diff --git a/clang/include/clang/Basic/Builtins.def b/clang/include/clang/Basic/Builtins.def
index ec39e926889b9..e222ba8943f74 100644
--- a/clang/include/clang/Basic/Builtins.def
+++ b/clang/include/clang/Basic/Builtins.def
@@ -633,6 +633,7 @@ BUILTIN(__builtin_vsscanf, "icC*RcC*Ra", "FS:1:")
 BUILTIN(__builtin_thread_pointer, "v*", "nc")
 BUILTIN(__builtin_launder, "v*v*", "ntE")
 LANGBUILTIN(__builtin_is_constant_evaluated, "b", "nE", CXX_LANG)
+LANGBUILTIN(__builtin_clear_padding, "v.", "n", CXX_LANG)
 
 // GCC exception builtins
 BUILTIN(__builtin_eh_return, "vzv*", "r") // FIXME: Takes intptr_t, not size_t!
diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
index 65d9862621061..006afcd42c9e4 100644
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -60,9 +60,14 @@
 #include "llvm/Support/ScopedPrinter.h"
 #include "llvm/TargetParser/AArch64TargetParser.h"
 #include "llvm/TargetParser/X86TargetParser.h"
+#include <algorithm>
 #include <optional>
 #include <sstream>
 
+
+
+#include <iostream>
+
 using namespace clang;
 using namespace CodeGen;
 using namespace llvm;
@@ -2456,6 +2461,139 @@ static RValue EmitHipStdParUnsupportedBuiltin(CodeGenFunction *CGF,
   return RValue::get(CGF->Builder.CreateCall(UBF, Args));
 }
 
+template <class T>
+void RecursivelyClearPaddingImpl(CodeGenFunction &CGF, Value *Ptr, QualType Ty, size_t CurrentStartOffset, size_t &RunningOffset, T&& WriteZeroAtOffset);
+
+template <class T>
+void ClearPaddingStruct(CodeGenFunction &CGF, Value *Ptr, QualType Ty, StructType *ST, 
+                        size_t CurrentStartOffset, size_t &RunningOffset, T&& WriteZeroAtOffset) {
+  std::cerr << "\n struct! " << ST->getName().data() << std::endl;
+  auto SL = CGF.CGM.getModule().getDataLayout().getStructLayout(ST);
+
+  auto R = dyn_cast<CXXRecordDecl>(Ty->getAsRecordDecl());
+  if(!R) {
+    std::cerr << "\n not a CXXRecordDecl" << std::endl;
+
+  }
+  const ASTRecordLayout &ASTLayout = CGF.getContext().getASTRecordLayout(R);
+  for (auto Base : R->bases()) {
+    std::cerr << "\n\n base!"  << std::endl;
+    // Zero padding between base elements.
+    auto BaseRecord = cast<CXXRecordDecl>(Base.getType()->getAsRecordDecl());
+    auto Offset = static_cast<size_t>(
+        ASTLayout.getBaseClassOffset(BaseRecord).getQuantity());
+    // Recursively zero out base classes.
+    auto Index = SL->getElementContainingOffset(Offset);
+    Value *Idx = CGF.Builder.getSize(Index);
+    llvm::Type *CurrentBaseType = CGF.ConvertTypeForMem(Base.getType());
+    Value *BaseElement = CGF.Builder.CreateGEP(CurrentBaseType, Ptr, Idx);
+    RecursivelyClearPaddingImpl(CGF, BaseElement, Base.getType(), CurrentStartOffset + Offset, RunningOffset, WriteZeroAtOffset);
+  }
+
+  size_t NumFields = std::distance(R->field_begin(), R->field_end());
+  auto CurrentField = R->field_begin();
+  for (size_t I = 0; I < NumFields; ++I, ++CurrentField) {
+    // Size needs to be in bytes so we can compare it later.
+    auto Offset = ASTLayout.getFieldOffset(I) / 8;
+    auto Index = SL->getElementContainingOffset(Offset);
+    Value *Idx = CGF.Builder.getSize(Index);
+    llvm::Type *CurrentFieldType = CGF.ConvertTypeForMem(CurrentField->getType());
+    Value *Element = CGF.Builder.CreateGEP(CurrentFieldType, Ptr, Idx);
+    RecursivelyClearPaddingImpl(CGF, Element, CurrentField->getType(), CurrentStartOffset + Offset, RunningOffset, WriteZeroAtOffset);
+  }
+}
+
+template <class T>
+void ClearPaddingConstantArray(CodeGenFunction &CGF, Value *Ptr, llvm::Type *Type, ConstantArrayType const *AT, 
+                               size_t CurrentStartOffset, size_t &RunningOffset, T&& WriteZeroAtOffset) {
+  for (size_t ArrIndex = 0; ArrIndex < AT->getSize().getLimitedValue();
+       ++ArrIndex) {
+
+    QualType ElementQualType = AT->getElementType();
+
+    auto *ElementRecord = ElementQualType->getAsRecordDecl();
+    if(!ElementRecord){
+      std::cerr<< "\n\n null!" << std::endl;
+    }
+    auto ElementAlign =ElementRecord?
