diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -4430,11 +4430,24 @@
     return RValue::get(EmitMSVCBuiltinExpr(MSVCIntrin::__fastfail, E));
 
   case Builtin::BI__builtin_coro_size: {
-    auto & Context = getContext();
+    auto &Context = getContext();
     auto SizeTy = Context.getSizeType();
     auto T = Builder.getIntNTy(Context.getTypeSize(SizeTy));
-    Function *F = CGM.getIntrinsic(Intrinsic::coro_size, T);
-    return RValue::get(Builder.CreateCall(F));
+    Function *CoroSize = CGM.getIntrinsic(Intrinsic::coro_size, T);
+    // Over-allocate to handle overaligned coro frame.
+    Function *CoroAlign = CGM.getIntrinsic(Intrinsic::coro_align, T);
+    auto &TI = CGM.getContext().getTargetInfo();
+    unsigned AlignOfNew = TI.getNewAlign() / TI.getCharWidth();
+    Value *SizeCall = Builder.CreateCall(CoroSize);
+    Value *AlignCall = Builder.CreateCall(CoroAlign);
+    // int x = coro_align - AlignOfNew;
+    // coro_size + (x > 0 ? x : 0)
+    Value *Diff =
+        Builder.CreateNSWSub(AlignCall, ConstantInt::get(T, AlignOfNew, true));
+    Value *Zero = ConstantInt::getSigned(T, 0);
+    Value *Cmp = Builder.CreateICmp(llvm::CmpInst::ICMP_SGT, Diff, Zero);
+    Value *Extra = Builder.CreateSelect(Cmp, Diff, Zero);
+    return RValue::get(Builder.CreateAdd(SizeCall, Extra));
   }
 
   case Builtin::BI__builtin_coro_id:
diff --git a/clang/lib/CodeGen/CGCoroutine.cpp b/clang/lib/CodeGen/CGCoroutine.cpp
--- a/clang/lib/CodeGen/CGCoroutine.cpp
+++ b/clang/lib/CodeGen/CGCoroutine.cpp
@@ -548,7 +548,7 @@
   auto *EntryBB = Builder.GetInsertBlock();
   auto *AllocBB = createBasicBlock("coro.alloc");
   // Align to overaligned boundary in this block. The over-allocation is
-  // handled by aligning up frame's store size.
+  // handled performed during lowering BI__builtin_coro_size.
   auto *AlignAllocBB = createBasicBlock("coro.alloc.align");
   auto *InitBB = createBasicBlock("coro.init");
   auto *FinalBB = createBasicBlock("coro.final");
diff --git a/clang/test/CodeGenCoroutines/coro-alloc.cpp b/clang/test/CodeGenCoroutines/coro-alloc.cpp
--- a/clang/test/CodeGenCoroutines/coro-alloc.cpp
+++ b/clang/test/CodeGenCoroutines/coro-alloc.cpp
@@ -60,9 +60,14 @@
   // CHECK: br i1 %[[NeedAlloc]], label %[[AllocBB:.+]], label %[[InitBB:.+]]
 
   // CHECK: [[AllocBB]]:
-  // CHECK: %[[SIZE:.+]] = call i64 @llvm.coro.size.i64()
-  // CHECK: %[[MEM:.+]] = call noalias nonnull i8* @_Znwm(i64 %[[SIZE]])
-  // CHECK: br label %[[AlignAllocBB:.+]]
+  // CHECK-NEXT: %[[SIZE:.+]] = call i64 @llvm.coro.size.i64()
+  // CHECK-NEXT: %[[ALIGN:.+]] = call i64 @llvm.coro.align.i64()
+  // CHECK-NEXT: %[[DIFF:.+]] = sub nsw i64 %[[ALIGN]], 16
+  // CHECK-NEXT: %[[CMP:.+]] = icmp sgt i64 %[[DIFF]], 0
+  // CHECK-NEXT: %[[SEL:.+]] = select i1 %[[CMP]], i64 %[[DIFF]], i64 0
+  // CHECK-NEXT: %[[NEWSIZE:.+]] = add i64 %[[SIZE]], %[[SEL]]
+  // CHECK-NEXT: %[[MEM:.+]] = call noalias nonnull i8* @_Znwm(i64 %[[NEWSIZE]])
+  // CHECK-NEXT: br label %[[AlignAllocBB:.+]]
 
