The following valid code snippet triggers an ICE since GCC 4.4.0:
===
templatetypename T struct A
{
operator T();
A() { T (A::*f)() = A::operator T; }
};
Aint a;
===
bug.cc: In constructor 'AT::A() [with
--
reichelt at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.4.3
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41994
--- Comment #8 from rootkit85 at yahoo dot it 2009-11-09 08:55 ---
1) define real geode
2) what CPU do I have?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41989
--- Comment #9 from rootkit85 at yahoo dot it 2009-11-09 09:01 ---
Read here pag. 15:
http://www.amd.com/files/connectivitysolutions/geode/geode_lx/33234F_LX_databook.pdf
The instruction set supported by the core is a combination
of Intel Pentium® processor, AMD Athlon processor, and
--- Comment #10 from ubizjak at gmail dot com 2009-11-09 09:34 ---
(In reply to comment #8)
1) define real geode
2) what CPU do I have?
http://en.wikipedia.org/wiki/Geode_%28processor%29#AMD_Geode
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41989
--- Comment #5 from rguenth at gcc dot gnu dot org 2009-11-09 10:09 ---
It might be a latent issue even with MPC installed. Kaveh should investigate.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #1 from rguenth at gcc dot gnu dot org 2009-11-09 10:16 ---
Which is because the gimplifier thinks this is an error for the frontend to
pass through:
Breakpoint 1, fancy_abort (
file=0x10b1ab8 /space/rguenther/src/svn/trunk/gcc/gimplify.c, line=7173,
--- Comment #1 from rguenth at gcc dot gnu dot org 2009-11-09 10:18 ---
Confirmed.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #11 from ubizjak at gmail dot com 2009-11-09 10:28 ---
OK, according to your benchmarks and documentation qoute, it looks we have to
split geode-lx out of generic geode option.
So, can you confirm, that the difference between generic geode and geode-lx is
presence of CMOV
--- Comment #12 from ubizjak at gmail dot com 2009-11-09 10:28 ---
Reopened to clear this geode mess.
--
ubizjak at gmail dot com changed:
What|Removed |Added
--- Comment #13 from ubizjak at gmail dot com 2009-11-09 10:34 ---
Created an attachment (id=18997)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18997action=view)
Patch that introduces geode-lx CPU option
Can you patch the compiler with attached patch?
gcc -march=native -###
--- Comment #2 from jakub at gcc dot gnu dot org 2009-11-09 10:45 ---
Caused by http://gcc.gnu.org/viewcvs?root=gccview=revrev=144618
--
jakub at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #14 from ubizjak at gmail dot com 2009-11-09 10:47 ---
(In reply to comment #11)
There is also geode NX, IIRC it represents itself as Athlon, but someone
should
confirm this.
According to [1], cpuid for Geode NX returns AMD Geode NX 1750, and this
string doesn't
--- Comment #33 from davek at gcc dot gnu dot org 2009-11-09 11:05 ---
This has been working fine for some time now, so closing. Verified by building
r154011: no debuginfo problems.
--
davek at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #9 from davek at gcc dot gnu dot org 2009-11-09 11:09 ---
Inactive for over a year, and was most likely a system or environment problem
rather than a bug in gcc itself, so closing. HEAD currently builds fine.
--
davek at gcc dot gnu dot org changed:
What
I know this is nitpicky, but I haven't found references to it (or the whys) in
the database.
I am not a lawyer, but I think that according to [0][1], the following program
should not compile:
#include iostream
template typename T
void f (T t) {
g (t);
}
int main () {
fint (42);
} //
--- Comment #15 from ubizjak at gmail dot com 2009-11-09 11:48 ---
(In reply to comment #7)
-march=geode disables cmov because the real geode does not have cmov :).
