--- Comment #2 from pinskia at gcc dot gnu dot org 2009-10-01 06:55 ---
x86_64 and x86 has issues with pre-RA scheduling pass because of register
requirements of the target.
*** This bug has been marked as a duplicate of 24319 ***
--
pinskia at gcc dot gnu dot org changed:
--- Comment #24 from pinskia at gcc dot gnu dot org 2009-10-01 06:55
---
*** Bug 41531 has been marked as a duplicate of this bug. ***
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
-
--- Comment #1 from t7 at gmail dot com 2009-10-01 06:50 ---
Created an attachment (id=18684)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18684&action=view)
Preprocessed test case.
gcc version 4.5.0 20090928 Revision 152257
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id
x86_64-w64-mingw32-gcc -DHAVE_AV_CONFIG_H -I.. -D_ISOC99_SOURCE
-D_POSIX_C_SOURCE=200112 -O1 -fschedule-insns -MMD -MF swscale.d -MT swscale.o
-save-temps -c -o swscale.o swscale.c
Error on both 4_4-branch and trunk
The error message of 4_4-branch is different:
swscale.c: In function 'sws_scale':
--- Comment #8 from hutchinsonandy at gcc dot gnu dot org 2009-10-01 01:49
---
Created an attachment (id=18681)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18681&action=view)
Fix
After lengthy and productive correspondence with Michael Matz, we have arrived
at a conclusion.
Th
--- Comment #9 from jvdelisle at gcc dot gnu dot org 2009-10-01 01:37
---
Rounding modes are now implemented for formatted WRITE operations. I do not
have a clear enough idea of what the rounding modes really mean for READ
operations. For example:
>From Comment #4
"Rounding is done
--- Comment #15 from mrs at apple dot com 2009-10-01 00:54 ---
Yeah, the patch in #11 is about the right for half the problem (darwin10), if
it weren't for ld's warning message. I don't know quite why it is doing that,
so guess we'll just have to nix that idea. So, the next best thing
--- Comment #14 from howarth at nitro dot med dot uc dot edu 2009-10-01
00:29 ---
Jakub,
If you could propose a patch to the darwin_emit_unwind_label subroutine in
gcc/config/darwin.c to implement the changes you propose in Comment 12, I would
be happy to test it on darwin9/darwin10.
--- Comment #13 from howarth at nitro dot med dot uc dot edu 2009-10-01
00:18 ---
Actually, the patch in Comment 11 leaves two of the original failures...
FAIL: g++.dg/tree-prof/partition1.C compilation, -g -fprofile-use
FAIL: g++.dg/tree-prof/partition1.C compilation, -O3 -g -fpro
--- Comment #1 from paolo dot carlini at oracle dot com 2009-09-30 23:44
---
Thanks for the PR and the patch, which indeed makes sense to me (also regtests
fine). Before committing the change, let's wait a bit in case Chris F has some
comments...
--
paolo dot carlini at oracle dot c
--- Comment #54 from howarth at nitro dot med dot uc dot edu 2009-09-30
23:38 ---
The new libgcc_s.dylib appears to be only of the native target architecture...
file libgcc_s.dylib
libgcc_s.dylib: Mach-O 64-bit dynamically linked shared library x86_64
file libgcc_s.1.dylib
libgcc_s.1.
--- Comment #53 from howarth at nitro dot med dot uc dot edu 2009-09-30
23:36 ---
Iain,
We seem to be producing an extra libgcc shared library with the new patch.
In darwin_objdir/stage1-x86_64-apple-darwin10.0.0/libgcc, I see...
libgcc_ext.10.4.dylib libgcc_ext.10.5.dylib libg
--- Comment #12 from jakub at gcc dot gnu dot org 2009-09-30 22:46 ---
It is unclear what those labels are good for, but if darwin is to support
hot/cold partitioning and FDEs covering it, the emit unwind_label hook (which
is apparently darwin specific only) should be passed also the SEC
--- Comment #7 from danglin at gcc dot gnu dot org 2009-09-30 22:46 ---
I tried 152326 and the bug has gone away. So, I'm closing this PR.
