Hi,
I'am getting this error when compiling 2.6.9-mm1 :
KSYM .tmp_kallsyms3.S
AS .tmp_kallsyms3.o
LD vmlinux
SYSMAP System.map
SYSMAP .tmp_System.map
AS arch/i386/boot/setup.o
LD arch/i386/boot/setup
AS arch/i386/boot/compressed/head.o
CC arch/i386/boot/compressed/misc.o
OBJCOPY arch/i386/boot/compressed/vmlinux.bin
BFD: Warning: Writing section `.bss' to huge (ie negative) file offset
0xc05c6000.
objcopy: arch/i386/boot/compressed/vmlinux.bin: File truncated
make[2]: *** [arch/i386/boot/compressed/vmlinux.bin] Error 1
make[1]: *** [arch/i386/boot/compressed/vmlinux] Error 2
make: *** [bzImage] Error 2
[root@tigre01 im]# objdump -h ./vmlinux
./vmlinux: file format elf32-i386
Sections:
Idx Name Size VMA LMA File off Algn
0 .text 0037d80e c0100000 00100000 00001000 2**12
<-- something wrong here for LMA used to be 0xc010 000
CONTENTS, ALLOC, LOAD, READONLY, CODE
1 __ex_table 000013b8 c047d810 0047d810 0037e810 2**2
CONTENTS, ALLOC, LOAD, READONLY, DATA
2 .rodata 000a8309 c047ebe0 0047ebe0 0037fbe0 2**5
CONTENTS, ALLOC, LOAD, READONLY, DATA
...
23 .exit.text 00001a5b c06ab430 006ab430 005ac430 2**4
CONTENTS, ALLOC, LOAD, READONLY, CODE
24 .init.ramfs 00000086 c06ad000 006ad000 005ae000 2**0
CONTENTS, ALLOC, LOAD, READONLY, DATA
25 .data.percpu 00018938 c06ad100 006ad100 005ae100 2**7
CONTENTS, ALLOC, LOAD, DATA
26 .bss 0006dfa0 c06c6000 c06c6000 005c7000 2**12 <--
somthing wrong with LMA for .bss.
CONTENTS, ALLOC, LOAD, DATA
27 .comment 0000c36f 00000000 00000000 00634fa0 2**0
CONTENTS, READONLY
28 .debug_line 0021497b 00000000 00000000 0064130f 2**0
...
This is very strange. The .text LMA of the vmlinux file is at 0x__0__010
0000 whereas, it used to be at 0x__c__010 0000. But then, the .bss LMA
of the vmlinux file suddenly jump at 0x__c__06c 6000 instead of
0x__0__06c 6000 which would have seemed more natural -0). What can
cause this offset? I have look for the .bss section content. It contains
.bss data and .bss.page_aligned. The later seems to be the root of the
.bss LMA offset. .bss.page_aligned contains then empty_zero_page and
the swapper_pg_dir.
I finally succeded to compile the kernel, using the following
OBJCOPYFLAGS in arch/i386/Makefile.
OBJCOPYFLAGS := -O binary --change-section-lma .bss-0xc0000000 -R
.note -R .comment -S
I haven't yet tried to boot it.
Any idea to fix this objcopy error otherwise?
Why is the .text LMA set at 0x0010 0000 instead of 0xc010 0000 as it
used to be?
Thanks
Remi
Selon Vivek Goyal <[email protected]>:
> Hi Remi,
>
> On Mon, 2004-10-25 at 03:30, Remi Colinet wrote:
>
> > Sections:
> > Idx Name Size VMA LMA File off Algn
> > 0 .text 0037d80e c0100000 00100000 00001000 2**12
> > <-- something wrong here for LMA used to be 0xc010 000
>
>
> This has been changed so that LMA (Load memory address) reflects the
> physical address where sections shall be loaded. This is required to
> make sure that a kexec boot can load a vmlinux image.
Does it mean that the .text LMA start address is going to be definitively 0010
0000, even for those who disable kexec in their .config file?
