> It doesn't add additional semantic, the computation semantic stays the same, > it is a hint to the graph compiler.
My apologies, I had meant the semantics of a node from the perspective of a TIR transformation, not the semantics from the perspective of the computation being described. For a TIR transformation, if an object is replaced, whatever attributes describe that object must be updated to refer to the new object. So if constraints are added to the block annotation, I had been thinking of that as a change to the semantics of the `BlockRealizeNode::annotations` from "does not need to be updated when a buffer is replaced" to "must be updated when a buffer is replaced". -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/pull/77#issuecomment-1163510231 You are receiving this because you are subscribed to this thread. Message ID: <apache/tvm-rfcs/pull/77/c1163510...@github.com>
