https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80735
Martin Jambor <jamborm at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |NEW
Last reconfirmed| |2017-05-15
CC| |jamborm at gcc dot gnu.org
Ever confirmed|0 |1
--- Comment #1 from Martin Jambor <jamborm at gcc dot gnu.org> ---
Confirmed and probably not easily fixable in the current IPA-SRA
framework. IPA SRA changes the IRA in a way that would require jump
functions from aggregate to scalar arguments to work, and we don't
have those.
I would hate to raise overly optimistic expectation but I am working
on a rewrite of IPA-SRA that will make it a real IPA stage and that
should allow it and IPA-CP not to step on each other's toes, with a
little bit of effort.
Thanks for proving the example, it will be useful.
