codegen_llvm.cc 41.8 KB
Newer Older
1 2 3 4 5 6 7 8
/*
 * Licensed to the Apache Software Foundation (ASF) under one
 * or more contributor license agreements.  See the NOTICE file
 * distributed with this work for additional information
 * regarding copyright ownership.  The ASF licenses this file
 * to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance
 * with the License.  You may obtain a copy of the License at
9
 *
10
 *   http://www.apache.org/licenses/LICENSE-2.0
11
 *
12 13 14 15 16 17 18 19
 * Unless required by applicable law or agreed to in writing,
 * software distributed under the License is distributed on an
 * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
 * KIND, either express or implied.  See the License for the
 * specific language governing permissions and limitations
 * under the License.
 */

20 21 22 23 24
/*!
 *  Copyright (c) 2017 by Contributors
 * \file codegen_llvm.cc
 */
#ifdef TVM_LLVM_VERSION
25
// Part of the code are adapted from Halide's CodeGen_LLVM
26
#include <tvm/runtime/device_api.h>
27
#include <tvm/runtime/c_runtime_api.h>
28 29 30

#include <algorithm>

31 32
#include "codegen_llvm.h"
#include "codegen_cpu.h"
33
#include "../../pass/ir_util.h"
34 35 36 37 38
#include "../../arithmetic/compute_expr.h"

namespace tvm {
namespace codegen {

39 40 41 42 43 44 45 46
std::unique_ptr<CodeGenLLVM> CodeGenLLVM::Create(llvm::TargetMachine *tm) {
  std::string target = tm->getTarget().getName();
  std::string factory_name = "tvm.codegen.llvm.target_" + target;
  const PackedFunc* f = runtime::Registry::Get(factory_name);
  if (f != nullptr) {
    void* handle = (*f)();
    return std::unique_ptr<CodeGenLLVM>(static_cast<CodeGenLLVM*>(handle));
  } else {
47
    return std::unique_ptr<CodeGenLLVM>(new CodeGenCPU());
48 49 50
  }
}

51
void CodeGenLLVM::Init(const std::string& module_name,
52
                       llvm::TargetMachine* tm,
53
                       llvm::LLVMContext* ctx,
54 55
                       bool system_lib,
                       bool dynamic_lookup) {
56 57
  InitializeLLVM();
  ctx_ = ctx;
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74
  builder_.reset(new IRBuilder(*ctx_));
  module_.reset(new llvm::Module(module_name, *ctx_));
  md_builder_.reset(new llvm::MDBuilder(*ctx_));
  // types
  t_void_ = llvm::Type::getVoidTy(*ctx_);
  t_void_p_ = llvm::Type::getInt8Ty(*ctx_)->getPointerTo();
  t_int_ = llvm::Type::getInt32Ty(*ctx_);
  t_char_ = llvm::Type::getInt8Ty(*ctx_);
  t_int8_ = llvm::Type::getInt8Ty(*ctx_);
  t_int16_ = llvm::Type::getInt16Ty(*ctx_);
  t_int32_ = llvm::Type::getInt32Ty(*ctx_);
  t_int64_ = llvm::Type::getInt64Ty(*ctx_);
  t_float64_ = llvm::Type::getDoubleTy(*ctx_);
  // meta data
  md_very_likely_branch_ = md_builder_->createBranchWeights(1<<20, 1);
  md_tbaa_root_ = md_builder_->createTBAARoot("tvm-tbaa");
  md_tbaa_alias_set_ = md_builder_->createTBAANode("tvm-alias", md_tbaa_root_);
75
  this->InitTarget(tm);
76 77
}

78 79
void CodeGenLLVM::InitTarget(llvm::TargetMachine* tm) {
  module_->setTargetTriple(tm->getTargetTriple().str());
80 81
  module_->setDataLayout(tm->createDataLayout());
  data_layout_.reset(new llvm::DataLayout(module_.get()));
82
  target_machine_ = tm;
83 84 85 86 87 88 89 90 91 92 93 94 95
  if (native_vector_bits_ == 0) {
    const auto& arch = tm->getTargetTriple().getArch();
    if (arch == llvm::Triple::x86_64) {
      // for avx512
      native_vector_bits_ = 512;
    } else if (arch == llvm::Triple::x86) {
      native_vector_bits_ = 256;
    } else if (arch == llvm::Triple::arm || arch == llvm::Triple::aarch64) {
      native_vector_bits_ = 128;
    } else {
      native_vector_bits_ = 128;
      std::string arch_name = tm->getTargetTriple().getArchName();
      LOG(WARNING) << "Set native vector bits to be 128 for " << arch_name;
96
    }
97
  }
98 99
}

100 101 102 103
void CodeGenLLVM::AddFunction(const LoweredFunc& f) {
  this->AddFunctionInternal(f, false);
}

104
void CodeGenLLVM::InitFuncState() {
105
  var_map_.clear();
106
  alias_var_set_.clear();
107
  alloc_storage_info_.clear();
108
  volatile_buf_.clear();
109
  analyzer_.reset(new arith::Analyzer());
110 111 112
}

void CodeGenLLVM::AddFunctionInternal(const LoweredFunc& f, bool ret_void) {
113
  this->InitFuncState();
114
  std::vector<llvm::Type*> arg_types;
115
  is_restricted_ = f->is_restricted;
116 117
  for (Var arg : f->args) {
    Type t = arg.type();
118 119 120 121 122 123 124 125
    if (t.is_handle()) {
      auto it = f->handle_data_type.find(arg);
      if (it != f->handle_data_type.end()) {
        arg_types.push_back(LLVMType((*it).second.type())
                            ->getPointerTo(GetGlobalAddressSpace()));
      } else {
        arg_types.push_back(t_int8_->getPointerTo(GetGlobalAddressSpace()));
      }
126 127 128
      if (!is_restricted_) {
        alias_var_set_.insert(arg.get());
      }
129
    } else {
130
      arg_types.push_back(LLVMType(arg.type()));
131 132
    }
  }
133
  llvm::FunctionType* ftype = llvm::FunctionType::get(
134 135 136 137 138 139
      ret_void ? t_void_ : t_int_, arg_types, false);
  CHECK(module_->getFunction(f->name) == nullptr)
      << "Function " << f->name << " already exist in module";
  function_ = llvm::Function::Create(
      ftype, llvm::Function::ExternalLinkage,
      f->name, module_.get());
140
  function_->setCallingConv(llvm::CallingConv::C);
Hu Shiwen committed
141
  function_->setDLLStorageClass(llvm::GlobalValue::DLLStorageClassTypes::DLLExportStorageClass);
142 143 144 145 146 147 148 149 150
  // set var map and align information
  auto arg_it = function_->arg_begin();
  for (size_t i = 0; i < f->args.size(); ++i, ++arg_it) {
    llvm::Argument* v = &(*arg_it);
    const Var& var = f->args[i];
    var_map_[var.get()] = v;
    if (is_restricted_) {
      if (var.type().is_handle() && !alias_var_set_.count(var.get())) {
        // set non alias.
151 152 153
#if TVM_LLVM_VERSION >= 50
        function_->addParamAttr(i, llvm::Attribute::NoAlias);
#else
154
        function_->setDoesNotAlias(i + 1);
155
#endif
156 157 158
      }
    }
  }
159 160
  llvm::BasicBlock* entry = llvm::BasicBlock::Create(*ctx_, "entry", function_);
  builder_->SetInsertPoint(entry);
161
  this->VisitStmt(f->body);
162 163 164 165 166
  if (ret_void) {
    builder_->CreateRetVoid();
  } else {
    builder_->CreateRet(ConstInt32(0));
  }
167 168
}