No, all geodes have cmov.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41989
--- Comment #1 from redi at gcc dot gnu dot org 2009-11-09 12:01 ---
probably related to PR 23885 and/or PR 16635
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41995
--- Comment #2 from rguenth at gcc dot gnu dot org 2009-11-09 12:18 ---
EDG accepts it in strict mode though I also think it's invalid.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41995
--- Comment #9 from rguenth at gcc dot gnu dot org 2009-11-09 13:02 ---
Hm, a build issue. Let's make this P2 for now.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P2
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41619
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Status|WAITING |NEW
Ever Confirmed|0 |1
Last
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Status|UNCONFIRMED |NEW
Ever Confirmed|0 |1
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P2
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41883
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Keywords||ice-on-valid-code
Priority|P3 |P2
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P2
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41888
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P2
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41889
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P1
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41905
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Keywords|ice-on-valid-code |ice-on-invalid-code
Priority|P3
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P1
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41926
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P4
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41930
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P2
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41957
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P2
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41963
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P2
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41994
--- Comment #2 from joseph at codesourcery dot com 2009-11-09 13:16 ---
Subject: Re: ICE on invalid dereferencing of void *
On Mon, 9 Nov 2009, rguenth at gcc dot gnu dot org wrote:
the C standard doesn't claim dereferencing a void pointer is invalid, so
the gimplifier should deal
--- Comment #16 from rootkit85 at yahoo dot it 2009-11-09 13:17 ---
Yes, it seems that even old Geode has such instructions:
# cat /proc/cpuinfo
processor : 0
vendor_id : Geode by NSC
cpu family : 5
model : 9
model name : Unknown
stepping: 1
cpu
--- Comment #3 from rguenth at gcc dot gnu dot org 2009-11-09 13:44 ---
Neither testcase crashes for me anymore (x86-64, with and without -m32).
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #4 from shailen dot n dot jain at gmail dot com 2009-11-09
13:47 ---
(In reply to comment #1)
Created an attachment (id=15274)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=15274action=view) [edit]
Add AIX 64bit support
I was trying to compile Mozilla Firefox
--- Comment #5 from rguenth at gcc dot gnu dot org 2009-11-09 13:48 ---
Hm.
(gdb) call debug_generic_expr (chrec)
{{1, +, {2, +, 1}_2}_1, +, 1}_2
how can this happen? Why does evolution_function_is_affine_multivariate_p
not reject this chrec?
--
rguenth at gcc dot gnu dot org
--- Comment #11 from rguenth at gcc dot gnu dot org 2009-11-09 13:49
---
Does this still happen with current trunk, esp. after the EH rewrite?
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #6 from rguenth at gcc dot gnu dot org 2009-11-09 13:51 ---
Seems to be fixed with current trunk. Please open if it is not.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #6 from rguenth at gcc dot gnu dot org 2009-11-09 13:59 ---
*** Bug 41957 has been marked as a duplicate of this bug. ***
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #4 from rguenth at gcc dot gnu dot org 2009-11-09 13:59 ---
Indeed. Same symptoms at least:
(gdb) call debug_generic_expr (chrec)
{0, +, {2, +, 1}_4}_4
This one fails for me with -m32 only though.
The chrec isn't quite {2, +, 1}_4 - I think the IV evolves as 0, 3, 4, 5,
--
jason at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |jason at gcc dot gnu dot org
|dot org
--- Comment #17 from ubizjak at gmail dot com 2009-11-09 15:10 ---
(In reply to comment #16)
Yes, it seems that even old Geode has such instructions:
So, I guess they should be listed under config X86_CMOV in
linux-2.6/arch/x86/Kconfig.cpu.
--
--- Comment #7 from reichelt at gcc dot gnu dot org 2009-11-09 15:15
---
Unfortunately the problem still persists, see following testcase
which crashes in the same fashion when compiled with -O3 -m32
on x86_64-unknown-linux-gnu.
