterInfo &MRI,
MachineInstrBuilder &MIB)
: X86IncomingValueHandler(MIRBuilder, MRI), MIB(MIB) {}
void markPhysRegUsed(unsigned PhysReg) override {
MIB.addDef(PhysReg, RegState::Implicit);
}
protected:
MachineInstrBuilder &MIB;
};
} // end anonymous namespace
bool X86CallLowering::lowerFormalArguments(MachineIRBuilder &MIRBuilder,
const Function &F,
ArrayRef<ArrayRef<Register>> VRegs,
FunctionLoweringInfo &FLI) const {
if (F.arg_empty())
return true;
// TODO: handle variadic function
if (F.isVarArg())
return false;
MachineFunction &MF = MIRBuilder.getMF();
MachineRegisterInfo &MRI = MF.getRegInfo();
auto DL = MF.getDataLayout();
SmallVector<ArgInfo, 8> SplitArgs;
unsigned Idx = 0;
for (const auto &Arg : F.args()) {
// TODO: handle not simple cases.
if (Arg.hasAttribute(Attribute::ByVal) ||
Arg.hasAttribute(Attribute::InReg) ||
Arg.hasAttribute(Attribute::StructRet) ||
Arg.hasAttribute(Attribute::SwiftSelf) ||
Arg.hasAttribute(Attribute::SwiftError) ||
Arg.hasAttribute(Attribute::Nest) || VRegs[Idx].size() > 1)
return false;
ArgInfo OrigArg(VRegs[Idx], Arg.getType(), Idx);
setArgFlags(OrigArg, Idx + AttributeList::FirstArgIndex, DL, F);
splitToValueTypes(OrigArg, SplitArgs, DL, F.getCallingConv());
Idx++;
}
MachineBasicBlock &MBB = MIRBuilder.getMBB();
if (!MBB.empty())
MIRBuilder.setInstr(*MBB.begin());
X86OutgoingValueAssigner Assigner(CC_X86);
FormalArgHandler Handler(MIRBuilder, MRI);
if (!determineAndHandleAssignments(Handler, Assigner, SplitArgs, MIRBuilder,
F.getCallingConv(), F.isVarArg()))
return false;
// Move back to the end of the basic block.
MIRBuilder.setMBB(MBB);
return true;
}
bool X86CallLowering::lowerCall(MachineIRBuilder &MIRBuilder,
CallLoweringInfo &Info) const {
MachineFunction &MF = MIRBuilder.getMF();
const Function &F = MF.getFunction();
MachineRegisterInfo &MRI = MF.getRegInfo();
const DataLayout &DL = F.getParent()->getDataLayout();
const X86Subtarget &STI = MF.getSubtarget<X86Subtarget>();
const TargetInstrInfo &TII = *STI.getInstrInfo();
const X86RegisterInfo *TRI = STI.getRegisterInfo();
// Handle only Linux C, X86_64_SysV calling conventions for now.
if (!STI.isTargetLinux() || !(Info.CallConv == CallingConv::C ||
Info.CallConv == CallingConv::X86_64_SysV))
return false;
unsigned AdjStackDown = TII.getCallFrameSetupOpcode();
auto CallSeqStart = MIRBuilder.buildInstr(AdjStackDown);
// Create a temporarily-floating call instruction so we can add the implicit
// uses of arg registers.
bool Is64Bit = STI.is64Bit();
unsigned CallOpc = Info.Callee.isReg()
? (Is64Bit ? X86::CALL64r : X86::CALL32r)
: (Is64Bit ? X86::CALL64pcrel32 : X86::CALLpcrel32);
auto MIB = MIRBuilder.buildInstrNoInsert(CallOpc)
.add(Info.Callee)
.addRegMask(TRI->getCallPreservedMask(MF, Info.CallConv));
SmallVector<ArgInfo, 8> SplitArgs;
for (const auto &OrigArg : Info.OrigArgs) {
// TODO: handle not simple cases.
if (OrigArg.Flags[0].isByVal())
return false;
if (OrigArg.Regs.size() > 1)
return false;
splitToValueTypes(OrigArg, SplitArgs, DL, Info.CallConv);
}
// Do the actual argument marshalling.
X86OutgoingValueAssigner Assigner(CC_X86);
X86OutgoingValueHandler Handler(MIRBuilder, MRI, MIB);
if (!determineAndHandleAssignments(Handler, Assigner, SplitArgs, MIRBuilder,
Info.CallConv, Info.IsVarArg))
return false;
bool IsFixed = Info.OrigArgs.empty() ? true : Info.OrigArgs.back().IsFixed;
if (STI.is64Bit() && !IsFixed && !STI.isCallingConvWin64(Info.CallConv)) {
// From AMD64 ABI document:
// For calls that may call functions that use varargs or stdargs
// (prototype-less calls or calls to functions containing ellipsis (...) in
// the declaration) %al is used as hidden argument to specify the number
// of SSE registers used. The contents of %al do not need to match exactly
// the number of registers, but must be an ubound on the number of SSE
// registers used and is in the range 0 - 8 inclusive.
MIRBuilder.buildInstr(X86::MOV8ri)
.addDef(X86::AL)
.addImm(Assigner.getNumXmmRegs());
MIB.addUse(X86::AL, RegState::Implicit);
}
// Now we can add the actual call instruction to the correct basic block.
MIRBuilder.insertInstr(MIB);
// If Callee is a reg, since it is used by a target specific
// instruction, it must have a register class matching the
// constraint of that instruction.
if (Info.Callee.isReg())
MIB->getOperand(0).setReg(constrainOperandRegClass(
MF, *TRI, MRI, *MF.getSubtarget().getInstrInfo(),
*MF.getSubtarget().getRegBankInfo(), *MIB, MIB->getDesc(), Info.Callee,
0));
// Finally we can copy the returned value back into its virtual-register. In
// symmetry with the arguments, the physical register must be an
// implicit-define of the call instruction.
if (!Info.OrigRet.Ty->isVoidTy()) {
if (Info.OrigRet.Regs.size() > 1)
return false;
SplitArgs.clear();
SmallVector<Register, 8> NewRegs;
splitToValueTypes(Info.OrigRet, SplitArgs, DL, Info.CallConv);
X86OutgoingValueAssigner Assigner(RetCC_X86);
CallReturnHandler Handler(MIRBuilder, MRI, MIB);
if (!determineAndHandleAssignments(Handler, Assigner, SplitArgs, MIRBuilder,
Info.CallConv, Info.IsVarArg))
return false;
if (!NewRegs.empty())
MIRBuilder.buildMerge(Info.OrigRet.Regs[0], NewRegs);
}
CallSeqStart.addImm(Assigner.getStackSize())
.addImm(0 /* see getFrameTotalSize */)
.addImm(0 /* see getFrameAdjustment */);
unsigned AdjStackUp = TII.getCallFrameDestroyOpcode();
MIRBuilder.buildInstr(AdjStackUp)
.addImm(Assigner.getStackSize())
.addImm(0 /* NumBytesForCalleeToPop */);
return true;
}
|