--
danglin at gcc dot gnu dot org changed:
What|Removed |Added
-
--- Comment #11 from howarth at nitro dot med dot uc dot edu 2009-09-30
22:36 ---
Mike,
I am a fuzzy on the best way to avoid darwin_emit_unwind_label since it is
assigned to a define in darwin.h. The only thing that makes sense to me is...
Index: gcc/config/darwin.c
=
--- Comment #4 from pinskia at gcc dot gnu dot org 2009-09-30 22:17 ---
Confirmed.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
Status|UNCON
--- Comment #3 from pinskia at gcc dot gnu dot org 2009-09-30 22:16 ---
*** Bug 41527 has been marked as a duplicate of this bug. ***
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
---
--- Comment #2 from pinskia at gcc dot gnu dot org 2009-09-30 22:16 ---
This was filed earlier by the contributor of IPA-SRA :).
*** This bug has been marked as a duplicate of 41503 ***
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Adde
Consider the following:
=
#include
class A {};
class B : public A {};
int main()
{
std::tuple b;
std::tuple a1(b); // Compiles fine
std::tuple a2(std::move(b)); // Error
return 0;
}
=
Copy constructing a std::tuple from a std::tuple works fine, but
move-constructing it doesn't,
--- Comment #13 from davek at gcc dot gnu dot org 2009-09-30 21:26 ---
Created an attachment (id=18680)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18680&action=view)
"bash -x" log of configure script run
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41418
--- Comment #12 from davek at gcc dot gnu dot org 2009-09-30 21:25 ---
Bad news: the patch doesn't work. I edited the generated config.status so
that it would use "bash -x" when reinvoking configure as part of --recheck, and
it seems that although the code goes through the switch and
--- Comment #10 from mrs at apple dot com 2009-09-30 20:54 ---
When targeting darwin10, the .eh symbols don't have to be generated anymore.
They are from darwin.c, darwin_emit_unwind_label.
if (darwin_macosx_version_min
&& strverscmp (darwin_macosx_version_min, "10.6") >= 0)
ca
>From http://gcc.gnu.org/ml/gcc-patches/2009-09/msg02106.html
> You enable LTO by default if a suitable libelf is found. But what if
> libelf is present but the target for which GCC is being configured is not
> an ELF target? Presumably LTO will not work in that case (i.e., it
> requires the tar
Much of the documentation from design documents, papers and presentations need
to be ported to the internals manual.
Also, from http://gcc.gnu.org/ml/gcc-patches/2009-09/msg02134.html
(A) Details in the patch.
(B) Details missing from the patch.
(C) Lack of general explanation of how to use LTO
--- Comment #9 from howarth at nitro dot med dot uc dot edu 2009-09-30
20:26 ---
(In reply to comment #8)
> darwin has -fpic by default.
>
Oddly, if I build gcc trunk on x86_64 Fedora 10 with CPPFLAGS set to -fPIC and
run the testsuite with -fPIC, I don't see the .eh symbols...
grep
--- Comment #2 from dnovillo at gcc dot gnu dot org 2009-09-30 20:17
---
More portability concerns from
http://gcc.gnu.org/ml/gcc-patches/2009-09/msg02157.html
> +/* This needs to be included after config.h. Otherwise, _GNU_SOURCE will not
> + be defined in time to set __USE_GNU in
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P4
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41515
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P2
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41511
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P1
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41497
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P4
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41494
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P4
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41493
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Status|UNCONFIRMED |WAITING
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41491
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P2
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41485
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P2
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41279
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Priority|P3 |P2
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41217
--- Comment #1 from ubizjak at gmail dot com 2009-09-30 19:57 ---
Please attach preprocessed source and generated asm file (please follow the
instructions in http://gcc.gnu.org/bugs.html#report).
