Closed guillermo-oyarzun closed 3 months ago
Hey! Thanks a lot for this PR! I see you have modified the content of
bitnot_assign_async
, but actually you should replace the content ofunchecked_bitnot_assign_async
with your changes. The unchecked version does not perform the initial carry propagation. 🙂
ahh I followed the CPU implementation. On CPU there is no unchecked_bitnot, maybe we should not have too?
Hey! Thanks a lot for this PR! I see you have modified the content of
bitnot_assign_async
, but actually you should replace the content ofunchecked_bitnot_assign_async
with your changes. The unchecked version does not perform the initial carry propagation. 🙂ahh I followed the CPU implementation. On CPU there is no unchecked_bitnot, maybe we should not have too?
@tmontaigu is there a reason there is no unchecked_bitnot on the CPU?
No I don't think there is a particular reason
Ok so @guillermo-oyarzun let's keep the unchecked entry point for the GPU and do the modifications there, I think it's better as it follows the same pattern as the other functions then.
closes: please link all relevant issues
PR content/description
Check-list: