https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66896
Martin Jambor <jamborm at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |jamborm at gcc dot gnu.org
--- Comment #7 from Martin Jambor <jamborm at gcc dot gnu.org> ---
(In reply to Vittorio Zecca from comment #6)
> I just found the same issue at line 2479:
As even the the gcc-5 branch is changing slightly, line numbers on
their own are not are not the best indicator of the position of the
code you refer to.
> dst_ctx->combine_with (ctx);
> dst_ctx is again NULL
> Maybe the same patch should be applied here? Namely:
> if (!dst_ctx)
> {
> vec_safe_grow_cleared
> (args->polymorphic_call_contexts,
> count);
> dst_ctx =
> ipa_get_ith_polymorhic_call_context (args, i);
> }
> immediately before?
I have looked at both occurrences of dst_ctx->combine_with (ctx) in
function update_jump_functions_after_inlining and they are all
guarded with the code just as you describe.
> Do you need a reproducer?
Yes please, I cannot see the error you describe. Thanks.
