https://bugs.winehq.org/show_bug.cgi?id=54777
Nikolay Sivov bunglehead@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |NEW Ever confirmed|0 |1
--- Comment #3 from Nikolay Sivov bunglehead@gmail.com --- For floats they are using dp* instructions, that's easy to do for float vectors. For ints/bools path is different it seems. For matrices it's optimized with row/column vectors.
I think it's fine to limit it to float vector/scalar case for now.