Fix for tril, triu when kernel_sched returns LS, GS such that LS*GS < N #589
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
An issue arises from the following discussion:
https://groups.google.com/d/msg/theano-users/lc6uMHU2rn4/Soxjr5c1FQAJ
It is related to issue #438 . The current implementation of
tril,triuand do_call do not take into account when the case kernel_sched returns LS and GS such that LS*GS < N.This PR is a quick fix by computing LS and GS itself and it guarantees LS*GS >= N.
I know that it may not be the most efficient value of LS, GS.
However, I am not sure the logic behind
kernel_schedand I believe if I modify it, everything may be affected. Therefore I decided to lettrilandtriucompute their own value of LS and GS intriltotriuto fix it.