--
reichelt at gcc dot gnu dot org changed:
--- Comment #8 from reichelt at gcc dot gnu dot org 2009-11-09 15:16
---
Created an attachment (id=18998)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18998action=view)
Self-contained testcase that crashes with -O3 -m32
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41290
--- Comment #18 from rootkit85 at yahoo dot it 2009-11-09 15:26 ---
As I did here?
http://patchwork.kernel.org/patch/51410/
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41989
--
jason at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |jason at gcc dot gnu dot org
|dot org
--- Comment #1 from jakub at gcc dot gnu dot org 2009-11-09 15:47 ---
This got fixed on the trunk with PR41317 and on the 4.4 branch also works well
with the backport of that patch (which I'm using for roughly a month on
redhat/gcc-4_4-branch because it cured some __builtin_object_size
--- Comment #2 from rguenther at suse dot de 2009-11-09 15:51 ---
Subject: Re: [4.3/4.4/4.5 Regression] Spurious array
subscript is above bounds warning
On Mon, 9 Nov 2009, jakub at gcc dot gnu dot org wrote:
--- Comment #1 from jakub at gcc dot gnu dot org 2009-11-09 15:47
--- Comment #10 from rguenth at gcc dot gnu dot org 2009-11-09 15:55
---
Confirmed. It passes if I add -fno-ipa-cp, maybe martin can have a look.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #3 from pinskia at gcc dot gnu dot org 2009-11-09 16:00 ---
Actually this is a combition of two issues really. First is argument dependent
lookup for fundamental types, do they have an assoicated namespaces (I think
there is a bug report for that)?
And then the other bug
--- Comment #2 from ebotcazou at gcc dot gnu dot org 2009-11-09 17:11
---
I see that it fails on HP-UX as well. That's probably because there is
something missing in the fallback routines in config/pa, namely:
fs-signal_frame = 1;
just before
return _URC_NO_REASON;
--
Current mainline as of 20091106 fails to compile on IRIX 6.5:
/vol/gcc/src/gcc-dist/gcc/lto/lto-elf.c: In function 'lto_elf_file_open':
/vol/gcc/src/gcc-dist/gcc/lto/lto-elf.c:561:28: error: expected ')' before
'PRId64'
cc1: warnings being treated as errors
--- Comment #2 from jason at gcc dot gnu dot org 2009-11-09 17:18 ---
This is happening because in C++ mode the body of bar looks like
D.1729 = foo ();
D.1728 = D.1729;
whereas in the C front end it looks like
D.1986 = foo ();
and so gimple_seq_may_fallthru says that the body can
--- Comment #7 from navinkumar+bugs at gmail dot com 2009-11-09 17:35
---
(From update of attachment 18986)
#include iostream
#include stdint.h
#include tuple
#include type_traits
struct empty_t {
template typename Arg
explicit empty_t(Arg arg) { }
};
template int N,
--- Comment #3 from jakub at gcc dot gnu dot org 2009-11-09 17:36 ---
/* { dg-do compile } */
/* { dg-options -march=armv5te -g -O2 } */
extern int a;
extern char b;
extern int foo (void);
void
test (void)
{
int c;
b = foo () ? '~' : '\0';
while ((c = foo ()))
if (c == '7')
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
CC||pinskia at gcc dot gnu dot
|
--- Comment #8 from navinkumar+bugs at gmail dot com 2009-11-09 17:39
---
Created an attachment (id=19000)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19000action=view)
demonstrates scenarios when zero-size base class optimization fails, compile
with flag: -std=c++0x
--
--- Comment #9 from paolo dot carlini at oracle dot com 2009-11-09 17:51
---
For the record, the Intel compiler, also implementing the IA64 C++ ABI, behaves
exactly the same as GCC.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39981
--- Comment #6 from ghazi at gcc dot gnu dot org 2009-11-09 17:56 ---
The bug is not latent and also it is not in GCC code related MPC. It lies in
the tree-based fallback code I wrote to handle complex division when MPC is not
used. This code will still persist even after the switch
--- Comment #10 from navinkumar+bugs at gmail dot com 2009-11-09 18:08
---
After review, I agree it is not a bug.
