Lunderberg commented on issue #14100: URL: https://github.com/apache/tvm/issues/14100#issuecomment-1454160886
> Emmmm, I think the kSurjective flag not neccesarily indicate injectiveness. Since h_w_fused // 54 (in [0, 54)) and kh (in [0, 3)) are independently iterated, v0 could cover the full region of [0, 56) but not injective. That's definitely a good point, and makes me think we have a naming problem somewhere. If I'm reading the checks [here](https://github.com/apache/tvm/blob/main/src/arith/iter_affine_map.cc#L617) correctly, it looks like the `IterMapLevel::Surjective` option checks whether the transformation is *in*jective. Maybe it crept in when we replaced the `bool require_bijective` argument with `IterMapLevel` -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
