lib/Transforms/LoopUnrollAndJam.cpp - platform/external/tensorflow - Git at Google

 //===- LoopUnrollAndJam.cpp - Code to perform loop unroll and jam ---------===//
 //
 // Copyright 2019 The MLIR Authors.
 //
 // Licensed 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
 //
 //   http://www.apache.org/licenses/LICENSE-2.0
 //
 // 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.
 // =============================================================================
 //
 // This file implements loop unroll and jam. Unroll and jam is a transformation
 // that improves locality, in particular, register reuse, while also improving
 // instruction level parallelism. The example below shows what it does in nearly
 // the general case. Loop unroll and jam currently works if the bounds of the
 // loops inner to the loop being unroll-jammed do not depend on the latter.
 //
 // Before      After unroll and jam of i by factor 2:
 //
 //             for i, step = 2
 // for i         S1(i);
 //   S1;         S2(i);
 //   S2;         S1(i+1);
 //   for j       S2(i+1);
 //     S3;       for j
 //     S4;         S3(i, j);
 //   S5;           S4(i, j);
 //   S6;           S3(i+1, j)
 //                 S4(i+1, j)
 //               S5(i);
 //               S6(i);
 //               S5(i+1);
 //               S6(i+1);
 //
 // Note: 'if/else' blocks are not jammed. So, if there are loops inside if
 // inst's, bodies of those loops will not be jammed.
 //===----------------------------------------------------------------------===//
 #include "mlir/Transforms/Passes.h"

 #include "mlir/AffineOps/AffineOps.h"
 #include "mlir/Analysis/LoopAnalysis.h"
 #include "mlir/IR/AffineExpr.h"
 #include "mlir/IR/AffineMap.h"
 #include "mlir/IR/BlockAndValueMapping.h"
 #include "mlir/IR/Builders.h"
 #include "mlir/IR/BuiltinOps.h"
 #include "mlir/Pass/Pass.h"
 #include "mlir/Transforms/LoopUtils.h"
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/Support/CommandLine.h"

 using namespace mlir;

 #define DEBUG_TYPE "loop-unroll-jam"

 static llvm::cl::OptionCategory clOptionsCategory(DEBUG_TYPE " options");

 // Loop unroll and jam factor.
 static llvm::cl::opt<unsigned>
     clUnrollJamFactor("unroll-jam-factor", llvm::cl::Hidden,
                       llvm::cl::desc("Use this unroll jam factor for all loops"
                                      " (default 4)"),
                       llvm::cl::cat(clOptionsCategory));

 namespace {
 /// Loop unroll jam pass. Currently, this just unroll jams the first
 /// outer loop in a Function.
 struct LoopUnrollAndJam : public FunctionPass {
   Optional<unsigned> unrollJamFactor;
   static const unsigned kDefaultUnrollJamFactor = 4;

   explicit LoopUnrollAndJam(Optional<unsigned> unrollJamFactor = None)
       : FunctionPass(&LoopUnrollAndJam::passID),
         unrollJamFactor(unrollJamFactor) {}

   PassResult runOnFunction(Function *f) override;
   bool runOnAffineForOp(OpPointer<AffineForOp> forOp);

   static char passID;
 };
 } // end anonymous namespace

 char LoopUnrollAndJam::passID = 0;

 FunctionPass *mlir::createLoopUnrollAndJamPass(int unrollJamFactor) {
   return new LoopUnrollAndJam(
       unrollJamFactor == -1 ? None : Optional<unsigned>(unrollJamFactor));
 }

 PassResult LoopUnrollAndJam::runOnFunction(Function *f) {
   // Currently, just the outermost loop from the first loop nest is
   // unroll-and-jammed by this pass. However, runOnAffineForOp can be called on
   // any for Inst.
   auto &entryBlock = f->front();
   if (!entryBlock.empty())
     if (auto forOp = entryBlock.front().dyn_cast<AffineForOp>())
       runOnAffineForOp(forOp);

   return success();
 }

 /// Unroll and jam a 'for' inst. Default unroll jam factor is
 /// kDefaultUnrollJamFactor. Return false if nothing was done.
 bool LoopUnrollAndJam::runOnAffineForOp(OpPointer<AffineForOp> forOp) {
   // Unroll and jam by the factor that was passed if any.
   if (unrollJamFactor.hasValue())
     return loopUnrollJamByFactor(forOp, unrollJamFactor.getValue());
   // Otherwise, unroll jam by the command-line factor if one was specified.
   if (clUnrollJamFactor.getNumOccurrences() > 0)
     return loopUnrollJamByFactor(forOp, clUnrollJamFactor);

