[Bug c/25990] gomp ICE with -fopenmp and -O2
--- Comment #1 from perrin at msli dot com 2006-01-27 07:35 --- Created an attachment (id=10740) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=10740&action=view) fails with -fopenmp and -O2 contains two #pragama omp parallel for loops -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25990
[Bug c/25990] New: gomp ICE with -fopenmp and -O2
The attached code ICE's with -fopenmp and -O2, but not with just -fopenmp
I build both the recent trunk (svn 110282) and the gomp branch, and it ICE's
with both.
/home/perrin/GOMP/INSTALL/110296/bin/gcc -Wall -std=c99 -v --save-temps -O2
-fopenmp-m64 -o test_gomp_2 test_gomp_2.c
Using built-in specs.
Target: x86_64-unknown-linux-gnu
Configured with: /home/perrin/GOMP/gomp-20050608-branch/configure
--prefix=/home/perrin/GOMP/INSTALL/110296/ --enable-threads=posix
--enable-languages=c,c++,fortran
Thread model: posix
gcc version 4.2.0-gomp-20050608-branch 20060126 (experimental) (merged
20060126)
/home/perrin/GOMP/INSTALL/110296/bin/../libexec/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/cc1
-E -quiet -v -iprefix
/home/perrin/GOMP/INSTALL/110296/bin/../lib/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/
-D_REENTRANT test_gomp_2.c -m64 -mtune=generic -std=c99 -Wall -fopenmp -O2
-fpch-preprocess -o test_gomp_2.i
ignoring nonexistent directory
"/home/perrin/GOMP/INSTALL/110296/bin/../lib/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/../../../../x86_64-unknown-linux-gnu/include"
ignoring duplicate directory
"/home/perrin/GOMP/INSTALL/110296//lib/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/include"
ignoring nonexistent directory
"/home/perrin/GOMP/INSTALL/110296//lib/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/../../../../x86_64-unknown-linux-gnu/include"
#include "..." search starts here:
#include <...> search starts here:
/home/perrin/GOMP/INSTALL/110296/bin/../lib/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/include
/usr/local/include
/home/perrin/GOMP/INSTALL/110296//include
/usr/include
End of search list.
/home/perrin/GOMP/INSTALL/110296/bin/../libexec/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/cc1
-fpreprocessed test_gomp_2.i -quiet -dumpbase test_gomp_2.c -m64 -mtune=generic
-auxbase test_gomp_2 -O2 -Wall -std=c99 -version -fopenmp -o test_gomp_2.s
GNU C version 4.2.0-gomp-20050608-branch 20060126 (experimental) (merged
20060126) (x86_64-unknown-linux-gnu)
compiled by GNU C version 4.2.0-gomp-20050608-branch 20060126
(experimental) (merged 20060126).
GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096
Compiler executable checksum: b18775707d7fd6662765314b1efc240e
test_gomp_2.c: In function 'main':
test_gomp_2.c:121: warning: unused variable 'HH'
test_gomp_2.c:120: warning: unused variable 'GG'
test_gomp_2.c: In function 'main.omp_fn.1':
test_gomp_2.c:223: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html> for instructions.
make: *** [test_gomp] Error 1
--
Summary: gomp ICE with -fopenmp and -O2
Product: gcc
Version: 4.2.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: perrin at msli dot com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25990
[Bug c/25989] gomp ICE with -O2 and schedule(guided)
--- Comment #1 from perrin at msli dot com 2006-01-27 07:18 --- Created an attachment (id=10739) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=10739&action=view) code fails when compiled with -O2 -fopenmp -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25989
[Bug c/25989] New: gomp ICE with -O2 and schedule(guided)
/home/perrin/gcc_HEAD/INSTALL/110282/bin/gcc -Wall -std=c99 -v --save-temps -O2
-fopenmp-m64 -o test_gomp_2 test_gomp_2.c
Using built-in specs.
Target: x86_64-unknown-linux-gnu
Configured with: /home/perrin/gcc_HEAD/gcc/configure
--prefix=/home/perrin/gcc_HEAD/INSTALL/110282/ --enable-threads=posix
--enable-languages=c,c++,fortran
Thread model: posix
gcc version 4.2.0 20060126 (experimental)
/home/perrin/gcc_HEAD/INSTALL/110282/bin/../libexec/gcc/x86_64-unknown-linux-gnu/4.2.0/cc1
-E -quiet -v -iprefix
/home/perrin/gcc_HEAD/INSTALL/110282/bin/../lib/gcc/x86_64-unknown-linux-gnu/4.2.0/
-D_REENTRANT test_gomp_2.c -m64 -mtune=generic -std=c99 -Wall -fopenmp -O2
-fpch-preprocess -o test_gomp_2.i
ignoring nonexistent directory
"/home/perrin/gcc_HEAD/INSTALL/110282/bin/../lib/gcc/x86_64-unknown-linux-gnu/4.2.0/../../../../x86_64-unknown-linux-gnu/include"
ignoring duplicate directory
"/home/perrin/gcc_HEAD/INSTALL/110282//lib/gcc/x86_64-unknown-linux-gnu/4.2.0/include"
ignoring nonexistent directory
"/home/perrin/gcc_HEAD/INSTALL/110282//lib/gcc/x86_64-unknown-linux-gnu/4.2.0/../../../../x86_64-unknown-linux-gnu/include"
#include "..." search starts here:
#include <...> search starts here:
/home/perrin/gcc_HEAD/INSTALL/110282/bin/../lib/gcc/x86_64-unknown-linux-gnu/4.2.0/include
/usr/local/include
/home/perrin/gcc_HEAD/INSTALL/110282//include
/usr/include
End of search list.
/home/perrin/gcc_HEAD/INSTALL/110282/bin/../libexec/gcc/x86_64-unknown-linux-gnu/4.2.0/cc1
-fpreprocessed test_gomp_2.i -quiet -dumpbase test_gomp_2.c -m64 -mtune=generic
-auxbase test_gomp_2 -O2 -Wall -std=c99 -version -fopenmp -o test_gomp_2.s
GNU C version 4.2.0 20060126 (experimental) (x86_64-unknown-linux-gnu)
compiled by GNU C version 4.2.0 20060126 (experimental).
GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096
Compiler executable checksum: bafa485bbf5b021c2e63fce173a125c4
test_gomp_2.c: In function 'main.omp_fn.0':
test_gomp_2.c:10: error: invalid operand to unary operator
.istart0.2D.1817
test_gomp_2.c:10: error: invalid operand to unary operator
.iend0.3D.1818
test_gomp_2.c:10: internal compiler error: verify_stmts failed
Please submit a full bug report,
with preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html> for instructions.
make: *** [test_gomp] Error 1
int main(void) {
int ii,jj;
float a=1.0, b;
#pragma omp parallel for schedule(guided,1) private(jj)
for ( ii = 1 ; ii <= 9 ; ii++ ) {
for ( jj = 1 ; jj <= 9 ; jj++ ) {
a = b;
}
}
}
--
Summary: gomp ICE with -O2 and schedule(guided)
Product: gcc
Version: 4.2.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: perrin at msli dot com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25989
LAPACK regression status
I am getting the following on Trunk: At line 1162 of file schkee.f Fortran runtime error: Bad integer for item 1 in list input With -O3 -march=pentium4 -funroll-loops I think we reported this before but I am not finding the PR. IIRC StevenB was going to delve into this optimization bug. Can anyone recall the PR? I think I may have failed to submit a PR before. LAPACK passes fine with no optimizations so it is not a Fortran Frontend or Library problem. Jerry
[Bug ada/25988] New: visibility problem with generic child package
Compiling test.adb results in
test.adb:7:04: instantiation error at foo.adb:10
test.adb:7:04: "S_Two" is not visible
test.adb:7:04: instantiation error at foo.adb:10
test.adb:7:04: non-visible declaration at xy-z.ads:4
test.adb:7:04: instantiation error at foo.adb:10
test.adb:7:04: non-visible declaration at s_x-z.ads:6
even though S_Two should be perfectly visible.
Uncommenting the indicated line in foo.adb "fixes" the problem.
This error occurs with various versions of gnat and gcc.
Anyhow:
Using built-in specs.