169 170
std::unique_ptr<llvm::Module> CodeGenLLVM::Finish() {
  this->AddStartupFunction();
171 172 173 174 175 176 177
  // link modules
  for (size_t i = 0; i < link_modules_.size(); ++i) {
    CHECK(!llvm::Linker::linkModules(*module_, std::move(link_modules_[i])))
        << "Failed to link modules";
  }
  link_modules_.clear();
  // optimize
178 179 180 181
  this->Optimize();
  return std::move(module_);
}

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

void CodeGenLLVM::HandleImport(const std::string& code) {
  std::unique_ptr<llvm::Module> mlib;
  llvm::SMDiagnostic err;
  if (code.length() >= 3 &&
      (code.substr(code.length() - 3) == ".ll" ||
       code.substr(code.length() - 3) == ".bc")) {
    mlib = llvm::parseIRFile(code, err, *ctx_);
    if (mlib.get() == nullptr) {
      std::string msg = err.getMessage();
      LOG(FATAL) << "Fail to load bitcode file " << code << "\n"
                 << "line " << err.getLineNo() << ":" << msg;
    }
  } else {
    std::unique_ptr<llvm::MemoryBuffer> buf =
        llvm::MemoryBuffer::getMemBuffer(code);
    mlib = llvm::parseIR(*buf, err, *ctx_);
    if (mlib.get() == nullptr) {
      std::string msg = err.getMessage();
      LOG(FATAL) << "Fail to load llvm ir "
                 << "line " << err.getLineNo() << ":" << msg
                 << "\ncontent:\n"  << code;
    }
  }
  mlib->setTargetTriple(target_machine_->getTargetTriple().str());
  mlib->setDataLayout(target_machine_->createDataLayout());
  // mark all the functions as force inline
  for (llvm::Function &f : mlib->functions()) {
    f.removeFnAttr(llvm::Attribute::NoInline);
    f.addFnAttr(llvm::Attribute::AlwaysInline);
    f.setLinkage(llvm::GlobalValue::AvailableExternallyLinkage);
  }
  // add to linker libraries.
  this->AddLinkModule(std::move(mlib));
}

218 219 220 221
void CodeGenLLVM::AddLinkModule(std::unique_ptr<llvm::Module>&& mod) {
  link_modules_.emplace_back(std::move(mod));
}

222
void CodeGenLLVM::AddMainFunction(const std::string& entry_func_name) {
223 224 225 226 227 228 229 230 231 232 233
  LOG(FATAL) << "not implemented";
}

llvm::Value* CodeGenLLVM::GetThreadIndex(const IterVar& iv) {
  LOG(FATAL) << "not implemented";
  return nullptr;
}

llvm::Value* CodeGenLLVM::CreateStorageSync(const Call* op) {
  LOG(FATAL) << "not implemented";
  return nullptr;
234 235
}

236 237 238 239 240 241 242 243 244
class FPassManager : public llvm::legacy::FunctionPassManager {
 public:
  explicit FPassManager(llvm::Module* m)
      : llvm::legacy::FunctionPassManager(m) {}
  // override add to allow messaging
  void add(llvm::Pass* p) final {
    llvm::legacy::FunctionPassManager::add(p);
  }
};
245

246 247 248 249 250 251 252 253
class MPassManager : public llvm::legacy::PassManager {
 public:
  // override add to allow messaging
  void add(llvm::Pass* p) final {
    llvm::legacy::PassManager::add(p);
  }
};

254 255 256
void CodeGenLLVM::InitPassManagerBuilder(llvm::PassManagerBuilder* builder) {
}

257
void CodeGenLLVM::Optimize() {
258 259 260 261 262 263 264 265 266 267
  // pass manager
  FPassManager fpass(module_.get());
  MPassManager mpass;
  mpass.add(llvm::createTargetTransformInfoWrapperPass(
              target_machine_ ? target_machine_->getTargetIRAnalysis() :
                                llvm::TargetIRAnalysis()));
  fpass.add(llvm::createTargetTransformInfoWrapperPass(
              target_machine_ ? target_machine_->getTargetIRAnalysis() :
              llvm::TargetIRAnalysis()));

268 269 270
  // place optimization pass
  llvm::PassManagerBuilder builder;
  builder.OptLevel = 3;
271 272 273 274

#if TVM_LLVM_VERSION >= 50
  builder.Inliner = llvm::createFunctionInliningPass(builder.OptLevel, 0, false);
#else
275
  builder.Inliner = llvm::createFunctionInliningPass(builder.OptLevel, 0);
276
#endif
277 278
  builder.LoopVectorize = true;
  builder.SLPVectorize = true;
279 280 281 282 283 284
  this->InitPassManagerBuilder(&builder);

#if TVM_LLVM_VERSION >= 50
  target_machine_->adjustPassManager(builder);
#endif

285 286 287 288 289 290 291 292 293 294 295
  builder.populateFunctionPassManager(fpass);
  builder.populateModulePassManager(mpass);

  fpass.doInitialization();
  for (auto it = module_->begin(); it != module_->end(); ++it) {
    fpass.run(*it);
  }
  fpass.doFinalization();
  mpass.run(*module_);
}

296 297 298 299 300 301
int CodeGenLLVM::NativeVectorBits(const runtime::StorageScope& storage_scope) const {
  return native_vector_bits_;
}

unsigned CodeGenLLVM::GetGlobalAddressSpace() {
  return 0;
302 303 304
}

llvm::Type* CodeGenLLVM::LLVMType(const Type& t) const {
305 306 307 308
  if (t.is_handle()) {
    CHECK_EQ(t.lanes(), 1);
    return t_void_p_;
  }
309
  llvm::Type* etype = nullptr;
310 311
  if (t.is_int() || t.is_uint()) {
    etype = llvm::Type::getIntNTy(*ctx_, t.bits());
312 313
  } else if (t.is_float()) {
    switch (t.bits()) {
314 315 316 317
      case 16: etype = llvm::Type::getHalfTy(*ctx_); break;
      case 32: etype = llvm::Type::getFloatTy(*ctx_); break;
      case 64: etype = llvm::Type::getDoubleTy(*ctx_); break;
      default: LOG(FATAL) << "do not support " << t;
318 319 320
    }
  }
  if (t.lanes() != 1) {
321 322 323
    return llvm::VectorType::get(etype, t.lanes());
  } else {
    return etype;
324 325 326
  }
}

