--- Comment #9 from rakdver at gcc dot gnu dot org 2006-08-12 08:20 ---
Patch: http://gcc.gnu.org/ml/gcc-patches/2006-08/msg00386.html
--
rakdver at gcc dot gnu dot org changed:
What|Removed |Added
--
The code inserted by output_function_profiler clobbers the lr register, causing
the function to jump back to the beginning of the function when it returns.
.globl hello
.type hello, @function
hello:
mflr 0
stw 0,4(1)
bl _mcount
stwu 1,-16(1)
FAIL: gcc.c-torture/execute/pr28651.c execution, -O0
FAIL: gcc.c-torture/execute/pr28651.c execution, -O1
FAIL: gcc.c-torture/execute/pr28651.c execution, -O2
FAIL: gcc.c-torture/execute/pr28651.c execution, -O3 -fomit-frame-pointer
FAIL: gcc.c-torture/execute/pr28651.c execution, -O3 -g
FAIL:
--- Comment #1 from pinskia at gcc dot gnu dot org 2006-08-12 18:18 ---
Confirmed, it happens everywhere.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #2 from rguenth at gcc dot gnu dot org 2006-08-12 19:11 ---
It only happens on the mainline - is dejagnu passing arguments to the
testcases?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28703
--- Comment #3 from rguenth at gcc dot gnu dot org 2006-08-12 19:20 ---
I cannot reproduce this with
make check-gcc RUNTESTFLAGS="execute.exp=pr28651.c"
ii dejagnu 1.4.4.cvs20060709-2
ii expect 5.43.0-7
Janis, any idea?
--
rguenth at gcc dot gnu dot org
--- Comment #4 from pinskia at gcc dot gnu dot org 2006-08-12 19:22 ---
I use dejagnu 1.4.4 release.
Anyways just get rid of the check inside the testcase for argc, and mark the
other function as non inliniable.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28703
--- Comment #11 from jakub at gcc dot gnu dot org 2006-08-12 19:45 ---
Created an attachment (id=12071)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=12071&action=view)
gcc41-pr27793.patch
It certainly can be backported easily, I'm using this patch on top of
redhat/gcc-4_1-branch
--- Comment #5 from rguenth at gcc dot gnu dot org 2006-08-12 19:59 ---
Sure - that would be the obvious fix if passing arguments is the problem. But
I'd like someone to confirm that first.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28703
--- Comment #1 from lopezibanez at gmail dot com 2006-08-12 20:40 ---
I can confirm this in trunk rev 115951 on i686-pc-gnu-linux.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28261
In cp/decl.c there is this code.
error ("qualified function types cannot be used to declare %s functions",
(staticp? "static member" : "free"));
The strings "static member" and "free" are not available for translation in the
po-file. Even if they were, it surely will not work in all l
--- Comment #2 from janis at gcc dot gnu dot org 2006-08-12 21:22 ---
A regression hunt on powerpc-linux using the testcase from comment #1
identified this patch:
http://gcc.gnu.org/viewcvs?view=rev&rev=114119
r114119 | mmitchel | 2006-05-25 20:18:26 + (Thu, 25 May 2006)
--- Comment #1 from janis at gcc dot gnu dot org 2006-08-12 21:25 ---
A regression hunt on powerpc-linux identified this patch:
http://gcc.gnu.org/viewcvs?view=rev&rev=114245
r114245 | mmitchel | 2006-05-30 21:25:02 + (Tue, 30 May 2006)
--
janis at gcc dot gnu dot org c
--- Comment #2 from janis at gcc dot gnu dot org 2006-08-12 21:26 ---
A regression hunt on powerpc-linux identified the following patch:
http://gcc.gnu.org/viewcvs?view=rev&rev=113081
r113081 | mmitchel | 2006-04-19 16:58:23 + (Wed, 19 Apr 2006)
--
janis at gcc dot gnu
--- Comment #2 from janis at gcc dot gnu dot org 2006-08-12 21:28 ---
A regression hunt on powerpc-linux using the testcase from the submitter's
description identified the following patch:
http://gcc.gnu.org/viewcvs?view=rev&rev=114119
r114119 | mmitchel | 2006-05-25 20:18:26 +
--- Comment #5 from janis at gcc dot gnu dot org 2006-08-12 21:31 ---
A testcase on powerpc-linux with the first testcase, mainline, checking
enabled, identified the following patch:
http://gcc.gnu.org/viewcvs?view=rev&rev=113081
r113081 | mmitchel | 2006-04-19 16:58:23 + (
--- Comment #3 from janis at gcc dot gnu dot org 2006-08-12 21:36 ---
A regression hunt on powerpc-linux identified the following patch:
http://gcc.gnu.org/viewcvs?view=rev&rev=102182
r102182 | giovannibajo | 2005-07-20 01:19:59 + (Wed, 20 Jul 2005)
--
janis at gcc dot
--- Comment #1 from janis at gcc dot gnu dot org 2006-08-12 21:37 ---
A regression hunt on powerpc-linux identified the following patch:
http://gcc.gnu.org/viewcvs?view=rev&rev=93055
r93055 | nathan | 2005-01-07 17:09:15 + (Fri, 07 Jan 2005)
--
janis at gcc dot gnu dot
--- Comment #1 from janis at gcc dot gnu dot org 2006-08-12 21:40 ---
A regression hunt on powerpc-linux identified the following patch:
http://gcc.gnu.org/viewcvs?view=rev&rev=90059
r90059 | giovannibajo | 2004-11-04 13:07:35 + (Thu, 04 Nov 2004)
--
janis at gcc dot gn
--- Comment #2 from janis at gcc dot gnu dot org 2006-08-12 21:44 ---
A regression hunt on powerpc-linux identified the following patch:
http://gcc.gnu.org/viewcvs?view=rev&rev=87810
r87810 | sayle | 2004-09-21 16:20:12 + (Tue, 21 Sep 2004)
--
janis at gcc dot gnu dot o
--- Comment #2 from janis at gcc dot gnu dot org 2006-08-12 21:45 ---
A regression hunt on powerpc-linux identified the following patch:
http://gcc.gnu.org/viewcvs?view=rev&rev=84689
r84689 | mmitchel | 2004-07-14 15:34:30 + (Wed, 14 Jul 2004)
--
janis at gcc dot gnu do
As reported at:
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=1539256&group_id=2435
This testcase:
=
namespace N
{
struct s { };
template struct tplt { void mf(const T &) {} };
tplt &f();
}
--- Comment #1 from pinskia at gcc dot gnu dot org 2006-08-12 22:57 ---
3.4.0 gave an error so this is a regression:
t.cc: In function `bool g()':
t.cc:11: error: no match for call to `(N::s) (N::s*&)'
I don't know if this is invalid code or not, but we should not be ICEing.
--
pin
--- Comment #2 from bangerth at dealii dot org 2006-08-13 00:26 ---
The code is invalid (because there is no constructor s::s(s*)), but
we can make the code valid and still ICE:
--
namespace N
{
struct A { A(A*); };
void foo(const A &);
}
template void g()
{
--- Comment #1 from amodra at bigpond dot net dot au 2006-08-13 02:45
---
Indeed it does. _mcount is supposed to restore lr before returning.
--
amodra at bigpond dot net dot au changed:
What|Removed |Added
---
25 matches
Mail list logo
