[clang][CodeGen] Implicit Conversion Sanitizer: handle increment/decrement (PR44054)(take 2)

Summary:
Implicit Conversion Sanitizer is *almost* feature complete.
There aren't *that* much unsanitized things left,
two major ones are increment/decrement (this patch) and bit fields.

As it was discussed in
[[ https://bugs.llvm.org/show_bug.cgi?id=39519 | PR39519 ]],
unlike `CompoundAssignOperator` (which is promoted internally),
or `BinaryOperator` (for which we always have promotion/demotion in AST)
or parts of `UnaryOperator` (we have promotion/demotion but only for
certain operations), for inc/dec, clang omits promotion/demotion
altogether, under as-if rule.

This is technically correct: https://rise4fun.com/Alive/zPgD
As it can be seen in `InstCombineCasts.cpp` `canEvaluateTruncated()`,
`add`/`sub`/`mul`/`and`/`or`/`xor` operators can all arbitrarily
be extended or truncated:
https://github.com/llvm/llvm-project/blob/901cd3b3f62d0c700e5d2c3f97eff97d634bec5e/llvm/lib/Transforms/InstCombine/InstCombineCasts.cpp#L1320-L1334

But that has serious implications:
1. Since we no longer model implicit casts, do we pessimise
   their AST representation and everything that uses it?
2. There is no demotion, so lossy demotion sanitizer does not trigger :]

Now, i'm not going to argue about the first problem here,
but the second one **needs** to be addressed. As it was stated
in the report, this is done intentionally, so changing
this in all modes would be considered a penalization/regression.
Which means, the sanitization-less codegen must not be altered.

It was also suggested to not change the sanitized codegen
to the one with demotion, but i quite strongly believe
that will not be the wise choice here:
1. One will need to re-engineer the check that the inc/dec was lossy
   in terms of `@llvm.{u,s}{add,sub}.with.overflow` builtins
2. We will still need to compute the result we would lossily demote.
   (i.e. the result of wide `add`ition/`sub`traction)
3. I suspect it would need to be done right here, in sanitization.
   Which kinda defeats the point of
   using `@llvm.{u,s}{add,sub}.with.overflow` builtins:
   we'd have two `add`s with basically the same arguments,
   one of which is used for check+error-less codepath and other one
   for the error reporting. That seems worse than a single wide op+check.
4. OR, we would need to do that in the compiler-rt handler.
   Which means we'll need a whole new handler.
   But then what about the `CompoundAssignOperator`,
   it would also be applicable for it.
   So this also doesn't really seem like the right path to me.
5. At least X86 (but likely others) pessimizes all sub-`i32` operations
   (due to partial register stalls), so even if we avoid promotion+demotion,
   the computations will //likely// be performed in `i32` anyways.

So i'm not really seeing much benefit of
not doing the straight-forward thing.

While looking into this, i have noticed a few more LLVM middle-end
missed canonicalizations, and filed
[[ https://bugs.llvm.org/show_bug.cgi?id=44100 | PR44100 ]],
[[ https://bugs.llvm.org/show_bug.cgi?id=44102 | PR44102 ]].

Those are not specific to inc/dec, we also have them for
`CompoundAssignOperator`, and it can happen for normal arithmetics, too.
But if we take some other path in the patch, it will not be applicable
here, and we will have most likely played ourselves.

TLDR: front-end should emit canonical, easy-to-optimize yet
un-optimized code. It is middle-end's job to make it optimal.

I'm really hoping reviewers agree with my personal assessment
of the path this patch should take..

This originally landed in 9872ea4
but got immediately reverted in cbfa237
because the assertion was faulty. That fault ended up being caused
by the enum - while there will be promotion, both types are unsigned,
with same width. So we still don't need to sanitize non-signed cases.
So far. Maybe the assert will tell us this isn't so.

Fixes [[ https://bugs.llvm.org/show_bug.cgi?id=44054 | PR44054 ]].
Refs. google/sanitizers#940

Reviewers: rjmccall, erichkeane, rsmith, vsk

Reviewed By: erichkeane

Subscribers: mehdi_amini, dexonsmith, cfe-commits, #sanitizers, llvm-commits, aaron.ballman, t.p.northover, efriedma, regehr

Tags: #llvm, #clang, #sanitizers

Differential Revision: https://reviews.llvm.org/D70539

Author: LebedevRI

clang/docs/ReleaseNotes.rst

@@ -78,6 +78,10 @@ Non-comprehensive list of changes in this release
   been extended to detect these cases, so that code relying on them can be
   detected and fixed.
 
