diff options
| author | Florian Hahn <flo@fhahn.com> | 2023-01-06 10:56:17 +0000 |
|---|---|---|
| committer | Florian Hahn <flo@fhahn.com> | 2023-01-06 10:56:17 +0000 |
| commit | 68469a80cb7468c577701d725a56a86454b8f0cb (patch) | |
| tree | d2310c78ed45802de8204ba605cbd91ec0a3e3a1 /llvm/lib/IR/ConstantFold.cpp | |
| parent | 7ea47f9e41935e0d132499e8832bb8d45f39d698 (diff) | |
| download | llvm-68469a80cb7468c577701d725a56a86454b8f0cb.zip llvm-68469a80cb7468c577701d725a56a86454b8f0cb.tar.gz llvm-68469a80cb7468c577701d725a56a86454b8f0cb.tar.bz2 | |
[LV] Disable runtime unrolling for vectorized loops.
This patch adds metadata to disable runtime unrolling to the vectorized
loop. If runtime unrolling/interleaving is considered profitable, LV
will interleave the loop directly. There should be no need to perform
runtime unrolling at a later stage.
Note that we already add metadata to disable runtime unrolling to the
scalar loop after vectorization.
The additional unrolling unnecessarily increases code size and compile
time. In addition to that we have several bug reports of unncessary
runtime unrolling for vectorized loops, e.g. PR40961
Compile-time improvements:
NewPM-O3: -1.04%
NewPM-ReleaseThinLTO: -0.59%
NewPM-ReleaseLTO-g: -0.97%
https://llvm-compile-time-tracker.com/compare.php?from=ce1be13a868d0f8afa367975558c1a6175cce33a&to=78bc2e67f22e9e10e61cdb6cdac4bb857d95eb1b&stat=instructions:u
Fixes #40306.
Reviewed By: lebedev.ri, nikic
Differential Revision: https://reviews.llvm.org/D115261
Diffstat (limited to 'llvm/lib/IR/ConstantFold.cpp')
0 files changed, 0 insertions, 0 deletions