+        CGF.getContext().getASTRecordLayout(ElementRecord).getAlignment():
+        CGF.getContext().getTypeAlignInChars(ElementQualType);
+
+      std::cerr<< "\n\n align: " << ElementAlign.getQuantity() << std::endl;
+    
+    // Value *Idx = CGF.Builder.getSize(0);
+    // Value *ArrayElement = CGF.Builder.CreateGEP(ElementType, Ptr, Idx);
+
+    Address FieldElementAddr{Ptr, Type, ElementAlign};
+
+    auto Element =
+        CGF.Builder.CreateConstArrayGEP(FieldElementAddr, ArrIndex);
+    auto *ElementType = CGF.ConvertTypeForMem(ElementQualType);
+    auto AllocSize = CGF.CGM.getModule().getDataLayout().getTypeAllocSize(ElementType);
+    std::cerr << "\n\n clearing array index! " << ArrIndex << std::endl;
+    RecursivelyClearPaddingImpl(CGF, Element.getPointer(), ElementQualType, CurrentStartOffset + ArrIndex * AllocSize.getKnownMinValue(), RunningOffset, WriteZeroAtOffset);
+  }
+}
+
+template <class T>
+void RecursivelyClearPaddingImpl(CodeGenFunction &CGF, Value *Ptr, QualType Ty, size_t CurrentStartOffset, 
+                                 size_t& RunningOffset, T&& WriteZeroAtOffset) {
+
+  std::cerr << "\n\n  zero padding before current  ["<< RunningOffset << ", " << CurrentStartOffset<< ")"<< std::endl;
+  for (; RunningOffset < CurrentStartOffset; ++RunningOffset) {
+    WriteZeroAtOffset(RunningOffset);
+  }
+  auto Type = CGF.ConvertTypeForMem(Ty);
+  auto Size = CGF.CGM.getModule()
+                    .getDataLayout()
+                    .getTypeSizeInBits(Type)
+                    .getKnownMinValue() / 8;
+
+  if (auto *AT = dyn_cast<ConstantArrayType>(Ty)) {
+    ClearPaddingConstantArray(CGF, Ptr, Type, AT, CurrentStartOffset, RunningOffset, WriteZeroAtOffset);
+  }
+  else if (auto *ST = dyn_cast<StructType>(Type); ST && Ty->isRecordType()) {
+    ClearPaddingStruct(CGF, Ptr, Ty, ST, CurrentStartOffset, RunningOffset, WriteZeroAtOffset);
+  } else if (Ty->isAtomicType()) {
+    RecursivelyClearPaddingImpl(CGF, Ptr, Ty.getAtomicUnqualifiedType(), CurrentStartOffset, RunningOffset, WriteZeroAtOffset);
+  } else {
+    std::cerr << "\n\n increment running offset from: " << RunningOffset << " to " << RunningOffset + Size << std::endl;
+    RunningOffset = std::max(RunningOffset, CurrentStartOffset + static_cast<size_t>(Size));
+  }
+
+}
+
+static void RecursivelyClearPadding(CodeGenFunction &CGF, Value *Ptr,
+                                        QualType Ty) {
+  auto *I8Ptr = CGF.Builder.CreateBitCast(Ptr, CGF.Int8PtrTy);
+  auto *Zero = ConstantInt::get(CGF.Int8Ty, 0);
+  auto WriteZeroAtOffset = [&](size_t Offset) {
+    auto *Index = ConstantInt::get(CGF.IntTy, Offset);
+    auto *Element = CGF.Builder.CreateGEP(CGF.Int8Ty, I8Ptr, Index);
+    CGF.Builder.CreateAlignedStore(
+        Zero, Element,
+        CharUnits::One().alignmentAtOffset(CharUnits::fromQuantity(Offset)));
+  };
+
+
+  size_t RunningOffset = 0;
+
+  RecursivelyClearPaddingImpl(CGF, Ptr, Ty, 0, RunningOffset, WriteZeroAtOffset);
+
+  // Clear tail padding
+  auto Type = CGF.ConvertTypeForMem(Ty);
+  
+  auto Size = CGF.CGM.getModule()
+                    .getDataLayout()
+                    .getTypeAllocSize(Type)
+                    .getKnownMinValue();
+
+  std::cerr << "\n\n  zero tail padding  ["<< RunningOffset << ", " << Size << ")"<< std::endl;
+  for (; RunningOffset < Size; ++RunningOffset) {
+    WriteZeroAtOffset(RunningOffset);
+  }
+}
+