+* The Implicit Conversion Sanitizer (``-fsanitize=implicit-conversion``) has
+  learned to sanitize pre/post increment/decrement of types with bit width
+  smaller than ``int``.
+
 - For X86 target, -march=skylake-avx512, -march=icelake-client,
   -march=icelake-server, -march=cascadelake, -march=cooperlake will default to
   not using 512-bit zmm registers in vectorized code unless 512-bit intrinsics

clang/lib/CodeGen/CGExprScalar.cpp

@@ -976,6 +976,11 @@ EmitIntegerTruncationCheckHelper(Value *Src, QualType SrcType, Value *Dst,
   return std::make_pair(Kind, std::make_pair(Check, Mask));
 }
 
+static bool PromotionIsPotentiallyEligibleForImplicitIntegerConversionCheck(
+    QualType SrcType, QualType DstType) {
+  return SrcType->isIntegerType() && DstType->isIntegerType();
+}
+
 void ScalarExprEmitter::EmitIntegerTruncationCheck(Value *Src, QualType SrcType,
                                                    Value *Dst, QualType DstType,
                                                    SourceLocation Loc) {
@@ -984,7 +989,8 @@ void ScalarExprEmitter::EmitIntegerTruncationCheck(Value *Src, QualType SrcType,
 
   // We only care about int->int conversions here.
   // We ignore conversions to/from pointer and/or bool.
-  if (!(SrcType->isIntegerType() && DstType->isIntegerType()))
+  if (!PromotionIsPotentiallyEligibleForImplicitIntegerConversionCheck(SrcType,
+                                                                       DstType))
     return;
 
   unsigned SrcBits = Src->getType()->getScalarSizeInBits();
@@ -1095,7 +1101,8 @@ void ScalarExprEmitter::EmitIntegerSignChangeCheck(Value *Src, QualType SrcType,
 
   // We only care about int->int conversions here.
   // We ignore conversions to/from pointer and/or bool.
-  if (!(SrcType->isIntegerType() && DstType->isIntegerType()))
+  if (!PromotionIsPotentiallyEligibleForImplicitIntegerConversionCheck(SrcType,
+                                                                       DstType))
     return;
 
   bool SrcSigned = SrcType->isSignedIntegerOrEnumerationType();
@@ -2419,9 +2426,51 @@ ScalarExprEmitter::EmitScalarPrePostIncDec(const UnaryOperator *E, LValue LV,
 
   // Most common case by far: integer increment.
   } else if (type->isIntegerType()) {
-    // Note that signed integer inc/dec with width less than int can't
-    // overflow because of promotion rules; we're just eliding a few steps here.
-    if (E->canOverflow() && type->isSignedIntegerOrEnumerationType()) {
+    QualType promotedType;
+    bool canPerformLossyDemotionCheck = false;
+    if (type->isPromotableIntegerType()) {
+      promotedType = CGF.getContext().getPromotedIntegerType(type);
+      assert(promotedType != type && "Shouldn't promote to the same type.");
+      canPerformLossyDemotionCheck = true;
+      canPerformLossyDemotionCheck &=
+          CGF.getContext().getCanonicalType(type) !=
+          CGF.getContext().getCanonicalType(promotedType);
+      canPerformLossyDemotionCheck &=
+          PromotionIsPotentiallyEligibleForImplicitIntegerConversionCheck(
+              type, promotedType);
+      assert((!canPerformLossyDemotionCheck ||
+              type->isSignedIntegerOrEnumerationType() ||
+              promotedType->isSignedIntegerOrEnumerationType() ||
+              ConvertType(type)->getScalarSizeInBits() ==
+                  ConvertType(promotedType)->getScalarSizeInBits()) &&
+             "The following check expects that if we do promotion to different "
+             "underlying canonical type, at least one of the types (either "
+             "base or promoted) will be signed, or the bitwidths will match.");
+    }
+    if (CGF.SanOpts.hasOneOf(
+            SanitizerKind::ImplicitIntegerArithmeticValueChange) &&
+        canPerformLossyDemotionCheck) {
+      // While `x += 1` (for `x` with width less than int) is modeled as
+      // promotion+arithmetics+demotion, and we can catch lossy demotion with
+      // ease; inc/dec with width less than int can't overflow because of
+      // promotion rules, so we omit promotion+demotion, which means that we can
+      // not catch lossy "demotion". Because we still want to catch these cases
+      // when the sanitizer is enabled, we perform the promotion, then perform
+      // the increment/decrement in the wider type, and finally
+      // perform the demotion. This will catch lossy demotions.
+
+      value = EmitScalarConversion(value, type, promotedType, E->getExprLoc());
+      Value *amt = llvm::ConstantInt::get(value->getType(), amount, true);
+      value = Builder.CreateAdd(value, amt, isInc ? "inc" : "dec");
+      // Do pass non-default ScalarConversionOpts so that sanitizer check is
+      // emitted.
+      value = EmitScalarConversion(value, promotedType, type, E->getExprLoc(),
+                                   ScalarConversionOpts(CGF.SanOpts));
+
+      // Note that signed integer inc/dec with width less than int can't
+      // overflow because of promotion rules; we're just eliding a few steps
+      // here.
+    } else if (E->canOverflow() && type->isSignedIntegerOrEnumerationType()) {
       value = EmitIncDecConsiderOverflowBehavior(E, value, isInc);
     } else if (E->canOverflow() && type->isUnsignedIntegerType() &&
                CGF.SanOpts.has(SanitizerKind::UnsignedIntegerOverflow)) {