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41524
--- Comment #8 from burnus at gcc dot gnu dot org 2009-09-30 19:56 ---
Subject: Bug 40996
Author: burnus
Date: Wed Sep 30 19:55:45 2009
New Revision: 152345
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=152345
Log:
fortran/
2009-09-30 Janus Weil
* check.c (gfc_check
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.5.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41493
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Target Milestone|--- |4.5.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41485
--- Comment #52 from developer at sandoe-acoustics dot co dot uk
2009-09-30 19:54 ---
(In reply to comment #51)
> Looks much better than past versions... Seems like muse-shared-libgcc-ext
> should be the default, and possibly that we don't even need the switch?
thanks Mike, I'll do a
--- Comment #1 from dnovillo at gcc dot gnu dot org 2009-09-30 19:50
---
Other portability concerns:
http://gcc.gnu.org/ml/gcc-patches/2009-09/msg02157.html
> The host portability issue with __attribute__ ((visibility ("hidden")))
> has already been noted. I suggest conditioning the
--- Comment #1 from ubizjak at gmail dot com 2009-09-30 19:47 ---
Created an attachment (id=18679)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18679&action=view)
preprocessed testcase
Preprocessed testcase, segfaults with "-O2 -g -std=c++0x" on alpha-linux-gnu
(also with crossco
--- Comment #51 from mrs at apple dot com 2009-09-30 19:45 ---
Looks much better than past versions... Seems like muse-shared-libgcc-ext
should be the default, and possibly that we don't even need the switch?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39888
ntext arg-type
>
$2 = void
Please note the name of param_decl argument (ISRA.11).
I sincerely don't know which component is responsible for this failure, so I'll
mark this PR as c++. Looking at the backtrace and due to the fact that "-g" is
needed, recent debug changes shoul
--- Comment #34 from pinskia at gcc dot gnu dot org 2009-09-30 19:41
---
Fixed in 4.4.2.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
Statu
--- Comment #1 from pinskia at gcc dot gnu dot org 2009-09-30 19:40 ---
*** This bug has been marked as a duplicate of 41180 ***
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
--
--- Comment #33 from pinskia at gcc dot gnu dot org 2009-09-30 19:40
---
*** Bug 41525 has been marked as a duplicate of this bug. ***
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
-
>From http://gcc.gnu.org/ml/gcc-patches/2009-09/msg02148.html
I can see some issues here with output portability that should be
addressed for objects with LTO information to be portable between hosts.
1. There are some structures lto_*header that include 16-bit or 32-bit
integer fields (plus an
This bug looks very similar to an existing bug 41180. However, the discussion
on that seems to be for 64-bit systems. I'm experiencing the same problem,
except that my Mac is a first generation Mac (hence, a 32-bit kernel) running
Snow Leopard. Not sure if this is the same as that or not.
First, h
Source file from "TaoCrypt release 0.9.2 02/5/2007", which is part of MySQL
source tree.
Works with gcc 4.3.3. With gcc 4.4.1, g++ (or gcc) crashes in the compilation
process due to complain from as.
Command line:
% g++ -static-libgcc -DHAVE_CONFIG_H -I. -I../../../../include -I./../include
-I.
--- Comment #1 from twlevo at gmail dot com 2009-09-30 19:28 ---
Created an attachment (id=18678)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18678&action=view)
patched graph.c
graph.c for gcc-4.5 used in gcc-4.5.0 20090924 (experimental)
--
http://gcc.gnu.org/bugzilla/show
gcc-4.3.2 and gcc-4.5.0 20090924 (experimental) snapshot generate few vcg files
with syntax errors because of a extra '}' in the graph.
it appears in gcc/graph.c print_rtl_graph_with_bb() sometimes called end_bb()
without a start_bb() and this results in a extra '}' in the output generating
the par
--- Comment #8 from hjl dot tools at gmail dot com 2009-09-30 19:13 ---
It no longer hangs as of revision 152331:
http://gcc.gnu.org/ml/gcc-testresults/2009-09/msg02788.html
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
--
--- Comment #7 from ubizjak at gmail dot com 2009-09-30 18:24 ---
Fixed.
--
ubizjak at gmail dot com changed:
What|Removed |Added
Status|NEW
--- Comment #6 from uros at gcc dot gnu dot org 2009-09-30 18:23 ---
Subject: Bug 22093
Author: uros
Date: Wed Sep 30 18:22:48 2009
New Revision: 152344
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=152344
Log:
PR target/22093
* config/alpha/alpha.md (unaligned_
--- Comment #5 from uros at gcc dot gnu dot org 2009-09-30 18:03 ---
Subject: Bug 22093
Author: uros
Date: Wed Sep 30 18:03:17 2009
New Revision: 152343
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=152343
Log:
PR target/22093
* config/alpha/alpha.md (unaligned_
--- Comment #19 from ludovic at ludovic-brenta dot org 2009-09-30 17:35
---
I also applied the patch to GCC 4.4 and confirmed that it fixes this particular
bug. Barring any other blocking problems, Debian will be able to switch to GCC
4.4 as the default Ada compiler. Thank you all for
--- Comment #1 from pinskia at gcc dot gnu dot org 2009-09-30 17:18 ---
The easy fix is to change the inline-asm strings to "".
