Bruce Dubbs wrote:
> hauradou wrote:
>
>> hauradou wrote:
>>
>>> Bruce Dubbs wrote:
>>>
>>>> hauradou wrote:
>>>>
>>>>> Is it truely a bug, as stated in the dev book?
>>>>>
>>>>> After building gcc-4.5.0, found a macro: __i686, expanding to 1
>>>>>
>>>>> glibc built fine in this chapter after adding 'asm-CPPFLAGS += -U
>>>>> __i686' to configparms, without using any patch.
>>>>>
>>>> You don't say what your platform is. The problem only shows up on a
>>>> 32-bit system. It builds fine without the patch on a 64-bit system.
>>>>
> Looking at the macro you found, I don't think it is doing exactly the
> same thing as the patch. You are undefining __i686, but we want to make
> the definition of __i686 to be itself, not a non-existant define.
>
> It looks like your method is allowing the system to build, but with less
> optimal code than what the developers intended.
>
> I suspect that it doesn't make a practical difference.
>
> -- Bruce
>
well, the difference between a big patch and just one option to pass to
the compiler...
Is upstream aware of this "bug"?
regards,
boloco
--
http://linuxfromscratch.org/mailman/listinfo/lfs-dev
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page