327 328 329 330 331 332 333 334 335 336 337
// Add tbaa alias information for load
//
// use a binary tree typed system to declare information
// and allow alias to be distinguished across nodes.
//
// This trick comes from Halide's CodeGen_LLVM
//
void CodeGenLLVM::AddAliasInfo(llvm::Instruction* inst,
                               const Variable* buffer,
                               Expr index,
                               Type type) {
338 339 340 341 342 343 344 345
  if (alias_var_set_.count(buffer) != 0) {
    // Mark all possibly aliased pointer as same type.
    llvm::MDNode* meta = md_tbaa_alias_set_;
    inst->setMetadata(
        "tbaa",
        md_builder_->createTBAAStructTagNode(meta, meta, 0));
    return;
  }
346 347
  int base = 0, width = 0;
  // create meta-data for alias analysis
348
  // Use a group of binary tree ranges of memory banks.
349 350 351 352 353 354 355 356 357 358 359 360 361 362 363
  if (index.defined()) {
    const Ramp* ramp = index.as<Ramp>();
    if (ramp) {
      int base, stride;
      if (arith::GetConstInt(ramp->base, &base) &&
          arith::GetConstInt(ramp->stride, &stride)) {
        int xwith = ramp->lanes * stride;
        width = 1;
        while (width < xwith) {
          width *= 2;
        }
        while (base % width) {
          base -= base % width;
          width *= 2;
        }
364
      }
365 366
    } else {
      if (arith::GetConstInt(index, &base)) width = 1;
367 368 369
    }
  }
  llvm::MDNode* meta = md_tbaa_root_;
370
  std::ostringstream buffer_addr, buffer_type;
371 372
  buffer_addr << buffer;
  meta = md_builder_->createTBAAScalarTypeNode(buffer_addr.str(), meta);
373
  buffer_type << type.element_of();
374
  meta = md_builder_->createTBAAScalarTypeNode(buffer_type.str(), meta);
375 376 377 378 379 380 381 382 383 384 385 386
  // create a tree-shape access structure.
  if (width != 0) {
    for (int w = 1024; w >= width; w /= 2) {
      int b = (base / w) * w;
      std::stringstream os;
      os << buffer << ".w" << w << ".b" << b;
      meta = md_builder_->createTBAAScalarTypeNode(os.str(), meta);
    }
  }
  inst->setMetadata(
      "tbaa",
      md_builder_->createTBAAStructTagNode(meta, meta, 0));
387 388
}

389 390 391 392 393 394 395 396 397 398 399 400 401 402 403
void CodeGenLLVM::GetAlignment(Type t,
                               const Variable* buf_var,
                               const Expr& index,
                               int* p_alignment,
                               int* p_native_bits) {
  int max_align_bits = t.bits();
  auto it = alloc_storage_info_.find(buf_var);
  if (it != alloc_storage_info_.end()) {
    const StorageInfo& info = it->second;
    *p_native_bits = NativeVectorBits(info.scope);
    max_align_bits = info.alignment * 8;
  } else {
    *p_native_bits = native_vector_bits_;
  }

404 405 406
  arith::ModularSet me = analyzer_->modular_set(index);
  int64_t base = me->base;
  int64_t coeff = me->coeff;
407

408 409
  int align_bits = t.bits();
  while (align_bits < max_align_bits &&
410 411 412 413
         base % 2  == 0 &&
         coeff % 2 == 0) {
    base =  base / 2;
    coeff =  coeff / 2;
414 415 416 417 418 419 420 421
    align_bits *= 2;
  }
  if (align_bits < 8) {
    align_bits = 8;
  }
  *p_alignment = align_bits / 8;
}

422
llvm::Value* CodeGenLLVM::CreateBroadcast(llvm::Value* value, int lanes) {
423
  llvm::Constant* undef = llvm::UndefValue::get(
424 425
      llvm::VectorType::get(value->getType(), lanes));
  llvm::Constant* zero = ConstInt32(0);
426
  value = builder_->CreateInsertElement(undef, value, zero);
427
  llvm::Constant* mask = llvm::ConstantVector::getSplat(lanes, zero);
428
  return builder_->CreateShuffleVector(value, undef, mask);
429 430
}

431 432 433
llvm::Value* CodeGenLLVM::CreateVecSlice(llvm::Value* vec, int begin, int extent) {
  int num_elems = static_cast<int>(vec->getType()->getVectorNumElements());
  if (extent == num_elems && begin == 0) return vec;
434 435
  std::vector<llvm::Constant*> indices;
  indices.reserve(extent);
436
  for (int i = 0; i < extent; ++i) {
437 438 439 440 441
    if (begin + i >= 0 && begin + i < num_elems) {
      indices.push_back(llvm::ConstantInt::get(t_int32_, begin + i));
    } else {
      indices.push_back(llvm::UndefValue::get(t_int32_));
    }
442
  }
443
  return builder_->CreateShuffleVector(vec, vec, llvm::ConstantVector::get(indices));
444
}
445

446 447 448 449 450
llvm::Value* CodeGenLLVM::CreateVecFlip(llvm::Value* vec) {
  int num_elems = static_cast<int>(vec->getType()->getVectorNumElements());
  std::vector<unsigned> indices;
  for (int i = 0; i < num_elems; ++i) {
    indices.push_back(num_elems - i - 1);
451
  }
452 453 454 455 456 457 458 459 460 461
  return builder_->CreateShuffleVector(vec, vec, indices);
}

llvm::Value* CodeGenLLVM::CreateVecPad(llvm::Value* vec, int target_lanes) {
  llvm::Value* mask = llvm::UndefValue::get(LLVMType(Int(32, target_lanes)));
  int num_elems = static_cast<int>(vec->getType()->getVectorNumElements());
  if (num_elems == target_lanes) return vec;
  CHECK_LT(num_elems, target_lanes);
  for (int i = 0; i < num_elems; ++i) {
    mask = builder_->CreateInsertElement(mask, ConstInt32(i), ConstInt32(i));
462
  }
463
  return builder_->CreateShuffleVector(vec, vec, mask);
464 465
}