Target: i486-linux-gnu
Configured with: ../src/configure -v
--enable-languages=c,c++,java,f95,objc,ada,treelang --prefix=/usr
--with-gxx-include-dir=/usr/include/c++/4.0.2 --enable-shared
--with-system-zlib --libexecdir=/usr/lib --enable-nls
--without-included-gettext --enable-threads=posix --program-suffix=-4.0
--enable-__cxa_atexit --enable-libstdcxx-allocator=mt --enable-clocale=gnu
--enable-libstdcxx-debug --enable-java-gc=boehm --enable-java-awt=gtk
--enable-gtk-cairo --with-java-home=/usr/lib/jvm/java-1.4.2-gcj-4.0-1.4.2.0/jre
--enable-mpfr --disable-werror --enable-checking=release i486-linux-gnu
Thread model: posix
gcc version 4.0.2 20050808 (prerelease) (Ubuntu 4.0.1-4ubuntu9)
command line: gcc-4.0 -c test.adb
And here the files:
--- x.ads
package X is
type XX is new Integer;
procedure One(N: in out XX);
end X;
--- x.adb
package body X is
procedure One(N: in out XX) is
begin
N := 1;
end One;
end X;
--- x-z.ads
package X.Z is
procedure Two(N: in out XX);
end X.Z;
--- x-z.adb
package body X.Z is
procedure Two(N: in out XX) is
begin
N := 2;
end Two;
end X.Z;
--- s_x.ads
with X;
use X;
package S_X is
subtype XX is X.XX;
procedure S_One(N: in out XX) renames One;
end S_X;
--- s_x-z.ads
with X.Z;
use X.Z;
package S_X.Z is
procedure S_Two(N: in out XX) renames Two;
end S_X.Z;
--- xy.ads
generic
type Number is private;
with procedure S_One(N: in out Number) is <>;
package XY is
procedure One(N: in out Number) renames S_One;
end XY;
--- xy-z.ads
generic
with procedure S_Two(N: in out Number) is <>;
package XY.Z is
procedure Two(N: in out Number) renames S_Two;
end XY.Z;
--- foo.ads
generic
M: in Integer;
package Foo is
procedure Three(K: in out Integer);
end Foo;
--- foo.adb
with S_X, S_X.Z, XY, XY.Z;
use S_X, S_X.Z;
package body Foo is
-- use S_X.Z; -- uncommenting this "fixes" the problem
package Numbers is new XY (Number => XX);
use Numbers;
package Numbers_Z is new Numbers.Z;
procedure Three(K: in out Integer) is
begin
K := 3;
end Three;
end Foo;
--- test.adb
with Ada.Text_IO;
use Ada.Text_IO;
with Foo;
procedure Test is
package F is new Foo (M => 10);
begin
Put("hi");
end Test;
--
Summary: visibility problem with generic child package
Product: gcc
Version: unknown
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: ada
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: koch at math dot utexas dot edu
GCC build triplet: i686-pc-linux-gnu
GCC host triplet: i686-pc-linux-gnu
GCC target triplet: i686-pc-linux-gnu
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25988
[Bug c++/25986] return from funcrtion of void value allowed
--- Comment #11 from pinskia at gcc dot gnu dot org 2006-01-27 03:22 --- (In reply to comment #10) > I never advocated C. I only noted that language was unreasonably changed to > simplify few GP tricks. Changed when, before 1998 Almost 8 years now. I cannot remember what ARM stated but it might even had this special case too. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25986
[Bug c++/25986] return from funcrtion of void value allowed
--- Comment #10 from yuri at tsoft dot com 2006-01-27 03:19 --- I never advocated C. I only noted that language was unreasonably changed to simplify few GP tricks. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25986
[Bug c++/25986] return from funcrtion of void value allowed
--- Comment #9 from pinskia at gcc dot gnu dot org 2006-01-27 03:12 --- (In reply to comment #8) > > Let's close this PR as we do not have place for trolls. > I am basing my opinion on the very practical experience w/out any intent of > trolling. Please at least know that the C++ standard allows this. Also please remind yourself that C++ is no longer C and that some rules are different in C++ than in C. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25986
[Bug c++/25986] return from funcrtion of void value allowed
--- Comment #8 from yuri at tsoft dot com 2006-01-27 03:09 --- > Let's close this PR as we do not have place for trolls. I am basing my opinion on the very practical experience w/out any intent of trolling. You either provide an example of wide-range successful GP use or take back your words. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25986
[Bug c++/25986] return from funcrtion of void value allowed
--- Comment #7 from gdr at cs dot tamu dot edu 2006-01-27 02:43 --- Subject: Re: return from funcrtion of void value allowed "yuri at tsoft dot com" <[EMAIL PROTECTED]> writes: | For templates it's sufficient to treat void as type in template | specializations. | No need to further change language. G++ implements the language as defined by the ISO committee. This is not a GNU C++ extension. | Especially because generic programming doesn't really scale much beyond it's | 'local' use like in STL. you're entitled to your opinion. Let's close this PR as we do not have place for trolls. -- Gaby -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25986
[Bug c++/25986] return from funcrtion of void value allowed
--- Comment #6 from gdr at cs dot tamu dot edu 2006-01-27 02:42 ---
Subject: Re: return from funcrtion of void value allowed
"yuri at tsoft dot com" <[EMAIL PROTECTED]> writes:
| > This functionality was *added* on
| > purpose to allow generic codes to be written seamlessly
| Sure!
| Then code "void f() { void x; retrun (x); }" should work.
If you want it to work, take it to the ISO C++ committee.
-- Gaby
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25986
[Bug c++/25986] return from funcrtion of void value allowed
--- Comment #5 from pinskia at gcc dot gnu dot org 2006-01-27 01:01 ---
(In reply to comment #4)
> No need to further change language.
Change what language? C++?
Well it is part of the C++ standard.
6.6.3/2:
A return statement with an expression of type "cv void" can be used only in
functions with a return type of cv void; the expression is evaluated just
before the function returns to its caller.
Which means:
int f(void);
void g(void)
{
return (void)f();
}
is also valid.
So the warning is useless as this is explicated in the C++ standard.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
Status|UNCONFIRMED |RESOLVED
Resolution||INVALID
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25986
[Bug bootstrap/25987] insn-automata.c:2433: warning: implicit declaration of function 'hppa_fpstore_bypass_p'
--- Comment #5 from dave at hiauly1 dot hia dot nrc dot ca 2006-01-27 00:55 --- Subject: Re: insn-automata.c:2433: warning: implicit declaration of function 'hppa_fpstore_bypass_p' > Rev 110274 is where the fix went in. Trying again. Dave -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25987
[Bug bootstrap/25987] insn-automata.c:2433: warning: implicit declaration of function 'hppa_fpstore_bypass_p'
--- Comment #4 from pinskia at gcc dot gnu dot org 2006-01-27 00:54 --- Rev 110274 is where the fix went in. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25987
[Bug bootstrap/25987] insn-automata.c:2433: warning: implicit declaration of function 'hppa_fpstore_bypass_p'
--- Comment #3 from dave at hiauly1 dot hia dot nrc dot ca 2006-01-27 00:53 --- Subject: Re: insn-automata.c:2433: warning: implicit declaration of function 'hppa_fpstore_bypass_p' > Is it before or after rev. 110274? Before (revision 110267M). Dave -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25987
[Bug c++/25986] return from funcrtion of void value allowed
--- Comment #4 from yuri at tsoft dot com 2006-01-27 00:44 --- For templates it's sufficient to treat void as type in template specializations. No need to further change language. Especially because generic programming doesn't really scale much beyond it's 'local' use like in STL. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25986
[Bug bootstrap/25987] insn-automata.c:2433: warning: implicit declaration of function 'hppa_fpstore_bypass_p'
--- Comment #2 from pinskia at gcc dot gnu dot org 2006-01-27 00:40 --- Is it before or after rev. 110274? -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25987
[Bug bootstrap/25987] insn-automata.c:2433: warning: implicit declaration of function 'hppa_fpstore_bypass_p'
--- Comment #1 from pinskia at gcc dot gnu dot org 2006-01-27 00:39 --- Which revision is this? -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25987
[Bug bootstrap/25987] New: insn-automata.c:2433: warning: implicit declaration of function 'hppa_fpstore_bypass_p'
/opt/gnu/gcc/gcc-4.2.0/hppa2.0w-hp-hpux11.11/bin/ -c -O2 -g -mpa-risc-2-0 -DIN _GCC -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedant ic -Wno-long-long -Wno-variadic-macros -Wold-style-definition -Wmissing-format-a ttribute -Werror -fno-common -DHAVE_CONFIG_H -I. -I. -I../../gcc/gcc -I../../g cc/gcc/. -I../../gcc/gcc/../include -I../../gcc/gcc/../libcpp/include -I/opt/gnu /gcc/gcc-4.2.0/include -I../../gcc/gcc/../libdecnumber -I../libdecnumber ins n-automata.c -o insn-automata.o cc1: warnings being treated as errors insn-automata.c: In function 'internal_insn_latency': insn-automata.c:2433: warning: implicit declaration of function 'hppa_fpstore_by pass_p' make[3]: *** [insn-automata.o] Error 1 make[3]: Leaving directory `/mnt/gnu/gcc-3.3/objdir/gcc' This has been around for a couple of days. I thought that this issue had been fixed ... -- Summary: insn-automata.c:2433: warning: implicit declaration of function 'hppa_fpstore_bypass_p' Product: gcc Version: 4.2.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: bootstrap AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: danglin at gcc dot gnu dot org GCC build triplet: hppa2.0w-hp-hpux11.11 GCC host triplet: hppa2.0w-hp-hpux11.11 GCC target triplet: hppa2.0w-hp-hpux11.11 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25987
[Bug other/25914] strsignal.c:558: warning: comparison between signed and unsigned
--- Comment #4 from dave at hiauly1 dot hia dot nrc dot ca 2006-01-27 00:12 --- Subject: Re: strsignal.c:558: warning: comparison between signed and unsigned > It is unsigned int on Solaris too. Seems like glibc/Linux is the out man out > really as this was developed by the BSD folks. The Linux manpage says: CONFORMING TO 4.3BSD I don't have access to a 4.3BSD system but DEC Ultrix is unsigned, so it looks to me like the Linux folks have messed up. It's unfortunate but the Linux version is what made it to ISO 23360. Dave -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25914
[Bug c++/25986] return from funcrtion of void value allowed
--- Comment #3 from pinskia at gcc dot gnu dot org 2006-01-27 00:07 --- http://www.glenmccl.com/ansi_023.htm I don't know how old that page is. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25986
[Bug c++/25986] return from funcrtion of void value allowed
--- Comment #2 from yuri at tsoft dot com 2006-01-27 00:02 ---
> This functionality was *added* on
> purpose to allow generic codes to be written seamlessly
Sure!
Then code "void f() { void x; retrun (x); }" should work. But it produces:
t.C: In function `void f()':
t.C:2: variable or field `x' declared void
t.C:2: return-statement with a value, in function declared with a void return
type
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25986
[Bug c++/25986] return from funcrtion of void value allowed
--- Comment #1 from gdr at cs dot tamu dot edu 2006-01-26 23:53 --- Subject: Re: New: return from funcrtion of void value allowed "yuri at tsoft dot com" <[EMAIL PROTECTED]> writes: | The following code is meaningless says you. | therefore should produce the WARNING. I'm unconvinced. C++ has evolved. This functionality was *added* on purpose to allow generic codes to be written seamlessly. Noting is wrong with it. | At least would be useful to have some pedantic warning mode when this should | produce warning. it certainly does not belong to -pedantic. -- Gaby -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25986
[Bug c++/24996] [4.0/4.1/4.2 Regression] ICE on throw code
--- Comment #15 from rakdver at gcc dot gnu dot org 2006-01-26 23:52 --- The patch fixes the problem by making gimplification of cleanups much more robust, and able to handle nested statements, at the expense of producing a bit worse code. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24996
[Bug c++/24996] [4.0/4.1/4.2 Regression] ICE on throw code
--- Comment #14 from rakdver at gcc dot gnu dot org 2006-01-26 23:51 --- Created an attachment (id=10738) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=10738&action=view) Possible patch -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24996
[Bug c++/25986] New: return from funcrtion of void value allowed
The following code is meaningless therefore should produce the WARNING.
At least would be useful to have some pedantic warning mode when this should
produce warning.