clang/test/CodeGen/catch-implicit-conversions-basics-negatives.c

@@ -0,0 +1,12 @@
+// RUN: %clang_cc1 -fsanitize=implicit-unsigned-integer-truncation,implicit-signed-integer-truncation,implicit-integer-sign-change -fsanitize-recover=implicit-unsigned-integer-truncation,implicit-signed-integer-truncation,implicit-integer-sign-change -emit-llvm %s -o - -triple x86_64-linux-gnu | FileCheck %s -implicit-check-not="call void @__ubsan_handle_implicit_conversion" --check-prefixes=CHECK
+
+// If we have an enum, it will be promoted to an unsigned integer.
+// But both types are unsigned, and have same bitwidth.
+// So we should not emit any sanitization. Also, for inc/dec we currently assume
+// (assert) that we will only have cases where at least one of the types
+// is signed, which isn't the case here.
+typedef enum { a } b;
+b t0(b c) {
+  c--;
+  return c;
+}

clang/test/CodeGen/catch-implicit-conversions-incdec-basics.c

@@ -0,0 +1,139 @@
+// RUN: %clang_cc1 -fsanitize=implicit-unsigned-integer-truncation,implicit-signed-integer-truncation,implicit-integer-sign-change -fsanitize-recover=implicit-unsigned-integer-truncation,implicit-signed-integer-truncation,implicit-integer-sign-change -emit-llvm %s -o - -triple x86_64-linux-gnu | FileCheck %s -implicit-check-not="call void @__ubsan_handle_implicit_conversion" --check-prefixes=CHECK
+
+// CHECK-DAG: @[[INT:.*]] = {{.*}} c"'int'\00" }
+// CHECK-DAG: @[[UNSIGNED_SHORT:.*]] = {{.*}} c"'unsigned short'\00" }
+// CHECK-DAG: @[[LINE_100:.*]] = {{.*}}, i32 100, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_SHORT]], i8 2 }
+// CHECK-DAG: @[[LINE_200:.*]] = {{.*}}, i32 200, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_SHORT]], i8 2 }
+// CHECK-DAG: @[[LINE_300:.*]] = {{.*}}, i32 300, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_SHORT]], i8 2 }
+// CHECK-DAG: @[[LINE_400:.*]] = {{.*}}, i32 400, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_SHORT]], i8 2 }
+// CHECK-DAG: @[[SHORT:.*]] = {{.*}} c"'short'\00" }
+// CHECK-DAG: @[[LINE_500:.*]] = {{.*}}, i32 500, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[SHORT]], i8 2 }
+// CHECK-DAG: @[[LINE_600:.*]] = {{.*}}, i32 600, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[SHORT]], i8 2 }
+// CHECK-DAG: @[[LINE_700:.*]] = {{.*}}, i32 700, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[SHORT]], i8 2 }
+// CHECK-DAG: @[[LINE_800:.*]] = {{.*}}, i32 800, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[SHORT]], i8 2 }
+// CHECK-DAG: @[[UNSIGNED_CHAR:.*]] = {{.*}} c"'unsigned char'\00" }
+// CHECK-DAG: @[[LINE_900:.*]] = {{.*}}, i32 900, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[LINE_1000:.*]] = {{.*}}, i32 1000, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[LINE_1100:.*]] = {{.*}}, i32 1100, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[LINE_1200:.*]] = {{.*}}, i32 1200, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[SIGNED_CHAR:.*]] = {{.*}} c"'signed char'\00" }