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41522
The test above fails on Solaris/SPARC when the native assembler is in use:
output is:
/usr/bin/as: "/var/tmp//cc83aiQ4.s", line 23: error: invalid character (0x72)
/usr/bin/as: "/var/tmp//cc83aiQ4.s", line 23: error: unknown opcode "register"
/usr/bin/as: "/var/tmp//cc83aiQ4.s", line 23: error:
--- Comment #3 from kargl at gcc dot gnu dot org 2009-09-30 17:06 ---
*** Bug 41520 has been marked as a duplicate of this bug. ***
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41515
--- Comment #1 from kargl at gcc dot gnu dot org 2009-09-30 17:06 ---
*** This bug has been marked as a duplicate of 41515 ***
--
kargl at gcc dot gnu dot org changed:
What|Removed |Added
--
--- Comment #50 from developer at sandoe-acoustics dot co dot uk
2009-09-30 16:58 ---
Created an attachment (id=18677)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18677&action=view)
further simplified lib ext patch
after some discussion with Ian Lance Taylor, I took another loo
While
gfortran -O1 -flto -c species.f90
gfortran -O1 -flto -c atom.f90
gfortran -O1 -flto -o species species.o atom.o
works, the following gives an ICE:
gfortran -O1 -flto -g -c species.f90
gfortran -O1 -flto -g -c atom.f90
gfortran -O1 -flto -g -o species species.o atom.o
lto1: error: mis
Using the PARAMETER statement in a module subroutine for assigning values to a
character array yields nonsense when printed at runtime. If, instead, the
PARAMETER attribute is used, everything is fine.
gfortran -v yields:
Using built-in specs.
Target: i686-pc-linux-gnu
Configured with: /home/jerr
--- Comment #5 from rguenth at gcc dot gnu dot org 2009-09-30 16:00 ---
Subject: Bug 41502
Author: rguenth
Date: Wed Sep 30 16:00:07 2009
New Revision: 152332
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=152332
Log:
2009-09-30 Richard Guenther
PR middle-end/41502
--- Comment #3 from rguenth at gcc dot gnu dot org 2009-09-30 15:48 ---
Fixed then.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
Status|WAIT
--- Comment #2 from ro at techfak dot uni-bielefeld dot de 2009-09-30
15:41 ---
Subject: Re: libobjc fails to configure on Solaris
rguenth at gcc dot gnu dot org writes:
> Can you check if this still applies? It should be broken on trunk as well
> then.
you're right: the problem is
--- Comment #2 from rguenth at gcc dot gnu dot org 2009-09-30 15:27 ---
Mine.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned
--- Comment #12 from rguenth at gcc dot gnu dot org 2009-09-30 15:26
---
Fixed by merging the patch from trunk.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #1 from rguenth at gcc dot gnu dot org 2009-09-30 15:25 ---
Can you check if this still applies? It should be broken on trunk as well
then.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #7 from rguenth at gcc dot gnu dot org 2009-09-30 15:24 ---
Needs re-confirming.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
St
--- Comment #2 from pinskia at gcc dot gnu dot org 2009-09-30 15:20 ---
As mentioned by Joseph, in C enums can hold all values of the undering type.
This is different from C++.
