[Bug tree-optimization/126398] Missed optimization in byte swapping
"rguenth at gcc dot gnu.org via Gcc-bugs" <[email protected]>
| Newsgroups | gmane.comp.gcc.bugs |
|---|---|
| Message-ID | <[email protected]/bugzilla/> |
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=126398
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Ever confirmed|0 |1
Last reconfirmed| |2026-07-27
Status|UNCONFIRMED |NEW
--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Note for 'byteswap' this would be loop distribution given it's memory-based.
For both loops complete unrolling would likely help expose this to our
scalar-code bswap matchers but even for -O3 this is considered only at late
cunroll.
I also see a _gfortran_internal_pack remaining for byteswap.
le_to_be should recognize this at store-merging time iff we'd run scalar
bswap recognition again (at -O3).