 RValue CodeGenFunction::EmitBuiltinExpr(const GlobalDecl GD, unsigned BuiltinID,
                                         const CallExpr *E,
                                         ReturnValueSlot ReturnValue) {
@@ -4315,6 +4453,13 @@ RValue CodeGenFunction::EmitBuiltinExpr(const GlobalDecl GD, unsigned BuiltinID,
 
     return RValue::get(Ptr);
   }
+  case Builtin::BI__builtin_clear_padding: {
+    const Expr *Op = E->getArg(0);
+    Value *Address = EmitScalarExpr(Op);
+    auto PointeeTy = Op->getType()->getPointeeType();
+    RecursivelyClearPadding(*this, Address, PointeeTy);
+    return RValue::get(nullptr);
+  }
   case Builtin::BI__sync_fetch_and_add:
   case Builtin::BI__sync_fetch_and_sub:
   case Builtin::BI__sync_fetch_and_or:
diff --git a/clang/lib/Sema/SemaChecking.cpp b/clang/lib/Sema/SemaChecking.cpp
index 77c8334f3ca25..cf34171a56399 100644
--- a/clang/lib/Sema/SemaChecking.cpp
+++ b/clang/lib/Sema/SemaChecking.cpp
@@ -2327,6 +2327,25 @@ Sema::CheckBuiltinFunctionCall(FunctionDecl *FDecl, unsigned BuiltinID,
   }
   case Builtin::BI__builtin_launder:
     return SemaBuiltinLaunder(*this, TheCall);
+  case Builtin::BI__builtin_clear_padding: {
+    const Expr *PtrArg = TheCall->getArg(0)->IgnoreParenImpCasts();
+    const QualType PtrArgType = PtrArg->getType();
+    if (!PtrArgType->isPointerType()) {
+      Diag(PtrArg->getBeginLoc(), diag::err_typecheck_convert_incompatible)
+          << PtrArgType << "pointer" << 1 << 0 << 3 << 1 << PtrArgType
+          << "pointer";
+      return ExprError();
+    }
+    if (PtrArgType->getPointeeType().isConstQualified()) {
+      Diag(PtrArg->getBeginLoc(), diag::err_typecheck_assign_const)
+          << TheCall->getSourceRange() << 5 /*ConstUnknown*/;
+      return ExprError();
+    }
+    if (RequireCompleteType(PtrArg->getBeginLoc(), PtrArgType->getPointeeType(),
+                            diag::err_typecheck_decl_incomplete_type))
+      return ExprError();
+    break;
+  }
   case Builtin::BI__sync_fetch_and_add:
   case Builtin::BI__sync_fetch_and_add_1:
   case Builtin::BI__sync_fetch_and_add_2:
diff --git a/clang/test/CodeGenCXX/builtin-clear-padding-codegen.cpp b/clang/test/CodeGenCXX/builtin-clear-padding-codegen.cpp
new file mode 100644
index 0000000000000..54455e6699849
--- /dev/null
+++ b/clang/test/CodeGenCXX/builtin-clear-padding-codegen.cpp
@@ -0,0 +1,112 @@
+// RUN: %clang_cc1 -triple=x86_64-linux-gnu -emit-llvm -o - %s | FileCheck %s
+
+struct alignas(4) Foo {
+  char a;
+  alignas(2) char b;
+};
+
+struct alignas(4) Bar {
+  char c;
+  alignas(2) char d;
+};
+
+struct alignas(4) Baz : Foo {
+  char e;
+  Bar f;
+};
+
+// Baz structure:
+// "a", PAD_1, "b", PAD_2, "c", PAD_3, PAD_4, PAD_5, "c", PAD_6, "d", PAD_7
+// %struct.Baz = type { %struct.Foo, i8, [3 x i8], %struct.Bar }
+// %struct.Foo = type { i8, i8, i8, i8 }
+// %struct.Bar = type { i8, i8, i8, i8 }
+
+// CHECK-LABEL: define void @_Z7testBazP3Baz(%struct.Baz* %baz)
+// CHECK: [[ADDR:%.*]] = alloca %struct.Baz*
+// CHECK: store %struct.Baz* %baz, %struct.Baz** [[ADDR]]
+// CHECK: [[BAZ:%.*]] = load %struct.Baz*, %struct.Baz** [[ADDR]]
+// CHECK: [[BAZ_RAW_PTR:%.*]] = bitcast %struct.Baz* [[BAZ]] to i8*
+
+// CHECK: [[FOO_BASE:%.*]] = getelementptr inbounds %struct.Baz, %struct.Baz* [[BAZ]], i32 0, i32 0
+// CHECK: [[FOO_RAW_PTR:%.*]] = bitcast %struct.Foo* [[FOO_BASE]] to i8*
+// CHECK: [[PAD_1:%.*]] = getelementptr i8, i8* [[FOO_RAW_PTR]], i32 1
+// CHECK: store i8 0, i8* [[PAD_1]]