Yuri
---testcase--
void a() { }
void f() {
return a();
}
--
Summary: return from funcrtion of void value allowed
Product: gcc
Version: 4.0.1
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c++
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: yuri at tsoft dot com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25986
[Bug middle-end/25985] [4.2 Regression] with optimization integer math fails
--- Comment #1 from pinskia at gcc dot gnu dot org 2006-01-26 23:28 --- Confirmed, quickly looking at this DOM gets it wrong. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added CC||pinskia at gcc dot gnu dot ||org Status|UNCONFIRMED |NEW Ever Confirmed|0 |1 Keywords||wrong-code Last reconfirmed|-00-00 00:00:00 |2006-01-26 23:28:04 date|| Summary|with optimization integer |[4.2 Regression] with |math fails |optimization integer math ||fails Target Milestone|--- |4.2.0 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25985
[Bug libgomp/25984] libgomp installs include/omp_lib.f90 even if Fortran is not built
--- Comment #2 from pinskia at gcc dot gnu dot org 2006-01-26 23:21 --- (In reply to comment #1) > should not this be versioned as other components do (e.g. C++ header > files)? That is a different bug which was already filed, PR 25938. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25984
[Bug libgomp/25984] libgomp installs include/omp_lib.f90 even if Fortran is not built
--- Comment #1 from gdr at cs dot tamu dot edu 2006-01-26 23:16 --- Subject: Re: New: libgomp installs include/omp_lib.f90 even if Fortran is not built "gerald at pfeifer dot com" <[EMAIL PROTECTED]> writes: | I found that libgomp installs $PREFIX/include/omp_lib.f90, even in case should not this be versioned as other components do (e.g. C++ header files)? -- Gaby -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25984
[Bug c/25985] New: with optimization integer math fails
gcc version 4.2.0 20060121
With optimzation the following code skips the last loop iteration:
#include
int main(void)
{
int bits = 25;
while (bits) {
printf("bits=%d\n",bits);
if (bits >= 8) {
bits -= 8;
} else {
bits = 0;
}
}
return 0;
}
[EMAIL PROTECTED]:~/pg/test$ ~/gcc/42/bin/gcc ip2.c
[EMAIL PROTECTED]:~/pg/test$ ./a.out
bits=25
bits=17
bits=9
bits=1
[EMAIL PROTECTED]:~/pg/test$ ~/gcc/42/bin/gcc -O ip2.c
[EMAIL PROTECTED]:~/pg/test$ ./a.out
bits=25
bits=17
bits=9
[EMAIL PROTECTED]:~/pg/test$
--
Summary: with optimization integer math fails
Product: gcc
Version: 4.2.0
Status: UNCONFIRMED
Severity: critical
Priority: P3
Component: c
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: jurka at ejurka dot com
GCC build triplet: i686-pc-linux-gnu
GCC host triplet: i686-pc-linux-gnu
GCC target triplet: i686-pc-linux-gnu
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25985
[Bug other/25914] strsignal.c:558: warning: comparison between signed and unsigned
--- Comment #3 from pinskia at gcc dot gnu dot org 2006-01-26 23:09 --- It is unsigned int on Solaris too. Seems like glibc/Linux is the out man out really as this was developed by the BSD folks. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25914
[Bug libgomp/25984] New: libgomp installs include/omp_lib.f90 even if Fortran is not built
I found that libgomp installs $PREFIX/include/omp_lib.f90, even in case the Fortran frontend has not been configured nor built, for example when configuring with --enable-languages=c,c++,objc,java --disable-libgcj. include/omp_lib.mod and include/omp_lib_kinds.mod are not installed in this case, so I assume installing include/omp_lib.f90 is just an oversight? -- Summary: libgomp installs include/omp_lib.f90 even if Fortran is not built Product: gcc Version: 4.2.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: libgomp AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: gerald at pfeifer dot com GCC host triplet: i686-pc-linux-gnu http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25984
[Bug bootstrap/25961] [4.2 Regression] Mainline failed to bootstrap on ia64
--- Comment #5 from pinskia at gcc dot gnu dot org 2006-01-26 22:43 --- Fixed. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|NEW |RESOLVED Keywords||build Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25961
[Bug testsuite/24962] gcc.target/ia64/20030811-1.c (test for excess errors) fails with -milp32
--- Comment #2 from sje at gcc dot gnu dot org 2006-01-26 22:40 --- Subject: Bug 24962 Author: sje Date: Thu Jan 26 22:40:15 2006 New Revision: 110283 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110283 Log: PR testsuite/24962 * gcc.target/ia64/20030811-1.c: Change 'long' to 'long long'. Modified: branches/gcc-4_1-branch/gcc/testsuite/ChangeLog branches/gcc-4_1-branch/gcc/testsuite/gcc.target/ia64/20030811-1.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24962
[Bug target/24831] [4.1/4.2 regression] gthr-dce.h:77: error: expected expression before '{' token
--- Comment #15 from danglin at gcc dot gnu dot org 2006-01-26 22:31 --- Fixed by patch. -- danglin at gcc dot gnu dot org changed: What|Removed |Added Status|NEW |RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24831
[Bug other/24829] [4.1/4.2 Regression] libobjc testsuite failures
--- Comment #18 from danglin at gcc dot gnu dot org 2006-01-26 22:30 --- Fixed by patch. -- danglin at gcc dot gnu dot org changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24829
[Bug c++/16021] Tests for container swap specialisations FAIL in debug mode
--- Comment #15 from jason at gcc dot gnu dot org 2006-01-26 22:29 --- Subject: Bug 16021 Author: jason Date: Thu Jan 26 22:29:12 2006 New Revision: 110282 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110282 Log: PR c++/16021 * name-lookup.c (parse_using_directive): Require strong using to name a nested namespace. Modified: trunk/gcc/cp/ChangeLog trunk/gcc/cp/name-lookup.c trunk/gcc/testsuite/g++.dg/lookup/strong-using-1.C trunk/gcc/testsuite/g++.dg/lookup/strong-using-2.C trunk/gcc/testsuite/g++.dg/lookup/strong-using-3.C -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16021
[Bug other/24829] [4.1/4.2 Regression] libobjc testsuite failures
--- Comment #17 from danglin at gcc dot gnu dot org 2006-01-26 22:26 --- Subject: Bug 24829 Author: danglin Date: Thu Jan 26 22:26:49 2006 New Revision: 110280 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110280 Log: 2006-01-26 Alexandre Oliva <[EMAIL PROTECTED]> PR other/24829 PR target/24831 * gthr-dce.h: Do not gthrw-ify pthread_once_init, pthread_key_delete. From John David Anglin. gthrw-ify pthread_getunique_np only if it it's not a macro. Delete UNUSED macro. (__gthread_objc_condition_allocate, __gthread_objc_condition_deallocate, __gthread_objc_condition_wait, __gthread_objc_condition_broadcast, __gthread_objc_condition_signal, __gthread_key_delete): Add "__attribute__ ((__unused__))" to unused arguments. (__gthread_recursive_mutex_init_function): Add missing return. * gthr-dce.h, gthr-tpf.h, gthr-solaris.h, gthr-posix.h, gthr-posix95.h: Use macro-based approach for name redirection on targets missing weakref. Modified: branches/gcc-4_1-branch/gcc/ChangeLog branches/gcc-4_1-branch/gcc/gthr-dce.h branches/gcc-4_1-branch/gcc/gthr-posix.h branches/gcc-4_1-branch/gcc/gthr-posix95.h branches/gcc-4_1-branch/gcc/gthr-solaris.h branches/gcc-4_1-branch/gcc/gthr-tpf.h -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24829
[Bug target/24831] [4.1/4.2 regression] gthr-dce.h:77: error: expected expression before '{' token
--- Comment #14 from danglin at gcc dot gnu dot org 2006-01-26 22:26 --- Subject: Bug 24831 Author: danglin Date: Thu Jan 26 22:26:49 2006 New Revision: 110280 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110280 Log: 2006-01-26 Alexandre Oliva <[EMAIL PROTECTED]> PR other/24829 PR target/24831 * gthr-dce.h: Do not gthrw-ify pthread_once_init, pthread_key_delete. From John David Anglin. gthrw-ify pthread_getunique_np only if it it's not a macro. Delete UNUSED macro. (__gthread_objc_condition_allocate, __gthread_objc_condition_deallocate, __gthread_objc_condition_wait, __gthread_objc_condition_broadcast, __gthread_objc_condition_signal, __gthread_key_delete): Add "__attribute__ ((__unused__))" to unused arguments. (__gthread_recursive_mutex_init_function): Add missing return. * gthr-dce.h, gthr-tpf.h, gthr-solaris.h, gthr-posix.h, gthr-posix95.h: Use macro-based approach for name redirection on targets missing weakref. Modified: branches/gcc-4_1-branch/gcc/ChangeLog branches/gcc-4_1-branch/gcc/gthr-dce.h branches/gcc-4_1-branch/gcc/gthr-posix.h branches/gcc-4_1-branch/gcc/gthr-posix95.h branches/gcc-4_1-branch/gcc/gthr-solaris.h branches/gcc-4_1-branch/gcc/gthr-tpf.h -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24831
[Bug target/24831] [4.1/4.2 regression] gthr-dce.h:77: error: expected expression before '{' token
--- Comment #13 from danglin at gcc dot gnu dot org 2006-01-26 22:00 --- Subject: Bug 24831 Author: danglin Date: Thu Jan 26 22:00:28 2006 New Revision: 110278 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110278 Log: PR other/24829 PR target/24831 * gthr-dce.h: Do not gthrw-ify pthread_once_init, pthread_key_delete. From John David Anglin. gthrw-ify pthread_getunique_np only if it it's not a macro. Delete UNUSED macro. (__gthread_objc_condition_allocate, __gthread_objc_condition_deallocate, __gthread_objc_condition_wait, __gthread_objc_condition_broadcast, __gthread_objc_condition_signal, __gthread_key_delete): Add "__attribute__ ((__unused__))" to unused arguments. (__gthread_recursive_mutex_init_function): Add missing return. * gthr-dce.h, gthr-tpf.h, gthr-solaris.h, gthr-posix.h, gthr-posix95.h: Use macro-based approach for name redirection on targets missing weakref. Modified: trunk/gcc/ChangeLog trunk/gcc/gthr-dce.h trunk/gcc/gthr-posix.h trunk/gcc/gthr-posix95.h trunk/gcc/gthr-solaris.h trunk/gcc/gthr-tpf.h -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24831
[Bug other/24829] [4.1/4.2 Regression] libobjc testsuite failures
