andykaylor wrote: > Did you have any response to the recursion discussion?
I acknowledge that it's a legitimate concern, but I don't have any ideas for representing cleanups in structured MLIR that would eliminate this concern. We could possibly have introduce some way of folding nested cleanups if some threshold is crossed, but I wouldn't want to try to design that before seeing the problem in action. Or maybe this is just something that would need to be accounted for at the point where the regions were being processed. https://github.com/llvm/llvm-project/pull/177625 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