+// CHECK-DAG: @[[LINE_1300:.*]] = {{.*}}, i32 1300, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[SIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[LINE_1400:.*]] = {{.*}}, i32 1400, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[SIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[LINE_1500:.*]] = {{.*}}, i32 1500, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[SIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[LINE_1600:.*]] = {{.*}}, i32 1600, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[SIGNED_CHAR]], i8 2 }
+
+// CHECK-LABEL: @t0(
+unsigned short t0(unsigned short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_100]] to i8*)
+#line 100
+  x++;
+  return x;
+}
+// CHECK-LABEL: @t1(
+unsigned short t1(unsigned short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_200]] to i8*)
+#line 200
+  x--;
+  return x;
+}
+// CHECK-LABEL: @t2(
+unsigned short t2(unsigned short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_300]] to i8*)
+#line 300
+  ++x;
+  return x;
+}
+// CHECK-LABEL: @t3(
+unsigned short t3(unsigned short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_400]] to i8*)
+#line 400
+  --x;
+  return x;
+}
+
+// CHECK-LABEL: @t4(
+signed short t4(signed short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_500]] to i8*)
+#line 500
+  x++;
+  return x;
+}
+// CHECK-LABEL: @t5(
+signed short t5(signed short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_600]] to i8*)
+#line 600
+  x--;
+  return x;
+}
+// CHECK-LABEL: @t6(
+signed short t6(signed short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_700]] to i8*)
+#line 700
+  ++x;
+  return x;
+}
+// CHECK-LABEL: @t7(
+signed short t7(signed short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_800]] to i8*)
+#line 800
+  --x;
+  return x;
+}
+
+// CHECK-LABEL: @t8(
+unsigned char t8(unsigned char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_900]] to i8*)
+#line 900
+  x++;
+  return x;
+}
+// CHECK-LABEL: @t9(
+unsigned char t9(unsigned char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1000]] to i8*)
+#line 1000
+  x--;
+  return x;
+}
+// CHECK-LABEL: @t10(
+unsigned char t10(unsigned char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1100]] to i8*)
+#line 1100
+  ++x;
+  return x;
+}
+// CHECK-LABEL: @t11(
+unsigned char t11(unsigned char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1200]] to i8*)
+#line 1200
+  --x;
+  return x;
+}
+
+// CHECK-LABEL: @t12(
+signed char t12(signed char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1300]] to i8*)
+#line 1300
+  x++;
+  return x;
+}
+// CHECK-LABEL: @t13(
+signed char t13(signed char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1400]] to i8*)
+#line 1400
+  x--;
+  return x;
+}
+// CHECK-LABEL: @t14(
+signed char t14(signed char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1500]] to i8*)
+#line 1500
+  ++x;
+  return x;
+}
+// CHECK-LABEL: @t15(
+signed char t15(signed char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1600]] to i8*)
+#line 1600
+  --x;
+  return x;
+}

clang/test/CodeGen/catch-implicit-integer-arithmetic-value-change-incdec-basics.c

