Index: llvm/include/llvm/IR/IRBuilder.h
===================================================================
--- llvm/include/llvm/IR/IRBuilder.h
+++ llvm/include/llvm/IR/IRBuilder.h
@@ -2543,6 +2543,9 @@
   CallInst *CreateAlignmentAssumption(const DataLayout &DL, Value *PtrValue,
                                       Value *Alignment,
                                       Value *OffsetValue = nullptr);
+  /// Create an assume intrinsic call that represents a nonnull
+  /// assumption on the provided pointer.
+  CallInst *CreateNonNullAssumption(Value *PtrValue);
 };
 
 /// This provides a uniform API for creating instructions and inserting
Index: llvm/lib/Analysis/InlineCost.cpp
===================================================================
--- llvm/lib/Analysis/InlineCost.cpp
+++ llvm/lib/Analysis/InlineCost.cpp
@@ -1732,7 +1732,7 @@
   // parameter attribute, but that's a less interesting case because hopefully
   // the callee would already have been simplified based on that.
   if (Argument *A = dyn_cast<Argument>(V))
-    if (paramHasAttr(A, Attribute::NonNull))
+    if (paramHasAttr(A, Attribute::NonNull) || paramHasAttr(A, Attribute::Dereferenceable))
       return true;
 
   // Is this an alloca in the caller?  This is distinct from the attribute case
Index: llvm/lib/IR/IRBuilder.cpp
===================================================================
--- llvm/lib/IR/IRBuilder.cpp
+++ llvm/lib/IR/IRBuilder.cpp
@@ -1410,6 +1410,12 @@
   return CreateAlignmentAssumptionHelper(DL, PtrValue, Alignment, OffsetValue);
 }
 
+CallInst *IRBuilderBase::CreateNonNullAssumption(Value *PtrValue) {
+  Value *Vals[] = { PtrValue };
+  OperandBundleDefT<Value *> NonNullOpB("nonnull", Vals);
+  return CreateAssumption(ConstantInt::getTrue(getContext()), {NonNullOpB});
+}
+
 IRBuilderDefaultInserter::~IRBuilderDefaultInserter() = default;
 IRBuilderCallbackInserter::~IRBuilderCallbackInserter() = default;
 IRBuilderFolder::~IRBuilderFolder() = default;
Index: llvm/lib/Transforms/Utils/InlineFunction.cpp
===================================================================
--- llvm/lib/Transforms/Utils/InlineFunction.cpp
+++ llvm/lib/Transforms/Utils/InlineFunction.cpp
@@ -1460,6 +1460,25 @@
   }
 }
 
+static void AddAssumptionsFromCallSiteAttrs(CallBase &CB, InlineFunctionInfo &IFI) {
+  if (!IFI.GetAssumptionCache)
+    return;
+
+  AssumptionCache *AC = &IFI.GetAssumptionCache(*CB.getCaller());
+  auto &DL = CB.getCaller()->getParent()->getDataLayout();
+  Function *CalledFunc = CB.getCalledFunction();
+  IRBuilder<> Builder(&CB);
+
+  for (Argument &Arg : CalledFunc->args()) {
+    unsigned ArgNo = Arg.getArgNo();
+    auto *ArgVal = CB.getArgOperand(ArgNo);
+    if (CB.getAttributes().hasParamAttr(ArgNo, Attribute::NonNull) && !isKnownNonZero(ArgVal, DL, 0, AC)) {
+      CallInst *NewAsmp = Builder.CreateNonNullAssumption(CB.getArgOperand(ArgNo));
+      AC->registerAssumption(cast<AssumeInst>(NewAsmp));
+    }
+  }
+}
+
 static void HandleByValArgumentInit(Type *ByValType, Value *Dst, Value *Src,
                                     Module *M, BasicBlock *InsertBlock,
                                     InlineFunctionInfo &IFI,
@@ -2130,6 +2149,8 @@
       VMap[&*I] = ActualArg;
     }
 
+    AddAssumptionsFromCallSiteAttrs(CB, IFI);
+
     // TODO: Remove this when users have been updated to the assume bundles.
     // Add alignment assumptions if necessary. We do this before the inlined
     // instructions are actually cloned into the caller so that we can easily
Index: llvm/test/Transforms/Coroutines/ex5.ll
===================================================================
--- llvm/test/Transforms/Coroutines/ex5.ll
+++ llvm/test/Transforms/Coroutines/ex5.ll
@@ -1,7 +1,30 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 2
 ; Fifth example from Doc/Coroutines.rst (final suspend)
 ; RUN: opt < %s -aa-pipeline=basic-aa -passes='default<O2>' -preserve-alignment-assumptions-during-inlining=false -S | FileCheck %s
 
