[AMDAIEFoldDmaWaits] Fold DMA wait operations across multi columns

compiler/plugins/target/AMD-AIE/iree-amd-aie/Transforms/AMDAIEControlCodeToTransaction.cpp

@@ -200,16 +200,46 @@ LogicalResult convertOp(AMDAIE::NpuAddressPatchOp op,
 }
 
 LogicalResult convertOp(AMDAIE::NpuDmaWaitOp op, TransactionBuilder &builder) {
-  for (Value token : op.getAsyncTokens()) {
-    auto pushToQueueOp =
-        dyn_cast_if_present<AMDAIE::NpuPushToQueueOp>(token.getDefiningOp());
+  // Collect all half DMA ops from the async tokens.
+  SmallVector<AMDAIE::NpuPushToQueueOp> pushToQueueOps;
+  for (Value asyncToken : op.getAsyncTokens()) {
+    auto pushToQueueOp = dyn_cast_if_present<AMDAIE::NpuPushToQueueOp>(
+        asyncToken.getDefiningOp());
     if (!pushToQueueOp) {
       return op.emitOpError()
-             << "should operate on an `amdaie.push_to_queue` op";
+             << "should operate on an `amdaie.push_to_queue` op async token";
     }
+    pushToQueueOps.push_back(pushToQueueOp);
+  }
+  // Sort the half DMA ops by channel, direction, row, and column.
+  std::sort(pushToQueueOps.begin(), pushToQueueOps.end(),
+            [](AMDAIE::NpuPushToQueueOp a, AMDAIE::NpuPushToQueueOp b) {
+              return std::make_tuple(a.getChannel(), a.getDirection(),
+                                     a.getRow(), a.getCol()) <
+                     std::make_tuple(b.getChannel(), b.getDirection(),
+                                     b.getRow(), b.getCol());
+            });
+  // Batch DMA operations with the same row, channel, and direction into a
+  // single TCT sync operation, as long as they have consecutive columns.
+  llvm::MapVector<AMDAIE::NpuPushToQueueOp, uint32_t> columnBatches;
+  for (auto pushToQueueOp : pushToQueueOps) {
+    if (!columnBatches.empty()) {
+      auto &[lastPushOp, lastColNum] = columnBatches.back();
+      if (lastPushOp.getRow() == pushToQueueOp.getRow() &&
+          lastPushOp.getCol() + lastColNum == pushToQueueOp.getCol() &&
+          lastPushOp.getDirection() == pushToQueueOp.getDirection() &&
+          lastPushOp.getChannel() == pushToQueueOp.getChannel()) {
+        ++lastColNum;
+        continue;
+      }
+    }
+    columnBatches.insert({pushToQueueOp, 1});
+  }
+  // Convert to TCT sync ops.
+  for (auto &[pushToQueueOp, colNum] : columnBatches) {
     if (failed(builder.appendTCTSync(
             pushToQueueOp.getCol(), pushToQueueOp.getRow(),
-            static_cast<uint32_t>(pushToQueueOp.getDirection()), 1, 1,
+            static_cast<uint32_t>(pushToQueueOp.getDirection()), 1, colNum,
             pushToQueueOp.getChannel()))) {
       return failure();
     }

