F2C, CLAPACK, scipy and numpy fixes for the upstream backend #1193
Conversation
|
I think this is OK. The patches related to the change of function return type can probably also be done with a conversion script with some regexp and a list of paths. Still as discussed in #1184 (review) we might as well do the migration to the upstream backend to unblock a number of points and then try to simplify the different patches. @dalcde Let me know if you have any other comments. Thanks! |
| diff --git a/scipy/sparse/linalg/dsolve/SuperLU/scipychanges.patch b/before/scipy/sparse/linalg/dsolve/SuperLU/scipychanges.patch | ||
| deleted file mode 100644 | ||
| index 7a0a3ba..0000000 | ||
| --- a/scipy/sparse/linalg/dsolve/SuperLU/scipychanges.patch | ||
| +++ /dev/null |
There was a problem hiding this comment.
@joemarshall why is this patch removed?
There was a problem hiding this comment.
I reverted the removal for now. Let's see if it will an issue in #1102 Thanks for the feedback.
There was a problem hiding this comment.
Weird, I can't for the life of me remember why I killed that. I think it was possibly accidental when I was going through superlu and fixing the missing externs. I think that patch may not be ever called, I think it just is there to show the diff between scipy and base superlu...
Taken from #1102