+// CHECK: [[PAD_2:%.*]] = getelementptr i8, i8* [[FOO_RAW_PTR]], i32 3
+// CHECK: store i8 0, i8* [[PAD_2]]
+
+// CHECK: [[PAD_3:%.*]] = getelementptr i8, i8* [[BAZ_RAW_PTR]], i32 5
+// CHECK: store i8 0, i8* [[PAD_3]]
+// CHECK: [[PAD_4:%.*]] = getelementptr i8, i8* [[BAZ_RAW_PTR]], i32 6
+// CHECK: store i8 0, i8* [[PAD_4]]
+// CHECK: [[PAD_5:%.*]] = getelementptr i8, i8* [[BAZ_RAW_PTR]], i32 7
+// CHECK: store i8 0, i8* [[PAD_5]]
+
+// CHECK: [[BAR_MEMBER:%.*]] = getelementptr inbounds %struct.Baz, %struct.Baz* [[BAZ]], i32 0, i32 3
+// CHECK: [[BAR_RAW_PTR:%.*]] = bitcast %struct.Bar* [[BAR_MEMBER]] to i8*
+// CHECK: [[PAD_6:%.*]] = getelementptr i8, i8* [[BAR_RAW_PTR]], i32 1
+// CHECK: store i8 0, i8* [[PAD_6]]
+// CHECK: [[PAD_7:%.*]] = getelementptr i8, i8* [[BAR_RAW_PTR]], i32 3
+// CHECK: store i8 0, i8* [[PAD_7]]
+// CHECK: ret void
+void testBaz(Baz *baz) {
+  __builtin_clear_padding(baz);
+}
+
+struct UnsizedTail {
+  int size;
+  alignas(8) char buf[];
+
+  UnsizedTail(int size) : size(size) {}
+};
+
+// UnsizedTail structure:
+// "size", PAD_1, PAD_2, PAD_3, PAD_4
+// %struct.UnsizedTail = type { i32, [4 x i8], [0 x i8] }
+
+// CHECK-LABEL: define void @_Z15testUnsizedTailP11UnsizedTail(%struct.UnsizedTail* %u)
+// CHECK: [[U_ADDR:%.*]] = alloca %struct.UnsizedTail*
+// CHECK: store %struct.UnsizedTail* %u, %struct.UnsizedTail** [[U_ADDR]]
+// CHECK: [[U:%.*]] = load %struct.UnsizedTail*, %struct.UnsizedTail** [[U_ADDR]]
+// CHECK: [[U_RAW_PTR:%.*]] = bitcast %struct.UnsizedTail* [[U]] to i8*
+// CHECK: [[PAD_1:%.*]] = getelementptr i8, i8* [[U_RAW_PTR]], i32 4
+// CHECK: store i8 0, i8* [[PAD_1]]
+// CHECK: [[PAD_2:%.*]] = getelementptr i8, i8* [[U_RAW_PTR]], i32 5
+// CHECK: store i8 0, i8* [[PAD_2]]
+// CHECK: [[PAD_3:%.*]] = getelementptr i8, i8* [[U_RAW_PTR]], i32 6
+// CHECK: store i8 0, i8* [[PAD_3]]
+// CHECK: [[PAD_4:%.*]] = getelementptr i8, i8* [[U_RAW_PTR]], i32 7
+// CHECK: store i8 0, i8* [[PAD_4]]
+// CHECK: ret void
+void testUnsizedTail(UnsizedTail *u) {
+  __builtin_clear_padding(u);
+}
+
+struct ArrOfStructsWithPadding {
+  Bar bars[2];
+};
+
+// ArrOfStructsWithPadding structure:
+// "c" (1), PAD_1, "d" (1), PAD_2, "c" (2), PAD_3, "d" (2), PAD_4
+// %struct.ArrOfStructsWithPadding = type { [2 x %struct.Bar] }
+
+// CHECK-LABEL: define void @_Z27testArrOfStructsWithPaddingP23ArrOfStructsWithPadding(%struct.ArrOfStructsWithPadding* %arr)
+// CHECK: [[ARR_ADDR:%.*]] = alloca %struct.ArrOfStructsWithPadding*
+// CHECK: store %struct.ArrOfStructsWithPadding* %arr, %struct.ArrOfStructsWithPadding** [[ARR_ADDR]]
+// CHECK: [[ARR:%.*]] = load %struct.ArrOfStructsWithPadding*, %struct.ArrOfStructsWithPadding** [[ARR_ADDR]]
+// CHECK: [[BARS:%.*]] = getelementptr inbounds %struct.ArrOfStructsWithPadding, %struct.ArrOfStructsWithPadding* [[ARR]], i32 0, i32 0
+// CHECK: [[FIRST:%.*]] = getelementptr inbounds [2 x %struct.Bar], [2 x %struct.Bar]* [[BARS]], i64 0, i64 0
+// CHECK: [[FIRST_RAW_PTR:%.*]] = bitcast %struct.Bar* [[FIRST]] to i8*
+// CHECK: [[PAD_1:%.*]] = getelementptr i8, i8* [[FIRST_RAW_PTR]], i32 1
+// CHECK: store i8 0, i8* [[PAD_1]]
+// CHECK: [[PAD_2:%.*]] = getelementptr i8, i8* %4, i32 3
+// CHECK: store i8 0, i8* [[PAD_2]]
+// CHECK: [[SECOND:%.*]] = getelementptr inbounds [2 x %struct.Bar], [2 x %struct.Bar]* [[BARS]], i64 0, i64 1
+// CHECK: [[SECOND_RAW_PTR:%.*]] = bitcast %struct.Bar* [[SECOND]] to i8*
+// CHECK: [[PAD_3:%.*]] = getelementptr i8, i8* [[SECOND_RAW_PTR]], i32 1
