Florian Hahn 004c67ea25
[LV] Vectorize maxnum/minnum w/o fast-math flags. (#148239)
Update LV to vectorize maxnum/minnum reductions without fast-math flags,
by adding an extra check in the loop if any inputs to maxnum/minnum are
NaN, due to maxnum/minnum behavior w.r.t to signaling NaNs. Signed-zeros 
are already handled consistently by maxnum/minnum.

If any input is NaN,
 *exit the vector loop,
 *compute the reduction result up to the vector iteration that contained
   NaN inputs and
 * resume in the scalar loop


New recurrence kinds are added for reductions using maxnum/minnum
without fast-math flags.

PR: https://github.com/llvm/llvm-project/pull/148239
2025-07-18 21:58:19 +01:00
..