   // Unroll and jam by four otherwise.
   return loopUnrollJamByFactor(forOp, kDefaultUnrollJamFactor);
 }

 bool mlir::loopUnrollJamUpToFactor(OpPointer<AffineForOp> forOp,
                                    uint64_t unrollJamFactor) {
   Optional<uint64_t> mayBeConstantTripCount = getConstantTripCount(forOp);

   if (mayBeConstantTripCount.hasValue() &&
       mayBeConstantTripCount.getValue() < unrollJamFactor)
     return loopUnrollJamByFactor(forOp, mayBeConstantTripCount.getValue());
   return loopUnrollJamByFactor(forOp, unrollJamFactor);
 }

 /// Unrolls and jams this loop by the specified factor.
 bool mlir::loopUnrollJamByFactor(OpPointer<AffineForOp> forOp,
                                  uint64_t unrollJamFactor) {
   // Gathers all maximal sub-blocks of instructions that do not themselves
   // include a for inst (a instruction could have a descendant for inst though
   // in its tree).
   struct JamBlockGatherer {
     // Store iterators to the first and last inst of each sub-block found.
     std::vector<std::pair<Block::iterator, Block::iterator>> subBlocks;

     // This is a linear time walk.
     void walk(Instruction *inst) {
       for (auto &blockList : inst->getBlockLists())
         for (auto &block : blockList)
           walk(block);
     }
     void walk(Block &block) {
       for (auto it = block.begin(), e = block.end(); it != e;) {
         auto subBlockStart = it;
         while (it != e && !it->isa<AffineForOp>())
           ++it;
         if (it != subBlockStart)
           subBlocks.push_back({subBlockStart, std::prev(it)});
         // Process all for insts that appear next.
         while (it != e && it->isa<AffineForOp>())
           walk(&*it++);
       }
     }
   };

   assert(unrollJamFactor >= 1 && "unroll jam factor should be >= 1");

   if (unrollJamFactor == 1 || forOp->getBody()->empty())
     return false;

   Optional<uint64_t> mayBeConstantTripCount = getConstantTripCount(forOp);

   if (!mayBeConstantTripCount.hasValue() &&
       getLargestDivisorOfTripCount(forOp) % unrollJamFactor != 0)
     return false;

   auto lbMap = forOp->getLowerBoundMap();
   auto ubMap = forOp->getUpperBoundMap();

   // Loops with max/min expressions won't be unrolled here (the output can't be
   // expressed as a Function in the general case). However, the right way to
   // do such unrolling for a Function would be to specialize the loop for the
   // 'hotspot' case and unroll that hotspot.
   if (lbMap.getNumResults() != 1 || ubMap.getNumResults() != 1)
     return false;

   // Same operand list for lower and upper bound for now.
   // TODO(bondhugula): handle bounds with different sets of operands.
   if (!forOp->matchingBoundOperandList())
     return false;

   // If the trip count is lower than the unroll jam factor, no unroll jam.
   // TODO(bondhugula): option to specify cleanup loop unrolling.
   if (mayBeConstantTripCount.hasValue() &&
       mayBeConstantTripCount.getValue() < unrollJamFactor)
     return false;

   auto *forInst = forOp->getInstruction();

   // Gather all sub-blocks to jam upon the loop being unrolled.
   JamBlockGatherer jbg;
   jbg.walk(forInst);
   auto &subBlocks = jbg.subBlocks;

   // Generate the cleanup loop if trip count isn't a multiple of
   // unrollJamFactor.
   if (mayBeConstantTripCount.hasValue() &&
       mayBeConstantTripCount.getValue() % unrollJamFactor != 0) {
     // Insert the cleanup loop right after 'forOp'.
     FuncBuilder builder(forInst->getBlock(),
                         std::next(Block::iterator(forInst)));
     auto cleanupAffineForOp = builder.clone(*forInst)->cast<AffineForOp>();
     cleanupAffineForOp->setLowerBoundMap(
         getCleanupLoopLowerBound(forOp, unrollJamFactor, &builder));