Because _0_emptyB and _2_emptyB both inherent from empty_t, the compiler is
obligated to ensure that
static_castempty_t*(static_cast_0_emptyB*(compositeB2)) !=
--- Comment #19 from ubizjak at gmail dot com 2009-11-09 18:16 ---
(In reply to comment #18)
As I did here?
http://patchwork.kernel.org/patch/51410/
Yes, but I don't know if perhaps -march=k6-2 should be used as a fallback, as
suggested in [1].
BTW: You can change optimization
--- Comment #3 from jason at gcc dot gnu dot org 2009-11-09 18:32 ---
Subject: Bug 41994
Author: jason
Date: Mon Nov 9 18:32:44 2009
New Revision: 154041
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=154041
Log:
PR c++/41994
* pt.c (tsubst_baselink): tsubst the
--- Comment #4 from jason at gcc dot gnu dot org 2009-11-09 18:33 ---
Subject: Bug 41972
Author: jason
Date: Mon Nov 9 18:32:53 2009
New Revision: 154042
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=154042
Log:
PR c++/41972
* parser.c
--- Comment #4 from jakub at gcc dot gnu dot org 2009-11-09 18:42 ---
For:
(insn 78 77 16 2 rh528639.i:12 (cond_exec (eq (reg:CC 24 cc)
(const_int 0 [0x0]))
(set (reg:SI 0 r0 [orig:133 iftmp.0 ] [133])
(reg:SI 4 r4))) 2367 {neon_vornv2di+77}
--- Comment #3 from jakub at gcc dot gnu dot org 2009-11-09 19:38 ---
Subject: Bug 40946
Author: jakub
Date: Mon Nov 9 19:38:29 2009
New Revision: 154046
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=154046
Log:
PR middle-end/40946
Backport from mainline
--- Comment #5 from jakub at gcc dot gnu dot org 2009-11-09 19:38 ---
Subject: Bug 41317
Author: jakub
Date: Mon Nov 9 19:38:29 2009
New Revision: 154046
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=154046
Log:
PR middle-end/40946
Backport from mainline
--- Comment #4 from jakub at gcc dot gnu dot org 2009-11-09 19:41 ---
Subject: Bug 40946
Author: jakub
Date: Mon Nov 9 19:40:56 2009
New Revision: 154047
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=154047
Log:
PR middle-end/40946
* gcc.dg/pr40946.c: New test.
--- Comment #5 from jakub at gcc dot gnu dot org 2009-11-09 19:44 ---
Fixed on the trunk and 4.4 branch.
--
jakub at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #9 from pinskia at gcc dot gnu dot org 2009-11-09 19:45 ---
Fixed so closing.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #20 from rootkit85 at yahoo dot it 2009-11-09 20:25 ---
Yes K6 is the best fallback for geode-lx, while pentium-mmx is the best one for
geode.
I need to know if this new -march argument will be added so I edit the kernel
patch.
--
--- Comment #4 from jason at gcc dot gnu dot org 2009-11-09 20:28 ---
Subject: Bug 41994
Author: jason
Date: Mon Nov 9 20:28:11 2009
New Revision: 154049
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=154049
Log:
PR c++/41994
* pt.c (tsubst_baselink): tsubst the
--- Comment #5 from jason at gcc dot gnu dot org 2009-11-09 20:28 ---
Subject: Bug 41972
Author: jason
Date: Mon Nov 9 20:28:18 2009
New Revision: 154050
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=154050
Log:
PR c++/41972
* parser.c
--- Comment #15 from reichelt at gcc dot gnu dot org 2009-11-09 21:42
---
The bug is not a regression, because the code snippet is invalid and was never
correctly rejected since the introduction of __builtin_offsetof.
Since it's fixed now we can close it.
--
reichelt at gcc dot gnu
--- Comment #5 from reichelt at gcc dot gnu dot org 2009-11-09 21:43
---
Fixed for GCC 4.4.3.
