[Mlir-commits] [mlir] [mlir][dataflow] Drop the firstIndex argument of old visitNonControlFlowArguments, add new visitNonControlFlowArguments API and use it in IntegerRangeAnalysis (PR #175210)

[lonely eagle]

================
@@ -187,8 +187,8 @@ void AbstractSparseForwardDataFlowAnalysis::visitBlock(Block *block) {
 
     // Otherwise, we can't reason about the data-flow.
     return visitNonControlFlowArgumentsImpl(
-        block->getParentOp(), RegionSuccessor(block->getParent()), ValueRange(),
-        argLattices, /*firstIndex=*/0);
+        block->getParentOp(), block->getParent(), block->getParent()->getArguments(),
----------------
linuxlonelyeagle wrote:

I'd like to explain why I introduced the new visitNonControlFlowArgumentsImpl. Since the op in question is not a RegionBranchOp, I believe it’s conceptually inconsistent to use its regions to construct a RegionSuccessor. Doing so might cause confusion for users, and I'd love to hear your thoughts on this. cc: @matthias-springer @ftynse @joker-eph 

https://github.com/llvm/llvm-project/pull/175210