[PATCH] D157591: [AArch64][GlobalISel] Optimize Combine Funnel Shift

[Dave Green via Phabricator via llvm-commits]

dmgreen added a comment.

Is it possible to add a vector test to the mir tests too?



================
Comment at: llvm/include/llvm/Target/GlobalISel/Combine.td:387
+// Fold fshr x, y, 0 -> y
+def funnel_shift_right_zero: GICombineRule<
+  (defs root:$root),
----------------
I would move these down to next to the other funnel shift combines


================
Comment at: llvm/include/llvm/Target/GlobalISel/Combine.td:843
 def funnel_shift_combines : GICombineGroup<[funnel_shift_from_or_shift,
-                                            funnel_shift_to_rotate]>;
+                                            funnel_shift_to_rotate, funnel_shift_right_zero, funnel_shift_left_zero, funnel_shift_overshift]>;
 
----------------
Can you format this to be a shorter line.


================
Comment at: llvm/lib/CodeGen/GlobalISel/CombinerHelper.cpp:2623
+
+  // Get the information needed
+  Register ConstReg = MI.getOperand(3).getReg();
----------------
This comment maybe doesn't add much


================
Comment at: llvm/lib/CodeGen/GlobalISel/CombinerHelper.cpp:2631
+
+  APInt NewConst = VRegAndVal->Value.urem(
+      APInt(ConstTy.getSizeInBits(), DstTy.getSizeInBits()));
----------------
It would be good to have a comment here that explains what this is doing.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D157591/new/

https://reviews.llvm.org/D157591