[Bug ld/13287] g++ -Wl,--as-needed -flto doesn't work with the BFD linker, but with gold

2011-10-20 Thread amodra at gmail dot com
http://sourceware.org/bugzilla/show_bug.cgi?id=13287

Alan Modra  changed:

   What|Removed |Added

 Status|NEW |ASSIGNED
 CC|amodra at gmail dot com |
 AssignedTo|unassigned at sourceware|amodra at gmail dot com
   |dot org |

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug ld/13287] g++ -Wl,--as-needed -flto doesn't work with the BFD linker, but with gold

2011-10-20 Thread cvs-commit at gcc dot gnu.org
http://sourceware.org/bugzilla/show_bug.cgi?id=13287

--- Comment #4 from cvs-commit at gcc dot gnu.org  2011-10-20 09:51:03 UTC ---
CVSROOT:/cvs/src
Module name:src
Changes by:amo...@sourceware.org2011-10-20 09:50:59

Modified files:
ld : ChangeLog plugin.c plugin.h ldlang.c 

Log message:
PR ld/13287
* plugin.c (plugin_should_reload): New function.
* plugin.h (plugin_should_reload): Declare.
* ldlang.c (open_input_bfds): Use above function.

Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/src/ld/ChangeLog.diff?cvsroot=src&r1=1.2383&r2=1.2384
http://sourceware.org/cgi-bin/cvsweb.cgi/src/ld/plugin.c.diff?cvsroot=src&r1=1.42&r2=1.43
http://sourceware.org/cgi-bin/cvsweb.cgi/src/ld/plugin.h.diff?cvsroot=src&r1=1.9&r2=1.10
http://sourceware.org/cgi-bin/cvsweb.cgi/src/ld/ldlang.c.diff?cvsroot=src&r1=1.378&r2=1.379

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug binutils/13292] Binutils uses "struct stat" instead of "#include "

2011-10-20 Thread pgquiles at elpauer dot org
http://sourceware.org/bugzilla/show_bug.cgi?id=13292

--- Comment #2 from Pau Garcia i Quiles  
2011-10-20 10:06:32 UTC ---
(In reply to comment #1)
> (In reply to comment #0)
> > Created attachment 5984 [details]
> > Fix
> > 
> > mingw-w64 recently implemented support for large files using
> > _FILE_OFFSET_BITS=64.
> > 
> > Binutils uses "struct stat" instead of "#include ", which makes 
> > it
> > fail to compile with mingw-w64 trunk.
> > 
> > The attached patch fixes the issue.
> 
> I think you shouldn't edit bfd/bfd-in2.h since it is auto generated.
> The bfd-in.h is the file you want to patch, I guess.

You are right. I'm attaching a new patch against bfd/bfd-in.h (which is
essentially the same).


> The patch doesn't seem to solve the problem for me.

What is the problem you are experiencing now? What compiler are you trying?


> Do you have the issue bfd/archive.c?

Sorry, I don't understand :-?

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug binutils/13292] Binutils uses "struct stat" instead of "#include "

2011-10-20 Thread pgquiles at elpauer dot org
http://sourceware.org/bugzilla/show_bug.cgi?id=13292

Pau Garcia i Quiles  changed:

   What|Removed |Added

   Attachment #5984|0   |1
is obsolete||

--- Comment #3 from Pau Garcia i Quiles  
2011-10-20 10:07:43 UTC ---
Created attachment 6020
  --> http://sourceware.org/bugzilla/attachment.cgi?id=6020
Fix (patch bfd-in.h)

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug binutils/13292] Binutils uses "struct stat" instead of "#include "

2011-10-20 Thread askondro at gmail dot com
http://sourceware.org/bugzilla/show_bug.cgi?id=13292

--- Comment #4 from Alen Skondro  2011-10-20 
12:02:39 UTC ---
(In reply to comment #2)
> (In reply to comment #1)
> > (In reply to comment #0)
> > > Created attachment 5984 [details]
> > > Fix
> > > 
> > > mingw-w64 recently implemented support for large files using
> > > _FILE_OFFSET_BITS=64.
> > > 
> > > Binutils uses "struct stat" instead of "#include ", which 
> > > makes it
> > > fail to compile with mingw-w64 trunk.
> > > 
> > > The attached patch fixes the issue.
> > 
> > I think you shouldn't edit bfd/bfd-in2.h since it is auto generated.
> > The bfd-in.h is the file you want to patch, I guess.
> 
> You are right. I'm attaching a new patch against bfd/bfd-in.h (which is
> essentially the same).
> 
> 
> > The patch doesn't seem to solve the problem for me.
> 
> What is the problem you are experiencing now? What compiler are you trying?

I'm using gcc 4.7 (trunk).

