On Tue, 2022-10-04 at 11:11 +0200, Jakub Jelinek wrote:
> Hi!
>
> In these spots, the error/error_at has some inform afterwards which
> are
> explanation part of the same diagnostics, so should be tied with
> auto_diagnostic_group with it.
>
> Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?
OK for trunk; I think future such patches can be self-approved as per
the "obvious" rule.
Thanks
Dave
>
> 2022-10-04 Jakub Jelinek <ja...@redhat.com>
>
> * attribs.cc (handle_ignored_attributes_option,
> decl_attributes,
> common_function_versions): Use auto_diagnostic_group.
>
> --- gcc/attribs.cc.jj 2022-10-03 18:31:16.130032566 +0200
> +++ gcc/attribs.cc 2022-10-03 21:27:02.364104260 +0200
> @@ -251,6 +251,7 @@ handle_ignored_attributes_option (vec<ch
> /* We don't accept '::attr'. */
> if (cln == nullptr || cln == opt)
> {
> + auto_diagnostic_group d;
> error ("wrong argument to ignored attributes");
> inform (input_location, "valid format is %<ns::attr%> or
> %<ns::%>");
> continue;
> @@ -732,6 +733,7 @@ decl_attributes (tree *node, tree attrib
> || (spec->max_length >= 0
> && nargs > spec->max_length))
> {
> + auto_diagnostic_group d;
> error ("wrong number of arguments specified for %qE
> attribute",
> name);
> if (spec->max_length < 0)
> @@ -1167,6 +1169,7 @@ common_function_versions (tree fn1, tree
> std::swap (fn1, fn2);
> attr1 = attr2;
> }
> + auto_diagnostic_group d;
> error_at (DECL_SOURCE_LOCATION (fn2),
> "missing %<target%> attribute for multi-versioned
> %qD",
> fn2);
>
> Jakub
>
