Files
Raptor/src/PIM/Conversion/ONNXToSpatial/Utils/SpatialReducer.cpp
NiccoloN 6e1de865bb add constant folding and verification pass for pim host operations
better validation scripts output
big refactors
2026-03-20 12:08:12 +01:00

349 lines
15 KiB
C++

#include "mlir/IR/BuiltinAttributes.h"
#include "mlir/IR/Value.h"
#include "llvm/Support/raw_ostream.h"
#include <cassert>
#include <unordered_map>
#include <utility>
#include "SpatialReducer.hpp"
#include "src/Accelerators/PIM/Dialect/Spatial/SpatialOps.hpp"
#define GET_COMP(computeOpAndResNum) std::get<0>(computeOpAndResNum)
#define GET_RES_NUM(computeOpAndResNum) std::get<1>(computeOpAndResNum)
namespace onnx_mlir {
llvm::SmallPtrSet<mlir::Operation*, 16> onnx_mlir::SpatialReducer::oldComputeOpsReplaced;
ResNum SpatialReducer::applyResultProcessing(ComputeAndResNum computeOpAndResNum,
std::function<mlir::Value(const mlir::Value&)> processFun,
mlir::ConversionPatternRewriter& rewriter) {
assert(processFun);
auto computeOp = GET_COMP(computeOpAndResNum);
auto resultNum = GET_RES_NUM(computeOpAndResNum);
spatial::SpatYieldOp yieldOp = mlir::cast<spatial::SpatYieldOp>(computeOp.getBody().front().getTerminator());
mlir::Value result = yieldOp->getOperand(resultNum);
rewriter.setInsertionPointAfterValue(result);
mlir::Value processedResult = processFun(result);
if (processedResult == result) {
// Sometimes we want processedResult to return the same value but do
// something else with it (e.g. in softmax we want to broadcast the value
// using a channel). In this case, we can just return the same value.
return resultNum;
}
yieldOp->insertOperands(yieldOp->getNumOperands(), processedResult);
return yieldOp.getNumOperands() - 1;
}
OpAndResNum
SpatialReducer::applyReducePattern(llvm::SmallVector<ComputeAndResNum>& computeOpsAndResNum,
std::function<mlir::Value(const mlir::Value&, const mlir::Value&)> reduce,
std::function<mlir::Value(const mlir::Value&)> preprocess,
std::function<mlir::Value(const mlir::Value&)> postprocess) {
if (preprocess)
for (auto& computeOpAndResNum : computeOpsAndResNum)
GET_RES_NUM(computeOpAndResNum) = applyResultProcessing(computeOpAndResNum, preprocess, rewriter);
// It is possible that `computeOpsAndResNum` contains two entries for the same
// computeOp. In this case, we need to apply the reduction within-computef
// Keep a map between a computeOp and the last Value for this reduction
std::unordered_map<mlir::Operation*, mlir::Value> lastValueForCompute;
for (auto& computeOpAndResNum : computeOpsAndResNum) {
auto computeOp = GET_COMP(computeOpAndResNum);
auto yieldOp = mlir::cast<spatial::SpatYieldOp>(computeOp.getBody().front().getTerminator());
mlir::Value valueWithinCompute = yieldOp->getOperand(GET_RES_NUM(computeOpAndResNum));
auto it = lastValueForCompute.find(computeOp.getOperation());
if (it != lastValueForCompute.end()) {
// If we have already seen this computeOp, apply the reduction
// within-compute
mlir::Value lastWithinComputeValue = it->second;
assert(valueWithinCompute.getDefiningOp() && lastWithinComputeValue.getDefiningOp());
if (valueWithinCompute.getDefiningOp()->isBeforeInBlock(lastWithinComputeValue.getDefiningOp()))
rewriter.setInsertionPointAfterValue(lastWithinComputeValue);
else
rewriter.setInsertionPointAfterValue(valueWithinCompute);
valueWithinCompute = reduce(lastWithinComputeValue, valueWithinCompute);
lastValueForCompute[computeOp.getOperation()] = valueWithinCompute;
}
lastValueForCompute[computeOp.getOperation()] = valueWithinCompute;