     // The upper bound needs to be adjusted.
     forOp->setUpperBoundMap(
         getUnrolledLoopUpperBound(forOp, unrollJamFactor, &builder));

     // Promote the loop body up if this has turned into a single iteration loop.
     promoteIfSingleIteration(cleanupAffineForOp);
   }

   // Scale the step of loop being unroll-jammed by the unroll-jam factor.
   int64_t step = forOp->getStep();
   forOp->setStep(step * unrollJamFactor);

   auto *forOpIV = forOp->getInductionVar();
   for (auto &subBlock : subBlocks) {
     // Builder to insert unroll-jammed bodies. Insert right at the end of
     // sub-block.
     FuncBuilder builder(subBlock.first->getBlock(), std::next(subBlock.second));

     // Unroll and jam (appends unrollJamFactor-1 additional copies).
     for (unsigned i = 1; i < unrollJamFactor; i++) {
       BlockAndValueMapping operandMapping;

       // If the induction variable is used, create a remapping to the value for
       // this unrolled instance.
       if (!forOpIV->use_empty()) {
         // iv' = iv + i, i = 1 to unrollJamFactor-1.
         auto d0 = builder.getAffineDimExpr(0);
         auto bumpMap = builder.getAffineMap(1, 0, {d0 + i * step}, {});
         auto ivUnroll =
             builder.create<AffineApplyOp>(forInst->getLoc(), bumpMap, forOpIV);
         operandMapping.map(forOpIV, ivUnroll);
       }
       // Clone the sub-block being unroll-jammed.
       for (auto it = subBlock.first; it != std::next(subBlock.second); ++it) {
         builder.clone(*it, operandMapping);
       }
     }
   }

   // Promote the loop body up if this has turned into a single iteration loop.
   promoteIfSingleIteration(forOp);

   return true;
 }

 static PassRegistration<LoopUnrollAndJam> pass("loop-unroll-jam",
                                                "Unroll and jam loops");
	//===- LoopUnrollAndJam.cpp - Code to perform loop unroll and jam ---------===//
	//
	// Copyright 2019 The MLIR Authors.
	//
	// Licensed 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
	//
	// http://www.apache.org/licenses/LICENSE-2.0
	//
	// 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.
	// =============================================================================
	//
	// This file implements loop unroll and jam. Unroll and jam is a transformation
	// that improves locality, in particular, register reuse, while also improving
	// instruction level parallelism. The example below shows what it does in nearly
	// the general case. Loop unroll and jam currently works if the bounds of the
	// loops inner to the loop being unroll-jammed do not depend on the latter.
	//
	// Before After unroll and jam of i by factor 2:
	//
	// for i, step = 2
	// for i S1(i);
	// S1; S2(i);
	// S2; S1(i+1);
	// for j S2(i+1);
	// S3; for j
	// S4; S3(i, j);
	// S5; S4(i, j);
	// S6; S3(i+1, j)
	// S4(i+1, j)
	// S5(i);
	// S6(i);
	// S5(i+1);
	// S6(i+1);
	//
	// Note: 'if/else' blocks are not jammed. So, if there are loops inside if
	// inst's, bodies of those loops will not be jammed.
	//===----------------------------------------------------------------------===//
	#include "mlir/Transforms/Passes.h"

	#include "mlir/AffineOps/AffineOps.h"
	#include "mlir/Analysis/LoopAnalysis.h"
	#include "mlir/IR/AffineExpr.h"
	#include "mlir/IR/AffineMap.h"
	#include "mlir/IR/BlockAndValueMapping.h"
	#include "mlir/IR/Builders.h"
	#include "mlir/IR/BuiltinOps.h"
	#include "mlir/Pass/Pass.h"
	#include "mlir/Transforms/LoopUtils.h"
	#include "llvm/ADT/DenseMap.h"
	#include "llvm/Support/CommandLine.h"

	using namespace mlir;

	#define DEBUG_TYPE "loop-unroll-jam"

	static llvm::cl::OptionCategory clOptionsCategory(DEBUG_TYPE " options");

