diff options
author | Florian Hahn <flo@fhahn.com> | 2025-07-18 22:58:19 +0200 |
---|---|---|
committer | GitHub <noreply@github.com> | 2025-07-18 21:58:19 +0100 |
commit | 004c67ea257039e4e98abc26dd4ac6e8f3d7a171 (patch) | |
tree | 24d7ae02719807bb7d7aa3b7a6ffeb57c75c0913 /llvm/tools/llvm-profdata/llvm-profdata.cpp | |
parent | 695660cdfd1ca65cd6e02e6950d10c990dfa0036 (diff) | |
download | llvm-004c67ea257039e4e98abc26dd4ac6e8f3d7a171.zip llvm-004c67ea257039e4e98abc26dd4ac6e8f3d7a171.tar.gz llvm-004c67ea257039e4e98abc26dd4ac6e8f3d7a171.tar.bz2 |
[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
Diffstat (limited to 'llvm/tools/llvm-profdata/llvm-profdata.cpp')
0 files changed, 0 insertions, 0 deletions