[Clang] [NFC] Prevent null pointer dereference in Sema::InstantiateFunctionDefinition

[smanna12]

In the lambda function within clang::Sema::InstantiateFunctionDefinition, the return value of a function that may return null is now checked before dereferencing to avoid potential null pointer dereference issues which can lead to crashes or undefined behavior in the program.

[llvmbot]

@llvm/pr-subscribers-clang

Author: None (smanna12)

Changes

…nctionDefinition

In the lambda function within clang::Sema::InstantiateFunctionDefinition, the return value of a function that may return null is now checked before dereferencing to avoid potential null pointer dereference issues.

---

Full diff: https://github.com/llvm/llvm-project/pull/89801.diff

1 Files Affected:

• (modified) clang/lib/Sema/SemaTemplateInstantiateDecl.cpp (+2)

diff --git a/clang/lib/Sema/SemaTemplateInstantiateDecl.cpp b/clang/lib/Sema/SemaTemplateInstantiateDecl.cpp
index 787a485e0b2f8c..2490ec4ca614a8 100644
--- a/clang/lib/Sema/SemaTemplateInstantiateDecl.cpp
+++ b/clang/lib/Sema/SemaTemplateInstantiateDecl.cpp
@@ -5183,7 +5183,9 @@ void Sema::InstantiateFunctionDefinition(SourceLocation PointOfInstantiation,
     Function->setTypeSourceInfo(NewSI);
 
     ParmVarDecl *Parm = Function->getParamDecl(0);
+    assert(Parm && "First parameter not found in function declaration.");
     TypeSourceInfo *NewParmSI = IR.TransformType(Parm->getTypeSourceInfo());
+    assert(NewParmSI && "Type transformation failed.");
     Parm->setType(NewParmSI->getType());
     Parm->setTypeSourceInfo(NewParmSI);
   };

[Sirraide]

A pr that just adds assertions and nothing else seems like a very weird thing to do. Moreover, a few lines above in the same function, we return if this isn’t a copy/move ctor/assignment operator, which always has at least one parameter; I don’t think it is possible for Parm to be nullptr here. Do you have any code examples that would trigger this asertion here?

[Sirraide]

And moreover, getParamDecl() already asserts that the index is in bounds.

Thanks @Sirraide for reviews and the pointer. Yes, it is not possible for Param to be nullptr. I have removed that assert

[tahonermann]

Looks good to me, thanks @smanna12!

[smanna12]

Thank you @tahonermann for reviews!