[PATCH] D122968: [AArch64][SelectionDAG] Add target-specific implementation of srem

[Eli Friedman via Phabricator via llvm-commits]

efriedma added inline comments.


================
Comment at: llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp:4582
+        return OptimizedRem;
+    }
     SDValue OptimizedDiv =
----------------
Do we need to do something special for something like the following, to match the existing handling for div/rem pairs?  I guess it's only a couple instructions different either way, but maybe worth considering.

```
define void @sdivrem(i32 %x, i32* %ap, i32* %bp) {
  %a = sdiv i32 %x, 4
  %b = srem i32 %x, 4
  store i32 %a, i32* %ap
  store i32 %b, i32* %bp
  ret void
}
```


================
Comment at: llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp:23899
+/// Given an ISD::SREM node expressing a remainder by constant power of 2,
+/// return a DAG expression that will generate the same value by right shifting.
+SDValue DAGCombiner::BuildSREMPow2(SDNode *N) {
----------------
Fix this comment?  Not sure how you compute srem "by right shifting"


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D122968/new/

https://reviews.llvm.org/D122968