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
@@ -159,6 +159,7 @@
     DCHECK_GT(B->getCount(), 0);
     C->Count = B->getCount();
     B->copyToArray(C->Chunks);
+    B->clear();
     destroyBatch(ClassId, B);
     return true;
   }
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
@@ -444,6 +444,18 @@
     if (BytesPushed < PageSize)
       return 0; // Nothing new to release.
 
+    // Releasing smaller blocks is expensive, so we want to make sure that a
+    // significant amount of bytes are free, and that there has been a good
+    // amount of batches pushed to the freelist before attempting to release.
+    if (BlockSize < PageSize / 16U) {
+      if (BytesPushed < Sci->AllocatedUser / 16U)
+        return 0;
+      // We want 8x% to 9x% free bytes (the larger the bock, the lower the %).
+      if ((BytesInFreeList * 100U) / Sci->AllocatedUser <
+          (100U - 1U - BlockSize / 16U))
+        return 0;
+    }
+
     if (!Force) {
       const s32 IntervalMs = getReleaseToOsIntervalMs();
       if (IntervalMs < 0)
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
@@ -213,9 +213,7 @@
     return reinterpret_cast<const char *>(RegionInfoArray);
   }
 
-  static uptr getRegionInfoArraySize() {
-    return sizeof(RegionInfoArray);
-  }
+  static uptr getRegionInfoArraySize() { return sizeof(RegionInfoArray); }
 
   static BlockInfo findNearestBlock(const char *RegionInfoData, uptr Ptr) {
     const RegionInfo *RegionInfoArray =
@@ -458,6 +456,18 @@
     if (BytesPushed < PageSize)
       return 0; // Nothing new to release.
 
+    // Releasing smaller blocks is expensive, so we want to make sure that a
+    // significant amount of bytes are free, and that there has been a good
+    // amount of batches pushed to the freelist before attempting to release.
+    if (BlockSize < PageSize / 16U) {
+      if (BytesPushed < Region->AllocatedUser / 16U)
+        return 0;
+      // We want 8x% to 9x% free bytes (the larger the bock, the lower the %).
+      if ((BytesInFreeList * 100U) / Region->AllocatedUser <
+          (100U - 1U - BlockSize / 16U))
+        return 0;
+    }
+
     if (!Force) {
       const s32 IntervalMs = getReleaseToOsIntervalMs();
       if (IntervalMs < 0)