Thanks, Jason!
--
reichelt at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #27 from reichelt at gcc dot gnu dot org 2009-11-09 21:54
---
After the reversion of the patch for on the 4.3 branch, it is only fixed for
GCC 4.4.3 and later. But let's keep it closed since this is just an
ice-on-invalid-code bug.
--
reichelt at gcc dot gnu dot org
--- Comment #21 from rootkit85 at yahoo dot it 2009-11-09 21:59 ---
Created an attachment (id=19001)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=19001action=view)
A patch which adds Geode LX support to GCC
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41989
--- Comment #4 from reichelt at gcc dot gnu dot org 2009-11-09 21:59
---
The second testcase stopped crashing in August or September. Because the first
testcase stopped crashing although the bug wasn't really fixed, I thought it
would maybe pop up again. But apparently it didn't. So
--- Comment #6 from jason at gcc dot gnu dot org 2009-11-09 22:01 ---
Subject: Bug 41972
Author: jason
Date: Mon Nov 9 22:01:09 2009
New Revision: 154053
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=154053
Log:
PR c++/41972
* parser.c
--- Comment #7 from jason at gcc dot gnu dot org 2009-11-09 22:02 ---
Fixed.
--
jason at gcc dot gnu dot org changed:
What|Removed |Added
Status|ASSIGNED
--- Comment #12 from reichelt at gcc dot gnu dot org 2009-11-09 22:13
---
The testcase from comment #7 doesn't crash since GCC 4.0.0.
The testcase from comment #4 stopped crashing in August or September,
so let's close the bug for good.
--
reichelt at gcc dot gnu dot org changed:
--- Comment #5 from jason at gcc dot gnu dot org 2009-11-09 22:43 ---
Yeah, there's no way to get this right in general with our unconditional
cloning.
--
jason at gcc dot gnu dot org changed:
What|Removed |Added
In C++0x, constant initializer_lists do not seem visible to the optimiser,
leading to sub-optimal code. Consider as a simple example:
#include initializer_list
int max_val(std::initializer_listint il)
{
int i = *(il.begin());
int j = *(il.begin() + 1);
return (i j ? i :
--- Comment #1 from paolo dot carlini at oracle dot com 2009-11-09 23:47
---
Let's CC Jason about this...
--
paolo dot carlini at oracle dot com changed:
What|Removed |Added
--- Comment #4 from lucier at math dot purdue dot edu 2009-11-10 00:28
---
This is fixed, at least by the time of
gcc version 4.5.0 20091109 (experimental) [trunk revision 154037] (GCC)
--
lucier at math dot purdue dot edu changed:
What|Removed
--
jason at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |jason at gcc dot gnu dot org
|dot org
--
jason at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |jason at gcc dot gnu dot org
|dot org
--- Comment #6 from aaron at aarongraham dot com 2009-11-10 04:38 ---
So it appears that the problem is gthreads. The monotonic_clock support is
purely superficial in gcc until gthreads supports such a concept. Developers
will need to create their own clock and modify the standard
--- Comment #12 from oliver dot kellogg at eads dot com 2009-11-10 05:38
---
Does this still happen with current trunk, esp. after the EH rewrite?
Does not happen with trunk r154034.
Thanks.
--
oliver dot kellogg at eads dot com changed:
What|Removed
--- Comment #7 from ghazi at gcc dot gnu dot org 2009-11-10 05:56 ---
Patch posted here:
http://gcc.gnu.org/ml/gcc-patches/2009-11/msg00488.html
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41987
--- Comment #22 from ubizjak at gmail dot com 2009-11-10 07:45 ---
(In reply to comment #20)
Yes K6 is the best fallback for geode-lx, while pentium-mmx is the best one
for
geode.
BTW, recommended fallback is K6-2.
I need to know if this new -march argument will be added so I
93 matches
Mail list logo