466 467 468 469 470 471 472 473
llvm::Value* CodeGenLLVM::CreateVecConcat(std::vector<llvm::Value*> vecs) {
  // concat vector, tree shape reduction
  int total_lanes = 0;
  for (llvm::Value* v : vecs) {
    total_lanes += static_cast<int>(
        v->getType()->getVectorNumElements());
  }
  while (vecs.size() > 1) {
474 475
    std::vector<llvm::Value*> new_vecs;
    for (size_t i = 0; i < vecs.size() - 1; i += 2) {
476 477
      llvm::Value* lhs = vecs[i];
      llvm::Value* rhs = vecs[i + 1];
478 479 480 481 482 483 484 485
      const size_t lhs_lanes = lhs->getType()->getVectorNumElements();
      const size_t rhs_lanes = rhs->getType()->getVectorNumElements();
      if (lhs_lanes < rhs_lanes) {
        lhs = CreateVecPad(lhs, rhs_lanes);
      } else if (rhs_lanes < lhs_lanes) {
        rhs = CreateVecPad(rhs, lhs_lanes);
      }
      const size_t shared_lanes = std::max(lhs_lanes, rhs_lanes);
486
      std::vector<unsigned> mask;
487
      for (size_t i = 0; i < lhs_lanes; ++i) {
488 489
        mask.push_back(i);
      }
490 491 492 493 494 495 496
      for (size_t i = 0; i < rhs_lanes; ++i) {
        mask.push_back(shared_lanes + i);
      }
      new_vecs.push_back(builder_->CreateShuffleVector(lhs, rhs, mask));
    }
    if (vecs.size() % 2 != 0) {
      new_vecs.push_back(vecs.back());
497
    }
498
    vecs.swap(new_vecs);
499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534
  }
  return CreateVecSlice(vecs[0], 0, total_lanes);
}


void CodeGenLLVM::CreateSerialFor(llvm::Value* begin,
                                  llvm::Value* end,
                                  llvm::Value* stride,
                                  const VarExpr& loop_var,
                                  const Stmt& body) {
  using llvm::BasicBlock;
  BasicBlock* pre_block = builder_->GetInsertBlock();
  BasicBlock* for_begin = BasicBlock::Create(
      *ctx_, "for_begin", function_);
  BasicBlock* for_body = BasicBlock::Create(
      *ctx_, "for_body", function_);
  BasicBlock* for_end = BasicBlock::Create(
      *ctx_, "for_end", function_);
  builder_->CreateBr(for_begin);
  builder_->SetInsertPoint(for_begin);
  llvm::PHINode* loop_value = builder_->CreatePHI(begin->getType(), 2);
  loop_value->addIncoming(begin, pre_block);
  CHECK(!var_map_.count(loop_var.get()));
  var_map_[loop_var.get()] = loop_value;
  builder_->CreateCondBr(CreateLT(loop_var.type(), loop_value, end),
                         for_body, for_end, md_very_likely_branch_);
  builder_->SetInsertPoint(for_body);
  this->VisitStmt(body);
  var_map_.erase(loop_var.get());
  llvm::Value* loop_next = CreateAdd(loop_var.type(), loop_value, stride);
  loop_value->addIncoming(loop_next, builder_->GetInsertBlock());
  builder_->CreateBr(for_begin);
  builder_->SetInsertPoint(for_end);
}

// cast operatpr
535 536 537
llvm::Value* CodeGenLLVM::CreateCast(Type from, Type to, llvm::Value* value) {
  llvm::Type * target = LLVMType(to);
  if (value->getType() == target) return value;
538
  if (to.is_handle()) {
539
    return builder_->CreateBitCast(value, target);
540 541 542 543 544 545 546 547
  } else if (to.is_uint() && to.bits() == 1) {
    if (from.is_float()) {
      llvm::Constant* zero = llvm::ConstantFP::get(LLVMType(from), 0.);
      return builder_->CreateFCmpONE(value, zero);
    } else {
      llvm::Constant* zero = llvm::ConstantInt::get(LLVMType(from), 0);
      return builder_->CreateICmpNE(value, zero);
    }
548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566
  } else if (!from.is_float() && !to.is_float()) {
    return builder_->CreateIntCast(value, target, from.is_int());
  } else if (from.is_float() && to.is_int()) {
    return builder_->CreateFPToSI(value, target);
  } else if (from.is_float() && to.is_uint()) {
    if (to.bits() < 8) {
      value = builder_->CreateFPToUI(value, LLVMType(to.with_bits(8)));
      return builder_->CreateIntCast(value, target, false);
    } else {
      return builder_->CreateFPToUI(value, target);
    }
  } else if (from.is_int() && to.is_float()) {
    return builder_->CreateSIToFP(value, target);
  } else if (from.is_uint() && to.is_float()) {
    return builder_->CreateUIToFP(value, target);
  } else {
    CHECK(from.is_float() && to.is_float());
    return builder_->CreateFPCast(value, target);
  }
567 568
}

569 570 571 572 573 574 575 576 577 578 579 580 581 582
llvm::Value* CodeGenLLVM::GetConstString(const std::string& str) {
  auto it = str_map_.find(str);
  if (it != str_map_.end()) return it->second;
  llvm::Type* type = llvm::ArrayType::get(t_char_, str.length() + 1);
  llvm::GlobalVariable *global = new llvm::GlobalVariable(
      *module_, type, true, llvm::GlobalValue::PrivateLinkage, 0, ".str");
  global->setAlignment(1);
  global->setInitializer(llvm::ConstantDataArray::getString(*ctx_, str));
  llvm::Constant* zero = ConstInt32(0);
  llvm::Constant* indices[] = {zero, zero};
  llvm::Constant* ptr = llvm::ConstantExpr::getGetElementPtr(
      type, global, indices);
  str_map_[str] = ptr;
  return ptr;
583 584
}

585 586 587 588 589 590 591 592
llvm::Value* CodeGenLLVM::CreateBufferPtr(
    Type t, llvm::Value* buffer, llvm::Value* index) {
  CHECK_EQ(t.lanes(), 1);
  llvm::PointerType* btype = llvm::dyn_cast<llvm::PointerType>(buffer->getType());
  CHECK(btype != nullptr);
  llvm::PointerType* ptype = LLVMType(t)->getPointerTo(btype->getAddressSpace());
  if (btype != ptype) {
    buffer = builder_->CreatePointerCast(buffer, ptype);
593
  }
594

595
  return builder_->CreateInBoundsGEP(buffer, index);
596 597
}

598 599 600 601 602 603 604 605 606 607 608 609
llvm::Value* CodeGenLLVM::CreateBufferVecPtr(
    Type t, llvm::Value* buffer, llvm::Value* index) {
  CHECK_GT(t.lanes(), 1);
  llvm::PointerType* btype = llvm::dyn_cast<llvm::PointerType>(buffer->getType());
  CHECK(btype != nullptr);
  llvm::PointerType* ptype = LLVMType(t)->getPointerTo(btype->getAddressSpace());
  if (btype != ptype) {
    buffer = builder_->CreatePointerCast(buffer, ptype);
  }
  return builder_->CreateInBoundsGEP(buffer, index);
}

610 611
llvm::Value* CodeGenLLVM::GetVarValue(const Variable* v) const {
  auto it = var_map_.find(v);
612
  CHECK(it != var_map_.end()) << "cannot find variable " << v->name_hint;
613
  return it->second;
614 615
}

