[Bug ld/11956] relocation truncated to fit R_MIPS_TLS_GD linking xulrunner

2010-09-02 Thread r0bertz at gentoo dot org


-- 
   What|Removed |Added

 CC||r0bertz at gentoo dot org


http://sourceware.org/bugzilla/show_bug.cgi?id=11956

--- You are receiving this mail because: ---
You are on the CC list for the bug, or are watching someone who is.

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


bug of ld 2.20.1

2010-09-02 Thread 徐超

 Recently, I update the toolchain(binutils 2.20.1) I meet one issue on 
LD.
 When I use LD to link the object with  options “�Coformat binary “ and 
“--start-group   �Cend-group” simultaneously, 
LD fail to find the symbol.
 
When I remove “―oformat”, it can build the ELF file. I try to write some simple 
case to reproduce it, but in the simple
Case, it is OK.
 (host i686-pc-linux-gnu  target i686-cm-linux)
 
 The log file is as follow:
Linking: elf/rom.bin
../..//lib/elf/libcore.a(bios_keybd.o): In function `biosKeybdShiftStatus':
bios_keybd.c:(.text+0x67): undefined reference to `kbdmgrGetShiftFlagStatusByte'
../..//lib/elf/libcore.a(bios_keybd.o): In function 
`biosKeybdExtendedShiftStatus':
bios_keybd.c:(.text+0x8b): undefined reference to `kbdmgrGetShiftFlagStatusByte'
 I use readelf and objdump to check the symbol, the symbol can be found 
in the libcore.a.
 At the same time the LD (2.17.50, 2.18.50) both can work under this 
situation.
 So it should be bug. ___
bug-binutils mailing list
bug-binutils@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-binutils


[Bug binutils/11971] New: config.status produces incompatible defines.awk and config.h

2010-09-02 Thread dhoyt at llnl dot gov
config.h was not being created correctly on my Windows 7 (x86_64) machine 
using the latest released tools from msys/mingw, causing all sorts of build 
problems. After investigating the issue, it was narrowed down to a problem 
with how config.status produces and then uses defines.awk. That awk file uses 
a regular expression that looks for lines like "#undef VALUE_HERE" to 
eventually change them to "#define VALUE_HERE 1" if configure deems it so, but 
it wasn't finding them due to mismatches from Windows-style line endings.

This patch fixes the issue in defines.awk, but I'm unsure of where it really 
needs to go (because defines.awk is created by config.status, itself created 
through configure) since I'm not an autotools expert:

--- defines-orig.awkThu Sep  2 15:52:10 2010
+++ defines.awk Thu Sep  2 15:56:01 2010
@@ -51,8 +51,9 @@
   for (key in D) D_is_set[key] = 1
   FS = " "
 }
-/^[\t ]*#[\t ]*(define|undef)[\t ]+
[_abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ]
[_abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789]*([\t (]|$)/ {
+/^[\t ]*#[\t ]*(define|undef)[\t ]+
[_abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ]
[_abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789]*([\t (]
|\r$)/ {
   line = $ 0
+  sub(/\r$/, "", line)
   split(line, arg, " ")
   if (arg[1] == "#") {
 defundef = arg[2]

-- 
   Summary: config.status produces incompatible defines.awk and
config.h
   Product: binutils
   Version: 2.21 (HEAD)
Status: NEW
  Severity: normal
  Priority: P2
 Component: binutils
AssignedTo: unassigned at sources dot redhat dot com
ReportedBy: dhoyt at llnl dot gov
CC: bug-binutils at gnu dot org
 GCC build triplet: i686-pc-mingw32
  GCC host triplet: i686-pc-mingw32
GCC target triplet: i686-w64-mingw32


http://sourceware.org/bugzilla/show_bug.cgi?id=11971

--- You are receiving this mail because: ---
You are on the CC list for the bug, or are watching someone who is.

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


[Bug binutils/11971] config.status produces incompatible defines.awk and config.h

2010-09-02 Thread dhoyt at llnl dot gov

--- Additional Comments From dhoyt at llnl dot gov  2010-09-02 23:10 ---
Created an attachment (id=4963)
 --> (http://sourceware.org/bugzilla/attachment.cgi?id=4963&action=view)
Fix for defines.awk


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=11971

--- You are receiving this mail because: ---
You are on the CC list for the bug, or are watching someone who is.

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


Re: bug of ld 2.20.1

2010-09-02 Thread Alan Modra
On Thu, Sep 02, 2010 at 01:25:11PM +0800, 徐超 wrote:
> 
>  Recently, I update the toolchain(binutils 2.20.1) I meet one issue 
> on LD.
>  When I use LD to link the object with  options “–oformat binary “ 
> and “--start-group   –end-group” simultaneously, 
> LD fail to find the symbol.
>  
> When I remove “—oformat”, it can build the ELF file. I try to write some 
> simple case to reproduce it, but in the simple
> Case, it is OK.
>  (host i686-pc-linux-gnu  target i686-cm-linux)
>  
>  The log file is as follow:
> Linking: elf/rom.bin
> ../..//lib/elf/libcore.a(bios_keybd.o): In function `biosKeybdShiftStatus':
> bios_keybd.c:(.text+0x67): undefined reference to 
> `kbdmgrGetShiftFlagStatusByte'
> ../..//lib/elf/libcore.a(bios_keybd.o): In function 
> `biosKeybdExtendedShiftStatus':
> bios_keybd.c:(.text+0x8b): undefined reference to 
> `kbdmgrGetShiftFlagStatusByte'
>  I use readelf and objdump to check the symbol, the symbol can be 
> found in the libcore.a.
>  At the same time the LD (2.17.50, 2.18.50) both can work under this 
> situation.
>  So it should be bug.   

What happens if you do not use an archive?  ie. extract the files in
the archive then specify all of them on the ld command line?

-- 
Alan Modra
Australia Development Lab, IBM

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


[Bug gas/11972] New: Large immediate for neon vmov (parse_big_immediate) fails

2010-09-02 Thread rmansfield at qnx dot com
$ cat $/bignum.s 
.fpu neon
vmov.i64d16, #-4294967296  @ v2si

$ ./as-new ~/bignum.s
/home/ryan/bignum.s: Assembler messages:
/home/ryan/bignum.s:2: Error: expected  or  or  operand -- `vmov.i64
d16,#-4294967296'

Works in 2.20.1. Fails in head and this appears to be due to the following 
change:

http://sourceware.org/cgi-bin/cvsweb.cgi/src/gas/expr.c.diff?r1=1.83&r2=1.84&cvsroot=src

-- 
   Summary: Large immediate for neon vmov (parse_big_immediate)
fails
   Product: binutils
   Version: 2.21 (HEAD)
Status: NEW
  Severity: normal
  Priority: P2
 Component: gas
AssignedTo: unassigned at sources dot redhat dot com
ReportedBy: rmansfield at qnx dot com
CC: bug-binutils at gnu dot org
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: arm-unknown-linux-gnueabi


http://sourceware.org/bugzilla/show_bug.cgi?id=11972

--- You are receiving this mail because: ---
You are on the CC list for the bug, or are watching someone who is.

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