public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Gabriel Dos Reis <gdr@integrable-solutions.net>
To: Dodji Seketeli <dodji@redhat.com>
Cc: GCC Patches <gcc-patches@gcc.gnu.org>,
	Tom Tromey <tromey@redhat.com>, 	Jason Merrill <jason@redhat.com>
Subject: Re: [PATCH 10/13] Fix location for static class members
Date: Wed, 25 Apr 2012 14:13:00 -0000	[thread overview]
Message-ID: <CAAiZkiBeLdTxwFkp0gmkn=CrB8edWVgTmOuofT75Y4s+hjMysA@mail.gmail.com> (raw)
In-Reply-To: <m3liljq5km.fsf@redhat.com>

On Wed, Apr 25, 2012 at 8:55 AM, Dodji Seketeli <dodji@redhat.com> wrote:
> Consider the test case g++.dg/other/offsetof5.C:
>
>    #include <stddef.h>
>
>    struct A
>    {
>      char c;
>      int &i;
>    };
>
>    int j = offsetof (A, i);            // { dg-warning "invalid access|offsetof" }
>
>    template <typename T>
>    struct S
>    {
>      T h;
>      T &i;
>      static const int j = offsetof (S, i);     // { dg-warning "invalid access|offsetof" }
>    };
>
>    int k = S<int>::j;                  // { dg-message "required from here" }
>
> The second warning (that involves the instantiation of the S template)
> is not emitted when -ftrack-macro-expansion is on.
>
> This is because during the instantiation of the member j of S
> template, the location that is used for the warning is the one for the
> DECL j (set by instantiate_decl).  And that location is inaccurately
> set to the locus of 'offsetof', which is a macro defined in a system
> header, so it's discarded by the diagnostics machinery.
>
> Note that when we reach the point where we emit the warning in
> build_class_member_access_expr offsetof expression has long been
> folded, so we cannot use e.g, the location of the ')' token that would
> have been in the source code.  So I believe the location of 'j' is the
> best we can get at this point.
>
> The patch below sets the location of the DECL for 'j' to what I
> believe is its precise location; with that, the test case passes with
> and without -ftrack-macro-expansion.  But I had to adjust
> g++.dg/template/sfinae6_neg.C for that.
>
> Tested on x86_64-unknown-linux-gnu against trunk.

OK.

>
> gcc/cp
>
>        * decl.c (grokdeclarator): Use the location carried by the
>        declarator for the DECL of the static class member.
>
> gcc/testsuite/
>
>        * g++.dg/template/sfinae6_neg.C: Adjust.
> ---
>  gcc/cp/decl.c                               |    3 ++-
>  gcc/testsuite/g++.dg/template/sfinae6_neg.C |    4 ++--
>  2 files changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/gcc/cp/decl.c b/gcc/cp/decl.c
> index 28c7cee..40818a3 100644
> --- a/gcc/cp/decl.c
> +++ b/gcc/cp/decl.c
> @@ -10267,7 +10267,8 @@ grokdeclarator (const cp_declarator *declarator,
>              {
>                /* C++ allows static class members.  All other work
>                   for this is done by grokfield.  */
> -               decl = build_lang_decl (VAR_DECL, unqualified_id, type);
> +               decl = build_lang_decl_loc (declarator->id_loc,
> +                                           VAR_DECL, unqualified_id, type);
>                set_linkage_for_static_data_member (decl);
>                /* Even if there is an in-class initialization, DECL
>                   is considered undefined until an out-of-class
> diff --git a/gcc/testsuite/g++.dg/template/sfinae6_neg.C b/gcc/testsuite/g++.dg/template/sfinae6_neg.C
> index d4be5dd..9b7bdfd1 100644
> --- a/gcc/testsuite/g++.dg/template/sfinae6_neg.C
> +++ b/gcc/testsuite/g++.dg/template/sfinae6_neg.C
> @@ -21,9 +21,9 @@ no_type check_is_callable2(...);
>  template<typename F, typename T1, typename T2 = T1>
>  struct is_callable2
>  {
> -  static const bool value =
> +  static const bool value = // { dg-error "within this context" }
>     (sizeof(check_is_callable2(type<F>(), type<T1>(), type<T2>()))
> -     == sizeof(yes_type)); // { dg-error "within this context" }
> +     == sizeof(yes_type));
>  };
>
>  #define JOIN( X, Y ) DO_JOIN( X, Y )
> --
>                Dodji

  reply	other threads:[~2012-04-25 14:13 UTC|newest]

