[Bug ld/3243] New: Mips R_MIPS_16 relocation incorrectly handled
R_MIPS_16 is incorrectly handled when at a half-word offset: See below: : 8: 0001movfzero,zero,$fcc0 c: 0130tge zero,zero,0x4 ... 10: R_MIPS_32 kdebug_init() 14: R_MIPS_32 kdebug_entry(void*) 50: R_MIPS_32 _memory_descriptors_offset 56: R_MIPS_16 _memory_descriptors_size a4: 003f0x3f After linking, the values that should be at offset 0x56, is actually placed at 0x60 in the elf output. -- Summary: Mips R_MIPS_16 relocation incorrectly handled Product: binutils Version: 2.15 Status: NEW Severity: critical Priority: P2 Component: ld AssignedTo: unassigned at sources dot redhat dot com ReportedBy: navlrac at hotmail dot com CC: bug-binutils at gnu dot org GCC host triplet: ia32-linux GCC target triplet: mips-linux http://sourceware.org/bugzilla/show_bug.cgi?id=3243 --- 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/3235] objdump on AMD64 - wrong output for prefixed offset operands.
--- Additional Comments From fruffell at cs dot uwaterloo dot ca 2006-09-21 23:32 --- This bug is not resolved. Objdump still incorrectly displays a 32-bit memory offset as a 64-bit offset on x86_64. Even though the default offset size is 64-bits, the address can be overridden by a 0x67 prefix. If the script attached does not work for you, I can attached an already generated object file. -- What|Removed |Added Status|RESOLVED|REOPENED Resolution|WORKSFORME | http://sourceware.org/bugzilla/show_bug.cgi?id=3235 --- 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/3235] objdump on AMD64 - wrong output for prefixed offset operands.
--- Additional Comments From fruffell at cs dot uwaterloo dot ca 2006-09-21 23:29 --- Created an attachment (id=1315) --> (http://sourceware.org/bugzilla/attachment.cgi?id=1315&action=view) Automatic testcase. A new testcase, which is automatic for those who don't want to run a hex editor. The script attached does: - generates a C file: contains a guarded invalid 0xa2 mov instruction - runs the executable which outputs the address of the char array tstStr - uses this address to patch the 0xa2 mov instruction, inverts the guard - recompiles and executes the source file, 0xa2 mov instruction now modifies the string, output is changed from "hello world" to "zello world" (e.g. the mov instruction successfully updates tstStr with the 32-bit address). - runs objdump, greps the 0xa2 mov instruction which is incorrectly displayed with an invalid 64-bit address <--- this is the bug On my test machine, output is: [EMAIL PROTECTED] ~/memOffset $ ./gcc-memoffset.sh 00500900 tstStr is "zello world" 40051e: 67 a2 00 09 50 00 beaddr32 mov %al,0x500900be00500900 -- http://sourceware.org/bugzilla/show_bug.cgi?id=3235 --- 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 ld/3223] ld fails to link correct variables from linker script.
--- Additional Comments From hjl at lucon dot org 2006-09-21 23:06 ---
Linker behavior is changed. Empty output section in the linker script no
longer changes the output layout. Unless you change DOT with empty output
section, the output layout will be the same as if empty output section
doesn't exist in the linker script. To make the output layout depend on
empty output section, you can add
. = ADDR (.data) + SIZEOF (.data);
after empty output section:
[EMAIL PROTECTED] empty-8]$ cat foo.t
SECTIONS
{
.text 0x: { *(.text) }
.data 0x2000:
{
__data_start = . ;
*(.data)
}
. = ADDR (.data) + SIZEOF (.data);
_edata = .;
/DISCARD/ : { *(.*) }
}
[EMAIL PROTECTED] empty-8]$ make
as -o foo.o foo.s
./ld -Tfoo.t -o foo foo.o
nm foo | grep __data_start
2000 A __data_start
nm foo | grep _edata
2000 A _edata
[EMAIL PROTECTED] empty-8]$
--
http://sourceware.org/bugzilla/show_bug.cgi?id=3223
--- 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 ld/3223] ld fails to link correct variables from linker script.
--- Additional Comments From hjl at lucon dot org 2006-09-21 22:33 ---
A simple testcase:
[EMAIL PROTECTED] empty-8]$ cat foo.s
.text
.global _start
_start:
.long __data_start
.long _edata
[EMAIL PROTECTED] empty-8]$ cat foo.t
SECTIONS
{
.text 0x: { *(.text) }
.data 0x2000:
{
__data_start = . ;
*(.data)
}
/* _edata = ADDR (.data) + SIZEOF (.data); */
_edata = .;
/DISCARD/ : { *(.*) }
}
[EMAIL PROTECTED] empty-8]$ make
as -o foo.o foo.s
ld -Tfoo.t -o foo foo.o
nm foo | grep __data_start
2000 A __data_start
nm foo | grep _edata
0008 A _edata
[EMAIL PROTECTED] empty-8]$
--
What|Removed |Added
CC||hjl at lucon dot org
GCC build triplet|i686-pc-cygwin |
GCC host triplet|i686-pc-cygwin |
GCC target triplet|arm-elf |
http://sourceware.org/bugzilla/show_bug.cgi?id=3223
--- 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 ld/3191] Dwarf 2 reader in linker doesn't suppor DW_FORM_ref_addr
--- Additional Comments From hjl at lucon dot org 2006-09-21 21:19 --- The current patch is at http://sourceware.org/ml/binutils/2006-09/msg00224.html -- http://sourceware.org/bugzilla/show_bug.cgi?id=3191 --- 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/3235] objdump on AMD64 - wrong output for prefixed offset operands.
--- Additional Comments From hjl at lucon dot org 2006-09-21 21:18 --- Please submit a proper testcase without requiring a hex editor. You can use ".byte 0x67,0xa2,0x08,0x09,0x50,0x00\n\t" in asm statement instead of nop + hex editor. -- What|Removed |Added CC||hjl at lucon dot org Status|NEW |RESOLVED Resolution||WORKSFORME http://sourceware.org/bugzilla/show_bug.cgi?id=3235 --- 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
Binutils 2.17 makefile problem
"make distclean" leaves a config.cache file in the intl directory. As a result tarring up the sources after building on one system, doing make distclean, installing on another system and configuring fails to do the right thing. Should be easy to fix, I imagine. -- Jim James Cownie <[EMAIL PROTECTED]> SSG/DPD/PAT Tel: +44 117 9071438 ___ bug-binutils mailing list bug-binutils@gnu.org http://lists.gnu.org/mailman/listinfo/bug-binutils
