--- Comment #2 from ubizjak at gmail dot com 2009-02-16 08:12 ---
Works OK with latest trunk:
version 4.4.0 20090216 (experimental) [trunk revision 144197]
--
ubizjak at gmail dot com changed:
What|Removed |Added
--- Comment #5 from ubizjak at gmail dot com 2009-02-16 08:45 ---
*** Bug 39199 has been marked as a duplicate of this bug. ***
--
ubizjak at gmail dot com changed:
What|Removed |Added
--- Comment #3 from ubizjak at gmail dot com 2009-02-16 08:45 ---
This will be fixed by backport of [1] to 4.3 branch.
[1] http://gcc.gnu.org/ml/gcc-patches/2009-02/msg00059.html
*** This bug has been marked as a duplicate of 39058 ***
--
ubizjak at gmail dot com changed:
--- Comment #12 from Joey dot ye at intel dot com 2009-02-16 08:49 ---
Created an attachment (id=17305)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17305action=view)
New patch attached
Test finished. No regression with emx_avx_sim. Wait to checkin to 4.5
--
Joey dot ye at
--- Comment #38 from bonzini at gnu dot org 2009-02-16 09:01 ---
Outlandish statement: maybe no-condexec if conversion should be moved to the
tree-level?!? Doing this kind of hoisting at the same time as if conversion is
simpler on GIMPLE...
--
bonzini at gnu dot org changed:
--- Comment #18 from bonzini at gnu dot org 2009-02-16 09:04 ---
For whatever reason, we're now faster than GCC 3.3.
-O2 -fno-ivopts is still faster than -O2, will open an enhancement request for
that.
--
bonzini at gnu dot org changed:
What|Removed
Split off from PR21676, which is no longer a regression but where -O2
-fno-ivopts is still much faster than -O2.
For the testcase, see that PR.
--
Summary: ivopts slows down SciMark sparse matrix benchmark
Product: gcc
Version: 4.4.0
Status:
Metabug for ivopts issues
--
Summary: ivopts metabug
Product: gcc
Version: 4.4.0
Status: UNCONFIRMED
Keywords: missed-optimization, meta-bug
Severity: normal
Priority: P3
Component: tree-optimization
--- Comment #4 from mp2 at ladybridge dot com 2009-02-16 09:10 ---
(In reply to comment #3)
I would generate the assembly by using a local function name rather than
using main().
The original problem was deep in a complex program. I was trying to bring it
down to the shortest program
--- Comment #1 from bonzini at gnu dot org 2009-02-16 09:11 ---
I'm adding the regressions for now. Anyone can feel free to add more. :-)
--
bonzini at gnu dot org changed:
What|Removed |Added
--- Comment #5 from bonzini at gnu dot org 2009-02-16 09:12 ---
This looks target-dependent, isn't it?
--
bonzini at gnu dot org changed:
What|Removed |Added
--- Comment #27 from bonzini at gnu dot org 2009-02-16 09:14 ---
Added bugs corresponding to the patch fallout in case distros want to backport
it (it gave quite a nice boost and probably fixed PR21676 too)
--
bonzini at gnu dot org changed:
What|Removed
--- Comment #5 from bonzini at gnu dot org 2009-02-16 09:23 ---
Just came here by chance :-)
You can check if (this.class == ...) in the constructor. It will slow down
constructors for subclasses though.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22377
--- Comment #1 from bonzini at gnu dot org 2009-02-16 09:29 ---
No commit there for 3 years, can be closed.
--
bonzini at gnu dot org changed:
What|Removed |Added
--- Comment #2 from ubizjak at gmail dot com 2009-02-16 09:44 ---
I'm adding PR 34163 in the hope that ivopts pass will someday know how to
handle x[n] = f (x[n-1]) from ... very tight loops that calculate x[n] from
the value x[n-1], where x[n-1] is the result of a previous step.
This
--- Comment #3 from bonzini at gnu dot org 2009-02-16 09:56 ---
Subject: Re: [meta-bug] ivopts metabug
I'm adding PR 34163 in the hope that ivopts pass will someday know how to
handle x[n] = f (x[n-1]) from ... very tight loops that calculate x[n] from
the value x[n-1], where
--- Comment #13 from bonzini at gnu dot org 2009-02-16 10:23 ---
Predictive commoning does exactly what you want.