compiler/plugins/target/AMD-AIE/iree-amd-aie/Transforms/AMDAIEFoldDmaWaits.cpp

@@ -16,13 +16,14 @@ namespace mlir::iree_compiler::AMDAIE {
 
 namespace {
 
+using DmaBdIdKey = std::pair<AMDAIE::TileOp, AMDAIE::ConnectionOp>;
+
 /// Utility function to determine whether a DMA wait op can be folded based on
 /// its half DMA copy operation.
-FailureOr<bool> canFoldBasedOnHalfDmaCpy(
+FailureOr<bool> canFoldByQueue(
     const AMDAIE::AMDAIEDeviceModel &deviceModel,
     AMDAIE::NpuHalfDmaCpyNdOp &npuHalfDmaCpyNdOp,
-    DenseMap<std::pair<AMDAIE::TileOp, AMDAIE::ConnectionOp>,
-             SmallVector<uint32_t>> &tileConnectToBdIdQueue) {
+    DenseMap<DmaBdIdKey, SmallVector<uint32_t>> &tileConnectToBdIdQueue) {
   // Retrieve the connection op.
   std::optional<AMDAIE::ConnectionOp> maybeConnectionOp =
       npuHalfDmaCpyNdOp.getConnectionOp();
@@ -101,13 +102,11 @@ FailureOr<bool> canFoldBasedOnHalfDmaCpy(
 /// Reverse traversal simplifies handling duplicate BD IDs, preventing
 /// the need to revisit and modify earlier operations after processing later
 /// ones.
-LogicalResult foldDmaWaits(const AMDAIE::AMDAIEDeviceModel &deviceModel,
-                           AMDAIE::ControlCodeOp controlCodeOp) {
+LogicalResult foldDmaWaitsByQueue(const AMDAIE::AMDAIEDeviceModel &deviceModel,
+                                  AMDAIE::ControlCodeOp controlCodeOp) {
   IRRewriter rewriter(controlCodeOp->getContext());
   std::vector<AMDAIE::NpuDmaWaitOp> waitOpsToErase;
-  DenseMap<std::pair<AMDAIE::TileOp, AMDAIE::ConnectionOp>,
-           SmallVector<uint32_t>>
-      tileConnectToBdIdQueue;
+  DenseMap<DmaBdIdKey, SmallVector<uint32_t>> tileConnectToBdIdQueue;
   // Traverse the control code in reverse.
   WalkResult res = controlCodeOp->walk<WalkOrder::PostOrder, ReverseIterator>(
       [&](AMDAIE::NpuDmaWaitOp waitOp) {
@@ -116,7 +115,7 @@ LogicalResult foldDmaWaits(const AMDAIE::AMDAIEDeviceModel &deviceModel,
           if (auto npuHalfDmaCpyNdOp =
                   dyn_cast_if_present<AMDAIE::NpuHalfDmaCpyNdOp>(
                       token.getDefiningOp())) {
-            FailureOr<bool> result = canFoldBasedOnHalfDmaCpy(
+            FailureOr<bool> result = canFoldByQueue(
                 deviceModel, npuHalfDmaCpyNdOp, tileConnectToBdIdQueue);
             if (failed(result)) return WalkResult::interrupt();
             toErase &= *result;
@@ -152,6 +151,179 @@ LogicalResult foldDmaWaits(const AMDAIE::AMDAIEDeviceModel &deviceModel,
   return success();
 }
 