 define ptr @f(i32 %n) presplitcoroutine {
+; CHECK-LABEL: define noalias nonnull ptr @f
+; CHECK-SAME: (i32 [[N:%.*]]) local_unnamed_addr {
+; CHECK-NEXT:  while.cond:
+; CHECK-NEXT:    [[ALLOC:%.*]] = tail call ptr @malloc(i32 24)
+; CHECK-NEXT:    store ptr @f.resume, ptr [[ALLOC]], align 8
+; CHECK-NEXT:    [[DESTROY_ADDR:%.*]] = getelementptr inbounds [[F_FRAME:%.*]], ptr [[ALLOC]], i64 0, i32 1
+; CHECK-NEXT:    store ptr @f.destroy, ptr [[DESTROY_ADDR]], align 8
+; CHECK-NEXT:    [[N_VAL_SPILL_ADDR:%.*]] = getelementptr inbounds [[F_FRAME]], ptr [[ALLOC]], i64 0, i32 2
+; CHECK-NEXT:    store i32 [[N]], ptr [[N_VAL_SPILL_ADDR]], align 4
+; CHECK-NEXT:    [[CMP:%.*]] = icmp sgt i32 [[N]], 0
+; CHECK-NEXT:    br i1 [[CMP]], label [[AFTERCOROSUSPEND:%.*]], label [[AFTERCOROSUSPEND3:%.*]]
+; CHECK:       AfterCoroSuspend:
+; CHECK-NEXT:    tail call void @print(i32 [[N]])
+; CHECK-NEXT:    [[INDEX_ADDR5:%.*]] = getelementptr inbounds [[F_FRAME]], ptr [[ALLOC]], i64 0, i32 3
+; CHECK-NEXT:    store i1 false, ptr [[INDEX_ADDR5]], align 1
+; CHECK-NEXT:    br label [[AFTERCOROEND:%.*]]
+; CHECK:       AfterCoroSuspend3:
+; CHECK-NEXT:    store ptr null, ptr [[ALLOC]], align 8
+; CHECK-NEXT:    br label [[AFTERCOROEND]]
+; CHECK:       AfterCoroEnd:
+; CHECK-NEXT:    ret ptr [[ALLOC]]
+;
 entry:
   %id = call token @llvm.coro.id(i32 0, ptr null, ptr null, ptr null)
   %size = call i32 @llvm.coro.size.i32()
@@ -18,12 +41,12 @@
   call void @print(i32 %n.val) #4
   %s = call i8 @llvm.coro.suspend(token none, i1 false)
   switch i8 %s, label %suspend [i8 0, label %while.cond
-                                i8 1, label %cleanup]
+  i8 1, label %cleanup]
 while.end:
   %s.final = call i8 @llvm.coro.suspend(token none, i1 true)
   switch i8 %s.final, label %suspend [i8 0, label %trap
-                                      i8 1, label %cleanup]
-trap: 
+  i8 1, label %cleanup]
+trap:
   call void @llvm.trap()
   unreachable
 cleanup:
@@ -48,8 +71,17 @@
 declare ptr @llvm.coro.free(token, ptr)
 declare i1 @llvm.coro.end(ptr, i1)
 
-; CHECK-LABEL: @main
 define i32 @main() {
+; CHECK-LABEL: define i32 @main() local_unnamed_addr {
+; CHECK-NEXT:  end:
+; CHECK-NEXT:    [[ALLOC_I:%.*]] = tail call ptr @malloc(i32 24)
+; CHECK-NEXT:    tail call void @print(i32 4)
+; CHECK-NEXT:    tail call void @print(i32 3)
+; CHECK-NEXT:    tail call void @print(i32 2)
+; CHECK-NEXT:    tail call void @print(i32 1)
+; CHECK-NEXT:    tail call void @free(ptr nonnull [[ALLOC_I]])
+; CHECK-NEXT:    ret i32 0
+;
 entry:
   %hdl = call ptr @f(i32 4)
   br label %while
@@ -61,11 +93,6 @@
   call void @llvm.coro.destroy(ptr %hdl)
   ret i32 0
 
-; CHECK:      call void @print(i32 4)
-; CHECK:      call void @print(i32 3)
-; CHECK:      call void @print(i32 2)
-; CHECK:      call void @print(i32 1)
-; CHECK:      ret i32 0
 }
 
