efriedma-quic wrote: > That's for instance useful if your architecture has memory segments whose > borders allocated objects cannot cross and where you can only fold offsets > into memory access instructions if they don't leave the memory segment of the > base address.
Hmm, I hadn't really thought about that. I guess that's true. That isn't really usable information on a conventional CPU target, but I guess on a target where the address-space has unusual properties, it could make sense. I think if we add something called "inbounds" to SelectionDAG, it will inevitably get misused for other purposes, though. https://github.com/llvm/llvm-project/pull/131862 _______________________________________________ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
