diff --git a/compiler-rt/lib/scudo/standalone/local_cache.h b/compiler-rt/lib/scudo/standalone/local_cache.h
--- a/compiler-rt/lib/scudo/standalone/local_cache.h
+++ b/compiler-rt/lib/scudo/standalone/local_cache.h
@@ -10,6 +10,7 @@
 #define SCUDO_LOCAL_CACHE_H_
 
 #include "internal_defs.h"
+#include "platform.h"
 #include "report.h"
 #include "stats.h"
 
@@ -20,8 +21,8 @@
   typedef typename SizeClassAllocator::CompactPtrT CompactPtrT;
 
   struct TransferBatch {
-    static const u32 MaxNumCached = SizeClassMap::MaxNumCachedHint;
-    void setFromArray(CompactPtrT *Array, u32 N) {
+    static const u16 MaxNumCached = SizeClassMap::MaxNumCachedHint;
+    void setFromArray(CompactPtrT *Array, u16 N) {
       DCHECK_LE(N, MaxNumCached);
       Count = N;
       memcpy(Batch, Array, sizeof(Batch[0]) * Count);
@@ -34,19 +35,19 @@
     void copyToArray(CompactPtrT *Array) const {
       memcpy(Array, Batch, sizeof(Batch[0]) * Count);
     }
-    u32 getCount() const { return Count; }
-    CompactPtrT get(u32 I) const {
+    u16 getCount() const { return Count; }
+    CompactPtrT get(u16 I) const {
       DCHECK_LE(I, Count);
       return Batch[I];
     }
-    static u32 getMaxCached(uptr Size) {
+    static u16 getMaxCached(uptr Size) {
       return Min(MaxNumCached, SizeClassMap::getMaxCachedHint(Size));
     }
     TransferBatch *Next;
 
   private:
-    u32 Count;
     CompactPtrT Batch[MaxNumCached];
+    u16 Count;
   };
 
   void init(GlobalStats *S, SizeClassAllocator *A) {
@@ -128,9 +129,9 @@
 private:
   static const uptr NumClasses = SizeClassMap::NumClasses;
   static const uptr BatchClassId = SizeClassMap::BatchClassId;
-  struct PerClass {
-    u32 Count;
-    u32 MaxCount;
+  struct alignas(SCUDO_CACHE_LINE_SIZE) PerClass {
+    u16 Count;
+    u16 MaxCount;
     // Note: ClassSize is zero for the transfer batch.
     uptr ClassSize;
     CompactPtrT Chunks[2 * TransferBatch::MaxNumCached];
@@ -180,7 +181,7 @@
   }
 
   NOINLINE void drain(PerClass *C, uptr ClassId) {
-    const u32 Count = Min(C->MaxCount / 2, C->Count);
+    const u16 Count = Min(static_cast<u16>(C->MaxCount / 2), C->Count);
     TransferBatch *B =
         createBatch(ClassId, Allocator->decompactPtr(ClassId, C->Chunks[0]));
     if (UNLIKELY(!B))
diff --git a/compiler-rt/lib/scudo/standalone/primary32.h b/compiler-rt/lib/scudo/standalone/primary32.h
--- a/compiler-rt/lib/scudo/standalone/primary32.h
+++ b/compiler-rt/lib/scudo/standalone/primary32.h
@@ -351,7 +351,7 @@
     }
 
     const uptr Size = getSizeByClassId(ClassId);
-    const u32 MaxCount = TransferBatch::getMaxCached(Size);
+    const u16 MaxCount = TransferBatch::getMaxCached(Size);
     DCHECK_GT(MaxCount, 0U);
     // The maximum number of blocks we should carve in the region is dictated
     // by the maximum number of batches we want to fill, and the amount of
@@ -382,7 +382,8 @@
           C->createBatch(ClassId, reinterpret_cast<void *>(ShuffleArray[I]));
       if (UNLIKELY(!B))
         return nullptr;
-      const u32 N = Min(MaxCount, NumberOfBlocks - I);
+      // `MaxCount` is u16 so the result will also fit in u16.
+      const u16 N = static_cast<u16>(Min<u32>(MaxCount, NumberOfBlocks - I));
       B->setFromArray(&ShuffleArray[I], N);
       Sci->FreeList.push_back(B);
       I += N;
diff --git a/compiler-rt/lib/scudo/standalone/primary64.h b/compiler-rt/lib/scudo/standalone/primary64.h
--- a/compiler-rt/lib/scudo/standalone/primary64.h
+++ b/compiler-rt/lib/scudo/standalone/primary64.h
@@ -333,7 +333,7 @@
   NOINLINE TransferBatch *populateFreeList(CacheT *C, uptr ClassId,
                                            RegionInfo *Region) {
     const uptr Size = getSizeByClassId(ClassId);
-    const u32 MaxCount = TransferBatch::getMaxCached(Size);
+    const u16 MaxCount = TransferBatch::getMaxCached(Size);
 
     const uptr RegionBeg = Region->RegionBeg;
     const uptr MappedUser = Region->MappedUser;
@@ -392,7 +392,8 @@
                                       CompactPtrBase, ShuffleArray[I])));
       if (UNLIKELY(!B))
         return nullptr;
-      const u32 N = Min(MaxCount, NumberOfBlocks - I);
+      // `MaxCount` is u16 so the result will also fit in u16.
+      const u16 N = static_cast<u16>(Min<u32>(MaxCount, NumberOfBlocks - I));
       B->setFromArray(&ShuffleArray[I], N);
       Region->FreeList.push_back(B);
       I += N;
diff --git a/compiler-rt/lib/scudo/standalone/release.h b/compiler-rt/lib/scudo/standalone/release.h
--- a/compiler-rt/lib/scudo/standalone/release.h
+++ b/compiler-rt/lib/scudo/standalone/release.h
@@ -242,7 +242,7 @@
   if (BlockSize <= PageSize && PageSize % BlockSize == 0) {
     // Each chunk affects one page only.
     for (const auto &It : FreeList) {
-      for (u32 I = 0; I < It.getCount(); I++) {
+      for (u16 I = 0; I < It.getCount(); I++) {
         const uptr P = DecompactPtr(It.get(I)) - Recorder->getBase();
         if (P >= RoundedSize)
           continue;
@@ -256,7 +256,7 @@
     DCHECK_GE(RegionSize, BlockSize);
     const uptr LastBlockInRegion = ((RegionSize / BlockSize) - 1U) * BlockSize;
     for (const auto &It : FreeList) {
-      for (u32 I = 0; I < It.getCount(); I++) {
+      for (u16 I = 0; I < It.getCount(); I++) {
         const uptr P = DecompactPtr(It.get(I)) - Recorder->getBase();
         if (P >= RoundedSize)
           continue;
diff --git a/compiler-rt/lib/scudo/standalone/size_class_map.h b/compiler-rt/lib/scudo/standalone/size_class_map.h
--- a/compiler-rt/lib/scudo/standalone/size_class_map.h
+++ b/compiler-rt/lib/scudo/standalone/size_class_map.h
@@ -23,7 +23,7 @@
 }
 
 template <typename Config> struct SizeClassMapBase {
-  static u32 getMaxCachedHint(uptr Size) {
+  static u16 getMaxCachedHint(uptr Size) {
     DCHECK_NE(Size, 0);
     u32 N;
     // Force a 32-bit division if the template parameters allow for it.
@@ -31,7 +31,10 @@
       N = static_cast<u32>((1UL << Config::MaxBytesCachedLog) / Size);
     else
       N = (1U << Config::MaxBytesCachedLog) / static_cast<u32>(Size);
-    return Max(1U, Min(Config::MaxNumCachedHint, N));
+
+    // Note that Config::MaxNumCachedHint is u16 so the result is guaranteed to
+    // fit in u16.
+    return static_cast<u16>(Max(1U, Min<u32>(Config::MaxNumCachedHint, N)));
   }
 };
 
@@ -65,7 +68,7 @@
   static const uptr M = (1UL << S) - 1;
 
 public:
-  static const u32 MaxNumCachedHint = Config::MaxNumCachedHint;
+  static const u16 MaxNumCachedHint = Config::MaxNumCachedHint;
 
   static const uptr MaxSize = (1UL << Config::MaxSizeLog) + Config::SizeDelta;
   static const uptr NumClasses =
@@ -99,7 +102,7 @@
     return MidClass + 1 + scaledLog2(Size - 1, Config::MidSizeLog, S);
   }
 