@@ -0,0 +1,139 @@
+// RUN: %clang_cc1 -fsanitize=implicit-signed-integer-truncation,implicit-integer-sign-change -fsanitize-recover=implicit-signed-integer-truncation,implicit-integer-sign-change -emit-llvm %s -o - -triple x86_64-linux-gnu | FileCheck %s -implicit-check-not="call void @__ubsan_handle_implicit_conversion" --check-prefixes=CHECK
+
+// CHECK-DAG: @[[INT:.*]] = {{.*}} c"'int'\00" }
+// CHECK-DAG: @[[UNSIGNED_SHORT:.*]] = {{.*}} c"'unsigned short'\00" }
+// CHECK-DAG: @[[LINE_100:.*]] = {{.*}}, i32 100, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_SHORT]], i8 2 }
+// CHECK-DAG: @[[LINE_200:.*]] = {{.*}}, i32 200, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_SHORT]], i8 2 }
+// CHECK-DAG: @[[LINE_300:.*]] = {{.*}}, i32 300, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_SHORT]], i8 2 }
+// CHECK-DAG: @[[LINE_400:.*]] = {{.*}}, i32 400, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_SHORT]], i8 2 }
+// CHECK-DAG: @[[SHORT:.*]] = {{.*}} c"'short'\00" }
+// CHECK-DAG: @[[LINE_500:.*]] = {{.*}}, i32 500, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[SHORT]], i8 2 }
+// CHECK-DAG: @[[LINE_600:.*]] = {{.*}}, i32 600, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[SHORT]], i8 2 }
+// CHECK-DAG: @[[LINE_700:.*]] = {{.*}}, i32 700, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[SHORT]], i8 2 }
+// CHECK-DAG: @[[LINE_800:.*]] = {{.*}}, i32 800, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[SHORT]], i8 2 }
+// CHECK-DAG: @[[UNSIGNED_CHAR:.*]] = {{.*}} c"'unsigned char'\00" }
+// CHECK-DAG: @[[LINE_900:.*]] = {{.*}}, i32 900, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[LINE_1000:.*]] = {{.*}}, i32 1000, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[LINE_1100:.*]] = {{.*}}, i32 1100, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[LINE_1200:.*]] = {{.*}}, i32 1200, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[UNSIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[SIGNED_CHAR:.*]] = {{.*}} c"'signed char'\00" }
+// CHECK-DAG: @[[LINE_1300:.*]] = {{.*}}, i32 1300, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[SIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[LINE_1400:.*]] = {{.*}}, i32 1400, i32 4 }, {{.*}}* @[[INT]], {{.*}}* @[[SIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[LINE_1500:.*]] = {{.*}}, i32 1500, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[SIGNED_CHAR]], i8 2 }
+// CHECK-DAG: @[[LINE_1600:.*]] = {{.*}}, i32 1600, i32 3 }, {{.*}}* @[[INT]], {{.*}}* @[[SIGNED_CHAR]], i8 2 }
+
+// CHECK-LABEL: @t0(
+unsigned short t0(unsigned short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_100]] to i8*)
+#line 100
+  x++;
+  return x;
+}
+// CHECK-LABEL: @t1(
+unsigned short t1(unsigned short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_200]] to i8*)
+#line 200
+  x--;
+  return x;
+}
+// CHECK-LABEL: @t2(
+unsigned short t2(unsigned short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_300]] to i8*)
+#line 300
+  ++x;
+  return x;
+}
+// CHECK-LABEL: @t3(
+unsigned short t3(unsigned short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_400]] to i8*)
+#line 400
+  --x;
+  return x;
+}
+
+// CHECK-LABEL: @t4(
+signed short t4(signed short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_500]] to i8*)
+#line 500
+  x++;
+  return x;
+}
+// CHECK-LABEL: @t5(
+signed short t5(signed short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_600]] to i8*)
+#line 600
+  x--;
+  return x;
+}
+// CHECK-LABEL: @t6(
+signed short t6(signed short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_700]] to i8*)
+#line 700
+  ++x;
+  return x;
+}
+// CHECK-LABEL: @t7(
+signed short t7(signed short x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_800]] to i8*)
+#line 800
+  --x;
+  return x;
+}
+
+// CHECK-LABEL: @t8(
+unsigned char t8(unsigned char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_900]] to i8*)
+#line 900
+  x++;
+  return x;
+}
+// CHECK-LABEL: @t9(
+unsigned char t9(unsigned char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1000]] to i8*)
+#line 1000
+  x--;
+  return x;
+}
+// CHECK-LABEL: @t10(
+unsigned char t10(unsigned char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1100]] to i8*)
+#line 1100
+  ++x;
+  return x;
+}
+// CHECK-LABEL: @t11(
+unsigned char t11(unsigned char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1200]] to i8*)
+#line 1200
+  --x;
+  return x;
+}
+
+// CHECK-LABEL: @t12(
+signed char t12(signed char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1300]] to i8*)
+#line 1300
+  x++;
+  return x;
+}
+// CHECK-LABEL: @t13(
+signed char t13(signed char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1400]] to i8*)
+#line 1400
+  x--;
+  return x;
+}
+// CHECK-LABEL: @t14(
+signed char t14(signed char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1500]] to i8*)
+#line 1500
+  ++x;
+  return x;
+}
+// CHECK-LABEL: @t15(
+signed char t15(signed char x) {
+  // CHECK: call void @__ubsan_handle_implicit_conversion(i8* bitcast ({ {{{.*}}}, {{{.*}}}*, {{{.*}}}*, i8 }* @[[LINE_1600]] to i8*)
+#line 1600
+  --x;
+  return x;
+}