1 //===- CodeEmitterGen.cpp - Code Emitter Generator ------------------------===//
2 //
3 //                     The LLVM Compiler Infrastructure
4 //
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
7 //
8 //===----------------------------------------------------------------------===//
9 //
10 // CodeEmitterGen uses the descriptions of instructions and their fields to
11 // construct an automated code emitter: a function that, given a MachineInstr,
12 // returns the (currently, 32-bit unsigned) value of the instruction.
13 //
14 //===----------------------------------------------------------------------===//
15 
16 #include "CodeGenTarget.h"
17 #include "llvm/ADT/StringExtras.h"
18 #include "llvm/Support/CommandLine.h"
19 #include "llvm/Support/Debug.h"
20 #include "llvm/TableGen/Record.h"
21 #include "llvm/TableGen/TableGenBackend.h"
22 #include <map>
23 #include <string>
24 #include <vector>
25 using namespace llvm;
26 
27 // FIXME: Somewhat hackish to use a command line option for this. There should
28 // be a CodeEmitter class in the Target.td that controls this sort of thing
29 // instead.
30 static cl::opt<bool>
31 MCEmitter("mc-emitter",
32           cl::desc("Generate CodeEmitter for use with the MC library."),
33           cl::init(false));
34 
35 namespace {
36 
37 class CodeEmitterGen {
38   RecordKeeper &Records;
39 public:
40   CodeEmitterGen(RecordKeeper &R) : Records(R) {}
41 
42   void run(raw_ostream &o);
43 private:
44   void emitMachineOpEmitter(raw_ostream &o, const std::string &Namespace);
45   void emitGetValueBit(raw_ostream &o, const std::string &Namespace);
46   int getVariableBit(const std::string &VarName, BitsInit *BI, int bit);
47   std::string getInstructionCase(Record *R, CodeGenTarget &Target);
48   void AddCodeToMergeInOperand(Record *R, BitsInit *BI,
49                                const std::string &VarName,
50                                unsigned &NumberedOp,
51                                std::string &Case, CodeGenTarget &Target);
52 
53 };
54 
55 // If the VarBitInit at position 'bit' matches the specified variable then
56 // return the variable bit position.  Otherwise return -1.
57 int CodeEmitterGen::getVariableBit(const std::string &VarName,
58                                    BitsInit *BI, int bit) {
59   if (VarBitInit *VBI = dyn_cast<VarBitInit>(BI->getBit(bit))) {
60     if (VarInit *VI = dyn_cast<VarInit>(VBI->getBitVar()))
61       if (VI->getName() == VarName)
62         return VBI->getBitNum();
63   } else if (VarInit *VI = dyn_cast<VarInit>(BI->getBit(bit))) {
64     if (VI->getName() == VarName)
65       return 0;
66   }
67 
68   return -1;
69 }
70 
71 void CodeEmitterGen::
72 AddCodeToMergeInOperand(Record *R, BitsInit *BI, const std::string &VarName,
73                         unsigned &NumberedOp,
74                         std::string &Case, CodeGenTarget &Target) {
75   CodeGenInstruction &CGI = Target.getInstruction(R);
76 
77   // Determine if VarName actually contributes to the Inst encoding.
78   int bit = BI->getNumBits()-1;
79 
80   // Scan for a bit that this contributed to.
81   for (; bit >= 0; ) {
82     if (getVariableBit(VarName, BI, bit) != -1)
83       break;
84 
85     --bit;
86   }
87 
88   // If we found no bits, ignore this value, otherwise emit the call to get the
89   // operand encoding.
90   if (bit < 0) return;
91 
92   // If the operand matches by name, reference according to that
93   // operand number. Non-matching operands are assumed to be in
94   // order.
95   unsigned OpIdx;
96   if (CGI.Operands.hasOperandNamed(VarName, OpIdx)) {
97     // Get the machine operand number for the indicated operand.
98     OpIdx = CGI.Operands[OpIdx].MIOperandNo;
99     assert(!CGI.Operands.isFlatOperandNotEmitted(OpIdx) &&
100            "Explicitly used operand also marked as not emitted!");
101   } else {
102     unsigned NumberOps = CGI.Operands.size();
103     /// If this operand is not supposed to be emitted by the
104     /// generated emitter, skip it.