   // CHECK: [[AlignAllocBB]]:
   // CHECK: %[[ALIGN:.+]] = call i64 @llvm.coro.align.i64()
@@ -104,7 +109,8 @@
 extern "C" void f1(promise_new_tag ) {
   // CHECK: %[[ID:.+]] = call token @llvm.coro.id(i32 16
   // CHECK: %[[SIZE:.+]] = call i64 @llvm.coro.size.i64()
-  // CHECK: call i8* @_ZNSt12experimental16coroutine_traitsIJv15promise_new_tagEE12promise_typenwEm(i64 %[[SIZE]])
+  // CHECK: %[[NEWSIZE:.+]] = add i64 %[[SIZE]], %{{.+}}
+  // CHECK: call i8* @_ZNSt12experimental16coroutine_traitsIJv15promise_new_tagEE12promise_typenwEm(i64 %[[NEWSIZE]])
 
   // CHECK: %[[FRAME:.+]] = call i8* @llvm.coro.begin(
   // CHECK: %[[MEM:.+]] = call i8* @llvm.coro.free(token %[[ID]], i8* %[[FRAME]])
@@ -133,10 +139,11 @@
   // CHECK: store double %z, double* %z.addr, align 8
   // CHECK: %[[ID:.+]] = call token @llvm.coro.id(i32 16
   // CHECK: %[[SIZE:.+]] = call i64 @llvm.coro.size.i64()
+  // CHECK: %[[NEWSIZE:.+]] = add i64 %[[SIZE]], %{{.+}}
   // CHECK: %[[INT:.+]] = load i32, i32* %x.addr, align 4
   // CHECK: %[[FLOAT:.+]] = load float, float* %y.addr, align 4
   // CHECK: %[[DOUBLE:.+]] = load double, double* %z.addr, align 8
-  // CHECK: call i8* @_ZNSt12experimental16coroutine_traitsIJv34promise_matching_placement_new_tagifdEE12promise_typenwEmS1_ifd(i64 %[[SIZE]], i32 %[[INT]], float %[[FLOAT]], double %[[DOUBLE]])
+  // CHECK: call i8* @_ZNSt12experimental16coroutine_traitsIJv34promise_matching_placement_new_tagifdEE12promise_typenwEmS1_ifd(i64 %[[NEWSIZE]], i32 %[[INT]], float %[[FLOAT]], double %[[DOUBLE]])
   co_return;
 }
 
@@ -183,7 +190,8 @@
 extern "C" void f2(promise_delete_tag) {
   // CHECK: %[[ID:.+]] = call token @llvm.coro.id(i32 16
   // CHECK: %[[SIZE:.+]] = call i64 @llvm.coro.size.i64()
-  // CHECK: call noalias nonnull i8* @_Znwm(i64 %[[SIZE]])
+  // CHECK: %[[NEWSIZE:.+]] = add i64 %[[SIZE]], %{{.+}}
+  // CHECK: call noalias nonnull i8* @_Znwm(i64 %[[NEWSIZE]])
 
   // CHECK: %[[FRAME:.+]] = call i8* @llvm.coro.begin(
   // CHECK: %[[MEM:.+]] = call i8* @llvm.coro.free(token %[[ID]], i8* %[[FRAME]])
@@ -208,12 +216,14 @@
 extern "C" void f3(promise_sized_delete_tag) {
   // CHECK: %[[ID:.+]] = call token @llvm.coro.id(i32 16
   // CHECK: %[[SIZE:.+]] = call i64 @llvm.coro.size.i64()
-  // CHECK: call noalias nonnull i8* @_Znwm(i64 %[[SIZE]])
+  // CHECK: %[[NEWSIZE:.+]] = add i64 %[[SIZE]], %{{.+}}
+  // CHECK: call noalias nonnull i8* @_Znwm(i64 %[[NEWSIZE]])
 
   // CHECK: %[[FRAME:.+]] = call i8* @llvm.coro.begin(
   // CHECK: %[[MEM:.+]] = call i8* @llvm.coro.free(token %[[ID]], i8* %[[FRAME]])
   // CHECK: %[[SIZE2:.+]] = call i64 @llvm.coro.size.i64()
-  // CHECK: call void @_ZNSt12experimental16coroutine_traitsIJv24promise_sized_delete_tagEE12promise_typedlEPvm(i8* %[[MEM]], i64 %[[SIZE2]])
+  // CHECK: %[[NEWSIZE2:.+]] = add i64 %[[SIZE2]], %{{.+}}
+  // CHECK: call void @_ZNSt12experimental16coroutine_traitsIJv24promise_sized_delete_tagEE12promise_typedlEPvm(i8* %[[MEM]], i64 %[[NEWSIZE2]])
   co_return;
 }
 
