diff options
| author | Richard Biener <rguenther@suse.de> | 2023-11-06 14:34:07 +0100 |
|---|---|---|
| committer | Richard Biener <rguenther@suse.de> | 2023-11-06 15:27:47 +0100 |
| commit | 9125969086bfa1bf804b246ea574a2329b06d2c5 (patch) | |
| tree | 57184ad8d9bc2224b6a863b8f8f3fbe33dca66ed /gcc/fortran | |
| parent | ffa0a73e152807a57fe4159a0d53ede25812fd1f (diff) | |
| download | gcc-9125969086bfa1bf804b246ea574a2329b06d2c5.zip gcc-9125969086bfa1bf804b246ea574a2329b06d2c5.tar.gz gcc-9125969086bfa1bf804b246ea574a2329b06d2c5.tar.bz2 | |
tree-optimization/112405 - SIMD clone calls with (loop) mask
The following fixes the mask argument generation for SIMD clone
calls under either loop masking or when the actual call is not
masked but only a inbranch simd clone is available. The issue
was that we tried to directly convert the vector mask to the
call argument type but SIMD clone masks require 1 or 0 (which
could be even float) values for mask elements so we have to
resort to a VEC_COND_EXPR to generate them just like we do for
regular passing of the mask.
PR tree-optimization/112405
* tree-vect-stmts.cc (vectorizable_simd_clone_call):
Properly handle invariant and/or loop mask passing.
Diffstat (limited to 'gcc/fortran')
0 files changed, 0 insertions, 0 deletions