105     while (NumberedOp < NumberOps &&
106            CGI.Operands.isFlatOperandNotEmitted(NumberedOp))
107       ++NumberedOp;
108 
109     OpIdx = NumberedOp++;
110   }
111 
112   std::pair<unsigned, unsigned> SO = CGI.Operands.getSubOperandNumber(OpIdx);
113   std::string &EncoderMethodName = CGI.Operands[SO.first].EncoderMethodName;
114 
115   // If the source operand has a custom encoder, use it. This will
116   // get the encoding for all of the suboperands.
117   if (!EncoderMethodName.empty()) {
118     // A custom encoder has all of the information for the
119     // sub-operands, if there are more than one, so only
120     // query the encoder once per source operand.
121     if (SO.second == 0) {
122       Case += "      // op: " + VarName + "\n" +
123               "      op = " + EncoderMethodName + "(MI, " + utostr(OpIdx);
124       if (MCEmitter)
125         Case += ", Fixups";
126       Case += ");\n";
127     }
128   } else {
129     Case += "      // op: " + VarName + "\n" +
130       "      op = getMachineOpValue(MI, MI.getOperand(" + utostr(OpIdx) + ")";
131     if (MCEmitter)
132       Case += ", Fixups";
133     Case += ");\n";
134   }
135 
136   for (; bit >= 0; ) {
137     int varBit = getVariableBit(VarName, BI, bit);
138 
139     // If this bit isn't from a variable, skip it.
140     if (varBit == -1) {
141       --bit;
142       continue;
143     }
144 
145     // Figure out the consecutive range of bits covered by this operand, in
146     // order to generate better encoding code.
147     int beginInstBit = bit;
148     int beginVarBit = varBit;
149     int N = 1;
150     for (--bit; bit >= 0;) {
151       varBit = getVariableBit(VarName, BI, bit);
152       if (varBit == -1 || varBit != (beginVarBit - N)) break;
153       ++N;
154       --bit;
155     }
156 
157     uint64_t opMask = ~(uint64_t)0 >> (64-N);
158     int opShift = beginVarBit - N + 1;
159     opMask <<= opShift;
160     opShift = beginInstBit - beginVarBit;
161 
162     if (opShift > 0) {
163       Case += "      Value |= (op & UINT64_C(" + utostr(opMask) + ")) << " +
164               itostr(opShift) + ";\n";
165     } else if (opShift < 0) {
166       Case += "      Value |= (op & UINT64_C(" + utostr(opMask) + ")) >> " +
167               itostr(-opShift) + ";\n";
168     } else {
169       Case += "      Value |= op & UINT64_C(" + utostr(opMask) + ");\n";
170     }
171   }
172 }
173 
174 
175 std::string CodeEmitterGen::getInstructionCase(Record *R,
176                                                CodeGenTarget &Target) {
177   std::string Case;
178 
179   BitsInit *BI = R->getValueAsBitsInit("Inst");
180   const std::vector<RecordVal> &Vals = R->getValues();
181   unsigned NumberedOp = 0;
182 
183   // Loop over all of the fields in the instruction, determining which are the
184   // operands to the instruction.
