[PATCH] Add support for __nvvm_reflect changes in libdevice in CUDA-7.0

[Eli Bendersky]

================
Comment at: lib/Target/NVPTX/NVVMReflect.cpp:162
@@ +161,3 @@
+    const Value *Operand = cast<Constant>(Sym)->getOperand(0);
+    if (isa<GlobalVariable>(Operand)) {
+      assert(cast<GlobalVariable>(Operand)->hasInitializer() &&
----------------
Artem, it definitely makes sense to document this change. For lack of a better place, a detailed comment somewhere here would do. Please describe the difference between the two formats and how it's handled here.

http://reviews.llvm.org/D8399

EMAIL PREFERENCES
  http://reviews.llvm.org/settings/panel/emailpreferences/