On Mon, Jul 27, 2020 at 04:12:09PM +0200, Martin Liška wrote:
> On 7/27/20 3:16 PM, Jakub Jelinek wrote:
> > On Mon, Jul 27, 2020 at 02:32:15PM +0200, Martin Liška wrote:
> > > As mentioned in the PR, we should not create a string constant for a type
> > > that is different from char_type_node. Looking at expr.c, I was inspired
> > > and used 'TYPE_MAIN_VARIANT (chartype) == char_type_node' to verify that
> > > underlying
> > > type is a character type.
> >
> > That doesn't look correct, there is char, signed char, unsigned char,
> > or say std::byte, and all of them are perfectly fine.
> > So, rather than requiring it is char and nothing else, you should instead
> > check that it is an INTEGRAL_TYPE_P (maybe better other than BOOLEAN_TYPE?),
> > which is complete and has the same TYPE_PRECISION as char_type_node.
>
> All right, the following survives tests and bootstraps.
LGTM.
Jakub