+// CHECK: store i8 0, i8* [[PAD_3]]
+// CHECK: [[PAD_4:%.*]] = getelementptr i8, i8* [[SECOND_RAW_PTR]], i32 3
+// CHECK: store i8 0, i8* [[PAD_4]]
+// CHECK: ret void
+void testArrOfStructsWithPadding(ArrOfStructsWithPadding *arr) {
+  __builtin_clear_padding(arr);
+}
diff --git a/clang/test/CodeGenCXX/builtin-clear-padding.cpp b/clang/test/CodeGenCXX/builtin-clear-padding.cpp
new file mode 100644
index 0000000000000..b9060e00d3203
--- /dev/null
+++ b/clang/test/CodeGenCXX/builtin-clear-padding.cpp
@@ -0,0 +1,556 @@
+// RUN: mkdir -p %t
+// RUN: %clang++ %s -o %t/run
+// RUN: %t/run
+
+#include <cassert>
+#include <cstdio>
+#include <cstring>
+#include <new>
+
+template <class T>
+void print_bytes(const T *object)
+{
+  auto size = sizeof(T);
+  const unsigned char * const bytes = reinterpret_cast<const unsigned char *>(object);
+  size_t i;
+
+  fprintf(stderr, "[ ");
+  for(i = 0; i < size; i++)
+  {
+    fprintf(stderr, "%02x ", bytes[i]);
+  }
+  fprintf(stderr, "]\n");
+}
+
+template <size_t A1, size_t A2, class T>
+struct alignas(A1) BasicWithPadding {
+  T x;
+  alignas(A2) T y;
+};
+
+template <size_t A1, size_t A2, size_t N, class T>
+struct alignas(A1) SpacedArrayMembers {
+  T x[N];
+  alignas(A2) char c;
+  T y[N];
+};
+
+template <size_t A1, size_t A2, class T>
+struct alignas(A1) PaddedPointerMembers {
+  T *x;
+  alignas(A2) T *y;
+};
+
+template <size_t A1, size_t A2, size_t A3, class T>
+struct alignas(A1) ThreeMembers {
+  T x;
+  alignas(A2) T y;
+  alignas(A3) T z;
+};
+
+template <class T>
+struct Normal {
+  T a;
+  T b;
+};
+
+template <class T>
+struct X {
+  T x;
+};
+
+template <class T>
+struct Z {
+  T z;
+};
+
+template <size_t A, class T>
+struct YZ : public Z<T> {
+  alignas(A) T y;
+};
+
+template <size_t A1, size_t A2, class T>
+struct alignas(A1) HasBase : public X<T>, public YZ<A2, T> {
+  T a;
+  alignas(A2) T b;
+};
+
+template <size_t A1, size_t A2, class T>
+void testAllForType(T a, T b, T c, T d) {
+  using B = BasicWithPadding<A1, A2, T>;
+  B basic1;
+  memset(&basic1, 0, sizeof(B));
+  basic1.x = a;
+  basic1.y = b;
+  B basic2;
+  memset(&basic2, 42, sizeof(B));
+  basic2.x = a;
+  basic2.y = b;
+  assert(memcmp(&basic1, &basic2, sizeof(B)) != 0);
+  __builtin_clear_padding(&basic2);
+  assert(memcmp(&basic1, &basic2, sizeof(B)) == 0);
+  using A = SpacedArrayMembers<A1, A2, 2, T>;
+  A arr1;
+  memset(&arr1, 0, sizeof(A));
+  arr1.x[0] = a;
+  arr1.x[1] = b;
+  arr1.y[0] = c;
+  arr1.y[1] = d;
+  A arr2;
+  memset(&arr2, 42, sizeof(A));
+  arr2.x[0] = a;
+  arr2.x[1] = b;
+  arr2.y[0] = c;
+  arr2.y[1] = d;
+  arr2.c = 0;
+  assert(memcmp(&arr1, &arr2, sizeof(A)) != 0);
+  __builtin_clear_padding(&arr2);
+  assert(memcmp(&arr1, &arr2, sizeof(A)) == 0);
+
+  using P = PaddedPointerMembers<A1, A2, T>;
+  P ptr1;
+  memset(&ptr1, 0, sizeof(P));
+  ptr1.x = &a;
+  ptr1.y = &b;
+  P ptr2;
+  memset(&ptr2, 42, sizeof(P));
+  ptr2.x = &a;
+  ptr2.y = &b;
+  assert(memcmp(&ptr1, &ptr2, sizeof(P)) != 0);
+  __builtin_clear_padding(&ptr2);
+  assert(memcmp(&ptr1, &ptr2, sizeof(P)) == 0);
+
+  using Three = ThreeMembers<A1, A2, A2, T>;
+  Three three1;
+  memset(&three1, 0, sizeof(Three));
+  three1.x = a;
+  three1.y = b;
+  three1.z = c;
+  Three three2;
+  memset(&three2, 42, sizeof(Three));
+  three2.x = a;
+  three2.y = b;
+  three2.z = c;
+  __builtin_clear_padding(&three2);
+  assert(memcmp(&three1, &three2, sizeof(Three)) == 0);