--- Comment #16 from danglin at gcc dot gnu dot org 2006-01-26 22:00 --- Subject: Bug 24829 Author: danglin Date: Thu Jan 26 22:00:28 2006 New Revision: 110278 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110278 Log: PR other/24829 PR target/24831 * gthr-dce.h: Do not gthrw-ify pthread_once_init, pthread_key_delete. From John David Anglin. gthrw-ify pthread_getunique_np only if it it's not a macro. Delete UNUSED macro. (__gthread_objc_condition_allocate, __gthread_objc_condition_deallocate, __gthread_objc_condition_wait, __gthread_objc_condition_broadcast, __gthread_objc_condition_signal, __gthread_key_delete): Add "__attribute__ ((__unused__))" to unused arguments. (__gthread_recursive_mutex_init_function): Add missing return. * gthr-dce.h, gthr-tpf.h, gthr-solaris.h, gthr-posix.h, gthr-posix95.h: Use macro-based approach for name redirection on targets missing weakref. Modified: trunk/gcc/ChangeLog trunk/gcc/gthr-dce.h trunk/gcc/gthr-posix.h trunk/gcc/gthr-posix95.h trunk/gcc/gthr-solaris.h trunk/gcc/gthr-tpf.h -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24829
[Bug other/25914] strsignal.c:558: warning: comparison between signed and unsigned
--- Comment #2 from sje at cup dot hp dot com 2006-01-26 21:56 --- It looks like linux (including LSB 3.1 standard) uses an int argument in psignal and Apple/BSD systems use an unsigned int argument. I don't see psignal in any ANSI or POSIX standards. The libiberty version is currently an unsigned int. -- sje at cup dot hp dot com changed: What|Removed |Added CC||sje at cup dot hp dot com http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25914
[Bug objc/25965] [4.1/4.2 Regression] Allows duplicate member names in objc subclasses
--- Comment #2 from pinskia at gcc dot gnu dot org 2006-01-26 21:30 --- Confirmed, I am going to say this is not expected at least for now until I found the patch which caused this. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|UNCONFIRMED |NEW Ever Confirmed|0 |1 Keywords||accepts-invalid Priority|P3 |P5 Last reconfirmed|-00-00 00:00:00 |2006-01-26 21:30:43 date|| Summary|Allows duplicate member |[4.1/4.2 Regression] Allows |names in objc subclasses|duplicate member names in ||objc subclasses Target Milestone|--- |4.1.0 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25965
[Bug tree-optimization/25962] Pointer (null) check after the use
--- Comment #3 from pinskia at gcc dot gnu dot org 2006-01-26 21:27 --- Confirmed. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|UNCONFIRMED |NEW Ever Confirmed|0 |1 Last reconfirmed|-00-00 00:00:00 |2006-01-26 21:27:52 date|| http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25962
[Bug bootstrap/25961] [4.2 Regression] Mainline failed to bootstrap on ia64
--- Comment #4 from sje at gcc dot gnu dot org 2006-01-26 21:14 --- Subject: Bug 25961 Author: sje Date: Thu Jan 26 21:14:16 2006 New Revision: 110274 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110274 Log: PR target/25961 * Makefile.in (insn-automata.o): Add header file dependencies. * genautomata.c: Add includes. Modified: trunk/gcc/ChangeLog trunk/gcc/Makefile.in trunk/gcc/genautomata.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25961
[Bug bootstrap/25961] [4.2 Regression] Mainline failed to bootstrap on ia64
--- Comment #3 from sje at gcc dot gnu dot org 2006-01-26 20:57 --- Subject: Bug 25961 Author: sje Date: Thu Jan 26 20:57:03 2006 New Revision: 110273 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110273 Log: PR target/25961 * Makefile.in (insn-automata.o): Add -Wno-error flag. Modified: trunk/gcc/ChangeLog trunk/gcc/Makefile.in -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25961
[Bug ada/21317] ACATS c954025 c954026 c99004a failures
--- Comment #7 from guerby at gcc dot gnu dot org 2006-01-26 20:42 --- Subject: Bug 21317 Author: guerby Date: Thu Jan 26 20:42:23 2006 New Revision: 110272 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110272 Log: 2006-01-26 Laurent GUERBY <[EMAIL PROTECTED]> PR ada/21317 * ada/acats/support/impdef.a: Add One_Long_Second. * ada/acats/tests/c9/c94025.a: Use One_Long_Second. * ada/acats/tests/c9/c94026.a: Likewise. * ada/acats/tests/c9/c97305c.ada: Likewise. * ada/acats/tests/c9/c99004a.ada: Likewise. Modified: branches/gcc-4_0-branch/gcc/testsuite/ChangeLog branches/gcc-4_0-branch/gcc/testsuite/ada/acats/support/impdef.a branches/gcc-4_0-branch/gcc/testsuite/ada/acats/tests/c9/c954025.a branches/gcc-4_0-branch/gcc/testsuite/ada/acats/tests/c9/c954026.a branches/gcc-4_0-branch/gcc/testsuite/ada/acats/tests/c9/c97305c.ada branches/gcc-4_0-branch/gcc/testsuite/ada/acats/tests/c9/c99004a.ada -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21317
[Bug c++/25983] [gomp] transient ICE in trunk, c++
--- Comment #1 from tbptbp at gmail dot com 2006-01-26 20:42 --- Created an attachment (id=10737) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=10737&action=view) Preprocessed offender -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25983
[Bug ada/21317] ACATS c954025 c954026 c99004a failures
--- Comment #6 from guerby at gcc dot gnu dot org 2006-01-26 20:40 --- Subject: Bug 21317 Author: guerby Date: Thu Jan 26 20:40:46 2006 New Revision: 110271 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110271 Log: 2006-01-26 Laurent GUERBY <[EMAIL PROTECTED]> PR ada/21317 * ada/acats/support/impdef.a: Add One_Long_Second. * ada/acats/tests/c9/c94025.a: Use One_Long_Second. * ada/acats/tests/c9/c94026.a: Likewise. * ada/acats/tests/c9/c97305c.ada: Likewise. * ada/acats/tests/c9/c99004a.ada: Likewise. Modified: branches/gcc-4_1-branch/gcc/testsuite/ChangeLog branches/gcc-4_1-branch/gcc/testsuite/ada/acats/support/impdef.a branches/gcc-4_1-branch/gcc/testsuite/ada/acats/tests/c9/c954025.a branches/gcc-4_1-branch/gcc/testsuite/ada/acats/tests/c9/c954026.a branches/gcc-4_1-branch/gcc/testsuite/ada/acats/tests/c9/c97305c.ada branches/gcc-4_1-branch/gcc/testsuite/ada/acats/tests/c9/c99004a.ada -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21317
[Bug c++/25983] New: [gomp] transient ICE in trunk, c++
Excuse the poor testcase but i was just trying to get the feel of gomp on non trivial code. It looks like some memory corruption (from the look of the nuked stack), the bigger the better. [time passes] It seems that if i leave exceptions on, an assert is hit. Here's how it looks with that minimal option set. Backtrace included. /usr/local/gomp/bin/g++ -v -save-temps -DNDEBUG -I./standalone -I/usr/include/lua50/ -I/usr/src/tbp/tolua++-1.0.5/include -I/usr/src/tbp/lib/FreeImage-cvs/Dist/ -fopenmp -c src/raytrace_packet.cpp -o bin/gcc/raytrace_packet.o Using built-in specs. Target: x86_64-unknown-linux-gnu Configured with: ../configure --prefix=/usr/local/gomp --enable-languages=c++ --enable-threads=posix --with-system-zlib --enable-__cxa_atexit --disable-multilib --enable-bootstrap --with-gcc --with-gnu-as --with-gnu-ld Thread model: posix gcc version 4.2.0-gomp-20050608-branch 20060119 (experimental) (merged 20060119) /usr/local/gomp/libexec/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/cc1plus -E -quiet -v -I./standalone -I/usr/include/lua50/ -I/usr/src/tbp/tolua++-1.0.5/include -I/usr/src/tbp/lib/FreeImage-cvs/Dist/ -D_GNU_SOURCE -D_REENTRANT -DNDEBUG src/raytrace_packet.cpp -mtune=generic -fopenmp -fpch-preprocess -o raytrace_packet.ii ignoring nonexistent directory "/usr/local/gomp/lib/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/../../../../x86_64-unknown-linux-gnu/include" ignoring nonexistent directory "/usr/src/tbp/tolua++-1.0.5/include" #include "..." search starts here: #include <...> search starts here: ./standalone /usr/include/lua50/ /usr/src/tbp/lib/FreeImage-cvs/Dist/ /usr/local/gomp/lib/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/../../../../include/c++/4.2.0-gomp-20050608-branch /usr/local/gomp/lib/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/../../../../include/c++/4.2.0-gomp-20050608-branch/x86_64-unknown-linux-gnu /usr/local/gomp/lib/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/../../../../include/c++/4.2.0-gomp-20050608-branch/backward /usr/local/include /usr/local/gomp/include /usr/local/gomp/lib/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/include /usr/include End of search list. /usr/local/gomp/libexec/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/cc1plus -fpreprocessed raytrace_packet.ii -quiet -dumpbase raytrace_packet.cpp -mtune=generic -auxbase-strip bin/gcc/raytrace_packet.o -version -fopenmp -o raytrace_packet.sGNU C++ version 4.2.0-gomp-20050608-branch 20060119 (experimental) (merged 20060119) (x86_64-unknown-linux-gnu) compiled by GNU C version 4.2.0-gomp-20050608-branch 20060119 (experimental) (merged 20060119). GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096 Compiler executable checksum: 89571ef7c91b12e8fe7e055d9792a102 src/raytrace_packet.cpp: In member function 'void rt::raytracer_t::prender()': src/raytrace_packet.cpp:1411: internal compiler error: in verify_eh_throw_stmt_node, at tree-eh.c:2085 Please submit a full bug report, with preprocessed source if appropriate. See http://gcc.gnu.org/bugs.html> for instructions. (gdb) break tree-eh.c:2085 Breakpoint 1 at 0x54fed4: file ../../gcc/tree-eh.c, line 2085. (gdb) run Starting program: /usr/local/gomp/libexec/gcc/x86_64-unknown-linux-gnu/4.2.0-gomp-20050608-branch/cc1plus -fpreprocessed raytrace_packet.ii -quiet -dumpbase raytrace_packet.cpp -mtune=generic -auxbase-strip bin/gcc/raytrace_packet.o -version -fopenmp -o raytrace_packet.s GNU C++ version 4.2.0-gomp-20050608-branch 20060119 (experimental) (merged 20060119) (x86_64-unknown-linux-gnu) compiled by GNU C version 4.2.0-gomp-20050608-branch 20060119 (experimental) (merged 20060119). GGC heuristics: --param ggc-min-expand=30 --param ggc-min-heapsize=4096 Compiler executable checksum: 89571ef7c91b12e8fe7e055d9792a102 Breakpoint 1, verify_eh_throw_stmt_node (slot=0x2cdfa000, data=0x0) at ../../gcc/tree-eh.c:2085 2085 gcc_assert (node->stmt->common.ann == NULL); (gdb) bt full #0 verify_eh_throw_stmt_node (slot=0x2cdfa000, data=0x0) at ../../gcc/tree-eh.c:2085 __FUNCTION__ = "verify_eh_throw_stmt_node" #1 0x009b23f8 in htab_traverse_noresize (htab=, callback=0x54fed0 , info=0x0) at ../../libiberty/hashtab.c:729 slot = (void **) 0x2cdfa000 limit = (void **) 0x2cdfafe8 #2 0x00889f1c in execute_one_pass (pass=0xc54f20) at ../../gcc/passes.c:849 __t = dname = 0x2b6c9100 "°5h«ª*" todo = 0 __FUNCTION__ = "execute_one_pass" #3 0x0088a03c in execute_pass_list (pass=0xc54f20) at ../../gcc/passes.c:881 No locals. #4 0x005581ca in tree_rest_of_compilation (fndecl=0x2b6c9100) at ../../gcc/tree-optimize.c:412 e = node = (struct cgraph_node *) 0x2b26a000 __FUNCTION__ = "tree_rest_of_compilation" #5 0x004d0848 in expand_body (fn=0x2b6c9100) at ../../gcc/cp/semantics.c:3015 saved_fun
[Bug ada/21317] ACATS c954025 c954026 c99004a failures
--- Comment #5 from guerby at gcc dot gnu dot org 2006-01-26 20:39 --- Subject: Bug 21317 Author: guerby Date: Thu Jan 26 20:39:16 2006 New Revision: 110270 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110270 Log: 2006-01-26 Laurent GUERBY <[EMAIL PROTECTED]> PR ada/21317 * ada/acats/support/impdef.a: Add One_Long_Second. * ada/acats/tests/c9/c94025.a: Use One_Long_Second. * ada/acats/tests/c9/c94026.a: Likewise. * ada/acats/tests/c9/c97305c.ada: Likewise. * ada/acats/tests/c9/c99004a.ada: Likewise. Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/ada/acats/support/impdef.a trunk/gcc/testsuite/ada/acats/tests/c9/c954025.a trunk/gcc/testsuite/ada/acats/tests/c9/c954026.a trunk/gcc/testsuite/ada/acats/tests/c9/c97305c.ada trunk/gcc/testsuite/ada/acats/tests/c9/c99004a.ada -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21317
[Bug c++/25980] Unexpected name conflict between symbols
--- Comment #3 from dwhorton at gmail dot com 2006-01-26 20:27 --- I see the issue. I am surprised that ADL causes non-functions from one namespace to pollute another. Nonetheless given the present state of the spec the confict as reported seems to be valid behaviour. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25980
[Bug fortran/25964] [4.1/4.2 Regression] NIST regression on fm311.f
--- Comment #4 from pault at gcc dot gnu dot org 2006-01-26 20:21 --- Fixed on trunk. Patch will be included in fix of global references in 4.1 (Sat. morning). Paul -- pault at gcc dot gnu dot org changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25964
[Bug fortran/25085] Array valued assumed length character function allowed
--- Comment #2 from pault at gcc dot gnu dot org 2006-01-26 20:19 --- Subject: Bug 25085 Author: pault Date: Thu Jan 26 20:19:09 2006 New Revision: 110269 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110269 Log: 2005-01-26 Paul Thomas <[EMAIL PROTECTED]> PR fortran/25964 * resolve.c (resolve_function): Exclude statement functions from global reference checking. PR fortran/25084 PR fortran/20852 PR fortran/25085 PR fortran/25086 * resolve.c (resolve_function): Declare a gfc_symbol to replace the references through the symtree to the symbol associated with the function expresion. Give error on reference to an assumed character length function is defined in an interface or an external function that is not a dummy argument. (resolve_symbol): Give error if an assumed character length function is array-valued, pointer-valued, pure or recursive. Emit warning that character(*) value functions are obsolescent in F95. PR fortran/25416 * trans-expr.c (gfc_conv_function_call): The above patch to resolve.c prevents any assumed character length function call from getting here except intrinsics such as SPREAD. In this case, ensure that no segfault occurs from referencing non-existent charlen->length-> expr_type and provide a backend_decl for the charlen from the charlen of the first actual argument. Cure temp name confusion. * trans-expr.c (gfc_get_interface_mapping_array): Change name of temporary from "parm" to "ifm" to avoid clash with temp coming from trans-array.c. 2005-01-26 Paul Thomas <[EMAIL PROTECTED]> PR fortran/25964 * gfortran.dg/global_references_2.f90: New test. PR fortran/25084 PR fortran/20852 PR fortran/25085 PR fortran/25086 * gfortran.dg/assumed_charlen_function_1.f90: New test. * gfortran.dg/assumed_charlen_function_3.f90: New test. PR fortran/25416 * gfortran.dg/assumed_charlen_function_2.f90: New test. Added: trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_1.f90 trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_2.f90 trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_3.f90 trunk/gcc/testsuite/gfortran.dg/global_references_2.f90 Modified: trunk/gcc/fortran/ChangeLog trunk/gcc/fortran/resolve.c trunk/gcc/fortran/trans-expr.c trunk/gcc/testsuite/ChangeLog -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25085
[Bug fortran/25964] [4.1/4.2 Regression] NIST regression on fm311.f
--- Comment #3 from pault at gcc dot gnu dot org 2006-01-26 20:19 --- Subject: Bug 25964 Author: pault Date: Thu Jan 26 20:19:09 2006 New Revision: 110269 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110269 Log: 2005-01-26 Paul Thomas <[EMAIL PROTECTED]> PR fortran/25964 * resolve.c (resolve_function): Exclude statement functions from global reference checking. PR fortran/25084 PR fortran/20852 PR fortran/25085 PR fortran/25086 * resolve.c (resolve_function): Declare a gfc_symbol to replace the references through the symtree to the symbol associated with the function expresion. Give error on reference to an assumed character length function is defined in an interface or an external function that is not a dummy argument. (resolve_symbol): Give error if an assumed character length function is array-valued, pointer-valued, pure or recursive. Emit warning that character(*) value functions are obsolescent in F95. PR fortran/25416 * trans-expr.c (gfc_conv_function_call): The above patch to resolve.c prevents any assumed character length function call from getting here except intrinsics such as SPREAD. In this case, ensure that no segfault occurs from referencing non-existent charlen->length-> expr_type and provide a backend_decl for the charlen from the charlen of the first actual argument. Cure temp name confusion. * trans-expr.c (gfc_get_interface_mapping_array): Change name of temporary from "parm" to "ifm" to avoid clash with temp coming from trans-array.c. 2005-01-26 Paul Thomas <[EMAIL PROTECTED]> PR fortran/25964 * gfortran.dg/global_references_2.f90: New test. PR fortran/25084 PR fortran/20852 PR fortran/25085 PR fortran/25086 * gfortran.dg/assumed_charlen_function_1.f90: New test. * gfortran.dg/assumed_charlen_function_3.f90: New test. PR fortran/25416 * gfortran.dg/assumed_charlen_function_2.f90: New test. Added: trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_1.f90 trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_2.f90 trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_3.f90 trunk/gcc/testsuite/gfortran.dg/global_references_2.f90 Modified: trunk/gcc/fortran/ChangeLog trunk/gcc/fortran/resolve.c trunk/gcc/fortran/trans-expr.c trunk/gcc/testsuite/ChangeLog -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25964
[Bug fortran/25416] Segmentation fault in gfc_conv_function_call
--- Comment #6 from pault at gcc dot gnu dot org 2006-01-26 20:19 --- Subject: Bug 25416 Author: pault Date: Thu Jan 26 20:19:09 2006 New Revision: 110269 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110269 Log: 2005-01-26 Paul Thomas <[EMAIL PROTECTED]> PR fortran/25964 * resolve.c (resolve_function): Exclude statement functions from global reference checking. PR fortran/25084 PR fortran/20852 PR fortran/25085 PR fortran/25086 * resolve.c (resolve_function): Declare a gfc_symbol to replace the references through the symtree to the symbol associated with the function expresion. Give error on reference to an assumed character length function is defined in an interface or an external function that is not a dummy argument. (resolve_symbol): Give error if an assumed character length function is array-valued, pointer-valued, pure or recursive. Emit warning that character(*) value functions are obsolescent in F95. PR fortran/25416 * trans-expr.c (gfc_conv_function_call): The above patch to resolve.c prevents any assumed character length function call from getting here except intrinsics such as SPREAD. In this case, ensure that no segfault occurs from referencing non-existent charlen->length-> expr_type and provide a backend_decl for the charlen from the charlen of the first actual argument. Cure temp name confusion. * trans-expr.c (gfc_get_interface_mapping_array): Change name of temporary from "parm" to "ifm" to avoid clash with temp coming from trans-array.c. 2005-01-26 Paul Thomas <[EMAIL PROTECTED]> PR fortran/25964 * gfortran.dg/global_references_2.f90: New test. PR fortran/25084 PR fortran/20852 PR fortran/25085 PR fortran/25086 * gfortran.dg/assumed_charlen_function_1.f90: New test. * gfortran.dg/assumed_charlen_function_3.f90: New test. PR fortran/25416 * gfortran.dg/assumed_charlen_function_2.f90: New test. Added: trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_1.f90 trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_2.f90 trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_3.f90 trunk/gcc/testsuite/gfortran.dg/global_references_2.f90 Modified: trunk/gcc/fortran/ChangeLog trunk/gcc/fortran/resolve.c trunk/gcc/fortran/trans-expr.c trunk/gcc/testsuite/ChangeLog -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25416
[Bug fortran/25084] Interface for assumed length character function allowed
--- Comment #2 from pault at gcc dot gnu dot org 2006-01-26 20:19 --- Subject: Bug 25084 Author: pault Date: Thu Jan 26 20:19:09 2006 New Revision: 110269 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110269 Log: 2005-01-26 Paul Thomas <[EMAIL PROTECTED]> PR fortran/25964 * resolve.c (resolve_function): Exclude statement functions from global reference checking. PR fortran/25084 PR fortran/20852 PR fortran/25085 PR fortran/25086 * resolve.c (resolve_function): Declare a gfc_symbol to replace the references through the symtree to the symbol associated with the function expresion. Give error on reference to an assumed character length function is defined in an interface or an external function that is not a dummy argument. (resolve_symbol): Give error if an assumed character length function is array-valued, pointer-valued, pure or recursive. Emit warning that character(*) value functions are obsolescent in F95. PR fortran/25416 * trans-expr.c (gfc_conv_function_call): The above patch to resolve.c prevents any assumed character length function call from getting here except intrinsics such as SPREAD. In this case, ensure that no segfault occurs from referencing non-existent charlen->length-> expr_type and provide a backend_decl for the charlen from the charlen of the first actual argument. Cure temp name confusion. * trans-expr.c (gfc_get_interface_mapping_array): Change name of temporary from "parm" to "ifm" to avoid clash with temp coming from trans-array.c. 2005-01-26 Paul Thomas <[EMAIL PROTECTED]> PR fortran/25964 * gfortran.dg/global_references_2.f90: New test. PR fortran/25084 PR fortran/20852 PR fortran/25085 PR fortran/25086 * gfortran.dg/assumed_charlen_function_1.f90: New test. * gfortran.dg/assumed_charlen_function_3.f90: New test. PR fortran/25416 * gfortran.dg/assumed_charlen_function_2.f90: New test. Added: trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_1.f90 trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_2.f90 trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_3.f90 trunk/gcc/testsuite/gfortran.dg/global_references_2.f90 Modified: trunk/gcc/fortran/ChangeLog trunk/gcc/fortran/resolve.c trunk/gcc/fortran/trans-expr.c trunk/gcc/testsuite/ChangeLog -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25084