--
bonzini at gnu dot org changed:
What|Removed |Added
--- Comment #5 from rob1weld at aol dot com 2009-02-16 13:11 ---
Results for 4.4.0 20090215 (experimental) [trunk revision 144190] (GCC)
testsuite on x86_64-pc-solaris2.11
http://gcc.gnu.org/ml/gcc-testresults/2009-02/msg01526.html
Rob
--
--- Comment #11 from rob1weld at aol dot com 2009-02-16 13:12 ---
Results for 4.4.0 20090215 (experimental) [trunk revision 144190] (GCC)
testsuite on x86_64-pc-solaris2.11
http://gcc.gnu.org/ml/gcc-testresults/2009-02/msg01526.html
Rob
--
testcase produces
internal compiler error: in collapse_rest_of_var, at
tree-ssa-structalias.c:3296
Version: trunk rev. 144201.
Reproduce: gcc-4.4 -O2 -c tst.c
typedef struct
{
union
{
int * aaa;
} u;
} t_a;
typedef struct
{
unsigned bbb : 1;
} t_b;
typedef struct
{
int ccc;
There needs to be a convenient way to specify that at link time the whole
program
is seen, or rather, that -fwhole-program -flto during compile-time doesn't have
-fwhole-program effects.
Building SPEC int 2000 with -flto -fwhole-program otherwise (obviously) results
in undefined symbols, like
--- Comment #1 from dominiq at lps dot ens dot fr 2009-02-16 15:39 ---
Confirmed on i686-apple-darwin9 with '-m64 -O2', works with '-m32 -O2' or '-m64
-O1'. This is a regression as the code compiles with 4.2.3 or 4.3.3 and '-m64
-O2'.
--
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
Status|UNCONFIRMED |NEW
Ever Confirmed|0 |1
Last
The testcase I'm going to attach ICEs in compute_attic, with checking enabled
num_iterations reaches 500, without checking enabled it is OOM killed after it
eats lots of memory. The .i file is from delta, so ignore the warnings, they
weren't in the original source.
--
Summary: [4.4
--- Comment #1 from jakub at gcc dot gnu dot org 2009-02-16 16:38 ---
Created an attachment (id=17306)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17306action=view)
rh485708.i
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39204
--- Comment #2 from jakub at gcc dot gnu dot org 2009-02-16 16:40 ---
./cc1 -O2 -m32 rh485708.i triggers this in 32-bit HWI build (but both in stage1
built by 4.3.x and later stages), in x86_64 gcc it isn't reproduceable with -O2
-m32.
--
--
jakub at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.4.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39204
--- Comment #2 from hjl dot tools at gmail dot com 2009-02-16 16:49 ---
This is caused by revision 137530:
http://gcc.gnu.org/ml/gcc-cvs/2008-07/msg00237.html
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
--- Comment #10 from hp at gcc dot gnu dot org 2009-02-16 17:03 ---
(In reply to comment #9)
The assertion at that revision as showed by libstdc++.log
Same at r144193; same log messages.
Result report at http://gcc.gnu.org/ml/gcc-testresults/2009-02/msg01541.html.
--
I can't think of a scenario where one would want to write x.f() over X::f()
when f() is static. I'd like a warning for this so I can catch with -Werror.
% g++ --version
g++ (GCC) 3.4.5 20051201 (Red Hat 3.4.5-2)
Copyright (C) 2004 Free Software Foundation, Inc.
This is free software; see the
(cross)gcc fails to create precompiled header-file, because for some
reason (v.pch_init != pch_init)
first time this failure occures when libstdc++/configure
tests whether to generate pch/gch-files or not.
Error message is:
conftest.cc:1:22: error: conftest.h.gch: had text segment at different
--- Comment #2 from jsm28 at gcc dot gnu dot org 2009-02-16 18:11 ---
Testing a patch.
--
jsm28 at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #3 from hjl dot tools at gmail dot com 2009-02-16 18:57 ---
This is caused by revision 142617:
http://gcc.gnu.org/ml/gcc-cvs/2008-12/msg00315.html
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
--- Comment #2 from hjl at gcc dot gnu dot org 2009-02-16 19:00 ---
Subject: Bug 37049
Author: hjl
Date: Mon Feb 16 19:00:39 2009
New Revision: 144207
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=144207
Log:
Mention PR target/37049 in ChangeLog entries.