+
+  using N = Normal<T>;
+  N normal1;
+  memset(&normal1, 0, sizeof(N));
+  normal1.a = a;
+  normal1.b = b;
+  N normal2;
+  memset(&normal2, 42, sizeof(N));
+  normal2.a = a;
+  normal2.b = b;
+  __builtin_clear_padding(&normal2);
+  assert(memcmp(&normal1, &normal2, sizeof(N)) == 0);
+
+  using H = HasBase<A1, A2, T>;
+  H base1;
+  memset(&base1, 0, sizeof(H));
+  base1.a = a;
+  base1.b = b;
+  base1.x = c;
+  base1.y = d;
+  base1.z = a;
+  H base2;
+  memset(&base2, 42, sizeof(H));
+  base2.a = a;
+  base2.b = b;
+  base2.x = c;
+  base2.y = d;
+  base2.z = a;
+  assert(memcmp(&base1, &base2, sizeof(H)) != 0);
+  __builtin_clear_padding(&base2);
+  unsigned i = 0;
+  assert(memcmp(&base1, &base2, sizeof(H)) == 0);
+}
+
+struct UnsizedTail {
+  int size;
+  alignas(8) char buf[];
+
+  UnsizedTail(int size) : size(size) {}
+};
+
+void otherTests() {
+  const size_t size1 = sizeof(UnsizedTail) + 4;
+  char buff1[size1];
+  char buff2[size1];
+  memset(buff1, 0, size1);
+  memset(buff2, 42, size1);
+  auto *u1 = new (buff1) UnsizedTail(4);
+  u1->buf[0] = 1;
+  u1->buf[1] = 2;
+  u1->buf[2] = 3;
+  u1->buf[3] = 4;
+  auto *u2 = new (buff2) UnsizedTail(4);
+  u2->buf[0] = 1;
+  u2->buf[1] = 2;
+  u2->buf[2] = 3;
+  u2->buf[3] = 4;
+  assert(memcmp(u1, u2, sizeof(UnsizedTail)) != 0);
+  __builtin_clear_padding(u2);
+  assert(memcmp(u1, u2, sizeof(UnsizedTail)) == 0);
+
+  using B = BasicWithPadding<8, 4, char>;
+  auto *basic1 = new B;
+  memset(basic1, 0, sizeof(B));
+  basic1->x = 1;
+  basic1->y = 2;
+  auto *basic2 = new B;
+  memset(basic2, 42, sizeof(B));
+  basic2->x = 1;
+  basic2->y = 2;
+  assert(memcmp(basic1, basic2, sizeof(B)) != 0);
+  __builtin_clear_padding(basic2);
+  assert(memcmp(basic1, basic2, sizeof(B)) == 0);
+  delete basic2;
+  delete basic1;
+
+  using B = BasicWithPadding<8, 4, char>;
+  B *basic3 = new B;
+  memset(basic3, 0, sizeof(B));
+  basic3->x = 1;
+  basic3->y = 2;
+  B *basic4 = new B;
+  memset(basic4, 42, sizeof(B));
+  basic4->x = 1;
+  basic4->y = 2;
+  assert(memcmp(basic3, basic4, sizeof(B)) != 0);
+  __builtin_clear_padding(const_cast<volatile B *>(basic4));
+  assert(memcmp(basic3, basic4, sizeof(B)) == 0);
+  delete basic4;
+  delete basic3;
+}
+
+struct Foo {
+  int x;
+  int y;
+};
+
+typedef float Float4Vec __attribute__((ext_vector_type(4)));
+typedef float Float3Vec __attribute__((ext_vector_type(3)));
+
+void primitiveTests()
+{
+  // no padding
+  {
+    int i1 = 42, i2 = 42;
+    __builtin_clear_padding(&i1); // does nothing
+    assert(i1 == 42);
+    assert(memcmp(&i1, &i2, sizeof(int)) == 0);
+  }
+
+  // long double
+  {
+    long double d1, d2;
+    memset(&d1, 42, sizeof(long double));
+    memset(&d2, 0, sizeof(long double));
+
+    d1 = 3.0L;
+    d2 = 3.0L;
+
+    __builtin_clear_padding(&d1);
+    assert(d1 == 3.0L);
+    assert(memcmp(&d1, &d2, sizeof(long double)) == 0);
+  }
+}
+
+void structTests()
+{
+  // no_unique_address
+  {
+    struct S1 {
+     int x;
+     char c;
+    };
+
+    struct S2{
+      [[no_unique_address]] S1 s;
+      bool b;
+    };
+
+    S2 s1, s2;
+    memset(&s1, 42, sizeof(S2));
+    memset(&s2, 0, sizeof(S2));
+
+    s1.s.x = 4;
+    s1.s.c = 'a';
+    s1.b = true;
+    s2.s.x = 4;
+    s2.s.c = 'a';
+    s2.b = true;
+
+    assert(memcmp(&s1, &s2, sizeof(S2)) != 0);
+    __builtin_clear_padding(&s1);
+    assert(s1.s.x == 4);
+    assert(s1.s.c == 'a');
+    assert(s1.b == true);
+