+/// For each batch, combine the async tokens into a single NpuDmaWaitOp.
+LogicalResult eraseBatchOperations(IRRewriter &rewriter,
+                                   SmallVector<AMDAIE::NpuDmaWaitOp> &waitOps) {
+  // Skip if there are less than two DMA wait operations.
+  if (waitOps.size() < 2) return success();
+
+  SmallVector<Value> asyncTokens;
+  Operation *parentOp = waitOps[0]->getParentOp();
+  for (AMDAIE::NpuDmaWaitOp waitOp : waitOps) {
+    if (waitOp->getParentOp() != parentOp) {
+      return waitOp.emitError(
+          "DMA operations to be batched must belong to the same scope");
+    }
+    asyncTokens.append(waitOp.getAsyncTokens().begin(),
+                       waitOp.getAsyncTokens().end());
+  }
+
+  rewriter.setInsertionPointAfter(waitOps.back());
+  rewriter.create<AMDAIE::NpuDmaWaitOp>(waitOps.back().getLoc(), asyncTokens);
+  for (AMDAIE::NpuDmaWaitOp waitOp : waitOps) rewriter.eraseOp(waitOp);
+  return success();
+}
+
+/// Utility function to determine if a DMA wait operation can be folded into a
+/// a batch based on its half DMA copy operation.
+FailureOr<bool> canFoldByBatch(
+    const Operation *batchParentOp,
+    const DenseSet<AMDAIE::ConnectionOp> &connectionOps,
+    const DenseMap<DmaBdIdKey, DenseSet<uint32_t>> &dmaBdIdsMap,
+    DmaBdIdKey &currBdIdKey, uint32_t &currBdIdVal,
+    AMDAIE::NpuHalfDmaCpyNdOp currHalfDmaCpyNdOp) {
+  // Check if the current operation is in the same scope as the rest of the
+  // batch.
+  bool isSameScope = currHalfDmaCpyNdOp->getParentOp() == batchParentOp;
+
+  // Retrieve the connection op.
+  std::optional<AMDAIE::ConnectionOp> maybeConnectionOp =
+      currHalfDmaCpyNdOp.getConnectionOp();
+  if (!maybeConnectionOp) {
+    return currHalfDmaCpyNdOp.emitOpError()
+           << "expected to operate on an `amdaie.connection`";
+  }
+  AMDAIE::ConnectionOp connectionOp = maybeConnectionOp.value();
+  bool isDuplicateConnection = connectionOps.contains(connectionOp);
+
+  // Retrieve the flow op.
+  std::optional<AMDAIE::FlowOp> maybeFlowOp = connectionOp.getFlowOp();
+  if (!maybeFlowOp) {
+    return connectionOp.emitOpError()
+           << "expected to operate on an `amdaie.flow`";
+  }
+  AMDAIE::FlowOp flowOp = maybeFlowOp.value();
+  bool isPacketFlow = flowOp.getIsPacketFlow();
+
+  // Retrieve the BD ID op.
+  std::optional<AMDAIE::BdIdOp> maybeBdIdOp = currHalfDmaCpyNdOp.getBdIdOp();
+  if (!maybeBdIdOp) {
+    return currHalfDmaCpyNdOp.emitOpError()
+           << "must have a BD ID op to lower to "
+              "`amdaie.npu.write_bd`";
+  }
+  AMDAIE::BdIdOp bdIdOp = maybeBdIdOp.value();
+  currBdIdVal = getConstantIndexOrAssert(bdIdOp.getValue());
+
+  // Retrieve the tile op.
+  AMDAIE::TileOp tileOp =
+      dyn_cast_if_present<AMDAIE::TileOp>(bdIdOp.getTile().getDefiningOp());
+  if (!tileOp) {
+    return bdIdOp.emitOpError() << "must operate on an `amdaie.tile`";
+  }
+  currBdIdKey = {tileOp, connectionOp};
+
+  bool isDuplicateBdId = llvm::any_of(dmaBdIdsMap, [&](const auto &entry) {
+    return entry.first.first == tileOp && entry.second.contains(currBdIdVal);
+  });
+
+  // Can't fold wait op if:
+  // (1) the current connection op already occurs in the batch, or
+  // (2) the current BD ID on the same tile already occurs in the batch, or
+  // (3) the current operation is a packet flow, or
+  // (4) the batch is empty, or
+  // (5) the current operation is not in the same scope as the batch.
+  return !(isDuplicateConnection || isDuplicateBdId || isPacketFlow ||
+           connectionOps.empty() || !isSameScope);
+}
+
+/// Traverses the control code in reverse, ensuring that only one DMA wait op is
+/// retained for every batch of DMA copy operations.
+///
+/// Example Input:
+///   %0 = dma_cpy_nd(connection0)
+///   dma_wait(%0)
+///   %1 = dma_cpy_nd(connection1)
+///   %2 = dma_cpy_nd(connection2)
+///   %3 = dma_cpy_nd(connection3)
+///   dma_wait(%1)
+///   dma_wait(%2)
+///   dma_wait(%3)