Modified:
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
Target Milestone|--- |4.4.0
Version|unknown |4.4.0
--- Comment #4 from jakub at gcc dot gnu dot org 2009-02-16 19:12 ---
This testcase is so sensitive to any uid changes (just removing unused u_long
typedef makes the bug go away), so I think doing a binary search is useless
here.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39204
#include string
#include map
const std::string usage ();
void
foo ()
{
std::map std::string, int best_scores;
const std::map std::string, int ::const_iterator current_best =
best_scores.find (usage ());
if (0 current_best-second) best_scores[usage ()] = 0;
}
yields:
LC_ALL=C g++ -S
--- Comment #1 from jakub at gcc dot gnu dot org 2009-02-16 19:30 ---
Created an attachment (id=17307)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17307action=view)
ai.C
LC_ALL=C ./cc1plus -O2 -Wall -quiet /tmp/ai.C
/tmp/ai.C: In function 'void foo()':
/tmp/ai.C:2615: warning:
--
jakub at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.4.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39207
--- Comment #5 from hjl dot tools at gmail dot com 2009-02-16 19:33 ---
Do we get overflow with 32-bit HWI in this testcase?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39204
--- Comment #3 from ubizjak at gmail dot com 2009-02-16 20:20 ---
(In reply to comment #1)
A patch is posted at
http://gcc.gnu.org/ml/gcc-patches/2008-08/msg00204.html
Although not a regression I think this patch is safe for mainline.
BTW: _If_ it happens that we segfault using
--- Comment #4 from hjl at gcc dot gnu dot org 2009-02-16 20:38 ---
Subject: Bug 37049
Author: hjl
Date: Mon Feb 16 20:37:49 2009
New Revision: 144209
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=144209
Log:
gcc/
2009-02-16 H.J. Lu hongjiu...@intel.com
PR
--- Comment #2 from jakub at gcc dot gnu dot org 2009-02-16 21:05 ---
I believe all 3 are the same issue.
static const_reference _S_value (_Const_Base_ptr __x)
{
return static_cast _Const_Link_type (__x)-_M_value_field;
}
casts pointer to _Rb_tree_node_base to
--- Comment #3 from paolo dot carlini at oracle dot com 2009-02-16 21:53
---
We have been through this already and I'm really surprised the spurious
warnings are still there. See middle-end/38477 and middle-end/38937.
--
paolo dot carlini at oracle dot com changed:
What
Given the following code:
templateint (p)(int, int) struct X { };
int f(int, int) { }
void g(Xf) { }
GCC produces the mangled name:
_Z1g1XILZ1fiiEE
for the function 'g'. However, the correct mangled name is:
_Z1g1XIL_Z1fiiEE
GCC is missing the '_' prior to the 'Z' in the mangled
Yes this known I forgot the bug number but try with -fabi-version=0.
Sent from my iPhone
On Feb 16, 2009, at 2:07 PM, dgregor at gcc dot gnu dot org gcc-bugzi...@gcc.gnu.org
wrote:
Given the following code:
templateint (p)(int, int) struct X { };
int f(int, int) { }
void g(Xf) { }
GCC
--- Comment #20 from jakub at gcc dot gnu dot org 2009-02-16 22:41 ---
*** Bug 39208 has been marked as a duplicate of this bug. ***
--
jakub at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #4 from jakub at gcc dot gnu dot org 2009-02-16 22:41 ---
to close as dup of PR16240.
*** This bug has been marked as a duplicate of 16240 ***
--
jakub at gcc dot gnu dot org changed:
What|Removed |Added
(Please note that in all the following code no standard libaries are linked -
this is for a bare-metal kernel project.)
The following code (full header/assembly attached below):
Iterator operator ++ ()
{
m_Node = (m_Node-*FunctionNext)();
return *this;
}
When
--- Comment #5 from hjl dot tools at gmail dot com 2009-02-16 22:49 ---
Fixed.
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
http://cimg.sourceforge.net/greycstoration/
g++ -I.. -I./plugins -o greycstoration4gimp greycstoration4gimp.cpp
`gimptool-2.0 --cflags` `gimptool-2.0 --libs` -lpthread -O3 -fno-math-errno
-ftree-vectorize -momit-leaf-frame-pointer -funroll-loops -march=native
-mtune=native
takes several hours CPU
--- Comment #3 from jsm28 at gcc dot gnu dot org 2009-02-16 22:56 ---
Subject: Bug 35446
Author: jsm28
Date: Mon Feb 16 22:55:56 2009
New Revision: 144216
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=144216
Log:
PR c/35446
* c-parser.c (c_parser_braced_init):
--- Comment #1 from steven at gcc dot gnu dot org 2009-02-16 23:13 ---
Hours is excessive alright.
