[PATCH] prevent folding a scalar FP load into a packed logical FP instruction (PR22371)

[Quentin Colombet]

Hi Sanjay,

> For example, what if there's a denorm FP value in that extra bit of memory that we were never supposed to load?


That's a good point, we may even generate exceptions! E.g., with a packed division.

I'll have a closer look to the patch.

Thanks,
-Quentin


================
Comment at: test/CodeGen/X86/stack-align-vector-load.ll:11
@@ +10,3 @@
+; stack must be part of the { double, double } struct, so it should be safe to read
+; the extra bytes, but we need to teach the backend to recognize this case. 
+
----------------
Following the logic that reading undef is evil, shouldn't we just make sure that indeed this test is failing to fold?

http://reviews.llvm.org/D7474

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