Re: 2.6.18-mm1 compile failure on x86_64

[Andy Whitcroft <apw@shadowen.org>]

Ian Campbell wrote:
> On Tue, 2006-09-26 at 01:29 -0700, Jeremy Fitzhardinge wrote:
>> Andy Whitcroft wrote:
>>> Martin Bligh wrote:
>>>   
>>>> http://test.kernel.org/abat/49037/debug/test.log.0   
>>>>
>>>>   AS      arch/x86_64/boot/bootsect.o
>>>>   LD      arch/x86_64/boot/bootsect
>>>>   AS      arch/x86_64/boot/setup.o
>>>>   LD      arch/x86_64/boot/setup
>>>>   AS      arch/x86_64/boot/compressed/head.o
>>>>   CC      arch/x86_64/boot/compressed/misc.o
>>>>   OBJCOPY arch/x86_64/boot/compressed/vmlinux.bin
>>>> BFD: Warning: Writing section `.data.percpu' to huge (ie negative) file
>>>> offset 0x804700c0.
>>>> /usr/local/autobench/sources/x86_64-cross/gcc-3.4.0-glibc-2.3.2/bin/x86_64-unknown-linux-gnu-objcopy:
>>>> arch/x86_64/boot/compressed/vmlinux.bin: File truncated
>>>> make[2]: *** [arch/x86_64/boot/compressed/vmlinux.bin] Error 1
>>>> make[1]: *** [arch/x86_64/boot/compressed/vmlinux] Error 2
>>>> make: *** [bzImage] Error 2
>>>> 09/25/06-09:13:48 Build the kernel. Failed rc = 2
>>>> 09/25/06-09:13:49 build: kernel build Failed rc = 1
>>>>
>>>> Wierd. Same box compiled 2.6.18 fine.
>>>>     
>>> Pretty sure this isn't a space problem, as we have just checked space
>>> before the build and I've taken no action since then.  Someone did
>>> mention "tool chain issue" when it was first spotted.  Will check with
>>> them and see why they thought that.
>>>   
>> Does this box have an older version of binutils (2.15?)?  If so, it 
>> might be getting upset over the patch "note-section" in Andi's queue.  I 
>> know it has been a bit problematic, but I don't know if the problems 
>> manifest in this way.
> 
> I've not seen it manifest like this but it would be worth trying Jan's
> patch from
> http://lists.xensource.com/archives/html/xen-devel/2006-08/msg01416.html
> to see if it helps.
> 
> Andi removed an identical patch (from someone else) from his queue due
> to http://marc.theaimsgroup.com/?l=linux-kernel&m=115629369729911&w=2 We
> have had the patch in the Xen tree for a couple of weeks now with no
> reported problems.

Ok.  To confirm, this seems to be a tool chain age issue.  This seems to
be triggered by the, x86_64-mm-note-section patch.  Reverting that patch
and installing the replacement as at the above URL seems to get us a
built kernel.

Not that it works yet ... but one step forward.

-apw