Fix some clang-tidy reports in MLIR (NFC)

Mostly replace uses of `container.size()` with `container.empty()` in
conditionals when applicable.

Author: joker-eph

mlir/lib/Transforms/LoopFusion.cpp

@@ -1565,7 +1565,7 @@ struct GreedyFusion {
             // producer scenarios will still go through profitability analysis
             // if only one of the stores is involved the producer-consumer
             // relationship of the candidate loops.
-            assert(producerStores.size() > 0 && "Expected producer store");
+            assert(!producerStores.empty() && "Expected producer store");
             if (producerStores.size() > 1)
               LLVM_DEBUG(llvm::dbgs() << "Skipping profitability analysis. Not "
                                          "supported for this case\n");

mlir/lib/Transforms/NormalizeMemRefs.cpp

@@ -460,7 +460,6 @@ void NormalizeMemRefs::normalizeFuncOpMemRefs(FuncOp funcOp,
       MemRefType newMemRefType = normalizeMemRefType(memrefType, b,
                                                      /*numSymbolicOperands=*/0);
       resultTypes.push_back(newMemRefType);
-      continue;
     }
 
     FunctionType newFuncType =

mlir/lib/Transforms/ParallelLoopCollapsing.cpp

@@ -28,11 +28,11 @@ struct ParallelLoopCollapsing
       // The common case for GPU dialect will be simplifying the ParallelOp to 3
       // arguments, so we do that here to simplify things.
       llvm::SmallVector<std::vector<unsigned>, 3> combinedLoops;
-      if (clCollapsedIndices0.size())
+      if (!clCollapsedIndices0.empty())
         combinedLoops.push_back(clCollapsedIndices0);
-      if (clCollapsedIndices1.size())
+      if (!clCollapsedIndices1.empty())
         combinedLoops.push_back(clCollapsedIndices1);
-      if (clCollapsedIndices2.size())
+      if (!clCollapsedIndices2.empty())
         combinedLoops.push_back(clCollapsedIndices2);
       collapseParallelLoops(op, combinedLoops);
     });