+/// Example Output:
+///   %0 = dma_cpy_nd(connection0)
+///   %1 = dma_cpy_nd(connection1)
+///   %2 = dma_cpy_nd(connection2)
+///   %3 = dma_cpy_nd(connection3)
+///   dma_wait(%0, %1, %2, %3)
+/// Reverse traversal simplifies handling duplicate connections, preventing
+/// the need to revisit and modify earlier operations after processing later
+/// ones.
+LogicalResult foldDmaWaitsByBatch(AMDAIE::ControlCodeOp controlCodeOp) {
+  IRRewriter rewriter(controlCodeOp->getContext());
+  SmallVector<AMDAIE::NpuDmaWaitOp> waitOps;
+  DenseSet<AMDAIE::ConnectionOp> connectionOps;
+  DenseMap<DmaBdIdKey, DenseSet<uint32_t>> dmaBdIdsMap;
+
+  auto updateWithCurrBdId =
+      [&](bool canFold, DenseSet<AMDAIE::ConnectionOp> &connectionOps,
+          DenseMap<DmaBdIdKey, DenseSet<uint32_t>> &dmaBdIdsMap,
+          DmaBdIdKey &currBdIdKey, uint32_t currBdIdVal) {
+        assert(currBdIdKey.first && "TileOp must not be null");
+        assert(currBdIdKey.second && "ConnectionOp must not be null");
+        if (!canFold) {
+          // Clear the BD IDs for all the connections in the batch.
+          for (auto &entry : dmaBdIdsMap) {
+            ConnectionOp connectionOp = entry.first.second;
+            DenseSet<uint32_t> &bdIds = entry.second;
+            if (connectionOps.contains(connectionOp)) bdIds.clear();
+          }
+          connectionOps.clear();
+        }
+        connectionOps.insert(currBdIdKey.second);
+        dmaBdIdsMap[currBdIdKey].insert(currBdIdVal);
+      };
+
+  // Traverse the control code in reverse.
+  WalkResult res = controlCodeOp->walk<WalkOrder::PostOrder, ReverseIterator>(
+      [&](AMDAIE::NpuDmaWaitOp waitOp) {
+        bool toBatch = true;
+        Operation *batchParentOp =
+            waitOps.empty() ? waitOp->getParentOp() : waitOps[0]->getParentOp();
+        for (Value token : waitOp.getAsyncTokens()) {
+          if (auto npuHalfDmaCpyNdOp =
+                  dyn_cast_if_present<AMDAIE::NpuHalfDmaCpyNdOp>(
+                      token.getDefiningOp())) {
+            DmaBdIdKey currBdIdKey = {nullptr, nullptr};
+            uint32_t currBdIdVal = 0;
+            FailureOr<bool> result =
+                canFoldByBatch(batchParentOp, connectionOps, dmaBdIdsMap,
+                               currBdIdKey, currBdIdVal, npuHalfDmaCpyNdOp);
+            if (failed(result)) return WalkResult::interrupt();
+            toBatch &= *result;
+            updateWithCurrBdId(*result, connectionOps, dmaBdIdsMap, currBdIdKey,
+                               currBdIdVal);
+          }
+        }
+        // Process the previous batch of wait ops, and start a new batch.
+        if (!toBatch) {
+          // Since the controlcode is traversed in reverse order, we need to
+          // restore the original order of the DMA operations.
+          std::reverse(waitOps.begin(), waitOps.end());
+          if (failed(eraseBatchOperations(rewriter, waitOps)))
+            return WalkResult::interrupt();
+          waitOps.clear();
+        }
+        waitOps.push_back(waitOp);
+        return WalkResult::advance();
+      });
+
+  if (res.wasInterrupted()) return failure();
+  // Process the remaining wait ops.
+  std::reverse(waitOps.begin(), waitOps.end());
+  if (failed(eraseBatchOperations(rewriter, waitOps))) return failure();
+  return success();
+}
+
 class AMDAIEFoldDmaWaitsPass
     : public impl::AMDAIEFoldDmaWaitsBase<AMDAIEFoldDmaWaitsPass> {
  public:
@@ -181,7 +353,10 @@ void AMDAIEFoldDmaWaitsPass::runOnOperation() {
 
   WalkResult res = parentOp->walk([&](AMDAIE::WorkgroupOp workgroupOp) {
     AMDAIE::ControlCodeOp controlCodeOp = workgroupOp.getControlCode();
-    if (failed(foldDmaWaits(deviceModel, controlCodeOp))) {
+    if (failed(foldDmaWaitsByQueue(deviceModel, controlCodeOp))) {
+      return WalkResult::interrupt();
+    }
+    if (failed(foldDmaWaitsByBatch(controlCodeOp))) {
       return WalkResult::interrupt();
     }
     return WalkResult::advance();