RE: vmlinux parse error 2.6.12.5

["Dave Korn" <dave dot korn at artimi dot com>]

----Original Message----
>From: Nicolas Pitre
>Sent: 25 August 2005 18:14

> On Thu, 25 Aug 2005, Dave Korn wrote:
> 
>>   I'm not sure yet how to test for the section size being zero, but I
>> just don't think ASSERT is up to the job.  It may or may not still be
>> possible in the linker script, or it may need to be done in the external
>> build machinery (e.g. objcopy out the section to a raw binary and test
>> the file size, or similary).
> 
> I don't know about your setup but ...

  Kernel 2.6.12.5, Binutils 2.15, Gcc 3.3.3, armcore target.

> So in this case everything works as expected for me at least.  Of course
> uncommenting the machine record in arch/arm/mach-pxa/mainstone.c results
> in a compiling again kernel.  

  That is very strange then.  I've been using binutils 2.15 all along.  I
didn't see this problem building 2.6.7, but it appeared when I tried moving
up to 2.6.12.5, without changing the toolchain in any way at all.

  Nonetheless, I still don't see _how_ it could possibly work, given the
known facts about the order in which ld does things.

> And given the time the ASSERT has been
> present in the kernel tree 

  How long _has_ the ASSERT been present?  It wasn't in 2.6.7.  As far as I
can see (assuming the file dates at
http://www.kernel.org/pub/linux/kernel/v2.6/ are actually correct and
more-or-less when the versions were released), that implies that it cannot
have been there any earlier than june last year.  It may have only been
present for a few versions and then be working again in 2.6.13.x

> it certainly must have worked fine for most
> people (and for those who had the problem then moving to binutils >=
> 2.15 seemed to have done some good too).

  Hmm, still hasn't helped me any.  Very strange.

    cheers,
      DaveK
-- 
Can't think of a witty .sigline today....