[PATCH] D63731: [AMDGPU] Prevent VGPR copies from moving across the EXEC mask definitions

[Quentin Colombet via Phabricator via llvm-commits]

qcolombet added inline comments.


================
Comment at: lib/CodeGen/PeepholeOptimizer.cpp:1811
   // If someone breaks this assumption, bad things will happen everywhere.
-  assert(Def->getNumOperands() == 2 && "Invalid number of operands");
+  assert(Def->getNumExplicitOperands() == 2 && "Invalid number of operands");
 
----------------
We should check that we don't have any implicit def and that would be good.
Generally speaking, I am somewhat uncomfortable that we attach thing to COPY, but as long as this is just implicit use, that's less critical. Otherwise, like the previous comment said, this will break in subtle way.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D63731/new/

https://reviews.llvm.org/D63731