[PATCH] D68360: PR41162 Implement LKK remainder and divisibility algorithms [urem]

[Tim Gymnich via Phabricator via llvm-commits]

TG908 marked 2 inline comments as done and an inline comment as not done.
TG908 added a comment.





================
Comment at: llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp:3940
   // makes sense since the simplification results in fatter code.
   if (DAG.isKnownNeverZero(N1) && !TLI.isIntDivCheap(VT, Attr)) {
     SDValue OptimizedDiv =
----------------
>>! In D68360#1698334, @lebedev.ri wrote:
> Where is that lowering being performed?

Right here


================
Comment at: llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp:4941-4943
+    if (!D.isStrictlyPositive() || D.isMaxValue() || D.isOneValue() ||
+        D.isPowerOf2()) {
+      // Divisor must be in the range of (1,2^N)
----------------
lebedev.ri wrote:
> TG908 wrote:
> > lebedev.ri wrote:
> > > There is no such restriction.
> > What do you mean?
> > In my code?
> > In LKK?
> I haven't read the paper, so i'm only looking at this code,
> and they don't incur any restrictions on the divisor: https://rise4fun.com/Alive/HiT
Oops. You are right.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D68360/new/

https://reviews.llvm.org/D68360