616 617 618
llvm::Value* CodeGenLLVM::CreateCallExtern(const Call* op) {
  std::vector<llvm::Value*> arg_value;
  std::vector<llvm::Type*> arg_type;
619 620
  for (size_t i = 0; i < op->args.size(); ++i) {
    arg_value.push_back(MakeValue(op->args[i]));
621
    arg_type.push_back(arg_value.back()->getType());
622
  }
623 624 625 626 627 628 629 630 631 632
  llvm::FunctionType* ftype = llvm::FunctionType::get(
      LLVMType(op->type), arg_type, false);
  llvm::Function* f = module_->getFunction(op->name);
  if (f == nullptr) {
    f = llvm::Function::Create(
        ftype, llvm::Function::ExternalLinkage,
        op->name, module_.get());
  }
  llvm::CallInst* call = builder_->CreateCall(f, arg_value);
  return call;
633 634
}

635
llvm::Value* CodeGenLLVM::CreateIntrinsic(const Call* op) {
636
  if (op->is_intrinsic("llvm_intrin")) {
637
    CHECK_GE(op->args.size(), 2U);
638 639
    llvm::Intrinsic::ID id = static_cast<llvm::Intrinsic::ID>(
        op->args[0].as<UIntImm>()->value);
640
    uint64_t num_signature = op->args[1].as<UIntImm>()->value;
641
    std::vector<llvm::Value*> arg_value;
642 643
    std::vector<llvm::Type*> sig_type;
    for (size_t i = 2; i < op->args.size(); ++i) {
644
      arg_value.push_back(MakeValue(op->args[i]));
645 646 647
      if (i - 2 < num_signature) {
        sig_type.push_back(arg_value.back()->getType());
      }
648
    }
649 650 651 652
    llvm::Type *return_type = LLVMType(op->type);
    if (sig_type.size() > 0 && return_type != sig_type[0]) {
      sig_type.insert(sig_type.begin(), return_type);
    }
653
    llvm::Function* f = llvm::Intrinsic::getDeclaration(
654
        module_.get(), id, sig_type);
655
    return builder_->CreateCall(f, arg_value);
656
  } else if (op->is_intrinsic(Call::bitwise_and)) {
657
    return builder_->CreateAnd(MakeValue(op->args[0]), MakeValue(op->args[1]));
658
  } else if (op->is_intrinsic(Call::bitwise_or)) {
659
    return builder_->CreateOr(MakeValue(op->args[0]), MakeValue(op->args[1]));
660 661
  } else if (op->is_intrinsic(Call::bitwise_not)) {
    return builder_->CreateNot(MakeValue(op->args[0]));
662 663
  } else if (op->is_intrinsic(Call::bitwise_xor)) {
    return builder_->CreateXor(MakeValue(op->args[0]), MakeValue(op->args[1]));
664
  } else if (op->is_intrinsic(Call::shift_left)) {
665
    return builder_->CreateShl(MakeValue(op->args[0]), MakeValue(op->args[1]));
666
  } else if (op->is_intrinsic(Call::shift_right)) {
667 668
    if (op->args[0].type().is_int()) {
      return builder_->CreateAShr(MakeValue(op->args[0]), MakeValue(op->args[1]));
669
    } else {
670
      return builder_->CreateLShr(MakeValue(op->args[0]), MakeValue(op->args[1]));
671
    }
672 673
  } else if (op->is_intrinsic(intrinsic::tvm_storage_sync)) {
    return CreateStorageSync(op);
674
  } else if (op->is_intrinsic(intrinsic::tvm_address_of)) {
675 676
    const Load *l = op->args[0].as<Load>();
    CHECK(op->args.size() == 1 && l);
677 678 679 680 681 682 683 684 685 686 687 688 689 690 691
    const Ramp *r = l->index.as<Ramp>();
    llvm::Value* ptr;
    unsigned addrspace;
    if (!r) {
        ptr = CreateBufferPtr(
          l->type, MakeValue(l->buffer_var), MakeValue(l->index));
        addrspace = llvm::dyn_cast<llvm::PointerType>(
          ptr->getType())->getAddressSpace();
    } else {
        Expr index = r->base / make_const(Int(32), r->lanes);
        ptr = CreateBufferVecPtr(
          l->type, MakeValue(l->buffer_var), MakeValue(index));
        addrspace = llvm::dyn_cast<llvm::PointerType>(
          ptr->getType())->getAddressSpace();
    }
692
    return builder_->CreatePointerCast(ptr, t_char_->getPointerTo(addrspace));
693 694
  } else if (op->is_intrinsic(Call::reinterpret) && is_zero(op->args[0])) {
    return llvm::Constant::getNullValue(t_void_p_);
695
  } else if (op->is_intrinsic(intrinsic::tvm_handle_is_null)) {
696
    return builder_->CreateIsNull(MakeValue(op->args[0]));
697
  } else if (op->is_intrinsic(intrinsic::tvm_if_then_else)) {
698 699
    CHECK_EQ(op->args[0].type().lanes(), 1)
        << "if_then_else can only take scalar condition";
700 701 702 703 704 705 706
    using llvm::BasicBlock;
    BasicBlock* then_block = BasicBlock::Create(
        *ctx_, "if_then", function_);
    BasicBlock* else_block = BasicBlock::Create(
        *ctx_, "if_else", function_);
    BasicBlock* end_block = BasicBlock::Create(
        *ctx_, "if_end", function_);
707
    builder_->CreateCondBr(MakeValue(op->args[0]), then_block, else_block);
708 709
    builder_->SetInsertPoint(then_block);
    llvm::Value* then_value = MakeValue(op->args[1]);
710
    BasicBlock* then_value_block = builder_->GetInsertBlock();
711 712 713
    builder_->CreateBr(end_block);
    builder_->SetInsertPoint(else_block);
    llvm::Value* else_value = MakeValue(op->args[2]);
714
    BasicBlock* else_value_block = builder_->GetInsertBlock();
715 716
    builder_->CreateBr(end_block);
    builder_->SetInsertPoint(end_block);
717
    llvm::PHINode* value = builder_->CreatePHI(then_value->getType(), 2);
718 719
    value->addIncoming(then_value, then_value_block);
    value->addIncoming(else_value, else_value_block);
720
    return value;
721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740
  } else if (op->is_intrinsic(Call::reinterpret)) {
    llvm::Type * target = LLVMType(op->type);
    return builder_->CreateBitCast(MakeValue(op->args[0]), target);
  } else if (op->is_intrinsic("vectorlow")) {
    llvm::Value *v = MakeValue(op->args[0]);
    int l = v->getType()->getVectorNumElements();
    return CreateVecSlice(v, 0, l/2);
  } else if (op->is_intrinsic("vectorhigh")) {
    llvm::Value *v = MakeValue(op->args[0]);
    int l = v->getType()->getVectorNumElements();
    return CreateVecSlice(v, l/2, l/2);
  } else if (op->is_intrinsic("vectorcombine")) {
    llvm::Value *v0 = MakeValue(op->args[0]);
    llvm::Value *v1 = MakeValue(op->args[1]);
    int num_elems = static_cast<int>(v0->getType()->getVectorNumElements()) * 2;
    std::vector<unsigned> indices;
    for (int i = 0; i < num_elems; ++i) {
      indices.push_back(i);
    }
    return builder_->CreateShuffleVector(v0, v1, indices);
741
  } else {
742 743
    LOG(FATAL) << "unknown intrinsic " << op->name;
    return nullptr;
744
  }
745 746
}

