From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-patches-return-294263-listarch-gcc-patches=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 8032 invoked by alias); 12 Jun 2011 16:27:49 -0000
Received: (qmail 8022 invoked by uid 22791); 12 Jun 2011 16:27:48 -0000
X-SWARE-Spam-Status: No, hits=-2.3 required=5.0	tests=AWL,BAYES_00,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FROM,RCVD_IN_DNSWL_LOW,RFC_ABUSE_POST
X-Spam-Check-By: sourceware.org
Received: from mail-qy0-f182.google.com (HELO mail-qy0-f182.google.com) (209.85.216.182)    by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Sun, 12 Jun 2011 16:27:26 +0000
Received: by qyk27 with SMTP id 27so2296353qyk.20        for <gcc-patches@gcc.gnu.org>; Sun, 12 Jun 2011 09:27:25 -0700 (PDT)
MIME-Version: 1.0
Received: by 10.229.17.148 with SMTP id s20mr3168758qca.149.1307896044408; Sun, 12 Jun 2011 09:27:24 -0700 (PDT)
Received: by 10.229.47.78 with HTTP; Sun, 12 Jun 2011 09:27:24 -0700 (PDT)
In-Reply-To: <BANLkTimpPtQfbrLPNgqh_WSSNq-VRPEpuQ@mail.gmail.com>
References: <20110611125552.GA25068@kam.mff.cuni.cz>	<BANLkTim23fOqkBS+QKfmmpExBrAXfTNdAw@mail.gmail.com>	<BANLkTik8zti5+FwGHX=j3_MKqEgAgxPZAA@mail.gmail.com>	<20110612145443.GA7920@kam.mff.cuni.cz>	<BANLkTimpPtQfbrLPNgqh_WSSNq-VRPEpuQ@mail.gmail.com>
Date: Sun, 12 Jun 2011 16:33:00 -0000
Message-ID: <BANLkTi=kdiDYqHNYXMcjJZYL6cCtfTAJTQ@mail.gmail.com>
Subject: Re: Cgraph alias reorg 15/14 (New infrastructure for same body aliases)
From: "H.J. Lu" <hjl.tools@gmail.com>
To: Jan Hubicka <hubicka@ucw.cz>
Cc: gcc-patches@gcc.gnu.org
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
X-IsSubscribed: yes
Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-patches.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-patches/>
List-Post: <mailto:gcc-patches@gcc.gnu.org>
List-Help: <mailto:gcc-patches-help@gcc.gnu.org>
Sender: gcc-patches-owner@gcc.gnu.org
X-SW-Source: 2011-06/txt/msg00948.txt.bz2

On Sun, Jun 12, 2011 at 8:01 AM, H.J. Lu <hjl.tools@gmail.com> wrote:
> On Sun, Jun 12, 2011 at 7:54 AM, Jan Hubicka <hubicka@ucw.cz> wrote:
>>>
>>> This also pretty much destroyed C++ for ia32:
>>>
>>> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D49378
>>> http://gcc.gnu.org/ml/gcc-regression/2011-06/msg00159.html
>>
>> Hi,
>> It seems somewhat amazing that we hit kernel sensitive miscompilation he=
re.
>> The problem most probably is the fact that thunks and functions with thu=
nks can become
>> local. This is correct since thunks are represented as direct calls now,=
 but this
>> makes i386 to use local ABI when calling or compiling them.
>>
>> Does the following patch help? We may also need to look for the presence=
 of thunk
>> callers.
>>
>> Index: ipa.c
>> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
>> --- ipa.c =A0 =A0 =A0 (revision 174958)
>> +++ ipa.c =A0 =A0 =A0 (working copy)
>> @@ -120,6 +120,7 @@ static bool
>> =A0cgraph_non_local_node_p_1 (struct cgraph_node *node, void *data ATTRI=
BUTE_UNUSED)
>> =A0{
>> =A0 =A0return !(cgraph_only_called_directly_or_aliased_p (node)
>> + =A0 =A0 =A0 =A0 =A0 && !ipa_ref_has_aliases_p (&node->ref_list)
>> =A0 =A0 =A0 =A0 =A0 =A0&& node->analyzed
>> =A0 =A0 =A0 =A0 =A0 =A0&& !DECL_EXTERNAL (node->decl)
>> =A0 =A0 =A0 =A0 =A0 =A0&& !node->local.externally_visible
>> @@ -132,7 +133,11 @@ cgraph_non_local_node_p_1 (struct cgraph
>> =A0static bool
>> =A0cgraph_local_node_p (struct cgraph_node *node)
>> =A0{
>> - =A0 return !cgraph_for_node_and_aliases (cgraph_function_or_thunk_node=
 (node, NULL),
>> + =A0 struct cgraph_node *n =3D cgraph_function_or_thunk_node (node, NUL=
L);
>> +
>> + =A0 if (n->thunk.thunk_p)
>> + =A0 =A0 return false;
>> + =A0 return !cgraph_for_node_and_aliases (n,
>> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0cgraph_non_local_node_p_1, NULL, true);
>>
>> =A0}
>>
>
> I am testing it now. =A0Will know the results in 2 hours.
>

Yes, it fixes C++ regressions.

Thanks.

--=20
H.J.