 declare i1 @llvm.coro.done(ptr)
Index: llvm/test/Transforms/Inline/assumptions-from-callsite-attrs.ll
===================================================================
--- llvm/test/Transforms/Inline/assumptions-from-callsite-attrs.ll
+++ llvm/test/Transforms/Inline/assumptions-from-callsite-attrs.ll
@@ -8,6 +8,8 @@
 define void @f(ptr %p, ptr %q, ptr %z) {
 ; CHECK-LABEL: define void @f
 ; CHECK-SAME: (ptr [[P:%.*]], ptr [[Q:%.*]], ptr [[Z:%.*]]) {
+; CHECK-NEXT:    call void @llvm.assume(i1 true) [ "nonnull"(ptr [[P]]) ]
+; CHECK-NEXT:    call void @llvm.assume(i1 true) [ "nonnull"(ptr [[Z]]) ]
 ; CHECK-NEXT:    call void @h(ptr [[P]], ptr [[Q]], ptr [[Z]])
 ; CHECK-NEXT:    ret void
 ;
Index: llvm/test/Transforms/Inline/nonnull.ll
===================================================================
--- llvm/test/Transforms/Inline/nonnull.ll
+++ llvm/test/Transforms/Inline/nonnull.ll
@@ -75,6 +75,7 @@
 define void @caller3(ptr %arg) {
 ; CHECK-LABEL: define void @caller3
 ; CHECK-SAME: (ptr [[ARG:%.*]]) {
+; CHECK-NEXT:    call void @llvm.assume(i1 true) [ "nonnull"(ptr [[ARG]]) ]
 ; CHECK-NEXT:    [[CMP_I:%.*]] = icmp eq ptr [[ARG]], null
 ; CHECK-NEXT:    br i1 [[CMP_I]], label [[EXPENSIVE_I:%.*]], label [[DONE_I:%.*]]
 ; CHECK:       expensive.i:
@@ -99,10 +100,11 @@
   ret void
 }
 
+; Positive test - arg is known non null
 define void @caller4(ptr dereferenceable(8) %arg) {
 ; CHECK-LABEL: define void @caller4
 ; CHECK-SAME: (ptr dereferenceable(8) [[ARG:%.*]]) {
-; CHECK-NEXT:    call void @callee(ptr dereferenceable(8) [[ARG]])
+; CHECK-NEXT:    call void @bar()
 ; CHECK-NEXT:    ret void
 ;
   call void @callee(ptr dereferenceable(8) %arg)
@@ -122,7 +124,24 @@
 define void @caller6(ptr %arg) {
 ; CHECK-LABEL: define void @caller6
 ; CHECK-SAME: (ptr [[ARG:%.*]]) {
-; CHECK-NEXT:    call void @callee(ptr dereferenceable(8) [[ARG]])
+; CHECK-NEXT:    [[CMP_I:%.*]] = icmp eq ptr [[ARG]], null
+; CHECK-NEXT:    br i1 [[CMP_I]], label [[EXPENSIVE_I:%.*]], label [[DONE_I:%.*]]
+; CHECK:       expensive.i:
+; CHECK-NEXT:    call void @foo()
+; CHECK-NEXT:    call void @foo()
+; CHECK-NEXT:    call void @foo()
+; CHECK-NEXT:    call void @foo()
+; CHECK-NEXT:    call void @foo()
+; CHECK-NEXT:    call void @foo()
+; CHECK-NEXT:    call void @foo()
+; CHECK-NEXT:    call void @foo()
+; CHECK-NEXT:    call void @foo()
+; CHECK-NEXT:    call void @foo()
+; CHECK-NEXT:    br label [[CALLEE_EXIT:%.*]]
+; CHECK:       done.i:
+; CHECK-NEXT:    call void @bar()
+; CHECK-NEXT:    br label [[CALLEE_EXIT]]
+; CHECK:       callee.exit:
 ; CHECK-NEXT:    ret void
 ;
   call void @callee(ptr dereferenceable(8) %arg)
Index: llvm/test/Transforms/PhaseOrdering/dae-dce.ll
===================================================================
--- llvm/test/Transforms/PhaseOrdering/dae-dce.ll
+++ llvm/test/Transforms/PhaseOrdering/dae-dce.ll
@@ -14,9 +14,14 @@
 }
 
 define internal void @capture_and_trap(ptr %ptr) noinline {
-; CHECK-LABEL: @capture_and_trap(
-; CHECK-NEXT:    tail call void @llvm.trap()
-; CHECK-NEXT:    unreachable
+; DEFAULT-LABEL: @capture_and_trap(
+; DEFAULT-NEXT:    tail call void @llvm.trap()
+; DEFAULT-NEXT:    unreachable
+;
+; LTO-LABEL: @capture_and_trap(
+; LTO-NEXT:    call void @llvm.assume(i1 true) [ "nonnull"(ptr poison) ]
+; LTO-NEXT:    tail call void @llvm.trap()
+; LTO-NEXT:    unreachable
 ;
   %alloca = alloca ptr, align 4
   store ptr %ptr, ptr %alloca, align 4
@@ -47,6 +52,3 @@
   call void @capture_and_trap(ptr @dead_fn2)
   unreachable
 }
-;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
-; DEFAULT: {{.*}}
-; LTO: {{.*}}