[PATCH] D19226: [X86] enable PIE for functions

[Asaf Badouh via llvm-commits]

AsafBadouh marked 2 inline comments as done.

================
Comment at: ../llvmOrg/lib/Target/X86/X86Subtarget.h:559
@@ +558,3 @@
+  /// for the current subtarget.
+  /// isFast - indicate fastLowerCall, ignore NonLazyBind attribute 
+  unsigned char classifyGlobalFunctionReference(const GlobalValue *GV,
----------------
rnk wrote:
> Is this parameter really necessary? I expect you can remove it, even if you have to change a few test expectations.
I can remove isFast flag but in case it's called from "fastLowerCall" and the return value is "MO_GOTPCREL" I will need to reset it back to zero.
that's because in the fast path we don't handle "NonLazyBind" case in the fast path (see lib/Target/X86/X86ISelLowering.cpp:3333).

do you think we should handle it in the fast path?
or should I remove the flag and reset the return value inside "fastLowerCall"?


Repository:
  rL LLVM

http://reviews.llvm.org/D19226