[Bug fortran/25086] Pointer valued assumed length character function allowed
--- Comment #2 from pault at gcc dot gnu dot org 2006-01-26 20:19 --- Subject: Bug 25086 Author: pault Date: Thu Jan 26 20:19:09 2006 New Revision: 110269 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110269 Log: 2005-01-26 Paul Thomas <[EMAIL PROTECTED]> PR fortran/25964 * resolve.c (resolve_function): Exclude statement functions from global reference checking. PR fortran/25084 PR fortran/20852 PR fortran/25085 PR fortran/25086 * resolve.c (resolve_function): Declare a gfc_symbol to replace the references through the symtree to the symbol associated with the function expresion. Give error on reference to an assumed character length function is defined in an interface or an external function that is not a dummy argument. (resolve_symbol): Give error if an assumed character length function is array-valued, pointer-valued, pure or recursive. Emit warning that character(*) value functions are obsolescent in F95. PR fortran/25416 * trans-expr.c (gfc_conv_function_call): The above patch to resolve.c prevents any assumed character length function call from getting here except intrinsics such as SPREAD. In this case, ensure that no segfault occurs from referencing non-existent charlen->length-> expr_type and provide a backend_decl for the charlen from the charlen of the first actual argument. Cure temp name confusion. * trans-expr.c (gfc_get_interface_mapping_array): Change name of temporary from "parm" to "ifm" to avoid clash with temp coming from trans-array.c. 2005-01-26 Paul Thomas <[EMAIL PROTECTED]> PR fortran/25964 * gfortran.dg/global_references_2.f90: New test. PR fortran/25084 PR fortran/20852 PR fortran/25085 PR fortran/25086 * gfortran.dg/assumed_charlen_function_1.f90: New test. * gfortran.dg/assumed_charlen_function_3.f90: New test. PR fortran/25416 * gfortran.dg/assumed_charlen_function_2.f90: New test. Added: trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_1.f90 trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_2.f90 trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_3.f90 trunk/gcc/testsuite/gfortran.dg/global_references_2.f90 Modified: trunk/gcc/fortran/ChangeLog trunk/gcc/fortran/resolve.c trunk/gcc/fortran/trans-expr.c trunk/gcc/testsuite/ChangeLog -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25086
[Bug fortran/20852] *-length character function may not be recursive
--- Comment #2 from pault at gcc dot gnu dot org 2006-01-26 20:19 --- Subject: Bug 20852 Author: pault Date: Thu Jan 26 20:19:09 2006 New Revision: 110269 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110269 Log: 2005-01-26 Paul Thomas <[EMAIL PROTECTED]> PR fortran/25964 * resolve.c (resolve_function): Exclude statement functions from global reference checking. PR fortran/25084 PR fortran/20852 PR fortran/25085 PR fortran/25086 * resolve.c (resolve_function): Declare a gfc_symbol to replace the references through the symtree to the symbol associated with the function expresion. Give error on reference to an assumed character length function is defined in an interface or an external function that is not a dummy argument. (resolve_symbol): Give error if an assumed character length function is array-valued, pointer-valued, pure or recursive. Emit warning that character(*) value functions are obsolescent in F95. PR fortran/25416 * trans-expr.c (gfc_conv_function_call): The above patch to resolve.c prevents any assumed character length function call from getting here except intrinsics such as SPREAD. In this case, ensure that no segfault occurs from referencing non-existent charlen->length-> expr_type and provide a backend_decl for the charlen from the charlen of the first actual argument. Cure temp name confusion. * trans-expr.c (gfc_get_interface_mapping_array): Change name of temporary from "parm" to "ifm" to avoid clash with temp coming from trans-array.c. 2005-01-26 Paul Thomas <[EMAIL PROTECTED]> PR fortran/25964 * gfortran.dg/global_references_2.f90: New test. PR fortran/25084 PR fortran/20852 PR fortran/25085 PR fortran/25086 * gfortran.dg/assumed_charlen_function_1.f90: New test. * gfortran.dg/assumed_charlen_function_3.f90: New test. PR fortran/25416 * gfortran.dg/assumed_charlen_function_2.f90: New test. Added: trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_1.f90 trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_2.f90 trunk/gcc/testsuite/gfortran.dg/assumed_charlen_function_3.f90 trunk/gcc/testsuite/gfortran.dg/global_references_2.f90 Modified: trunk/gcc/fortran/ChangeLog trunk/gcc/fortran/resolve.c trunk/gcc/fortran/trans-expr.c trunk/gcc/testsuite/ChangeLog -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20852
[Bug other/25982] New: writer written by write_writer doesn't quote where necessary
When you have an insn condition like: "TARGET_SH1 && ! (TARGET_SHCOMPACT && (current_function_args_info.call_cookie & CALL_COOKIE_RET_TRAMP (1))) && reload_completed && lookup_attribute (\"trap_exit\", DECL_ATTRIBUTES (current_function_decl)) == NULL_TREE", build/gencondmd.c writes a file, which ultimately becomes insn-conditions.md, where "trap_exit" appears inside the define_conditions construct without the proper quoting of the quotes. -- Summary: writer written by write_writer doesn't quote where necessary Product: gcc Version: 4.2.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: other AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: amylaar at gcc dot gnu dot org http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25982
[Bug c/25892] -Wpointer-sign creates problems for Emacs
--- Comment #9 from aoliva at gcc dot gnu dot org 2006-01-26 19:09 --- Fixed -- aoliva at gcc dot gnu dot org changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25892
[Bug c/25892] -Wpointer-sign creates problems for Emacs
--- Comment #8 from aoliva at gcc dot gnu dot org 2006-01-26 19:06 --- Subject: Bug 25892 Author: aoliva Date: Thu Jan 26 19:06:06 2006 New Revision: 110267 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110267 Log: gcc/ChangeLog: PR c/25892 * c.opt (Wpointer-sign): Init to -1. * c-opts.c (c_common_handle_option): Set to 1 on OPT_Wall and OPT_pedantic, to 0 otherwise. * doc/invoke.texi: Update. gcc/testsuite/ChangeLog: PR c/25892 * gcc.dg/Wpointer-sign.c: New. * gcc.dg/Wpointer-sign-Wall.c: New. * gcc.dg/Wpointer-sign-Wall-no.c: New. * gcc.dg/Wpointer-sign-pedantic.c: New. * gcc.dg/Wpointer-sign-pedantic-no.c: New. * gcc.dg/conv-2.c: Use -Wpointer-sign. Added: branches/gcc-4_1-branch/gcc/testsuite/gcc.dg/Wpointer-sign-Wall-no.c - copied unchanged from r110265, trunk/gcc/testsuite/gcc.dg/Wpointer-sign-Wall-no.c branches/gcc-4_1-branch/gcc/testsuite/gcc.dg/Wpointer-sign-Wall.c - copied unchanged from r110265, trunk/gcc/testsuite/gcc.dg/Wpointer-sign-Wall.c branches/gcc-4_1-branch/gcc/testsuite/gcc.dg/Wpointer-sign-pedantic-no.c - copied unchanged from r110265, trunk/gcc/testsuite/gcc.dg/Wpointer-sign-pedantic-no.c branches/gcc-4_1-branch/gcc/testsuite/gcc.dg/Wpointer-sign-pedantic.c - copied unchanged from r110265, trunk/gcc/testsuite/gcc.dg/Wpointer-sign-pedantic.c branches/gcc-4_1-branch/gcc/testsuite/gcc.dg/Wpointer-sign.c - copied unchanged from r110265, trunk/gcc/testsuite/gcc.dg/Wpointer-sign.c Modified: branches/gcc-4_1-branch/gcc/ChangeLog branches/gcc-4_1-branch/gcc/c-opts.c branches/gcc-4_1-branch/gcc/c.opt branches/gcc-4_1-branch/gcc/doc/invoke.texi branches/gcc-4_1-branch/gcc/testsuite/ChangeLog branches/gcc-4_1-branch/gcc/testsuite/gcc.dg/conv-2.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25892
[Bug fortran/25964] [4.1/4.2 Regression] NIST regression on fm311.f
--- Comment #2 from pault at gcc dot gnu dot org 2006-01-26 19:01 --- The patch is on its way in the next hour or so Statement functions = internal procedures, as far as the standard is concerned. Paul -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25964
[Bug c++/25980] Unexpected name conflict between symbols
--- Comment #2 from gdr at cs dot tamu dot edu 2006-01-26 18:56 ---
Subject: Re: New: Unexpected name conflict between symbols
"dwhorton at gmail dot com" <[EMAIL PROTECTED]> writes:
| The following fails to compile, it would appear that symbols from seperate
| namespaces are colliding in error:
|
| [EMAIL PROTECTED] tmp]$ cat test2.cpp
| namespace ns1 {
| class c {};
| typedef int f;
|
| static c x;
| }
|
| namespace ns2 {
| void f(ns1::c& cc)
| {
| }
|
| void g()
| {
| f(ns1::x);
| }
| }
[...]
| test2.cpp: In function `void ns2::g()':
| test2.cpp:3: error: `ns1::f' is not a function,
| test2.cpp:10: error: conflict with `void ns2::f(ns1::c&)'
| test2.cpp:15: error: in call to `f'
| [EMAIL PROTECTED] tmp]$
| [EMAIL PROTECTED] tmp]$ cat test2.*i*
| # 1 "test2.cpp"
| # 1 ""
| # 1 ""
| # 1 "test2.cpp"
| namespace ns1 {
| class c {};
| typedef int f;
|
| static c x;
| }
|
| namespace ns2 {
| void f(ns1::c& cc)
| {
| }
|
| void g()
| {
| f(ns1::x);
| }
| }
| [EMAIL PROTECTED] tmp]$
|
| I don't see why symbols ns1::f and ns2::f are in conflict here, as when f()
in
| ns2::g is being resolved ns1::f should not be visible, should it?
This is a known issue. Name lookup apply uniformly. In particular,
ADL finds in this case both function and non-function.
Since we don't have (yet) overload set that can include functions and
non-functions, the program is ill-formed. See Core Issue 218
http://www.open-std.org/JTC1/SC22/WG21/docs/cwg_active.html#218
Notice that the last note was quite a long time ago -- before C++03
and now we're in the "extensions" timeframe. There is strong feeling
that name lookup should apply uniformly, and we should also have
overload resolution with function objects included.
I believe there alreay is a PR for this. The proper action is to
suspend it until the issue is completely resolved.
