https://gcc.gnu.org/bugzilla/show_bug.cgi?id=48696
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |pinskia at gcc dot gnu.org
Last reconfirmed|2012-02-01 00:00:00 |2021-2-11
--- Comment #18 from Richard Biener <rguenth at gcc dot gnu.org> ---
And re-confirmed a few years later. Sorry :/
Sofar we've improved the middle-end representation for bitfields, we have an
underlying field covering what memory models require to see as a single entity
and we've gained simpler IL for bitfield inserts.
But the actual lowering hasn't materialized yet and we'd still need to deal
with optimizations turning larger to smaller stores if we know we can do it
(which is difficult - we've seen STLF fails across TUs in benchmarks in the
context of vectorization).
IIRC Andrew was most recently working on sth like bitfield lowering.
