[clang][UBSan] Sanitization for alignment assumptions.

Summary:
UB isn't nice. It's cool and powerful, but not nice.
Having a way to detect it is nice though.
[[ https://wg21.link/p1007r3 | P1007R3: std::assume_aligned ]] / http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2018/p1007r2.pdf says:
```
We propose to add this functionality via a library function instead of a core language attribute.
...
If the pointer passed in is not aligned to at least N bytes, calling assume_aligned results in undefined behaviour.
```

This differential teaches clang to sanitize all the various variants of this assume-aligned attribute.

Requires D54588 for LLVM IRBuilder changes.
The compiler-rt part is D54590.

This is a second commit, the original one was r351105,
which was mass-reverted in r351159 because 2 compiler-rt tests were failing.

Reviewers: ABataev, craig.topper, vsk, rsmith, rnk, #sanitizers, erichkeane, filcab, rjmccall

Reviewed By: rjmccall

Subscribers: chandlerc, ldionne, EricWF, mclow.lists, cfe-commits, bkramer

Tags: #sanitizers

Differential Revision: https://reviews.llvm.org/D54589

llvm-svn: 351177

1 files changed, 98 insertions, 0 deletions

diff --git a/clang/lib/CodeGen/CodeGenFunction.cpp b/clang/lib/CodeGen/CodeGenFunction.cpp
index 2b25fb469d7..1713e40c312 100644
--- a/clang/lib/CodeGen/CodeGenFunction.cpp
+++ b/clang/lib/CodeGen/CodeGenFunction.cpp
@@ -2207,6 +2207,49 @@ void CodeGenFunction::unprotectFromPeepholes(PeepholeProtection protection) {
   protection.Inst->eraseFromParent();
 }
 
+void CodeGenFunction::EmitAlignmentAssumption(llvm::Value *PtrValue,
+                                              QualType Ty, SourceLocation Loc,
+                                              SourceLocation AssumptionLoc,
+                                              llvm::Value *Alignment,
+                                              llvm::Value *OffsetValue) {
+  llvm::Value *TheCheck;
+  llvm::Instruction *Assumption = Builder.CreateAlignmentAssumption(
+      CGM.getDataLayout(), PtrValue, Alignment, OffsetValue, &TheCheck);
+  if (SanOpts.has(SanitizerKind::Alignment)) {
+    EmitAlignmentAssumptionCheck(PtrValue, Ty, Loc, AssumptionLoc, Alignment,
+                                 OffsetValue, TheCheck, Assumption);
+  }
+}
+
+void CodeGenFunction::EmitAlignmentAssumption(llvm::Value *PtrValue,
+                                              QualType Ty, SourceLocation Loc,
+                                              SourceLocation AssumptionLoc,
+                                              unsigned Alignment,
+                                              llvm::Value *OffsetValue) {
+  llvm::Value *TheCheck;
+  llvm::Instruction *Assumption = Builder.CreateAlignmentAssumption(
+      CGM.getDataLayout(), PtrValue, Alignment, OffsetValue, &TheCheck);
+  if (SanOpts.has(SanitizerKind::Alignment)) {
+    llvm::Value *AlignmentVal = llvm::ConstantInt::get(IntPtrTy, Alignment);
+    EmitAlignmentAssumptionCheck(PtrValue, Ty, Loc, AssumptionLoc, AlignmentVal,
+                                 OffsetValue, TheCheck, Assumption);
+  }
+}
+
+void CodeGenFunction::EmitAlignmentAssumption(llvm::Value *PtrValue,
+                                              const Expr *E,
+                                              SourceLocation AssumptionLoc,
+                                              unsigned Alignment,
+                                              llvm::Value *OffsetValue) {
+  if (auto *CE = dyn_cast<CastExpr>(E))
+    E = CE->getSubExprAsWritten();
+  QualType Ty = E->getType();
+  SourceLocation Loc = E->getExprLoc();
+
+  EmitAlignmentAssumption(PtrValue, Ty, Loc, AssumptionLoc, Alignment,
+                          OffsetValue);
+}
+
 llvm::Value *CodeGenFunction::EmitAnnotationCall(llvm::Value *AnnotationFn,
                                                  llvm::Value *AnnotatedVal,
                                                  StringRef AnnotationStr,
@@ -2459,6 +2502,61 @@ void CodeGenFunction::EmitMultiVersionResolver(
   Builder.ClearInsertionPoint();
 }
 
+// Loc - where the diagnostic will point, where in the source code this
+//  alignment has failed.
+// SecondaryLoc - if present (will be present if sufficiently different from
+//  Loc), the diagnostic will additionally point a "Note:" to this location.
+//  It should be the location where the __attribute__((assume_aligned))
+//  was written e.g.
+void CodeGenFunction::EmitAlignmentAssumptionCheck(
+    llvm::Value *Ptr, QualType Ty, SourceLocation Loc,
+    SourceLocation SecondaryLoc, llvm::Value *Alignment,
+    llvm::Value *OffsetValue, llvm::Value *TheCheck,
+    llvm::Instruction *Assumption) {
+  assert(Assumption && isa<llvm::CallInst>(Assumption) &&
+         cast<llvm::CallInst>(Assumption)->getCalledValue() ==
+             llvm::Intrinsic::getDeclaration(
+                 Builder.GetInsertBlock()->getParent()->getParent(),
+                 llvm::Intrinsic::assume) &&
+         "Assumption should be a call to llvm.assume().");
+  assert(&(Builder.GetInsertBlock()->back()) == Assumption &&
+         "Assumption should be the last instruction of the basic block, "
+         "since the basic block is still being generated.");
+
+  if (!SanOpts.has(SanitizerKind::Alignment))
+    return;
+
+  // Don't check pointers to volatile data. The behavior here is implementation-
+  // defined.
+  if (Ty->getPointeeType().isVolatileQualified())
+    return;
+
+  // We need to temorairly remove the assumption so we can insert the
+  // sanitizer check before it, else the check will be dropped by optimizations.
+  Assumption->removeFromParent();
+
+  {
+    SanitizerScope SanScope(this);
+
+    if (!OffsetValue)
+      OffsetValue = Builder.getInt1(0); // no offset.
+
+    llvm::Constant *StaticData[] = {EmitCheckSourceLocation(Loc),
+                                    EmitCheckSourceLocation(SecondaryLoc),
+                                    EmitCheckTypeDescriptor(Ty)};
+    llvm::Value *DynamicData[] = {EmitCheckValue(Ptr),
+                                  EmitCheckValue(Alignment),
+                                  EmitCheckValue(OffsetValue)};
+    EmitCheck({std::make_pair(TheCheck, SanitizerKind::Alignment)},
+              SanitizerHandler::AlignmentAssumption, StaticData, DynamicData);
+  }
+
+  // We are now in the (new, empty) "cont" basic block.
+  // Reintroduce the assumption.
+  Builder.Insert(Assumption);
+  // FIXME: Assumption still has it's original basic block as it's Parent.
+}
+
 llvm::DebugLoc CodeGenFunction::SourceLocToDebugLoc(SourceLocation Location) {
   if (CGDebugInfo *DI = getDebugInfo())
     return DI->SourceLocToDebugLoc(Location);