Thread overview: 61+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-04-10 14:53 Patches to enable -ftrack-macro-expansion by default Dodji Seketeli
2012-04-10 14:55 ` [PATCH 01/11] Fix cpp_sys_macro_p with -ftrack-macro-expansion Dodji Seketeli
2012-04-11 13:40   ` Jason Merrill
2012-04-14 18:05     ` Dodji Seketeli
2012-04-15  3:52       ` Jason Merrill
2012-04-10 14:57 ` [PATCH 02/11] Fix token pasting " Dodji Seketeli
2012-04-11 13:41   ` Jason Merrill
2012-04-14 18:07     ` Dodji Seketeli
2012-04-10 15:08 ` [PATCH 03/11] Fix PCH crash on GTYed pointer-to-scalar field of a Dodji Seketeli
2012-04-11  3:15   ` Laurynas Biveinis
2012-04-10 19:43 ` [PATCH 04/11] Fix expansion point loc for macro-like tokens Dodji Seketeli
2012-04-11 13:46   ` Jason Merrill
2012-04-25  9:07     ` Dodji Seketeli
2012-04-29  4:08       ` Jason Merrill
2012-04-29 16:55         ` Dodji Seketeli
2012-04-30  3:20           ` Jason Merrill
2012-04-10 19:50 ` [PATCH 05/11] Make expand_location resolve to locus in main source file Dodji Seketeli
2012-04-11 13:49   ` Jason Merrill
2012-04-25  9:50     ` Dodji Seketeli
2012-04-25 15:31       ` Dodji Seketeli
2012-04-29  4:11         ` Jason Merrill
2012-04-29 16:57           ` Dodji Seketeli
2012-04-10 19:56 ` [PATCH 06/11] Strip "<built-in>" loc from displayed expansion context Dodji Seketeli
2012-04-11 13:50   ` Jason Merrill
2012-04-10 20:31 ` [PATCH 07/11] Fix -Wuninitialized for -ftrack-macro-expansion Dodji Seketeli
2012-04-11 13:52   ` Jason Merrill
2012-04-11  9:00 ` [PATCH 09/11] Fix va_arg type location Dodji Seketeli
2012-04-11 13:36   ` Gabriel Dos Reis
2012-04-11  9:26 ` [PATCH 08/11] Make conversion warnings work on NULL with -ftrack-macro-expansion Dodji Seketeli
2012-04-11 13:33   ` Gabriel Dos Reis
2012-04-25 13:42     ` Dodji Seketeli
2012-04-25 14:07       ` Gabriel Dos Reis
2012-04-25 14:50         ` Dodji Seketeli
2012-04-25 15:22           ` Gabriel Dos Reis
2012-04-25 13:55 ` [PATCH 10/13] Fix location for static class members Dodji Seketeli
2012-04-25 14:13   ` Gabriel Dos Reis [this message]
2012-04-25 14:04 ` [PATCH 11/13] Fix va_start related location Dodji Seketeli
2012-04-25 14:11   ` Gabriel Dos Reis
2012-04-25 15:20     ` Dodji Seketeli
2012-04-25 15:23       ` Gabriel Dos Reis
2012-04-27 15:06         ` Dodji Seketeli
2012-04-27 21:46           ` Dodji Seketeli
2012-04-28 23:30             ` Gabriel Dos Reis
2012-04-25 14:17 ` [PATCH 12/13] Adjust relevant test cases wrt -ftrack-macro-expansion=[0|2] Dodji Seketeli
2012-04-25 15:25   ` Gabriel Dos Reis
2012-04-29 17:38     ` Dodji Seketeli
2012-04-30  6:23       ` Dodji Seketeli
2012-04-30  7:34         ` Gabriel Dos Reis
2012-04-30 16:09       ` Mike Stump
2012-05-02 17:22         ` Greta Yorsh
2012-04-25 23:23   ` Benjamin Kosnik
2012-04-25 14:33 ` [PATCH 13/13] Switch -ftrack-macro-expansion=2 on by default Dodji Seketeli
2012-04-25 15:27   ` Gabriel Dos Reis
2012-04-25 17:50     ` Tom Tromey
2012-04-30 11:47 ` Patches to enable -ftrack-macro-expansion " Dodji Seketeli
2012-05-08 10:31   ` Andreas Krebbel
2012-08-26  0:28   ` Gerald Pfeifer
2012-08-26  0:41     ` Gabriel Dos Reis
2012-08-26  8:32       ` Dodji Seketeli
2012-08-26 20:01         ` Gerald Pfeifer
2012-08-26 20:14           ` Gabriel Dos Reis

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAAiZkiBeLdTxwFkp0gmkn=CrB8edWVgTmOuofT75Y4s+hjMysA@mail.gmail.com' \
    --to=gdr@integrable-solutions.net \
    --cc=dodji@redhat.com \
    --cc=gcc-patches@gcc.gnu.org \
    --cc=jason@redhat.com \
    --cc=tromey@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).