Hi,
On 08/29/2013 09:40 PM, Jason Merrill wrote:
On 08/29/2013 11:24 AM, Paolo Carlini wrote:
+ if ((complain & tf_error)
+ && (flags & LOOKUP_DELEGATING_CONS)
+ && name == complete_ctor_identifier
+ && TREE_CODE (ret) == CALL_EXPR
+ && (DECL_ABSTRACT_ORIGIN (TREE_OPERAND (CALL_EXPR_FN (ret), 0))
+ == current_function_decl))
+ error ("constructor delegates to itself");
How about doing this check in perform_target_ctor instead, so you
don't need another LOOKUP flag?
I should have explained some of that in better detail. The main issue I
had yesterday, is that the pattern matching can easily become very
difficult if not impossible: if you look at the second half of
expand_default_init, in some cases we wrap the returned CALL_EXPR in a
COND_EXPR, also, even more difficult, in case of constexpr constructors,
we don't have CALL_EXPRs at all (yesterday my first drafts failed very
badly for those). Only early enough, around the end of the first
build_special_member_call call we uniformly have CALL_EXPRs.
I could, for example pass down a separate bit, instead of playing again
with the LOOKUP_* bits. At some point yesterday I even had that version
tested ;)
What do you think?
Thanks,
Paolo.