}
// Now, reconstruct from the map the computeOpsAndResNum list
computeOpsAndResNum.clear();
computeOpsAndResNum.reserve(lastValueForCompute.size());
for (auto& entry : lastValueForCompute) {
auto computeOp = mlir::cast<spatial::SpatWeightedCompute>(entry.first);
auto valueWithinCompute = entry.second;
// We check if `valueWithinCompute` is already used by the yieldOp, in that
// case no need to add it
auto yieldOp = mlir::cast<spatial::SpatYieldOp>(computeOp.getBody().front().getTerminator());
bool yieldOpUseFound = false;
for (auto& use : valueWithinCompute.getUses()) {
if (use.getOwner() == yieldOp.getOperation()) {
// If the value is already used by the yieldOp, we can just use it
computeOpsAndResNum.push_back({computeOp, use.getOperandNumber()});
yieldOpUseFound = true;
break;
}
}
if (yieldOpUseFound)
continue;
// If this result is not used within a yieldOp, then add it
auto resultNum = yieldOp->getNumOperands();
yieldOp->insertOperands(resultNum, valueWithinCompute);
computeOpsAndResNum.push_back({computeOp, resultNum});
}
mlir::Location loc = GET_COMP(computeOpsAndResNum[0])->getLoc();
// Recursive algorithm to reduce the inputs to a single one:
// - Take two inputs at a time, and reduce them into a single one, updating
// the computeOpsAndResNum list which becomes half the size.
// - Repeat until there is only one input left.
llvm::OwningArrayRef<ComputeAndResNum> computeOpsRef(computeOpsAndResNum);
while (computeOpsRef.size() > 1) {
llvm::SmallVector<ComputeAndResNum> nextComputeOps;
nextComputeOps.reserve(computeOpsRef.size() / 2);
for (size_t i = 0; i < computeOpsRef.size() - 1; i += 2) {
auto [firstCompute, firstResultNum] = computeOpsRef[i];
auto [secondCompute, secondResultNum] = computeOpsRef[i + 1];
if (secondCompute->isBeforeInBlock(firstCompute)) {
std::swap(firstCompute, secondCompute);
std::swap(firstResultNum, secondResultNum);
}
// We do not immediately alter the computeOps results/operands, instead we
// do it in a delayed manner, to avoid invalidating the references to the
// computeOps (which must be replaced by a cloned ComputeOp when changing
// the number of results)
// See below `reducerChanges.push_back` and `finalizeReduceUpdates`
auto yieldOpFirstCompute = mlir::cast<spatial::SpatYieldOp>(firstCompute.getBody().front().getTerminator());
// Add a new operand to the block of the second computeOp
mlir::Block& secondBlock = secondCompute.getBody().front();
mlir::Value formerRes1 = secondBlock.addArgument(yieldOpFirstCompute->getOperand(firstResultNum).getType(), loc);
auto secondComputeWeightsNum =
secondCompute->getAttrOfType<mlir::DenseI32ArrayAttr>(secondCompute.getOperandSegmentSizesAttrName())[0];
auto secondComputeOperandNum = secondComputeWeightsNum + secondBlock.getNumArguments() - 1;
// Take the "former-result" from the second computeOp
spatial::SpatYieldOp secondYield = mlir::cast<spatial::SpatYieldOp>(secondBlock.getTerminator());
mlir::Value formerRes2 = secondYield.getOperand(secondResultNum);
// Apply reduction operation
rewriter.setInsertionPoint(secondYield);
mlir::Value reduced = reduce(formerRes2, formerRes1);
// Unfortunately, it is not possible to update the result in place,
// because we may have already referenced it by <computeOp, resultNum>
// outside of this function, thus replacing it would invalidate the
// reference. Therefore, we need to append a new result to the yieldOp,
// and then at a later stage update the computeOp accordingly.