@@ -236,7 +246,8 @@
   // CHECK: %[[Gro:.+]] = alloca i32
   // CHECK: %[[ID:.+]] = call token @llvm.coro.id(i32 16
   // CHECK: %[[SIZE:.+]] = call i64 @llvm.coro.size.i64()
-  // CHECK: %[[MEM:.+]] = call noalias i8* @_ZnwmRKSt9nothrow_t(i64 %[[SIZE]], %"struct.std::nothrow_t"* nonnull align 1 dereferenceable(1) @_ZStL7nothrow)
+  // CHECK: %[[NEWSIZE:.+]] = add i64 %[[SIZE]], %{{.+}}
+  // CHECK: %[[MEM:.+]] = call noalias i8* @_ZnwmRKSt9nothrow_t(i64 %[[NEWSIZE]], %"struct.std::nothrow_t"* nonnull align 1 dereferenceable(1) @_ZStL7nothrow)
   // CHECK: %[[OK:.+]] = icmp ne i8* %[[MEM]], null
   // CHECK: br i1 %[[OK]], label %[[OKBB:.+]], label %[[ERRBB:.+]]
 
diff --git a/clang/test/CodeGenCoroutines/coro-builtins.c b/clang/test/CodeGenCoroutines/coro-builtins.c
--- a/clang/test/CodeGenCoroutines/coro-builtins.c
+++ b/clang/test/CodeGenCoroutines/coro-builtins.c
@@ -21,7 +21,12 @@
   __builtin_coro_noop();
 
   // CHECK-NEXT: %[[SIZE:.+]] = call i64 @llvm.coro.size.i64()
-  // CHECK-NEXT: %[[MEM:.+]] = call i8* @myAlloc(i64 %[[SIZE]])
+  // CHECK-NEXT: %[[ALIGN:.+]] = call i64 @llvm.coro.align.i64()
+  // CHECK-NEXT: %[[DIFF:.+]] = sub nsw i64 %[[ALIGN]], 16
+  // CHECK-NEXT: %[[CMP:.+]] = icmp sgt i64 %[[DIFF]], 0
+  // CHECK-NEXT: %[[SEL:.+]] = select i1 %[[CMP]], i64 %[[DIFF]], i64 0
+  // CHECK-NEXT: %[[NEWSIZE:.+]] = add i64 %[[SIZE]], %[[SEL]]
+  // CHECK-NEXT: %[[MEM:.+]] = call i8* @myAlloc(i64 %[[NEWSIZE]])
   // CHECK-NEXT: %[[FRAME:.+]] = call i8* @llvm.coro.begin(token %[[COROID]], i8* %[[MEM]])
   __builtin_coro_begin(myAlloc(__builtin_coro_size()));
 
diff --git a/clang/test/CodeGenCoroutines/coro-gro.cpp b/clang/test/CodeGenCoroutines/coro-gro.cpp
--- a/clang/test/CodeGenCoroutines/coro-gro.cpp
+++ b/clang/test/CodeGenCoroutines/coro-gro.cpp
@@ -49,7 +49,8 @@
   // CHECK: %[[GroActive:.+]] = alloca i1
 
   // CHECK: %[[Size:.+]] = call i64 @llvm.coro.size.i64()
-  // CHECK: call noalias nonnull i8* @_Znwm(i64 %[[Size]])
+  // CHECK: %[[NewSize:.+]] = add i64 %[[Size]], %{{.+}}
+  // CHECK: call noalias nonnull i8* @_Znwm(i64 %[[NewSize]])
   // CHECK: store i1 false, i1* %[[GroActive]]
   // CHECK: call void @_ZNSt12experimental16coroutine_traitsIJiEE12promise_typeC1Ev(
   // CHECK: call void @_ZNSt12experimental16coroutine_traitsIJiEE12promise_type17get_return_objectEv(
diff --git a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
--- a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
+++ b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
@@ -1004,14 +1004,10 @@
   if (!Shape.CoroSizes.empty()) {
     // In the same function all coro.sizes should have the same result type.
     auto *SizeIntrin = Shape.CoroSizes.back();
-    auto *SizeConstant =
-        ConstantInt::get(SizeIntrin->getType(), Shape.FrameSize);
-    assert(
-        Shape.FrameSize ==
-            alignTo(SizeIntrin->getModule()->getDataLayout().getTypeStoreSize(
-                        Shape.FrameTy),
-                    Align(Shape.FrameAlign)) &&
-        "FrameSize should already be aligned up to FrameAlign");
+    Module *M = SizeIntrin->getModule();
+    const DataLayout &DL = M->getDataLayout();
+    auto Size = DL.getTypeAllocSize(Shape.FrameTy);
+    auto *SizeConstant = ConstantInt::get(SizeIntrin->getType(), Size);
 
     for (CoroSizeInst *CS : Shape.CoroSizes) {
       CS->replaceAllUsesWith(SizeConstant);