	// Loop unroll and jam factor.
	static llvm::cl::opt<unsigned>
	clUnrollJamFactor("unroll-jam-factor", llvm::cl::Hidden,
	llvm::cl::desc("Use this unroll jam factor for all loops"
	" (default 4)"),
	llvm::cl::cat(clOptionsCategory));

	namespace {
	/// Loop unroll jam pass. Currently, this just unroll jams the first
	/// outer loop in a Function.
	struct LoopUnrollAndJam : public FunctionPass {
	Optional<unsigned> unrollJamFactor;
	static const unsigned kDefaultUnrollJamFactor = 4;

	explicit LoopUnrollAndJam(Optional<unsigned> unrollJamFactor = None)
	: FunctionPass(&LoopUnrollAndJam::passID),
	unrollJamFactor(unrollJamFactor) {}

	PassResult runOnFunction(Function *f) override;
	bool runOnAffineForOp(OpPointer<AffineForOp> forOp);

	static char passID;
	};
	} // end anonymous namespace

	char LoopUnrollAndJam::passID = 0;

	FunctionPass *mlir::createLoopUnrollAndJamPass(int unrollJamFactor) {
	return new LoopUnrollAndJam(
	unrollJamFactor == -1 ? None : Optional<unsigned>(unrollJamFactor));
	}

	PassResult LoopUnrollAndJam::runOnFunction(Function *f) {
	// Currently, just the outermost loop from the first loop nest is
	// unroll-and-jammed by this pass. However, runOnAffineForOp can be called on
	// any for Inst.
	auto &entryBlock = f->front();
	if (!entryBlock.empty())
	if (auto forOp = entryBlock.front().dyn_cast<AffineForOp>())
	runOnAffineForOp(forOp);

	return success();
	}

	/// Unroll and jam a 'for' inst. Default unroll jam factor is
	/// kDefaultUnrollJamFactor. Return false if nothing was done.
	bool LoopUnrollAndJam::runOnAffineForOp(OpPointer<AffineForOp> forOp) {
	// Unroll and jam by the factor that was passed if any.
	if (unrollJamFactor.hasValue())
	return loopUnrollJamByFactor(forOp, unrollJamFactor.getValue());
	// Otherwise, unroll jam by the command-line factor if one was specified.
	if (clUnrollJamFactor.getNumOccurrences() > 0)
	return loopUnrollJamByFactor(forOp, clUnrollJamFactor);

	// Unroll and jam by four otherwise.
	return loopUnrollJamByFactor(forOp, kDefaultUnrollJamFactor);
	}

	bool mlir::loopUnrollJamUpToFactor(OpPointer<AffineForOp> forOp,
	uint64_t unrollJamFactor) {
	Optional<uint64_t> mayBeConstantTripCount = getConstantTripCount(forOp);

	if (mayBeConstantTripCount.hasValue() &&
	mayBeConstantTripCount.getValue() < unrollJamFactor)
	return loopUnrollJamByFactor(forOp, mayBeConstantTripCount.getValue());
	return loopUnrollJamByFactor(forOp, unrollJamFactor);
	}

	/// Unrolls and jams this loop by the specified factor.
	bool mlir::loopUnrollJamByFactor(OpPointer<AffineForOp> forOp,
	uint64_t unrollJamFactor) {
	// Gathers all maximal sub-blocks of instructions that do not themselves
	// include a for inst (a instruction could have a descendant for inst though
	// in its tree).
	struct JamBlockGatherer {
	// Store iterators to the first and last inst of each sub-block found.
	std::vector<std::pair<Block::iterator, Block::iterator>> subBlocks;

	// This is a linear time walk.
	void walk(Instruction *inst) {
	for (auto &blockList : inst->getBlockLists())
	for (auto &block : blockList)
	walk(block);
	}
	void walk(Block &block) {
	for (auto it = block.begin(), e = block.end(); it != e;) {
	auto subBlockStart = it;
	while (it != e && !it->isa<AffineForOp>())
	++it;
	if (it != subBlockStart)
	subBlocks.push_back({subBlockStart, std::prev(it)});
	// Process all for insts that appear next.
	while (it != e && it->isa<AffineForOp>())
	walk(&*it++);
	}
	}
	};

	assert(unrollJamFactor >= 1 && "unroll jam factor should be >= 1");

	if (unrollJamFactor == 1 \|\| forOp->getBody()->empty())
	return false;