>
>
>
> > CONTENTS, ALLOC, LOAD, DATA
> > 26 .bss 0006dfa0 c06c6000 c06c6000 005c7000 2**12 <--
> > somthing wrong with LMA for .bss.
> > CONTENTS, ALLOC, LOAD, DATA
> > 27 .comment 0000c36f 00000000 00000000 00634fa0 2**0
> > CONTENTS, READONLY
> > 28 .debug_line 0021497b 00000000 00000000 0064130f 2**0
> > ...
> >
> > This is very strange. The .text LMA of the vmlinux file is at 0x__0__010
> > 0000 whereas, it used to be at 0x__c__010 0000. But then, the .bss LMA
> > of the vmlinux file suddenly jump at 0x__c__06c 6000 instead of
> > 0x__0__06c 6000 which would have seemed more natural -0). What can
> > cause this offset? I have look for the .bss section content. It contains
> > .bss data and .bss.page_aligned. The later seems to be the root of the
> > .bss LMA offset. .bss.page_aligned contains then empty_zero_page and
> > the swapper_pg_dir.
>
> This looks like a problem with the older binutils package. I had faced
> similar problem on one of the machines but it was resolved as soon as I
> switched to a newer binutils package.
My distro is effectively becoming old (FC1). Going to upgrade it.
Thanks
Remi
Hi Remi,
On Mon, 2004-10-25 at 03:30, Remi Colinet wrote:
> Sections:
> Idx Name Size VMA LMA File off Algn
> 0 .text 0037d80e c0100000 00100000 00001000 2**12
> <-- something wrong here for LMA used to be 0xc010 000
This has been changed so that LMA (Load memory address) reflects the
physical address where sections shall be loaded. This is required to
make sure that a kexec boot can load a vmlinux image.
> CONTENTS, ALLOC, LOAD, DATA
> 26 .bss 0006dfa0 c06c6000 c06c6000 005c7000 2**12 <--
> somthing wrong with LMA for .bss.
> CONTENTS, ALLOC, LOAD, DATA
> 27 .comment 0000c36f 00000000 00000000 00634fa0 2**0
> CONTENTS, READONLY
> 28 .debug_line 0021497b 00000000 00000000 0064130f 2**0
> ...
>
> This is very strange. The .text LMA of the vmlinux file is at 0x__0__010
> 0000 whereas, it used to be at 0x__c__010 0000. But then, the .bss LMA
> of the vmlinux file suddenly jump at 0x__c__06c 6000 instead of
> 0x__0__06c 6000 which would have seemed more natural -0). What can
> cause this offset? I have look for the .bss section content. It contains
> .bss data and .bss.page_aligned. The later seems to be the root of the
> .bss LMA offset. .bss.page_aligned contains then empty_zero_page and
> the swapper_pg_dir.
This looks like a problem with the older binutils package. I had faced
similar problem on one of the machines but it was resolved as soon as I
switched to a newer binutils package.
> I finally succeded to compile the kernel, using the following
> OBJCOPYFLAGS in arch/i386/Makefile.
>
> OBJCOPYFLAGS := -O binary --change-section-lma .bss-0xc0000000 -R
> .note -R .comment -S
>
> I haven't yet tried to boot it.
>
> Any idea to fix this objcopy error otherwise?
Switching to latest binutils package should help.
> Why is the .text LMA set at 0x0010 0000 instead of 0xc010 0000 as it
> used to be?
To reflect the physical address where .text section shall be loaded. As
mentioned earlier, vmlinux booting using kexec requires this.
Thanks
Vivek
Hi Remi,
On Mon, 2004-10-25 at 14:45, [email protected] wrote:
> Does it mean that the .text LMA start address is going to be definitively 0010
> 0000, even for those who disable kexec in their .config file?
Yes.
> >
> > This looks like a problem with the older binutils package. I had faced
> > similar problem on one of the machines but it was resolved as soon as I
> > switched to a newer binutils package.
>
> My distro is effectively becoming old (FC1). Going to upgrade it.
We are trying to find out why older versions of binutils create this
problem. Also looking into if compatibility can be ascertained between
this new patch and older version of binutils.
Thanks
Vivek
>
> Thanks
> Remi
>
>
>