747 748 749 750 751 752 753 754 755 756 757 758 759 760
void CodeGenLLVM::Scalarize(const Expr& e,
                            std::function<void(int i, llvm::Value* v)> f) {
  if (const Ramp* ramp = e.as<Ramp>()) {
    for (int i = 0; i < ramp->type.lanes(); ++i) {
      Expr offset = arith::ComputeExpr<Add>(
          ramp->base,
          arith::ComputeExpr<Mul>(ramp->stride, i));
      f(i, MakeValue(offset));
    }
  } else {
    llvm::Value* value = MakeValue(e);
    for (int i = 0; i < e.type().lanes(); ++i) {
      f(i, builder_->CreateExtractElement(value, i));
    }
761 762 763
  }
}

764 765

// Visitors
766 767
llvm::Value* CodeGenLLVM::VisitExpr_(const Variable* op) {
  return GetVarValue(op);
768
}
769 770 771

llvm::Value* CodeGenLLVM::VisitExpr_(const Cast* op) {
  return CreateCast(op->value.type(), op->type, MakeValue(op->value));
772
}
773 774
llvm::Value* CodeGenLLVM::VisitExpr_(const IntImm* op) {
  return llvm::ConstantInt::getSigned(LLVMType(op->type), op->value);
775 776
}

777 778 779 780 781 782 783 784 785 786 787 788
llvm::Value* CodeGenLLVM::VisitExpr_(const UIntImm* op) {
  return llvm::ConstantInt::get(LLVMType(op->type), op->value);
}

llvm::Value* CodeGenLLVM::VisitExpr_(const FloatImm* op) {
  return llvm::ConstantFP::get(LLVMType(op->type), op->value);
}

llvm::Value* CodeGenLLVM::VisitExpr_(const StringImm* op) {
  return GetConstString(op->value);
}

789 790
#define DEFINE_CODEGEN_BINARY_OP(Op)                                    \
  llvm::Value* CodeGenLLVM::Create ## Op(                               \
791
      Type t, llvm::Value* a, llvm::Value *b) {                         \
792 793 794 795 796 797 798 799 800 801 802 803
    if (t.is_int()) {                                                   \
      if (t.bits() >= 32) {                                             \
        return builder_->CreateNSW ## Op (a, b);                        \
      } else {                                                          \
        return builder_->Create ## Op (a, b);                           \
      }                                                                 \
    } else if (t.is_uint()) {                                           \
      if (t.bits() >= 32) {                                             \
        return builder_->CreateNUW ## Op (a, b);                        \
      } else {                                                          \
        return builder_->Create ## Op (a, b);                           \
      }                                                                 \
804
    } else {                                                            \
805 806
      CHECK(t.is_float());                                              \
      return builder_->CreateF ## Op (a, b);                            \
807 808
    }                                                                   \
  }                                                                     \
809 810 811
  llvm::Value* CodeGenLLVM::VisitExpr_(const Op* op) {                  \
    return Create ## Op(op->type, MakeValue(op->a), MakeValue(op->b));  \
  }
812 813 814 815 816

DEFINE_CODEGEN_BINARY_OP(Add);
DEFINE_CODEGEN_BINARY_OP(Sub);
DEFINE_CODEGEN_BINARY_OP(Mul);

817 818 819 820 821 822 823 824 825 826 827 828 829 830 831
#define DEFINE_CODEGEN_CMP_OP(Op)                                       \
  llvm::Value* CodeGenLLVM::Create ## Op(                               \
      Type t, llvm::Value* a, llvm::Value* b) {                         \
    if (t.is_int()) {                                                   \
      return builder_->CreateICmpS ## Op (a, b);                        \
    } else if (t.is_uint()) {                                           \
      return builder_->CreateICmpU ## Op (a, b);                        \
    } else {                                                            \
      CHECK(t.is_float());                                              \
      return builder_->CreateFCmpO ## Op (a, b);                        \
    }                                                                   \
}                                                                       \
  llvm::Value* CodeGenLLVM::VisitExpr_(const Op* op) {                  \
    return Create ## Op(op->a.type(), MakeValue(op->a), MakeValue(op->b)); \
  }
832

833 834 835 836
DEFINE_CODEGEN_CMP_OP(LT);
DEFINE_CODEGEN_CMP_OP(LE);
DEFINE_CODEGEN_CMP_OP(GT);
DEFINE_CODEGEN_CMP_OP(GE);
837 838 839

llvm::Value* CodeGenLLVM::VisitExpr_(const Div* op) {
  llvm::Value* a = MakeValue(op->a);
840
  llvm::Value* b = MakeValue(op->b);
841
  if (op->type.is_int()) {
842 843 844
    return builder_->CreateSDiv(a, b);
  } else if (op->type.is_uint()) {
    return builder_->CreateUDiv(a, b);
845
  } else {
846 847
    CHECK(op->type.is_float());
    return builder_->CreateFDiv(a, b);
848 849 850 851
  }
}

llvm::Value* CodeGenLLVM::VisitExpr_(const Mod* op) {
852 853
  llvm::Value* a = MakeValue(op->a);
  llvm::Value* b = MakeValue(op->b);
854
  if (op->type.is_int()) {
855 856 857
    return builder_->CreateSRem(a, b);
  } else if (op->type.is_uint()) {
    return builder_->CreateURem(a, b);
858
  } else {
859 860
    CHECK(op->type.is_float());
    return builder_->CreateFRem(a, b);
861 862 863 864 865 866
  }
}

llvm::Value* CodeGenLLVM::VisitExpr_(const Min* op) {
  llvm::Value* a = MakeValue(op->a);
  llvm::Value* b = MakeValue(op->b);
867
  return builder_->CreateSelect(CreateLT(op->a.type(), a, b), a, b);
868 869 870 871 872
}

llvm::Value* CodeGenLLVM::VisitExpr_(const Max* op) {
  llvm::Value* a = MakeValue(op->a);
  llvm::Value* b = MakeValue(op->b);
873
  return builder_->CreateSelect(CreateGT(op->a.type(), a, b), a, b);
874 875 876
}

llvm::Value* CodeGenLLVM::VisitExpr_(const EQ* op) {
877 878 879 880
  llvm::Value* a = MakeValue(op->a);
  llvm::Value* b = MakeValue(op->b);
  if (op->a.type().is_int() || op->a.type().is_uint()) {
    return builder_->CreateICmpEQ(a, b);
881
  } else {
882
    return builder_->CreateFCmpOEQ(a, b);
883 884 885 886
  }
}

llvm::Value* CodeGenLLVM::VisitExpr_(const NE* op) {
887 888 889 890
  llvm::Value* a = MakeValue(op->a);
  llvm::Value* b = MakeValue(op->b);
  if (op->a.type().is_int() || op->a.type().is_uint()) {
    return builder_->CreateICmpNE(a, b);
891
  } else {
892
    return builder_->CreateFCmpONE(a, b);
893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916
  }
}