185   for (unsigned i = 0, e = Vals.size(); i != e; ++i) {
186     // Ignore fixed fields in the record, we're looking for values like:
187     //    bits<5> RST = { ?, ?, ?, ?, ? };
188     if (Vals[i].getPrefix() || Vals[i].getValue()->isComplete())
189       continue;
190 
191     AddCodeToMergeInOperand(R, BI, Vals[i].getName(), NumberedOp, Case, Target);
192   }
193 
194   std::string PostEmitter = R->getValueAsString("PostEncoderMethod");
195   if (!PostEmitter.empty())
196     Case += "      Value = " + PostEmitter + "(MI, Value);\n";
197 
198   return Case;
199 }
200 
201 void CodeEmitterGen::run(raw_ostream &o) {
202   CodeGenTarget Target(Records);
203   std::vector<Record*> Insts = Records.getAllDerivedDefinitions("Instruction");
204 
205   // For little-endian instruction bit encodings, reverse the bit order
206   Target.reverseBitsForLittleEndianEncoding();
207 
208   const std::vector<const CodeGenInstruction*> &NumberedInstructions =
209     Target.getInstructionsByEnumValue();
210 
211   // Emit function declaration
212   o << "uint64_t " << Target.getName();
213   if (MCEmitter)
214     o << "MCCodeEmitter::getBinaryCodeForInstr(const MCInst &MI,\n"
215       << "    SmallVectorImpl<MCFixup> &Fixups) const {\n";
216   else
217     o << "CodeEmitter::getBinaryCodeForInstr(const MachineInstr &MI) const {\n";
218 
219   // Emit instruction base values
220   o << "  static const uint64_t InstBits[] = {\n";
221   for (std::vector<const CodeGenInstruction*>::const_iterator
222           IN = NumberedInstructions.begin(),
223           EN = NumberedInstructions.end();
224        IN != EN; ++IN) {
225     const CodeGenInstruction *CGI = *IN;
226     Record *R = CGI->TheDef;
227 
228     if (R->getValueAsString("Namespace") == "TargetOpcode" ||
229         R->getValueAsBit("isPseudo")) {
230       o << "    UINT64_C(0),\n";
231       continue;
232     }
233 
234     BitsInit *BI = R->getValueAsBitsInit("Inst");
235 
236     // Start by filling in fixed values.
237     uint64_t Value = 0;
238     for (unsigned i = 0, e = BI->getNumBits(); i != e; ++i) {
239       if (BitInit *B = dyn_cast<BitInit>(BI->getBit(e-i-1)))
240         Value |= (uint64_t)B->getValue() << (e-i-1);
241     }
242     o << "    UINT64_C(" << Value << ")," << '\t' << "// " << R->getName() << "\n";
243   }
244   o << "    UINT64_C(0)\n  };\n";
245 
246   // Map to accumulate all the cases.
247   std::map<std::string, std::vector<std::string> > CaseMap;
248 
249   // Construct all cases statement for each opcode
250   for (std::vector<Record*>::iterator IC = Insts.begin(), EC = Insts.end();
251         IC != EC; ++IC) {
252     Record *R = *IC;
253     if (R->getValueAsString("Namespace") == "TargetOpcode" ||
254         R->getValueAsBit("isPseudo"))
255       continue;
256     const std::string &InstName = R->getValueAsString("Namespace") + "::"
257       + R->getName();
258     std::string Case = getInstructionCase(R, Target);
259 
260     CaseMap[Case].push_back(InstName);
261   }
262 
263   // Emit initial function code
264   o << "  const unsigned opcode = MI.getOpcode();\n"
265     << "  uint64_t Value = InstBits[opcode];\n"
266     << "  uint64_t op = 0;\n"
267     << "  (void)op;  // suppress warning\n"
268     << "  switch (opcode) {\n";
269 
270   // Emit each case statement
271   std::map<std::string, std::vector<std::string> >::iterator IE, EE;
272   for (IE = CaseMap.begin(), EE = CaseMap.end(); IE != EE; ++IE) {
273     const std::string &Case = IE->first;
274     std::vector<std::string> &InstList = IE->second;
275 
276     for (int i = 0, N = InstList.size(); i < N; i++) {
277       if (i) o << "\n";
278       o << "    case " << InstList[i]  << ":";
279     }
280     o << " {\n";
281     o << Case;
282     o << "      break;\n"
283       << "    }\n";
284   }
285 
286   // Default case: unhandled opcode
287   o << "  default:\n"
288     << "    std::string msg;\n"
289     << "    raw_string_ostream Msg(msg);\n"
290     << "    Msg << \"Not supported instr: \" << MI;\n"
291     << "    report_fatal_error(Msg.str());\n"
292     << "  }\n"
293     << "  return Value;\n"
294     << "}\n\n";
295 }
296 
297 } // End anonymous namespace
298 
299 namespace llvm {
300 
301 void EmitCodeEmitter(RecordKeeper &RK, raw_ostream &OS) {
302   emitSourceFileHeader("Machine Code Emitter", OS);
303   CodeEmitterGen(RK).run(OS);
304 }
305 
306 } // End llvm namespace
307