On Tue, Jun 03, 2025 at 11:02:30AM -0400, Jason Merrill wrote:
> > On the potential_constant_expression_1 side, important change was to
> > set *jump_target conservatively on calls that could throw for C++26 (the
> > patch uses magic void_node for potential_constant_expression* instead of
> > VAR_DECL, so that we don't have to create new VAR_DECLs there uselessly).
> > Without that change, several methods in libstdc++ wouldn't work correctly.
> > I'm not sure what exactly potential_constant_expression_1 maps to in the
> > C++26 standard wording now and whether doing that is ok, because basically
> > after the first call to non-noexcept function it stops checking stuff.
>
> I would expect that we could do better since for constexpr we need to be
> able to see the definitions of all the functions and know whether or not
> they can actually throw.
You mean also check TREE_NOTHROW (fndecl) which we currently propagate
conservatively (if we can prove something never throws, we set it)
or something else?
If TREE_NOTHROW, that would be twice
+ tree callee_fn = cp_get_fndecl_from_callee (callee, false);
if (!flag_enforce_eh_specs
|| type_dependent_expression_p (callee)
|| !POINTER_TYPE_P (TREE_TYPE (callee))
- || !type_noexcept_p (TREE_TYPE (TREE_TYPE (callee))))
+ || (!type_noexcept_p (TREE_TYPE (TREE_TYPE (callee)))
+ && (callee_fn == NULL_TREE
+ || !TREE_NOTHROW (callee_fn))))
*jump_target = void_node;
or so.
> But that's not necessary if it's difficult; at this point
> potential_constant_expression is just a conservative early filter to avoid
> later trying to evaluate thigs that can't possibly succeed.
>
> > And, in some spots where I know potential_constant_expression_1 didn't
> > check some subexpressions (e.g. the EH only cleanups or TRY_BLOCK handlers)
> > I've added *potential_constant_expression* calls during cxx_eval_constant*,
> > not sure if I need to do that because potential_constant_expression_1 is
> > very conservative and just doesn't recurse on subexpressions in many cases.
>
> I think that shouldn't be necessary.
Ok, I can remove those then.
It is
+ /* potential_constant_expression_1 doesn't recurse into
+ TRY_BLOCK handlers, so call it now. */
+ if (!ctx->quiet
+ ? !require_potential_constant_expression (HANDLER_BODY (h))
+ : !potential_constant_expression (HANDLER_BODY (h)))
+ *non_constant_p = true;
+ else
hunk and
+ if (CLEANUP_EH_ONLY (t)
+ && (!ctx->quiet
+ ? !require_potential_constant_expression (CLEANUP_EXPR (t))
+ : !potential_constant_expression (CLEANUP_EXPR (t))))
+ *non_constant_p = true;
+ else
hunk (plus reindent the else parts of those).
Jakub