llvm::Value* CodeGenLLVM::VisitExpr_(const And* op) {
  return builder_->CreateAnd(MakeValue(op->a), MakeValue(op->b));
}

llvm::Value* CodeGenLLVM::VisitExpr_(const Or* op) {
  return builder_->CreateOr(MakeValue(op->a), MakeValue(op->b));
}

llvm::Value* CodeGenLLVM::VisitExpr_(const Not* op) {
  return builder_->CreateNot(MakeValue(op->a));
}

llvm::Value* CodeGenLLVM::VisitExpr_(const Select* op) {
  return builder_->CreateSelect(
      MakeValue(op->condition),
      MakeValue(op->true_value),
      MakeValue(op->false_value));
}

llvm::Value* CodeGenLLVM::VisitExpr_(const Let* op) {
  CHECK(!var_map_.count(op->var.get()));
917
  var_map_[op->var.get()] = MakeValue(op->value);
918
  analyzer_->Bind(op->var, op->value);
919 920 921
  return MakeValue(op->body);
}

922
llvm::Value* CodeGenLLVM::VisitExpr_(const Load* op) {
923
  Type t = op->type;
924 925 926 927 928
  bool is_volatile = volatile_buf_.count(op->buffer_var.get());
  llvm::Value* buffer = MakeValue(op->buffer_var);
  llvm::Value* index = MakeValue(op->index);

  if (t.lanes() == 1) {
929 930
    int alignment, native_bits;
    GetAlignment(t, op->buffer_var.get(), op->index, &alignment, &native_bits);
931 932 933 934
    llvm::Value* ptr = CreateBufferPtr(t, buffer, index);
    llvm::LoadInst* load = builder_->CreateAlignedLoad(ptr, alignment, is_volatile);
    AddAliasInfo(load, op->buffer_var.get(), op->index, t);
    return load;
935
  } else {
936 937 938 939 940
    // vector load
    unsigned addrspace = llvm::dyn_cast<llvm::PointerType>(
      buffer->getType())->getAddressSpace();
    if (const Ramp* ramp = op->index.as<Ramp>()) {
      if (is_one(ramp->stride)) {
941 942
        int alignment, native_bits;
        GetAlignment(t, op->buffer_var.get(), ramp->base, &alignment, &native_bits);
943 944 945 946 947 948 949 950
        CHECK_EQ(ramp->lanes, t.lanes());
        llvm::Value* ptr = CreateBufferPtr(
            t.element_of(), buffer, MakeValue(ramp->base));
        ptr = builder_->CreatePointerCast(ptr, LLVMType(t)->getPointerTo(addrspace));
        llvm::LoadInst* load = builder_->CreateAlignedLoad(ptr, alignment, is_volatile);
        AddAliasInfo(load, op->buffer_var.get(), op->index, t);
        return load;
      }
951 952
    }
  }
953 954 955 956 957 958 959 960 961 962 963 964
  // scalarized load.
  int basic_align = t.bits() / 8;
  llvm::Value* ret = llvm::UndefValue::get(LLVMType(t));
  auto f = [&](int i, llvm::Value* index) {
    llvm::Value* ptr = CreateBufferPtr(t.element_of(), buffer, index);
    llvm::LoadInst* load = builder_->CreateAlignedLoad(
        ptr, basic_align, is_volatile);
    ret = builder_->CreateInsertElement(ret, load, ConstInt32(i));
    AddAliasInfo(load, op->buffer_var.get(), Expr(), t);
  };
  this->Scalarize(op->index, f);
  return ret;
965 966
}

967 968 969 970 971 972 973 974 975 976
llvm::Value* CodeGenLLVM::VisitExpr_(const Call* op) {
  if (op->call_type == Call::Intrinsic ||
      op->call_type == Call::PureIntrinsic) {
    return CreateIntrinsic(op);
  } else if (op->call_type == Call::Extern ||
             op->call_type == Call::PureExtern) {
    return CreateCallExtern(op);
  } else {
    LOG(FATAL) << "Unknown call type ";
    return nullptr;
977 978 979
  }
}

980 981 982 983 984 985
llvm::Value* CodeGenLLVM::VisitExpr_(const Ramp* op) {
  llvm::Value* vec = llvm::UndefValue::get(LLVMType(op->type));
  for (int i = 0; i < op->lanes; ++i) {
    vec = builder_->CreateInsertElement(
        vec, MakeValue(op->base + op->stride * make_const(op->stride.type(), i)),
        ConstInt32(i));
986
  }
987
  return vec;
988 989
}

990 991
llvm::Value* CodeGenLLVM::VisitExpr_(const Broadcast* op) {
  return CreateBroadcast(MakeValue(op->value), op->lanes);
992 993 994
}

void CodeGenLLVM::VisitStmt_(const Store* op) {
995
  CHECK(is_one(op->predicate));
996
  Type t = op->value.type();
997 998 999 1000
  bool is_volatile = volatile_buf_.count(op->buffer_var.get());
  llvm::Value* buffer = MakeValue(op->buffer_var);
  llvm::Value* index = MakeValue(op->index);
  llvm::Value* value = MakeValue(op->value);
1001

1002
  if (t.lanes() == 1) {
1003 1004
    int alignment, native_bits;
    GetAlignment(t, op->buffer_var.get(), op->index, &alignment, &native_bits);
1005 1006 1007 1008
    llvm::Value* ptr = CreateBufferPtr(t, buffer, index);
    llvm::StoreInst* store = builder_->CreateAlignedStore(value, ptr, alignment, is_volatile);
    AddAliasInfo(store, op->buffer_var.get(), op->index, op->value.type());
    return;
1009
  } else {
1010 1011 1012 1013 1014
    // vector store
    unsigned addrspace = llvm::dyn_cast<llvm::PointerType>(
        buffer->getType())->getAddressSpace();
    if (const Ramp* ramp = op->index.as<Ramp>()) {
      if (is_one(ramp->stride)) {
1015 1016
        int alignment, native_bits;
        GetAlignment(t, op->buffer_var.get(), ramp->base, &alignment, &native_bits);
1017 1018 1019 1020 1021 1022 1023 1024 1025
        CHECK_EQ(ramp->lanes, t.lanes());
        llvm::Value* ptr = CreateBufferPtr(
            t.element_of(), buffer, MakeValue(ramp->base));
        ptr = builder_->CreatePointerCast(ptr, LLVMType(t)->getPointerTo(addrspace));
        llvm::StoreInst* store = builder_->CreateAlignedStore(value, ptr, alignment, is_volatile);
        AddAliasInfo(store, op->buffer_var.get(), op->index, op->value.type());
        return;
      }
    }
1026
  }
1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037
  CHECK_GE(t.bits(), 8);
  // scalarized store.
  int basic_align = t.bits() / 8;
  auto f = [&](int i, llvm::Value* index) {
    llvm::Value* ptr = CreateBufferPtr(t.element_of(), buffer, index);
    llvm::StoreInst* store = builder_->CreateAlignedStore(
        builder_->CreateExtractElement(value, i),
        ptr, basic_align, is_volatile);
    AddAliasInfo(store, op->buffer_var.get(), Expr(), op->value.type());
  };
  this->Scalarize(op->index, f);
1038 1039 1040 1041
}