+    assert(memcmp(&s1, &s2, sizeof(S2)) == 0);
+  }
+
+  // struct with long double
+  {
+    struct S{
+      long double l;
+      bool b;
+    };
+
+    S s1, s2;
+    memset(&s1, 42, sizeof(S));
+    memset(&s2, 0, sizeof(S));
+
+    s1.l = 3.0L;
+    s1.b = true;
+    s2.l = 3.0L;
+    s2.b = true;
+
+    assert(memcmp(&s1, &s2, sizeof(S)) != 0);
+    __builtin_clear_padding(&s1);
+    assert(s1.l == 3.0L);
+    assert(s1.b == true);
+    assert(memcmp(&s1, &s2, sizeof(S)) == 0);
+  }
+
+  // EBO
+  {
+    struct Empty{};
+    struct B {
+      int i;
+    };
+    struct S : Empty, B {
+      bool b;
+    }; 
+
+    S s1, s2;
+    memset(&s1, 42, sizeof(S));
+    memset(&s2, 0, sizeof(S));
+
+    s1.i = 4;
+    s1.b = true;
+    s2.i = 4;
+    s2.b = true;
+
+    assert(memcmp(&s1, &s2, sizeof(S)) != 0);
+    __builtin_clear_padding(&s1);
+    assert(s1.i == 4);
+    assert(s1.b == true);
+    assert(memcmp(&s1, &s2, sizeof(S)) == 0);
+  }
+
+  // padding between bases
+  {
+    struct B1 {
+      char c1;
+    };
+    struct B2 {
+      alignas(4) char c2;
+    };
+
+    struct S: B1, B2 {};
+
+    S s1, s2;
+    memset(&s1, 42, sizeof(S));
+    memset(&s2, 0, sizeof(S));
+
+    s1.c1 = 'a';
+    s1.c2 = 'b';
+    s2.c1 = 'a';
+    s2.c2 = 'b';
+
+    assert(memcmp(&s1, &s2, sizeof(S)) != 0);
+    __builtin_clear_padding(&s1);
+    assert(s1.c1 == 'a');
+    assert(s1.c2 == 'b');
+    assert(memcmp(&s1, &s2, sizeof(S)) == 0);
+  }
+
+  // padding after last base
+  {
+    struct B1 {
+      char c1;
+    };
+    struct B2 {
+      char c2;
+    };
+
+    struct S: B1, B2 {
+      alignas(4) char c3;
+    };
+
+    S s1, s2;
+    memset(&s1, 42, sizeof(S));
+    memset(&s2, 0, sizeof(S));
+
+    s1.c1 = 'a';
+    s1.c2 = 'b';
+    s1.c3 = 'c';
+    s2.c1 = 'a';
+    s2.c2 = 'b';
+    s2.c3 = 'c';
+
+    assert(memcmp(&s1, &s2, sizeof(S)) != 0);
+    __builtin_clear_padding(&s1);
+    assert(s1.c1 == 'a');
+    assert(s1.c2 == 'b');
+    assert(s1.c3 == 'c');
+    assert(memcmp(&s1, &s2, sizeof(S)) == 0);
+  }
+
+  testAllForType<32, 16, char>(11, 22, 33, 44);
+  testAllForType<64, 32, char>(4, 5, 6, 7);
+  testAllForType<32, 16, volatile char>(11, 22, 33, 44);
+  testAllForType<64, 32, volatile char>(4, 5, 6, 7);
+  testAllForType<32, 16, int>(0, 1, 2, 3);
+  testAllForType<64, 32, int>(4, 5, 6, 7);
+  testAllForType<32, 16, volatile int>(0, 1, 2, 3);
+  testAllForType<64, 32, volatile int>(4, 5, 6, 7);
+  testAllForType<32, 16, double>(0, 1, 2, 3);
+  testAllForType<64, 32, double>(4, 5, 6, 7);
+  testAllForType<32, 16, _ExtInt(28)>(0, 1, 2, 3);
+  testAllForType<64, 32, _ExtInt(28)>(4, 5, 6, 7);
+  testAllForType<32, 16, _ExtInt(60)>(0, 1, 2, 3);
+  testAllForType<64, 32, _ExtInt(60)>(4, 5, 6, 7);
+  testAllForType<32, 16, _ExtInt(64)>(0, 1, 2, 3);
+  testAllForType<64, 32, _ExtInt(64)>(4, 5, 6, 7);
+  testAllForType<32, 16, Foo>(Foo{1, 2}, Foo{3, 4}, Foo{1, 2}, Foo{3, 4});
+  testAllForType<64, 32, Foo>(Foo{1, 2}, Foo{3, 4}, Foo{1, 2}, Foo{3, 4});
+  testAllForType<256, 128, Float3Vec>(0, 1, 2, 3);
+  testAllForType<128, 128, Float3Vec>(4, 5, 6, 7);
+  testAllForType<256, 128, Float4Vec>(0, 1, 2, 3);
+  testAllForType<128, 128, Float4Vec>(4, 5, 6, 7);
+
+  otherTests();
+}
+
+void unionTests() {
+  // different length, do not clear object repr bits of non-active member
+  {
+    union u {
+      int i;
+      char c;
+    };
+
+    u u1, u2;
+    memset(&u1, 42, sizeof(u));
