On Tue, Dec 6, 2011 at 4:10 AM, Dodji Seketeli <do...@redhat.com> wrote:
> Hello,
>
> Consider this short invalid example:
>
> typedef struct _GMutex GMutex; // #1
> typedef union _GMutex GMutex; // #2 <-- _GMutex named again using a
> // "union" class-key even though it was
> // named using a "struct" class-key in #1.
>
> This is a simple diagnostics tweak to add a note, telling the user
> where the struct _GMutex was first named using a different class-key
> previously. This can greatly help understand the error in cases where
> e.g, #1 and #2 are in different header files.
>
> Bootstrapped and tested on x86_64-unknown-linux-gnu against trunk,
> approved in the audit trail of the PR and applied to trunk.
>
> Thanks.
>
>
> From: Dodji Seketeli <do...@redhat.com>
> Date: Mon, 5 Dec 2011 23:24:23 +0100
> Subject: [PATCH] PR c++/51427 - Better diagnostic when union/struct tags
> conflicts
>
> gcc/cp/
>
> PR c++/51427
> * parser.c (cp_parser_check_class_key): Add note about earlier
> declaration.
>
> gcc/testsuite/
>
> PR c++/51427
> * g++.dg/diagnostic/wrong-tag-1.C
> ---
> gcc/cp/ChangeLog | 6 ++++++
> gcc/cp/parser.c | 12 ++++++++----
> gcc/testsuite/ChangeLog | 5 +++++
> gcc/testsuite/g++.dg/diagnostic/wrong-tag-1.C | 4 ++++
> 4 files changed, 23 insertions(+), 4 deletions(-)
> create mode 100644 gcc/testsuite/g++.dg/diagnostic/wrong-tag-1.C
>
> diff --git a/gcc/cp/ChangeLog b/gcc/cp/ChangeLog
> index 8f26f23..eb9b163 100644
> --- a/gcc/cp/ChangeLog
> +++ b/gcc/cp/ChangeLog
> @@ -1,3 +1,9 @@
> +2011-12-06 Dodji Seketeli <do...@redhat.com>
> +
> + PR c++/51427
> + * parser.c (cp_parser_check_class_key): Add note about earlier
> + declaration.
> +
> 2011-12-03 Paolo Carlini <paolo.carl...@oracle.com>
>
> PR c++/51313
> diff --git a/gcc/cp/parser.c b/gcc/cp/parser.c
> index 2fdd675..6559116 100644
> --- a/gcc/cp/parser.c
> +++ b/gcc/cp/parser.c
> @@ -22520,10 +22520,14 @@ static void
> cp_parser_check_class_key (enum tag_types class_key, tree type)
> {
> if ((TREE_CODE (type) == UNION_TYPE) != (class_key == union_type))
> - permerror (input_location, "%qs tag used in naming %q#T",
> - class_key == union_type ? "union"
> - : class_key == record_type ? "struct" : "class",
> - type);
> + {
> + permerror (input_location, "%qs tag used in naming %q#T",
> + class_key == union_type ? "union"
> + : class_key == record_type ? "struct" : "class",
> + type);
> + inform (DECL_SOURCE_LOCATION (TYPE_NAME (type)),
> + "%q#T was previously declared here", type);
> + }
> }
Hmm, I am wondering if we shouldn't remove the permerror and make it
an unconditional error.
