When a function takes a variable number of pointer arguments, with a zero

pointer marking the end of the list, the zero *must* be cast to the pointer
type.  An un-cast zero is a 32-bit int, and at least on x86_64, gcc will
not extend the zero to 64 bits, thus allowing the upper 32 bits to be
random junk.

The new END_WITH_NULL macro may be used to annotate a such a function
so that GCC (version 4 or newer) will detect the use of un-casted zero
at compile time.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23888 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Jeff Cohen
2005-10-23 04:37:20 +00:00
parent 8b7f14e970
commit 66c5fd6c53
27 changed files with 290 additions and 262 deletions

View File

@ -110,7 +110,8 @@ void DefaultIntrinsicLowering::AddPrototypes(Module &M) {
case Intrinsic::memset:
M.getOrInsertFunction("memset", PointerType::get(Type::SByteTy),
PointerType::get(Type::SByteTy),
Type::IntTy, (--(--I->arg_end()))->getType(), 0);
Type::IntTy, (--(--I->arg_end()))->getType(),
(Type *)0);
break;
case Intrinsic::isunordered:
EnsureFunctionExists(M, "isunordered", I->arg_begin(), I->arg_end(),