// Add `reduced` to the second yieldOp
auto secondYieldOperandNum = secondYield.getNumOperands();
secondYield->insertOperands(secondYieldOperandNum, reduced);
secondResultNum = secondYieldOperandNum;
// We should also add an entry for updating the results of the last
// operation (the one which never becomes a `firstCompute`): because it is
// not tracked by reducerChanges as `fromOp`
reducerChanges.push_back(
{firstCompute.getOperation(), firstResultNum, secondCompute.getOperation(), secondComputeOperandNum});
nextComputeOps.push_back(std::make_pair(secondCompute, secondResultNum));
}
// If we have an odd number of inputs, we need to add the last one to the
// newInputs list.
if (computeOpsRef.size() % 2 == 1)
nextComputeOps.push_back(computeOpsRef.back());
// Replace the inputOps list with the new one.
computeOpsRef = llvm::OwningArrayRef<ComputeAndResNum>(std::move(nextComputeOps));
}
assert(computeOpsRef.size() == 1 && "Internal error: expected a single input at this point.");
auto finalComputeAndResNum = computeOpsRef[0];
// Force the update of the results of this computeOp, when finalizing
computeOpNeedingResUpdate.push_back(GET_COMP(finalComputeAndResNum));
if (postprocess)
GET_RES_NUM(finalComputeAndResNum) = applyResultProcessing(finalComputeAndResNum, postprocess, rewriter);
return std::make_pair(GET_COMP(finalComputeAndResNum).getOperation(), GET_RES_NUM(finalComputeAndResNum));
}
void SpatialReducer::finalizeReduceUpdates() {
assert(reducesFinalized == false && "Cannot finalize two times.");
reducesFinalized = true;
// First, add the results to the computeOps
for (auto& reduceChange : reducerChanges)
updateResultsOfCompute(reduceChange.fromOp);
for (auto& c : computeOpNeedingResUpdate)
updateResultsOfCompute(c.getOperation());
for (auto& reducerChange : this->reducerChanges) {
auto fromOp = reducerChange.fromOp;
auto toOp = reducerChange.toOp;
auto fromOpResNum = reducerChange.fromOpResNum;
auto toOpOperandNum = reducerChange.toOpOperandNum;
auto fromComputeOp = opToReplacedCompute[fromOp];
assert(fromComputeOp && "fromOp should have been mapped before!");
// toComputeOp could be the existing pointer, or we have to remap it with
// `opToReplacedCompute`
auto toComputeOp = opToReplacedCompute[toOp];
if (!toComputeOp)
toComputeOp = mlir::cast<spatial::SpatWeightedCompute>(toOp);
assert(toComputeOp != fromComputeOp && "Oops should have caught this earlier!");
assert(toComputeOp->getNumOperands() == toOpOperandNum
&& "toOpOperandNum should be the last operand of toComputeOp, are the "
"operations in the right order?");
// Add the new operand to `toComputeOp`
auto fromResult = fromComputeOp.getResult(fromOpResNum);
toComputeOp->insertOperands(toOpOperandNum, fromResult);
incrementWeightedComputeInputsSegmentSize(toComputeOp, 1);
}
}
mlir::Value SpatialReducer::resolveValueFromOpAndResNum(OpAndResNum& opAndResNum) {
assert(reducesFinalized && "Cannot create resolve values before finalizing the reduce updates.");
mlir::Operation* opToCast;
auto it = opToReplacedCompute.find(opAndResNum.first);
if (it != opToReplacedCompute.end())
opToCast = it->second;
else
opToCast = opAndResNum.first;
auto computeOp = mlir::cast<spatial::SpatWeightedCompute>(opToCast);
return computeOp.getResult(opAndResNum.second);
}
void SpatialReducer::updateResultsOfCompute(mlir::Operation* computeOp) {
if (opToReplacedCompute.find(computeOp) != opToReplacedCompute.end()) {