-- Gaby
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25980
[Bug target/23552] FAIL: gfortran.dg/large_real_kind_1.f90
--- Comment #3 from sje at cup dot hp dot com 2006-01-26 18:55 --- These tests pass for me on the 4.1 branch and on mainline. -- sje at cup dot hp dot com changed: What|Removed |Added Status|NEW |RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23552
[Bug c/25892] -Wpointer-sign creates problems for Emacs
--- Comment #7 from aoliva at gcc dot gnu dot org 2006-01-26 18:50 --- Subject: Bug 25892 Author: aoliva Date: Thu Jan 26 18:50:37 2006 New Revision: 110265 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110265 Log: gcc/ChangeLog: PR c/25892 * c.opt (Wpointer-sign): Init to -1. * c-opts.c (c_common_handle_option): Set to 1 on OPT_Wall and OPT_pedantic, to 0 otherwise. * doc/invoke.texi: Update. gcc/testsuite/ChangeLog: PR c/25892 * gcc.dg/Wpointer-sign.c: New. * gcc.dg/Wpointer-sign-Wall.c: New. * gcc.dg/Wpointer-sign-Wall-no.c: New. * gcc.dg/Wpointer-sign-pedantic.c: New. * gcc.dg/Wpointer-sign-pedantic-no.c: New. * gcc.dg/conv-2.c: Use -Wpointer-sign. Added: trunk/gcc/testsuite/gcc.dg/Wpointer-sign-Wall-no.c trunk/gcc/testsuite/gcc.dg/Wpointer-sign-Wall.c trunk/gcc/testsuite/gcc.dg/Wpointer-sign-pedantic-no.c trunk/gcc/testsuite/gcc.dg/Wpointer-sign-pedantic.c trunk/gcc/testsuite/gcc.dg/Wpointer-sign.c Modified: trunk/gcc/ChangeLog trunk/gcc/c-opts.c trunk/gcc/c.opt trunk/gcc/doc/invoke.texi trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/gcc.dg/conv-2.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25892
[Bug c++/25980] Unexpected name conflict between symbols
--- Comment #1 from pinskia at gcc dot gnu dot org 2006-01-26 18:49 --- I think you (and EDG) are missing what argument dependent lookup does for this case. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25980
[Bug c++/25980] New: Unexpected name conflict between symbols
The following fails to compile, it would appear that symbols from seperate
namespaces are colliding in error:
[EMAIL PROTECTED] tmp]$ cat test2.cpp
namespace ns1 {
class c {};
typedef int f;
static c x;
}
namespace ns2 {
void f(ns1::c& cc)
{
}
void g()
{
f(ns1::x);
}
}
[EMAIL PROTECTED] tmp]$ g++ -v -save-temps -Wall -c test2.cpp
Reading specs from /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.3.2/specs
Configured with: /usr/local/share/gcc-3.3.2/configure -march=pentium4 :
(reconfi
gured) /usr/local/share/gcc-3.3.2/configure -march=pentium4 : (reconfigured)
/us
r/local/share/gcc-3.3.2/configure --with-arch=pentium4
Thread model: posix
gcc version 3.3.2
/usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.3.2/cc1plus -E -D__GNUG__=3 -quiet
-
v -D__GNUC__=3 -D__GNUC_MINOR__=3 -D__GNUC_PATCHLEVEL__=2 -D_GNU_SOURCE
test2.cp
p -Wall test2.ii
ignoring nonexistent directory "NONE/include"
ignoring nonexistent directory "/usr/local/i686-pc-linux-gnu/include"
#include "..." search starts here:
#include <...> search starts here:
/usr/local/include/c++/3.3.2
/usr/local/include/c++/3.3.2/i686-pc-linux-gnu
/usr/local/include/c++/3.3.2/backward
/usr/local/include
/usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.3.2/include
/usr/include
End of search list.
/usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.3.2/cc1plus -fpreprocessed test2.ii
-quiet -dumpbase test2.cpp -auxbase test2 -Wall -version -o test2.s
GNU C++ version 3.3.2 (i686-pc-linux-gnu)
compiled by GNU C version 3.2.2 20030222 (Red Hat Linux 3.2.2-5).
GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
test2.cpp: In function `void ns2::g()':
test2.cpp:3: error: `ns1::f' is not a function,
test2.cpp:10: error: conflict with `void ns2::f(ns1::c&)'
test2.cpp:15: error: in call to `f'
[EMAIL PROTECTED] tmp]$
[EMAIL PROTECTED] tmp]$ cat test2.*i*
# 1 "test2.cpp"
# 1 ""
# 1 ""
# 1 "test2.cpp"
namespace ns1 {
class c {};
typedef int f;
static c x;
}
namespace ns2 {
void f(ns1::c& cc)
{
}
void g()
{
f(ns1::x);
}
}
[EMAIL PROTECTED] tmp]$
I don't see why symbols ns1::f and ns2::f are in conflict here, as when f() in
ns2::g is being resolved ns1::f should not be visible, should it?
If the parameter ns1::x is removed from the signature of ns2::f then there is
no error. It would appear that the ns1 namespace is being opened up as a
result of the parameter.
The online Comeau C/C++ compiler accepts this particular example.
--
Summary: Unexpected name conflict between symbols
Product: gcc
Version: 3.3.2
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c++
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: dwhorton at gmail dot com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25980
[Bug fortran/25964] [4.1/4.2 Regression] NIST regression on fm311.f
-- pinskia at gcc dot gnu dot org changed: What|Removed |Added Priority|P3 |P5 Target Milestone|4.2.0 |4.1.0 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25964
[Bug c++/25979] [4.0/4.1/4.2 Regression] incorrect codegen for conditional [SVO issue]
-- pinskia at gcc dot gnu dot org changed: What|Removed |Added Severity|normal |critical http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25979
[Bug tree-optimization/17064] -falias-noargument-global doesn't eliminate dead stores/loads
--- Comment #12 from pinskia at gcc dot gnu dot org 2006-01-26 18:15
---
This is the patch which helps the call clobering but does not fix it all the
way (because a different call clobbering issue):
Index: tree-ssa-alias.c
===
--- tree-ssa-alias.c(revision 110263)
+++ tree-ssa-alias.c(working copy)
@@ -328,7 +328,8 @@ set_initial_properties (struct alias_inf
}
else if (TREE_CODE (var) == PARM_DECL
&& default_def (var)
- && POINTER_TYPE_P (TREE_TYPE (var)))
+ && POINTER_TYPE_P (TREE_TYPE (var))
+ && flag_argument_noalias > 1)
{
tree def = default_def (var);
get_ptr_info (def)->value_escapes_p = 1;
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17064
[Bug bootstrap/25961] [4.2 Regression] Mainline failed to bootstrap on ia64
--- Comment #2 from sje at cup dot hp dot com 2006-01-26 18:00 --- See http://gcc.gnu.org/ml/gcc-patches/2006-01/msg01864.html http://gcc.gnu.org/ml/gcc-patches/2006-01/msg01779.html For the two patches needed to fix this. Waiting for approval. -- sje at cup dot hp dot com changed: What|Removed |Added CC||sje at cup dot hp dot com http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25961
[Bug tree-optimization/23619] Missed pre opportunity
--- Comment #12 from dberlin at gcc dot gnu dot org 2006-01-26 17:53 --- Fixed -- dberlin at gcc dot gnu dot org changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23619
[Bug tree-optimization/17064] -falias-noargument-global doesn't eliminate dead stores/loads
--- Comment #11 from pinskia at gcc dot gnu dot org 2006-01-26 17:51 --- The partial fix has been applied, I have another patch which fixes another part of this but still not fully. It just helps the call cloberring mechanism. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17064
[Bug tree-optimization/21470] -fargument-noalias-global has no effect for the vectorizer
--- Comment #8 from pinskia at gcc dot gnu dot org 2006-01-26 17:51 --- Subject: Bug 21470 Author: pinskia Date: Thu Jan 26 17:51:25 2006 New Revision: 110263 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110263 Log: 2006-01-26 Richard Guenther <[EMAIL PROTECTED]> Andrew Pinski <[EMAIL PROTECTED]> PR tree-opt/21470 partial PR tree-opt/17064 * tree-ssa-structalias.c (intra_create_variable_infos): Create heap variables for incoming parameters if flag_argument_noalias > 1. (find_what_p_points_to): Look through default defs of parameter decls. Modified: trunk/gcc/ChangeLog -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21470
[Bug tree-optimization/17064] -falias-noargument-global doesn't eliminate dead stores/loads
--- Comment #10 from pinskia at gcc dot gnu dot org 2006-01-26 17:51 --- Subject: Bug 17064 Author: pinskia Date: Thu Jan 26 17:51:25 2006 New Revision: 110263 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110263 Log: 2006-01-26 Richard Guenther <[EMAIL PROTECTED]> Andrew Pinski <[EMAIL PROTECTED]> PR tree-opt/21470 partial PR tree-opt/17064 * tree-ssa-structalias.c (intra_create_variable_infos): Create heap variables for incoming parameters if flag_argument_noalias > 1. (find_what_p_points_to): Look through default defs of parameter decls. Modified: trunk/gcc/ChangeLog -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17064
[Bug tree-optimization/21470] -fargument-noalias-global has no effect for the vectorizer
--- Comment #7 from pinskia at gcc dot gnu dot org 2006-01-26 17:50 --- Fixed applied. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution||FIXED Target Milestone|--- |4.2.0 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21470
[Bug middle-end/25977] [4.1/4.2 Regression] RSO vs NRV (twice)
--- Comment #5 from pinskia at gcc dot gnu dot org 2006-01-26 17:36 --- *** Bug 25979 has been marked as a duplicate of this bug. *** -- pinskia at gcc dot gnu dot org changed: What|Removed |Added CC||hhinnant at apple dot com http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25977
[Bug c++/25979] [4.0/4.1/4.2 Regression] incorrect codegen for conditional [SVO issue]
--- Comment #4 from pinskia at gcc dot gnu dot org 2006-01-26 17:43 --- Confirmed, very much related to PR 25977. Though I think this and PR 25977 are almost the same bug as we get: TARGET_EXPR >> We are losing some piece of information. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added BugsThisDependsOn||25977 Status|UNCONFIRMED |NEW Ever Confirmed|0 |1 Keywords||wrong-code Known to fail||4.0.3 4.1.0 4.2.0 Known to work||3.4.0 Last reconfirmed|-00-00 00:00:00 |2006-01-26 17:43:24 date|| Summary|incorrect codegen for |[4.0/4.1/4.2 Regression] |conditional |incorrect codegen for ||conditional [SVO issue] Target Milestone|--- |4.0.3 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25979
[Bug c++/25979] incorrect codegen for conditional
--- Comment #3 from pinskia at gcc dot gnu dot org 2006-01-26 17:39 --- Actually this is not a full dup as this one is also still broken on the 4.0 branch. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|RESOLVED|UNCONFIRMED Resolution|DUPLICATE | http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25979
[Bug c++/25979] incorrect codegen for conditional
--- Comment #2 from pinskia at gcc dot gnu dot org 2006-01-26 17:36 --- It was worked around in 4.0.2 (done on 2005-04-05 23:13:35) by: PR c++/19317 * calls.c (expand_call): Disable return slot optimization. Which just disabled the return slot optimization for 4.0.x Now 4.1.0 never got that patch but a different one which fixed a similar bug (PR 19317) but it did not fix this one or the duplicated one. *** This bug has been marked as a duplicate of 25977 *** -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution||DUPLICATE http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25979
[Bug c++/25979] incorrect codegen for conditional
--- Comment #1 from pinskia at gcc dot gnu dot org 2006-01-26 17:31 --- This is actually a dup of bug 25977. But I think it was worked around in 4.0.2 (or maybe just 4.0.3, I have to double check that). -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25979
[Bug c++/25979] New: incorrect codegen for conditional
I'm not positive whether or not this is a duplicate of 25895. I figured I'd
better enter it just in case it wasn't. Test case:
#include
struct A
{
A() : data1_(0), data2_(0) {}
A(int i, int j) : data1_(i), data2_(j) {}
A operator+(int);
friend A operator+(int, const A&);
~A() {}
//private:
int data1_;
int data2_;
};
extern bool x;
void display(const A& x)
{
printf("%d %d\n", x.data1_, x.data2_);
}
int main()
{
A a1(1,2);
a1 = (x ? a1 + 3 : 3 + a1);
display(a1);
}
bool x = false;
A
A::operator+(int i)
{
A a;
a = *this;
a.data2_ = i;
return a;
}
A
operator+(int i, const A& x)
{
A a;
a = x;
a.data1_ = i;
return a;
}
Output:
3 0
Expected output:
3 2
The gimple tree (-fdump-tree-gimple) is showing statements like:
operator+ (&a1, 3, &a1) [return slot addr];
which shoud instead be:
operator+ (temp, 3, &a1) [return slot addr];
...