	Optional<uint64_t> mayBeConstantTripCount = getConstantTripCount(forOp);

	if (!mayBeConstantTripCount.hasValue() &&
	getLargestDivisorOfTripCount(forOp) % unrollJamFactor != 0)
	return false;

	auto lbMap = forOp->getLowerBoundMap();
	auto ubMap = forOp->getUpperBoundMap();

	// Loops with max/min expressions won't be unrolled here (the output can't be
	// expressed as a Function in the general case). However, the right way to
	// do such unrolling for a Function would be to specialize the loop for the
	// 'hotspot' case and unroll that hotspot.
	if (lbMap.getNumResults() != 1 \|\| ubMap.getNumResults() != 1)
	return false;

	// Same operand list for lower and upper bound for now.
	// TODO(bondhugula): handle bounds with different sets of operands.
	if (!forOp->matchingBoundOperandList())
	return false;

	// If the trip count is lower than the unroll jam factor, no unroll jam.
	// TODO(bondhugula): option to specify cleanup loop unrolling.
	if (mayBeConstantTripCount.hasValue() &&
	mayBeConstantTripCount.getValue() < unrollJamFactor)
	return false;

	auto *forInst = forOp->getInstruction();

	// Gather all sub-blocks to jam upon the loop being unrolled.
	JamBlockGatherer jbg;
	jbg.walk(forInst);
	auto &subBlocks = jbg.subBlocks;

	// Generate the cleanup loop if trip count isn't a multiple of
	// unrollJamFactor.
	if (mayBeConstantTripCount.hasValue() &&
	mayBeConstantTripCount.getValue() % unrollJamFactor != 0) {
	// Insert the cleanup loop right after 'forOp'.
	FuncBuilder builder(forInst->getBlock(),
	std::next(Block::iterator(forInst)));
	auto cleanupAffineForOp = builder.clone(*forInst)->cast<AffineForOp>();
	cleanupAffineForOp->setLowerBoundMap(
	getCleanupLoopLowerBound(forOp, unrollJamFactor, &builder));

	// The upper bound needs to be adjusted.
	forOp->setUpperBoundMap(
	getUnrolledLoopUpperBound(forOp, unrollJamFactor, &builder));

	// Promote the loop body up if this has turned into a single iteration loop.
	promoteIfSingleIteration(cleanupAffineForOp);
	}

	// Scale the step of loop being unroll-jammed by the unroll-jam factor.
	int64_t step = forOp->getStep();
	forOp->setStep(step * unrollJamFactor);

	auto *forOpIV = forOp->getInductionVar();
	for (auto &subBlock : subBlocks) {
	// Builder to insert unroll-jammed bodies. Insert right at the end of
	// sub-block.
	FuncBuilder builder(subBlock.first->getBlock(), std::next(subBlock.second));

	// Unroll and jam (appends unrollJamFactor-1 additional copies).
	for (unsigned i = 1; i < unrollJamFactor; i++) {
	BlockAndValueMapping operandMapping;

	// If the induction variable is used, create a remapping to the value for
	// this unrolled instance.
	if (!forOpIV->use_empty()) {
	// iv' = iv + i, i = 1 to unrollJamFactor-1.
	auto d0 = builder.getAffineDimExpr(0);
	auto bumpMap = builder.getAffineMap(1, 0, {d0 + i * step}, {});
	auto ivUnroll =
	builder.create<AffineApplyOp>(forInst->getLoc(), bumpMap, forOpIV);
	operandMapping.map(forOpIV, ivUnroll);
	}
	// Clone the sub-block being unroll-jammed.
	for (auto it = subBlock.first; it != std::next(subBlock.second); ++it) {
	builder.clone(*it, operandMapping);
	}
	}
	}

	// Promote the loop body up if this has turned into a single iteration loop.
	promoteIfSingleIteration(forOp);

	return true;
	}

	static PassRegistration<LoopUnrollAndJam> pass("loop-unroll-jam",
	"Unroll and jam loops");