--
pinskia at gcc dot gnu dot org changed:
What|Removed |Added
---
--- Comment #11 from rguenther at suse dot de 2009-09-30 15:14 ---
Subject: Re: ICE in duplicate_node_data, at
ipa-pure-const.c:633
On Wed, 30 Sep 2009, hubicka at ucw dot cz wrote:
> --- Comment #10 from hubicka at ucw dot cz 2009-09-30 15:11 ---
> Subject: Re: ICE in dupl
--- Comment #10 from hubicka at ucw dot cz 2009-09-30 15:11 ---
Subject: Re: ICE in duplicate_node_data, at ipa-pure-const.c:633
> The merge decisions should be all settled at that point. I don't know
> if checking cgraph->decl is reliable though (I didn't trace to where
> we allocate
--- Comment #3 from joel at gcc dot gnu dot org 2009-09-30 15:06 ---
(In reply to comment #2)
> Looks like something has clobbered register A6.
>
I reran against the head yesterday. I have an objdump and a qemu trace
with register values between each block of assembly. I don't see w
--- Comment #7 from law at redhat dot com 2009-09-30 14:47 ---
Subject: Re: GCC choosing poor code sequence for certain
stores (x86)
On 09/30/09 03:22, jakub at gcc dot gnu dot org wrote:
> --- Comment #6 from jakub at gcc dot gnu dot org 2009-09-30 09:22 ---
> For x86-64 we
--- Comment #2 from burnus at gcc dot gnu dot org 2009-09-30 14:38 ---
(In reply to comment #1)
> As noted,
> character(3), parameter :: parm(5) = (/'xo ','yo ','ag ','xr ','yr '/)
> works.
Interestingly, it also works if one places "parm" into PROGRAM instead of using
it in a sub
--- Comment #9 from rguenther at suse dot de 2009-09-30 14:38 ---
Subject: Re: ICE in duplicate_node_data, at
ipa-pure-const.c:633
On Wed, 30 Sep 2009, hubicka at ucw dot cz wrote:
> --- Comment #8 from hubicka at ucw dot cz 2009-09-30 14:33 ---
> Subject: Re: ICE in duplic
--- Comment #8 from hubicka at ucw dot cz 2009-09-30 14:33 ---
Subject: Re: ICE in duplicate_node_data, at ipa-pure-const.c:633
>
>
> --- Comment #7 from rguenth at gcc dot gnu dot org 2009-09-30 13:01
> ---
> During pure_const_read_summary lto_cgraph_encoder_deref returns
--- Comment #6 from dave at hiauly1 dot hia dot nrc dot ca 2009-09-30
14:22 ---
Subject: Re: [4.5 Regressions] ld: (Warning) Unsatisfied symbol
"gomp_tls_data"
> --- Comment #3 from davek at gcc dot gnu dot org 2009-09-30 13:50 ---
> Comment 32 in bug 41357 says that that bug
--- Comment #1 from joseph at codesourcery dot com 2009-09-30 14:14 ---
Subject: Re: New: Unnecessary uninitialized warning
In C, enums may hold any value of the underlying integer type, so this
warning seems correct.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41519
--- Comment #5 from davek at gcc dot gnu dot org 2009-09-30 14:10 ---
(In reply to comment #4)
> It uses GCC's emulated TLS support (don't support HP's TLS implementation).
> There were no libgomp failures prior to the change:
> http://gcc.gnu.org/ml/gcc-testresults/2009-09/msg02555.htm
--- Comment #4 from dave at hiauly1 dot hia dot nrc dot ca 2009-09-30
13:55 ---
Subject: Re: [4.5 Regressions] ld: (Warning) Unsatisfied symbol
"gomp_tls_data"
> What kind of TLS do you have on your platform? Also, does reverting the patch
> help you any, or do you just end up with t
--- Comment #3 from davek at gcc dot gnu dot org 2009-09-30 13:50 ---
Comment 32 in bug 41357 says that that bug has now gone away again as of
r.152325; I haven't had time to verify that myself yet but perhaps that one
will work for you also?
--
http://gcc.gnu.org/bugzilla/show_bug.
--- Comment #4 from armin76 at gentoo dot org 2009-09-30 13:49 ---
(In reply to comment #3)
> Can you see what a backtrace gives you in a gdb session ? It sounds like you
> are out of memory. I'll second what Mikael said in Comment #2 with respect to
> being able to bootstrap .
Hrm...wel
--- Comment #32 from christian dot joensson at gmail dot com 2009-09-30
13:47 ---
I'm not sure how (or why) but... I checked out gcc trunk rev. 152325, and the
problems seems to have gone away (at least for now).