../../../binutils/bfd/archive.c: In function 'bfd_ar_hdr_from_filesystem':
../../../binutils/bfd/archive.c:1841:7: error: passing argument 1 of 'time'
from incompatible pointer type [-Werror]
In file included from
j:\cross-gcc\bin\../lib/gcc/i686-w64-mingw32/4.7.0/../../../../i686-w64-mingw32/include/sys/time.h:10:0,
 from ../../../binutils/bfd/sysdep.h:64,
 from ../../../binutils/bfd/archive.c:130:
j:\cross-gcc\bin\../lib/gcc/i686-w64-mingw32/4.7.0/../../../../i686-w64-mingw32/include/time.h:241:29:
note: expected 'time_t *' but argument is of type '__time64_t *'
cc1.exe: all warnings being treated as errors


> 
> 
> > Do you have the issue bfd/archive.c?
> 
> Sorry, I don't understand :-?

Sorry, I wanted to ask if you also experienced the compile error in
bfd/archive.c or somewhere else?

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug binutils/13292] Binutils uses "struct stat" instead of "#include "

2011-10-20 Thread pgquiles at elpauer dot org
http://sourceware.org/bugzilla/show_bug.cgi?id=13292

--- Comment #5 from Pau Garcia i Quiles  
2011-10-20 12:29:30 UTC ---
(In reply to comment #4)
> (In reply to comment #2)
> > (In reply to comment #1)
> > > (In reply to comment #0)
> > > > Created attachment 5984 [details]
> > > > Fix
> > > > 
> > > > mingw-w64 recently implemented support for large files using
> > > > _FILE_OFFSET_BITS=64.
> > > > 
> > > > Binutils uses "struct stat" instead of "#include ", which 
> > > > makes it
> > > > fail to compile with mingw-w64 trunk.
> > > > 
> > > > The attached patch fixes the issue.
> > > 
> > > I think you shouldn't edit bfd/bfd-in2.h since it is auto generated.
> > > The bfd-in.h is the file you want to patch, I guess.
> > 
> > You are right. I'm attaching a new patch against bfd/bfd-in.h (which is
> > essentially the same).
> > 
> > 
> > > The patch doesn't seem to solve the problem for me.
> > 
> > What is the problem you are experiencing now? What compiler are you trying?
> 
> I'm using gcc 4.7 (trunk).
> 
> ../../../binutils/bfd/archive.c: In function 'bfd_ar_hdr_from_filesystem':
> ../../../binutils/bfd/archive.c:1841:7: error: passing argument 1 of 'time'
> from incompatible pointer type [-Werror]
> In file included from
> j:\cross-gcc\bin\../lib/gcc/i686-w64-mingw32/4.7.0/../../../../i686-w64-mingw32/include/sys/time.h:10:0,
>  from ../../../binutils/bfd/sysdep.h:64,
>  from ../../../binutils/bfd/archive.c:130:
> j:\cross-gcc\bin\../lib/gcc/i686-w64-mingw32/4.7.0/../../../../i686-w64-mingw32/include/time.h:241:29:
> note: expected 'time_t *' but argument is of type '__time64_t *'
> cc1.exe: all warnings being treated as errors

I see.

I am compiling a cross binutils CVS with gcc 4.6.1 on Linux. No, I am not
seeing that problem.

> 
> > 
> > 
> > > Do you have the issue bfd/archive.c?
> > 
> > Sorry, I don't understand :-?
> 
> Sorry, I wanted to ask if you also experienced the compile error in
> bfd/archive.c or somewhere else?

No, I do not have that problem.

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug ld/13049] Generates R_ARM_NONE relocations for stub_* symbols when using -fdata-sections

2011-10-20 Thread nickc at redhat dot com
http://sourceware.org/bugzilla/show_bug.cgi?id=13049

Nick Clifton  changed:

   What|Removed |Added

   Attachment #6019|0   |1
is obsolete||

--- Comment #4 from Nick Clifton  2011-10-20 14:30:47 
UTC ---
Created attachment 6021
  --> http://sourceware.org/bugzilla/attachment.cgi?id=6021
Rename stub sections in ARM, PPC and HPPA

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug ld/13049] Generates R_ARM_NONE relocations for stub_* symbols when using -fdata-sections

2011-10-20 Thread nickc at redhat dot com
http://sourceware.org/bugzilla/show_bug.cgi?id=13049

--- Comment #5 from Nick Clifton  2011-10-20 14:38:45 
UTC ---
Hi Christophe,

  OK, I have uploaded a revised patch which contains an expanded comment, and
similar fixes for the elf64-ppc.c and elf32-hppa.c files.  I have also removed
the changes to the linker scripts, since these were not necessary.  (It was a
case of me being too paranoid).

  Are you happy with this version of the patch ?