// If we have already replaced the fromOp, we do not need to do it again
return;
}
auto oldComputeOp = mlir::cast<spatial::SpatWeightedCompute>(computeOp);
auto oldComputeOpNum = oldComputeOp->getNumOperands();
auto yieldOp = mlir::cast<spatial::SpatYieldOp>(oldComputeOp.getBody().front().getTerminator());
if (yieldOp.getNumOperands() == oldComputeOp->getNumResults()) {
// No result was added, just add itself to the map
opToReplacedCompute[oldComputeOp.getOperation()] = oldComputeOp;
return;
}
// Add the results by inspecting its YieldOp
auto newResultTypes = yieldOp.getOperandTypes();
// Create a new ComputeOp with the new result type, but same operands
rewriter.setInsertionPoint(oldComputeOp);
auto newComputeOp = rewriter.create<spatial::SpatWeightedCompute>(
oldComputeOp->getLoc(), newResultTypes, oldComputeOp.getWeights(), oldComputeOp.getInputs());
newComputeOp.getBody().takeBody(oldComputeOp.getBody());
auto newComputeOpNum = newComputeOp->getNumOperands();
assert(oldComputeOpNum == newComputeOpNum);
// Since we replaced the old ComputeOp with a new one, we need to replace
// all its results' uses
for (size_t i = 0; i < oldComputeOp.getNumResults(); i++) {
mlir::Value oldResult = oldComputeOp.getResult(i);
mlir::Value newResult = newComputeOp.getResult(i);
// Replace the uses, except the uses of the compute ops which got deleted
// previously
rewriter.replaceAllUsesExcept(oldResult, newResult, oldComputeOpsReplaced);
}
// Finally, erase the old computeOp and update the map
opToReplacedCompute[oldComputeOp.getOperation()] = newComputeOp;
oldComputeOpsReplaced.insert(oldComputeOp.getOperation());
rewriter.setInsertionPoint(oldComputeOp);
rewriter.eraseOp(oldComputeOp);
}
mlir::Value
SpatialReducer::createImgConcatOp(llvm::SmallVector<llvm::SmallVector<llvm::SmallVector<OpAndResNum>>>& outputTiles,
mlir::Location& loc,
mlir::Type outputType) {
assert(reducesFinalized && "Cannot create ImgConcatOp before finalizing the reduce updates.");
// outputTiles are indexed like this: [channelTile][x][y]
auto tilesCount = outputTiles.size();
auto width = outputTiles[0].size();
auto height = outputTiles[0][0].size();
llvm::SmallVector<llvm::SmallVector<llvm::SmallVector<mlir::Value>>> remappedOutputTiles(
tilesCount, llvm::SmallVector<llvm::SmallVector<mlir::Value>>(width, llvm::SmallVector<mlir::Value>(height)));
for (size_t t = 0; t < tilesCount; t++)
for (size_t x = 0; x < width; x++)
for (size_t y = 0; y < height; y++)
remappedOutputTiles[t][x][y] = resolveValueFromOpAndResNum(outputTiles[t][x][y]);
return ::onnx_mlir::createImgConcatOp(remappedOutputTiles, rewriter, loc, outputType);
}
OpAndResNum SpatialReducer::applyAddMapReduction(llvm::SmallVector<ComputeAndResNum>& computeOps,
mlir::ConversionPatternRewriter& rewriter,
mlir::Value biasTile,
MapOperations mapOp) {
std::function<mlir::Value(const mlir::Value&)> postprocessing = nullptr;
if (mapOp != MapOperations::None) {
postprocessing = [&](const mlir::Value a) {
mlir::Value mapOperand = a;
if (biasTile)
mapOperand = rewriter.create<spatial::SpatVAddOp>(a.getLoc(), a.getType(), a, biasTile);
return createMapOperation(rewriter, mapOp, mapOperand);
};
}
return this->applyReducePattern(
computeOps,
[&](mlir::Value a, mlir::Value b) { return rewriter.create<spatial::SpatVAddOp>(a.getLoc(), a.getType(), a, b); },
/* preprocess = */ nullptr,
postprocessing);
}
} // namespace onnx_mlir