-  static u32 getMaxCachedHint(uptr Size) {
+  static u16 getMaxCachedHint(uptr Size) {
     DCHECK_LE(Size, MaxSize);
     return Base::getMaxCachedHint(Size);
   }
@@ -178,7 +181,7 @@
   static constexpr LSBTable LTable = {};
 
 public:
-  static const u32 MaxNumCachedHint = Config::MaxNumCachedHint;
+  static const u16 MaxNumCachedHint = Config::MaxNumCachedHint;
 
   static const uptr NumClasses = ClassesSize + 1;
   static_assert(NumClasses < 256, "");
@@ -212,7 +215,7 @@
     return SzTable.Tab[scaledLog2(Size - 1, Config::MidSizeLog, S)];
   }
 
-  static u32 getMaxCachedHint(uptr Size) {
+  static u16 getMaxCachedHint(uptr Size) {
     DCHECK_LE(Size, MaxSize);
     return Base::getMaxCachedHint(Size);
   }
@@ -223,7 +226,7 @@
   static const uptr MinSizeLog = 5;
   static const uptr MidSizeLog = 8;
   static const uptr MaxSizeLog = 17;
-  static const u32 MaxNumCachedHint = 14;
+  static const u16 MaxNumCachedHint = 14;
   static const uptr MaxBytesCachedLog = 10;
   static const uptr SizeDelta = 0;
 };
@@ -235,7 +238,7 @@
   static const uptr MinSizeLog = 5;
   static const uptr MidSizeLog = 8;
   static const uptr MaxSizeLog = 17;
-  static const u32 MaxNumCachedHint = 12;
+  static const u16 MaxNumCachedHint = 12;
   static const uptr MaxBytesCachedLog = 10;
   static const uptr SizeDelta = Chunk::getHeaderSize();
 };
@@ -248,7 +251,7 @@
   static const uptr MinSizeLog = 4;
   static const uptr MidSizeLog = 6;
   static const uptr MaxSizeLog = 16;
-  static const u32 MaxNumCachedHint = 13;
+  static const u16 MaxNumCachedHint = 13;
   static const uptr MaxBytesCachedLog = 13;
 
   static constexpr u32 Classes[] = {
@@ -263,7 +266,7 @@
   static const uptr MinSizeLog = 4;
   static const uptr MidSizeLog = 7;
   static const uptr MaxSizeLog = 16;
-  static const u32 MaxNumCachedHint = 14;
+  static const u16 MaxNumCachedHint = 14;
   static const uptr MaxBytesCachedLog = 13;
 
   static constexpr u32 Classes[] = {
@@ -292,7 +295,7 @@
   static const uptr MinSizeLog = 4;
   static const uptr MidSizeLog = 8;
   static const uptr MaxSizeLog = 14;
-  static const u32 MaxNumCachedHint = 13;
+  static const u16 MaxNumCachedHint = 13;
   static const uptr MaxBytesCachedLog = 10;
   static const uptr SizeDelta = Chunk::getHeaderSize();
 #else
@@ -300,7 +303,7 @@
   static const uptr MinSizeLog = 3;
   static const uptr MidSizeLog = 7;
   static const uptr MaxSizeLog = 14;
-  static const u32 MaxNumCachedHint = 14;
+  static const u16 MaxNumCachedHint = 14;
   static const uptr MaxBytesCachedLog = 10;
   static const uptr SizeDelta = Chunk::getHeaderSize();
 #endif
@@ -315,7 +318,7 @@
   static const uptr MinSizeLog = 7;
   static const uptr MidSizeLog = 7;
   static const uptr MaxSizeLog = 7;
-  static const u32 MaxNumCachedHint = 8;
+  static const u16 MaxNumCachedHint = 8;
   static const uptr MaxBytesCachedLog = 10;
   static const uptr SizeDelta = 0;
 };
diff --git a/compiler-rt/lib/scudo/standalone/tests/combined_test.cpp b/compiler-rt/lib/scudo/standalone/tests/combined_test.cpp
--- a/compiler-rt/lib/scudo/standalone/tests/combined_test.cpp
+++ b/compiler-rt/lib/scudo/standalone/tests/combined_test.cpp
@@ -506,7 +506,7 @@
   static const scudo::uptr MinSizeLog = 10;
   static const scudo::uptr MidSizeLog = 10;
   static const scudo::uptr MaxSizeLog = 13;
-  static const scudo::u32 MaxNumCachedHint = 4;
+  static const scudo::u16 MaxNumCachedHint = 4;
   static const scudo::uptr MaxBytesCachedLog = 12;
   static const scudo::uptr SizeDelta = 0;
 };
diff --git a/compiler-rt/lib/scudo/standalone/tests/primary_test.cpp b/compiler-rt/lib/scudo/standalone/tests/primary_test.cpp
--- a/compiler-rt/lib/scudo/standalone/tests/primary_test.cpp
+++ b/compiler-rt/lib/scudo/standalone/tests/primary_test.cpp
@@ -176,7 +176,7 @@
       AllocationFailed = true;
       break;
     }
-    for (scudo::u32 J = 0; J < B->getCount(); J++)
+    for (scudo::u16 J = 0; J < B->getCount(); J++)
       memset(Allocator.decompactPtr(ClassId, B->get(J)), 'B', Size);
     Batches.push_back(B);
   }
diff --git a/compiler-rt/lib/scudo/standalone/tests/release_test.cpp b/compiler-rt/lib/scudo/standalone/tests/release_test.cpp
--- a/compiler-rt/lib/scudo/standalone/tests/release_test.cpp
+++ b/compiler-rt/lib/scudo/standalone/tests/release_test.cpp
@@ -130,22 +130,22 @@
 