void CodeGenLLVM::VisitStmt_(const For* op) {
  CHECK(is_zero(op->min));
1042
  analyzer_->Bind(op->loop_var, Range::make_by_min_extent(op->min, op->extent));
1043 1044 1045 1046 1047 1048
  if (op->for_type == ForType::Unrolled) {
    LOG(WARNING) << "Unroll hint get ignore at CodeGenLLVM backend, "
                 << " consider set unroll_explicit=True";
  } else {
    CHECK(op->for_type == ForType::Serial);
  }
1049 1050
  CreateSerialFor(MakeValue(op->min), MakeValue(op->extent),
                  ConstInt32(1), op->loop_var, op->body);
1051 1052
}

1053

1054 1055
void CodeGenLLVM::VisitStmt_(const IfThenElse* op) {
  using llvm::BasicBlock;
1056
  llvm::Value* cond = MakeValue(op->condition);
1057 1058 1059 1060 1061
  BasicBlock* then_block = BasicBlock::Create(
      *ctx_, "if_then", function_);
  BasicBlock* end_block = BasicBlock::Create(
      *ctx_, "if_end", function_);
  if (op->else_case.defined()) {
1062 1063 1064 1065 1066 1067
    BasicBlock* else_block = BasicBlock::Create(
        *ctx_, "if_else", function_);
    builder_->CreateCondBr(cond, then_block, else_block);
    builder_->SetInsertPoint(then_block);
    this->VisitStmt(op->then_case);
    builder_->CreateBr(end_block);
1068 1069 1070
    builder_->SetInsertPoint(else_block);
    this->VisitStmt(op->else_case);
    builder_->CreateBr(end_block);
1071 1072 1073 1074 1075
  } else {
    builder_->CreateCondBr(cond, then_block, end_block, md_very_likely_branch_);
    builder_->SetInsertPoint(then_block);
    this->VisitStmt(op->then_case);
    builder_->CreateBr(end_block);
1076 1077 1078 1079
  }
  builder_->SetInsertPoint(end_block);
}

1080

1081 1082 1083
void CodeGenLLVM::VisitStmt_(const Allocate* op) {
  CHECK(!is_zero(op->condition));
  llvm::Value* buf = nullptr;
1084 1085 1086 1087 1088 1089
  if (op->new_expr.defined()) {
    CHECK_EQ(op->free_function, "nop");
    buf = MakeValue(op->new_expr);
  } else {
    int32_t constant_size = op->constant_allocation_size();
    CHECK_GT(constant_size, 0)
1090
        << "Can only handle constant size stack allocation";
1091 1092
    StorageInfo& info = alloc_storage_info_[op->buffer_var.get()];
    if (constant_size % 4 == 0 && info.alignment == 0) {
1093
      info.alignment = GetTempAllocaAlignment(op->type, constant_size);
1094
    }
1095 1096 1097 1098
    // maximum necessary alignment in the NV devices
    if (info.alignment > 16) {
      info.alignment = 16;
    }
1099 1100 1101 1102
    llvm::AllocaInst* alloca = WithFunctionEntry([&]() {
        return builder_->CreateAlloca(
            LLVMType(op->type), ConstInt32(constant_size));
      });
1103 1104 1105 1106
    if (alloca->getAlignment() < static_cast<uint32_t>(info.alignment)) {
      alloca->setAlignment(info.alignment);
    }
    info.alignment = alloca->getAlignment();
1107
    buf = alloca;
1108
  }
1109 1110 1111
  buf = builder_->CreatePointerCast(
      buf, LLVMType(op->type)->getPointerTo(
          buf->getType()->getPointerAddressSpace()));
1112 1113
  CHECK(!var_map_.count(op->buffer_var.get()));
  var_map_[op->buffer_var.get()] = buf;
1114
  this->VisitStmt(op->body);
1115 1116
}

1117
void CodeGenLLVM::VisitStmt_(const AttrStmt* op) {
1118
  if (op->attr_key == attr::thread_extent) {
1119 1120 1121 1122
    IterVar iv(op->node.node_);
    if (iv->thread_tag.length() != 0) {
      if (!var_map_.count(iv->var.get())) {
        var_map_[iv->var.get()] = GetThreadIndex(iv);
1123
        analyzer_->Bind(iv->var, Range::make_by_min_extent(0, op->value));
1124 1125 1126
      }
    }
  } else if (op->attr_key == ir::attr::storage_scope) {
1127 1128
    const Variable* v = op->node.as<Variable>();
    CHECK(v);
1129 1130
    alloc_storage_info_[v].scope =
        runtime::StorageScope::make(op->value.as<StringImm>()->value);
1131 1132 1133 1134 1135
  } else if (op->attr_key == ir::attr::storage_alignment) {
    const Variable* v = op->node.as<Variable>();
    CHECK(v);
    alloc_storage_info_[v].alignment =
        static_cast<int>(op->value.as<IntImm>()->value);
1136 1137 1138 1139
  } else if (op->attr_key == ir::attr::volatile_scope) {
    const Variable* v = op->node.as<Variable>();
    CHECK(v);
    volatile_buf_.insert(v);
1140
  }
1141
  this->VisitStmt(op->body);
1142 1143
}

1144
void CodeGenLLVM::VisitStmt_(const AssertStmt* op) {
1145
  With<arith::ConstraintContext> cctx(analyzer_.get(), op->condition);
1146
  this->VisitStmt(op->body);
1147 1148
}

1149
void CodeGenLLVM::VisitStmt_(const LetStmt* op) {
1150
  CHECK(!var_map_.count(op->var.get()));
1151 1152 1153 1154 1155
  if (op->var.type().is_handle()) {
    if (!is_restricted_) {
      alias_var_set_.insert(op->var.get());
    }
  }
1156
  var_map_[op->var.get()] = MakeValue(op->value);
1157
  analyzer_->Bind(op->var, op->value);
1158
  this->VisitStmt(op->body);
1159
}
1160

1161
void CodeGenLLVM::VisitStmt_(const Block* op) {
1162 1163 1164 1165
  this->VisitStmt(op->first);
  if (op->rest.defined()) {
    this->VisitStmt(op->rest);
  }
1166
}
1167

1168
void CodeGenLLVM::VisitStmt_(const Evaluate* op) {
1169 1170
  MakeValue(op->value);
}
1171

1172
void CodeGenLLVM::VisitStmt_(const ProducerConsumer* op) {
1173
  this->VisitStmt(op->body);
1174 1175 1176 1177
}
}  // namespace codegen
}  // namespace tvm
#endif  // TVM_LLVM_VERSION