Unfortunately, there are a couple of things missing to reproduce your problem.
There has to be at least a test case (preprocessed source), and the output of
gcc when you compile the test
--- Comment #2 from kurt at garloff dot de 2009-02-16 23:34 ---
Created an attachment (id=17308)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17308action=view)
greycstoration.ii.bz2
Preprocessed file, as desired. (bzip2ed, as it's too large otherwise.)
--
--- Comment #1 from hjl dot tools at gmail dot com 2009-02-17 02:28 ---
Please provide a reproducible testcase. It could be related to
PR 36159.
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
--- Comment #4 from mmitchel at gcc dot gnu dot org 2009-02-17 02:53
---
The key question is whether the bogus warning could indicate a potential for
miscompilation. If it's just a bogus warning, then it's going to annoy and
confuse people, but not result in too much damage. But, if
--
mmitchel at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P1
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39202
--
mmitchel at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P1
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39204
//The following code works and there is no way it should. Seems like a bug
someone put in on purpose
#include iostream
using namespace std;
void swap(int *x, int *y)
{
int temp;
temp = *x;
*x = *y;
*y = temp;
}
int main()
{
int x = 10;
int y =
The following code works and it shouldn't. It only works with passing 2
parameters from what I tested. If I try swap(int *x) or swap(int *x, int *y,
int *z), it gives an error as expected. But with 2 parameters it compiles and
runs.
void swap(int *x, int *y)
{
int temp;
temp
This bug is seen using the avr-gcc that gets built by the script that comes
with FemtoOS 0.88.
reg...@john-home:~/volatile/tmp140$ avr-gcc -O1 small.c
small.c: In function 'uint32func':
small.c:22: error: unable to find a register to spill in class 'POINTER_REGS'
small.c:22: error: this is the
--- Comment #1 from eakrohn at cs dot uiowa dot edu 2009-02-17 03:53
---
My concern is this function
swap(int *x, int *y){ ... }
with this function call
int x = 5, y = 7;
swap(x,y);
compiles and works just fine. But this function
swap(int *x){ ... }
with this function call
int
--- Comment #2 from eakrohn at cs dot uiowa dot edu 2009-02-17 04:00
---
Problem fixed, didn't realize there was a built in swap that was getting
called. My mistake.
--
eakrohn at cs dot uiowa dot edu changed:
What|Removed |Added
--enable-checking=release --with-tune=k8 --with-cpu=k8 --with-arch=k8
--with-gnu-as --with-as=/usr/local/bin/as --without-gnu-ld
--with-ld=/usr/bin/amd64/ld --with-gmp=/usr/local --with-mpfr=/usr/local
--without-ppl
Thread model: posix
gcc version 4.4.0 20090216 (experimental) [trunk revision 144203
--- Comment #1 from eric dot weddington at atmel dot com 2009-02-17 04:43
---
Fails on 4.3.2 with -O1, successful with -O[023s].
--
eric dot weddington at atmel dot com changed:
What|Removed |Added
--- Comment #8 from terry at chem dot gu dot se 2009-02-17 04:51 ---
It's not as frivolous as it sounds. ;-)
If it is indeed illegal not to obey an explicit interface (and though I can't
see anywhere in the standard that says this, the standard is pretty stupid if
it's not) then this
--- Comment #2 from eric dot weddington at atmel dot com 2009-02-17 05:02
---
Success on 4.3.2 with -O1 -finline-small-functions.
Fails on 4.3.2 with -O1 -fno-inline-small-functions.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39212
--- Comment #3 from eric dot weddington at atmel dot com 2009-02-17 05:09
---
Fails on 4.3.2 with -O[123s] -fno-inline-small-functions.
Success on 4.3.2 with -O0 -fno-inline-small-functions.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39212
--with-gmp=/usr/local --with-mpfr=/usr/local
--without-ppl
Thread model: posix
gcc version 4.4.0 20090216 (experimental) [trunk revision 144203] (GCC)
Rob
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39161
--- Comment #4 from steven at gcc dot gnu dot org 2009-02-17 06:31 ---
Investigating...
--
steven at gcc dot gnu dot org changed:
What|Removed |Added
70 matches
Mail list logo