+    memset(&u2, 42, sizeof(u));
+    u1.c = '4';
+    u2.c = '4';
+
+    __builtin_clear_padding(&u1); // should have no effect
+    assert(u1.c == '4');
+
+    assert(memcmp(&u1, &u2, sizeof(u)) == 0);
+  }
+
+  // tail padding of longest member
+  {
+    struct s {
+      alignas(8) char c1;
+    };
+
+    union u {
+      s s1;
+      char c2;
+    };
+
+    u u1, u2;
+    memset(&u1, 42, sizeof(u));
+    memset(&u2, 0, sizeof(u));
+
+    u1.s1.c1 = '4';
+    u2.s1.c1 = '4';
+
+    assert(memcmp(&u1, &u2, sizeof(u)) != 0);
+    __builtin_clear_padding(&u1);
+    assert(u1.s1.c1 == '4');
+    assert(memcmp(&u1, &u2, sizeof(u)) == 0);
+  }
+}
+
+void arrayTests() {
+  // no padding
+  {
+    int i1[2] = {1, 2};
+    int i2[2] = {1, 2};
+
+    __builtin_clear_padding(&i1);
+    assert(i1[0]==1);
+    assert(i1[1]==2);
+    assert(memcmp(&i1, &i2, 2*sizeof(int)) == 0);
+  }
+
+  // long double
+  {
+    long double d1[2], d2[2];
+    memset(&d1, 42, 2*sizeof(long double));
+    memset(&d2, 0, 2*sizeof(long double));
+
+    d1[0] = 3.0L;
+    d1[1] = 4.0L;
+    d2[0] = 3.0L;
+    d2[1] = 4.0L;
+
+    __builtin_clear_padding(&d1);
+    assert(d1[0] == 3.0L);
+    assert(d2[1] == 4.0L);
+    assert(memcmp(&d1, &d2, 2*sizeof(long double)) == 0);
+  }
+
+  // struct
+  {
+    struct S {
+      int i1;
+      char c1;
+      int i2;
+      char c2;
+    };
+
+    S s1[2], s2[2];
+    memset(&s1, 42, 2*sizeof(S));
+    memset(&s2, 0, 2*sizeof(S));
+
+    s1[0].i1 = 1;
+    s1[0].c1 = 'a';
+    s1[0].i2 = 2;
+    s1[0].c2 = 'b';
+    s1[1].i1 = 3;
+    s1[1].c1 = 'c';
+    s1[1].i2 = 4;
+    s1[1].c2 = 'd';
+
+    s2[0].i1 = 1;
+    s2[0].c1 = 'a';
+    s2[0].i2 = 2;
+    s2[0].c2 = 'b';
+    s2[1].i1 = 3;
+    s2[1].c1 = 'c';
+    s2[1].i2 = 4;
+    s2[1].c2 = 'd';
+
+    assert(memcmp(&s1, &s2, 2*sizeof(S)) != 0);
+    __builtin_clear_padding(&s1);
+
+    assert(s1[0].i1 == 1);
+    assert(s1[0].c1 == 'a');
+    assert(s1[0].i2 == 2);
+    assert(s1[0].c2 == 'b');
+    assert(s1[1].i1 == 3);
+    assert(s1[1].c1 == 'c');
+    assert(s1[1].i2 == 4);
+    assert(s1[1].c2 == 'd');
+    assert(memcmp(&s1, &s2, 2*sizeof(S)) == 0);
+  }
+}
+
+int main(int, const char**) {
+  primitiveTests();
+  unionTests();
+  structTests();
+  arrayTests();
+
+  return 0;
+}
diff --git a/clang/test/SemaCXX/builtin-clear-padding.cpp b/clang/test/SemaCXX/builtin-clear-padding.cpp
new file mode 100644
index 0000000000000..ea87249c87b0a
--- /dev/null
+++ b/clang/test/SemaCXX/builtin-clear-padding.cpp
@@ -0,0 +1,15 @@
+// RUN: %clang_cc1 -fsyntax-only -verify %s
+
+struct Foo {};
+
+struct Incomplete; // expected-note {{forward declaration of 'Incomplete'}}
+
+void test(int a, Foo b, void *c, int *d, Foo *e, const Foo *f, Incomplete *g) {
+  __builtin_clear_padding(a); // expected-error {{passing 'int' to parameter of incompatible type pointer: type mismatch at 1st parameter ('int' vs pointer)}}
+  __builtin_clear_padding(b); // expected-error {{passing 'Foo' to parameter of incompatible type pointer: type mismatch at 1st parameter ('Foo' vs pointer)}}
+  __builtin_clear_padding(c); // expected-error {{variable has incomplete type 'void'}}
+  __builtin_clear_padding(d); // This should not error.
+  __builtin_clear_padding(e); // This should not error.
+  __builtin_clear_padding(f); // expected-error {{read-only variable is not assignable}}
+  __builtin_clear_padding(g); // expected-error {{variable has incomplete type 'Incomplete'}}
+}