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41357
[...@gnu-6 tmp]$ cat x.c
enum foo
{
foo1,
foo2
} foo;
int
bar ()
{
int x;
switch (foo)
{
case foo1:
x = 3;
break;
case foo2:
x = 8;
break;
}
return x;
}
[...@gnu-6 tmp]$ gcc -Wall -S x.c -O
x.c: In function âbarâ:
x.c:10: warning: âxâ may be used u
--- Comment #43 from ubizjak at gmail dot com 2009-09-30 13:34 ---
(In reply to comment #42)
> Created an attachment (id=18676)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18676&action=view) [edit]
> Making IPA-SRA ignore va_lists
>
> This patch prevents IPA-SRA from considering
--- Comment #3 from rguenth at gcc dot gnu dot org 2009-09-30 13:17 ---
The issue here is that loop header copying creates a new induction variable
by moving the use before the definition of D.2708_4 = start_3(D) + i_12.
So it just does what we teach PRE not to do ...
--
http://gcc
--- Comment #7 from rguenth at gcc dot gnu dot org 2009-09-30 13:01 ---
During pure_const_read_summary lto_cgraph_encoder_deref returns a reference
to a bogus cgraph node (it's all zeros, just the UID is filled in). Later
the cgraph node gets reused by cgraph_clone_node. So something l
--- Comment #4 from hjl dot tools at gmail dot com 2009-09-30 12:59 ---
Fixed.
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
Status|UNCONFIRM
Dear all,
I would like to post a fault report for the GNU C/C++ compiler 4.3.0.
Used invokation line for the GNU C++ compiler:
gcc -c -x c++ -ansi -Wall -Werror -mcpu=603e -fverbose-asm -mbig
-mmultiple -mstring -mstrict-align -meabi -msdata -fno-common
-fno-exceptions -fno-rtti -O3 -fno
--- Comment #6 from rguenth at gcc dot gnu dot org 2009-09-30 11:51 ---
More reduced:
a.c:
extern "C" double sqrt (double __x) throw ();
typedef double VECTOR[3];
enum { X = 0, Y = 1, Z = 2, T = 3 };
inline void VLength(double& a, const VECTOR b)
{
a = sqrt(b[X] * b[X] + b[Y] * b[Y
--- Comment #42 from jamborm at gcc dot gnu dot org 2009-09-30 11:37
---
Created an attachment (id=18676)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18676&action=view)
Making IPA-SRA ignore va_lists
This patch prevents IPA-SRA from considering va_list structs as candidates for
--- Comment #41 from jamborm at gcc dot gnu dot org 2009-09-30 11:35
---
(In reply to comment #38)
> (In reply to comment #37)
>
> > Can you please attach pre-processed source of it so that I can try
> > running it through a cross-compiler?
> >
> > Additionally, -fdump-tree-cddce-sl
--- Comment #5 from rguenth at gcc dot gnu dot org 2009-09-30 11:18 ---
Reduced from 453.povray:
lto-g/gcc> ./g++ -B. -o povray -nostdlib -shared -fPIC -O -flto torus.3.ii
bezier.3.ii
lto1: internal compiler error: in duplicate_node_data, at ipa-pure-const.c:633
Please submit a full bu
--- Comment #4 from rguenth at gcc dot gnu dot org 2009-09-30 11:18 ---
Created an attachment (id=18675)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18675&action=view)
reduced testcase, part2
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41487
--- Comment #3 from rguenth at gcc dot gnu dot org 2009-09-30 11:17 ---
Created an attachment (id=18674)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18674&action=view)
reduced testcase, part1
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41487
Gcc behaviour looks like it might be incorrect on thefollowing example.
It relates to the treatment of the pragma. (I cannot provide the .i file
here because then the pragma is stripped out, and this changes the
behaviour of the compiler.)
$ cat test22.c
#include
int x;
void foo (void)
{
int
--- Comment #3 from jakub at gcc dot gnu dot org 2009-09-30 10:07 ---
Try using 4U instead of 4 or (size_t) 4 and you'll see it not optimized out
either.
The problem is that the second comparison is done in size_t, unsigned type, so
if argc is say -3, the function doesn't return -1 early
1 - 100 of 116 matches
Mail list logo