 // Simplified version of a TransferBatch.
 template <class SizeClassMap> struct FreeBatch {
-  static const scudo::u32 MaxCount = SizeClassMap::MaxNumCachedHint;
+  static const scudo::u16 MaxCount = SizeClassMap::MaxNumCachedHint;
   void clear() { Count = 0; }
   void add(scudo::uptr P) {
     DCHECK_LT(Count, MaxCount);
     Batch[Count++] = P;
   }
-  scudo::u32 getCount() const { return Count; }
-  scudo::uptr get(scudo::u32 I) const {
+  scudo::u16 getCount() const { return Count; }
+  scudo::uptr get(scudo::u16 I) const {
     DCHECK_LE(I, Count);
     return Batch[I];
   }
   FreeBatch *Next;
 
 private:
-  scudo::u32 Count;
   scudo::uptr Batch[MaxCount];
+  scudo::u16 Count;
 };
 
 template <class SizeClassMap> void testReleaseFreeMemoryToOS() {
diff --git a/compiler-rt/lib/scudo/standalone/tests/size_class_map_test.cpp b/compiler-rt/lib/scudo/standalone/tests/size_class_map_test.cpp
--- a/compiler-rt/lib/scudo/standalone/tests/size_class_map_test.cpp
+++ b/compiler-rt/lib/scudo/standalone/tests/size_class_map_test.cpp
@@ -33,7 +33,7 @@
   static const scudo::uptr MinSizeLog = 5;
   static const scudo::uptr MidSizeLog = 5;
   static const scudo::uptr MaxSizeLog = 5;
-  static const scudo::u32 MaxNumCachedHint = 0;
+  static const scudo::u16 MaxNumCachedHint = 0;
   static const scudo::uptr MaxBytesCachedLog = 0;
   static const scudo::uptr SizeDelta = 0;
 };
@@ -48,7 +48,7 @@
   static const scudo::uptr MinSizeLog = 4;
   static const scudo::uptr MidSizeLog = 8;
   static const scudo::uptr MaxSizeLog = 63;
-  static const scudo::u32 MaxNumCachedHint = 128;
+  static const scudo::u16 MaxNumCachedHint = 128;
   static const scudo::uptr MaxBytesCachedLog = 16;
   static const scudo::uptr SizeDelta = 0;
 };
diff --git a/compiler-rt/lib/scudo/standalone/tools/compute_size_class_config.cpp b/compiler-rt/lib/scudo/standalone/tools/compute_size_class_config.cpp
--- a/compiler-rt/lib/scudo/standalone/tools/compute_size_class_config.cpp
+++ b/compiler-rt/lib/scudo/standalone/tools/compute_size_class_config.cpp
@@ -140,7 +140,7 @@
   static const uptr MinSizeLog = %zu;
   static const uptr MidSizeLog = %zu;
   static const uptr MaxSizeLog = %zu;
-  static const u32 MaxNumCachedHint = 14;
+  static const u16 MaxNumCachedHint = 14;
   static const uptr MaxBytesCachedLog = 14;
 
   static constexpr u32 Classes[] = {)",