diff options
| author | David Green <david.green@arm.com> | 2020-06-12 11:35:03 +0100 |
|---|---|---|
| committer | David Green <david.green@arm.com> | 2020-06-12 19:21:17 +0100 |
| commit | 46529978bfb3a05297db95c213c72556f3efeec3 (patch) | |
| tree | ee5d6680760ce1f40eb8d5f0edec34b7afdac252 /llvm/docs/tutorial/MyFirstLanguageFrontend | |
| parent | 4bf1124edab5cb489bdf21c05b083a95d263015e (diff) | |
| download | llvm-46529978bfb3a05297db95c213c72556f3efeec3.zip llvm-46529978bfb3a05297db95c213c72556f3efeec3.tar.gz llvm-46529978bfb3a05297db95c213c72556f3efeec3.tar.bz2 | |
[ARM] Always use reductions intrinsics under MVE
Similar to a recent change to the X86 backend, this changes things so
that we always produce a reduction intrinsics for all reduction types,
not just the legal ones. This gives a better chance in the backend to
custom lower them to something more suitable for MVE. Especially for
something like fadd the in-order reduction produced during DAG lowering
is already better than the shuffles produced in the midend, and we can
do even better with a bit of custom lowering.
Differential Revision: https://reviews.llvm.org/D81398
Diffstat (limited to 'llvm/docs/tutorial/MyFirstLanguageFrontend')
0 files changed, 0 insertions, 0 deletions