a1 = temp;
The bad codegen is sensitive to the presence or absence of special member
functions. For example if you comment out ~A(), you get the expected output.
--
Summary: incorrect codegen for conditional
Product: gcc
Version: 4.0.1
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c++
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: hhinnant at apple dot com
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25979
[Bug c/25975] Problems with -ffast-math and isnan
--- Comment #4 from terra at gnome dot org 2006-01-26 17:22 --- ok, thanks. I'll prevent this by checking for __FAST_MATH__ in my code. One could argue for a warning like nan.c:13: warning: comparison is always false due to limited range of data type for use of isnan and similarly (with true) for finite. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25975
[Bug ada/25900] [4.2 Regression] ICE on ACATS cxac004 in Tree-VRP
--- Comment #10 from law at redhat dot com 2006-01-26 17:14 --- Fixed with today's patch to tree-vrp.c -- law at redhat dot com changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25900
[Bug ada/25900] [4.2 Regression] ICE on ACATS cxac004 in Tree-VRP
--- Comment #8 from law at redhat dot com 2006-01-26 17:14 --- Subject: Re: [4.2 Regression] ACATS ICE cxac0004 in set_value_range, at tree-vrp.c:161 on x86-linux On Tue, 2006-01-24 at 15:12 -0700, Jeffrey A Law wrote: > That "fixes" cxac004, but causes c99004a to fail! As Laurent noted, c99004a has been failing without rhyme or reason and the failure is not an indication of a regression due to my patch. So I'm going to go forward with my patch. The basic problem is we query SCEV to get ranges in some cases and we query it on a different type than we ultimately care about. And I've been unable to convince myself that even if we passed it the right type that SCEV would do the right thing. This patch makes VRP check the result of querying SCEV against the type we actually care about. When SCEV's result is inconsistent with the type we car about, we simply ignore SCEV's result. This patch also refines my recent change in one case. Specifically if we are intersecting a VR_RANGE and VR_ANTI_RANGE where the VR_ANTI_RANGE is completely contained within the VR_RANGE, we have the option of using either the VR_RANGE or VR_ANTI_RANGE as the result. We were selecting the VR_ANTI_RANGE, but in practice the VR_RANGE is likely going to be more precise. So I've changed the code to give us the VR_RANGE instead. About the only time the VR_ANTI_RANGE might be more useful would be if it was something like -[0, 0] and referred to a pointer type. Note that I'll be looking more closely at the VRP null-pointer check elimination code shortly so if the need arises I may add that final bit of refinement. Anyway, bootstrapped and regression tested on i686-pc-linux-gnu (ada included). Jeff --- Comment #9 from law at redhat dot com 2006-01-26 17:14 --- Created an attachment (id=10735) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=10735&action=view) -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25900
[Bug ada/25900] [4.2 Regression] ICE on ACATS cxac004 in Tree-VRP
--- Comment #7 from law at gcc dot gnu dot org 2006-01-26 17:14 --- Subject: Bug 25900 Author: law Date: Thu Jan 26 17:14:22 2006 New Revision: 110261 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=110261 Log: PR ada/25900 * tree-vrp.c (extract_range_from_assert): When merging a VR_RANGE with a VR_ANTI_RANGE and the VR_ANTI_RANGEis completely contained within the VR_RANGE, use the VR_RANGE as the result, not the VR_ANTI_RANGE. (adjust_range_with_scev): Reject ranges from SCEV which are out of bounds for the type. Modified: trunk/gcc/ChangeLog trunk/gcc/tree-vrp.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25900
[Bug c/19606] wrong code for arith.expr: (((unsigned int)(signed int) a ) / 2LL) with signed char a=-4
--- Comment #9 from kazu at gcc dot gnu dot org 2006-01-26 17:07 --- Posted a new patch. -- kazu at gcc dot gnu dot org changed: What|Removed |Added URL|http://gcc.gnu.org/ml/gcc- |http://gcc.gnu.org/ml/gcc- |patches/2006- |patches/2006- |01/msg00845.html|01/msg01853.html http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19606
[Bug c++/25927] Spurious offsetof warnings with private members
--- Comment #7 from rcbilson at plg dot uwaterloo dot ca 2006-01-26 16:59 --- (In reply to comment #6) > class xxx is NOT a POD. Indeed not. Thank you for explaining, and sorry to waste your time. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25927
[Bug fortran/18937] quadratic behaviour with many label "spaghetti" code
--- Comment #10 from tobi at gcc dot gnu dot org 2006-01-26 16:46 --- I don't know when I will have time for this, so I'm unassigning myself. -- tobi at gcc dot gnu dot org changed: What|Removed |Added AssignedTo|tobi at gcc dot gnu dot org |unassigned at gcc dot gnu ||dot org Status|ASSIGNED|NEW http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18937
[Bug libstdc++/25608] g++ miscompiles gcjx
--- Comment #17 from gdr at cs dot tamu dot edu 2006-01-26 16:45 --- Subject: Re: g++ miscompiles gcjx "bkoz at gcc dot gnu dot org" <[EMAIL PROTECTED]> writes: | ps tom this kind of organization for compiling looks dreamy: Yup! I dream of when other GCC parts would be similarly organized. (the source code is well structured, with separate concepts developed separately.) -- Gaby -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25608
[Bug rtl-optimization/24626] [4.1 Regression] internal compiler error: verify_flow_info failed
--- Comment #66 from yanov at il dot ibm dot com 2006-01-26 16:45 --- Bootstrapped and tested on PowerPC linux. No regressions. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24626
[Bug libstdc++/25608] g++ miscompiles gcjx
--- Comment #16 from gdr at cs dot tamu dot edu 2006-01-26 16:44 --- Subject: Re: g++ miscompiles gcjx "bkoz at gcc dot gnu dot org" <[EMAIL PROTECTED]> writes: | Gaby, here's the one thing that I can think of that changed between 3.4.x and | mainline/4.1 in this code path: Thank you very much for looking into this. I'll do that, once I get to office. -- Gaby -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25608
[Bug libstdc++/25608] g++ miscompiles gcjx
--- Comment #15 from bkoz at gcc dot gnu dot org 2006-01-26 16:35 --- ps tom this kind of organization for compiling looks dreamy: #20 0x400afc18 in compiler::do_analyze_unit (this=0x8055248, unit=0x8089dc0) at exception.hh:63 #21 0x400afebf in compiler::semantic_analysis (this=0x8055248) NICE -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25608
[Bug libstdc++/25608] g++ miscompiles gcjx
--- Comment #14 from bkoz at gcc dot gnu dot org 2006-01-26 16:32 --- Gaby, here's the one thing that I can think of that changed between 3.4.x and mainline/4.1 in this code path: 2005-10-10 Ian Lance Taylor PR libstdc++/13583 * include/bits/locale_classes.h (locale::_Impl::_M_install_cache): Move out of line. * src/locale.cc: Define here, add mutex. You might want to try reverting this and seeing if it has any impact. I kind of doubt it, but it's the only thing that really changed WRT ios/locale init. I'm surprised to see this issue. I would imagine that if this is a libstdc++ issue everybody and their dog would have found it by now. best, benjamin -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25608
[Bug libgomp/25978] New: All libgomp tests timeout on ppc-darwin
Because Darwin does not implement sem_init correctly all the tests timeout. One way of fixing this is to change libgomp for Darwin to use the mach semaphores. Another way is to use named semaphores. -- Summary: All libgomp tests timeout on ppc-darwin Product: gcc Version: 4.2.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: libgomp AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: pinskia at gcc dot gnu dot org GCC target triplet: powerpc-darwin http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25978
[Bug bootstrap/21936] Ada bootstrap failure on Solaris 10/x86 with native as
--- Comment #2 from pinskia at gcc dot gnu dot org 2006-01-26 15:58 --- No feedback in 3 months (T-4 days). -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|WAITING |RESOLVED Resolution||INVALID http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21936
[Bug bootstrap/21739] ../../.././libstdc++-v3/libmath/stubs.c:268: error: parse error before '(' token
--- Comment #2 from pinskia at gcc dot gnu dot org 2006-01-26 15:58 --- No feedback in 3 months (T-4 days). -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|WAITING |RESOLVED Resolution||INVALID http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21739
[Bug middle-end/25977] [4.1/4.2 Regression] RSO vs NRV (twice)
--- Comment #4 from pinskia at gcc dot gnu dot org 2006-01-26 15:54 --- Hmm: (void) ( = *(struct A &) (struct A *) &TARGET_EXPR , (struct A &) (struct A *) &r)>) >>> >>; Maybe it is not fully related to PR 19317 but a different issue really. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25977
