[llvm] [WIP][InstCombine] Add assume-based optimizations for equality and AMDGPU ballot patterns (PR #160670)

[Pankaj Dwivedi via llvm-commits]

================
@@ -2206,6 +2207,23 @@ bool GVNPass::processAssumeIntrinsic(AssumeInst *IntrinsicI) {
           std::swap(CmpLHS, CmpRHS);
       }
 
+      // Optimize AMDGPU ballot pattern: assume(ballot(cmp) == -1) or
+      // assume(ballot(cmp) == exec_mask). This implies cmp is true on all
+      // active lanes and hence can be replaced with true.
+      if (isa<IntrinsicInst>(CmpLHS) && isa<Constant>(CmpRHS)) {
+        auto *IntrCall = cast<IntrinsicInst>(CmpLHS);
+        // Check if CmpLHS is a ballot intrinsic
+        if (IntrCall->getIntrinsicID() ==
+            Intrinsic::AMDGCNIntrinsics::amdgcn_ballot) {
+          Value *BallotArg = IntrCall->getArgOperand(0);
+          if (BallotArg->getType()->isIntegerTy(1) &&
+              (match(CmpRHS, m_AllOnes()) || !isa<Constant>(CmpRHS))) {
----------------
PankajDwivedi-25 wrote:

> What if it is simply a value that the programmer knows will match the current state of `cmp` across active threads?

In that case, it will be too complex to determine the state of cmp; at least, I can't see why `assume(ballot(cmp) == arbitrary)` is true.

Possibly we can we ignore it?

https://github.com/llvm/llvm-project/pull/160670