[Bug fortran/66189] Block loops for inline matmul

[jvdelisle at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66189

Jerry DeLisle <jvdelisle at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jvdelisle at gcc dot gnu.org

--- Comment #6 from Jerry DeLisle <jvdelisle at gcc dot gnu.org> ---
(In reply to Thomas Koenig from comment #4)
> Now we have a blocking version for the library, do
> you think we still need blocking in the front end?

Is there much to be gained? For example 3x3 and 4x4 problem sets may be more
common than 6x6. Can we gain much by blocking these smaller arrays?

If we know of key applications that could benefit from further optimizations on
the smaller array sizes then we could consider it further. For now, I would
consider closing this PR and 68600