https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117698
--- Comment #5 from GCC Commits ---
The master branch has been updated by Richard Biener :
https://gcc.gnu.org/g:2383ed144b54be6bae5ba5aec8859c6c2e215daf
commit r15-5523-g2383ed144b54be6bae5ba5aec8859c6c2e215daf
Author: Richard Biener
Date:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117698
Richard Biener changed:
What|Removed |Added
Status|ASSIGNED|RESOLVED
Resolution|---
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117698
--- Comment #4 from Richard Biener ---
So this is another case where single-element (in this case strided) accesses
go not VMAT_ELEMENTWISE where alignment is irrelevant but instead we use
VMAT_STRIDED_SLP which isn't exempt from alignment requi
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117698
--- Comment #3 from ro at CeBiTec dot Uni-Bielefeld.DE ---
> --- Comment #1 from Richard Biener ---
> Might be due to PR114189. Alternatively can you check whether --param
> vect-force-slp=0 makes the FAILs go away.
It does indeed for all aff
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117698
Richard Biener changed:
What|Removed |Added
Last reconfirmed||2024-11-20
Assignee|unassign
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117698
--- Comment #1 from Richard Biener ---
Might be due to PR114189. Alternatively can you check whether --param
vect-force-slp=0 makes the FAILs go away.
Note I wonder whether SPARC ends up using generic vectorization (aka
integer mode) here or V
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117698
Rainer Orth changed:
What|Removed |Added
Target Milestone|--- |15.0
