[Bug middle-end/58723] [4.9 Regression] ICE in lto_output_edge, at lto-cgraph.c:300 for OpenMP's simd reduction

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58723

--- Comment #14 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
(In reply to Richard Biener from comment #9)
> We fail to stream cfun->has_force_vect_loops.

Note we don't necessarily have to stream that (and has_simduid_loops), the
alternative is to compute it during streaming in the IL - if it sees any
force_vect loop, it can set fun->has_force_vect_loops for the loop, similarly
for non-zero simdlen and fun->has_simduid_loops.  Those two flags are just
(pessimistically correct) flags to avoid scanning the whole IL for it, but LTO
scans the whole IL anyway...