Cheers
  Nick

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug binutils/13295] Segfault and/or incorrect output in objdump when disassembling an ARM library

2011-10-20 Thread nickc at redhat dot com
http://sourceware.org/bugzilla/show_bug.cgi?id=13295

--- Comment #3 from Nick Clifton  2011-10-20 15:23:53 
UTC ---
Created attachment 6022
  --> http://sourceware.org/bugzilla/attachment.cgi?id=6022
Fix for 2.21 sources

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug binutils/13295] Segfault and/or incorrect output in objdump when disassembling an ARM library

2011-10-20 Thread nickc at redhat dot com
http://sourceware.org/bugzilla/show_bug.cgi?id=13295

Nick Clifton  changed:

   What|Removed |Added

 Status|NEW |WAITING

--- Comment #4 from Nick Clifton  2011-10-20 15:24:46 
UTC ---
Hi Zoltan,

  I have uploaded a patch for the 2.21 sources which should fix this problem. 
Please let me know if you have any problems with it.

Cheers
  Nick

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug ld/13049] Generates R_ARM_NONE relocations for stub_* symbols when using -fdata-sections

2011-10-20 Thread christophe.lyon at st dot com
http://sourceware.org/bugzilla/show_bug.cgi?id=13049

--- Comment #6 from christophe.lyon at st dot com  2011-10-20 16:08:24 UTC ---
Hi Nick,
The comment is much clearer indeed, thanks.
I have noticed a typo "defintion" near the end of the comment.

However, I am wondering whether prefixing with "__" is secure enough: in my
understanding, the C standard defines the "__" namespace as reserved for tools,
but as such it could very well be used by some GCC or GLIBC magic, or by a
end-user, couldn't it?

What about using something like "_._" which could still be used in assembly,
but much less likely? But I won't argue strongly on this point.

Thanks

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug ld/13049] Generates R_ARM_NONE relocations for stub_* symbols when using -fdata-sections

2011-10-20 Thread cvs-commit at gcc dot gnu.org
http://sourceware.org/bugzilla/show_bug.cgi?id=13049

--- Comment #7 from cvs-commit at gcc dot gnu.org  2011-10-20 16:27:16 UTC ---
CVSROOT:/cvs/src
Module name:src
Changes by:ni...@sourceware.org2011-10-20 16:27:04

Modified files:
bfd: ChangeLog elf32-arm.c elf64-ppc.c elf32-hppa.c 

Log message:
PR ld/13049
* elf32-arm.c (STUB_SUFFIX): Avoid collision with user namespace
symbol names.
* elf64-ppc.c (STUB_SUFFIX): Likewise.
* elf32-hppa.c (STUB_SUFFIX): Likewise.

Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/src/bfd/ChangeLog.diff?cvsroot=src&r1=1.5496&r2=1.5497
http://sourceware.org/cgi-bin/cvsweb.cgi/src/bfd/elf32-arm.c.diff?cvsroot=src&r1=1.278&r2=1.279
http://sourceware.org/cgi-bin/cvsweb.cgi/src/bfd/elf64-ppc.c.diff?cvsroot=src&r1=1.367&r2=1.368
http://sourceware.org/cgi-bin/cvsweb.cgi/src/bfd/elf32-hppa.c.diff?cvsroot=src&r1=1.179&r2=1.180

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug ld/13049] Generates R_ARM_NONE relocations for stub_* symbols when using -fdata-sections

2011-10-20 Thread nickc at redhat dot com
http://sourceware.org/bugzilla/show_bug.cgi?id=13049

Nick Clifton  changed:

   What|Removed |Added

 Status|WAITING |RESOLVED
 Resolution||FIXED

--- Comment #8 from Nick Clifton  2011-10-20 16:32:22 
UTC ---
Hi Christophe,

  Thanks for spelling typo catch.

  You are right of course.  System type tools, including the compiler, can
generate symbols with a double underscore prefix.  But I think that in this
context the stub suffix *is* a system type symbol, and so it is correct to use
the __ prefix.  Plus if they really want to a user could always generate a
symbol with a _._ prefix, or any other combination that we might choose, so we
cannot win  

  So - I have gone ahead and checked the patch in.

Cheers
  Nick

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug binutils/13302] IRELATIVE relocation should come last

2011-10-20 Thread hjl.tools at gmail dot com
http://sourceware.org/bugzilla/show_bug.cgi?id=13302

H.J. Lu  changed:

   What|Removed |Added

   Attachment #6016|0   |1
is obsolete||

--- Comment #2 from H.J. Lu  2011-10-20 23:21:27 
UTC ---
Created attachment 6023
  --> http://sourceware.org/bugzilla/attachment.cgi?id=6023
A patch

This patch can build glibc on ia32 and x86-64.

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are on the CC list for the bug.

___
bug-binutils mailing list
bug-binutils@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-binutils