GCC 4.9.2 Build Fails at cc1

[Cyd Haselton]

On Thu, Dec 4, 2014 at 11:32 AM, Andrew Haley <aph@redhat.com> wrote:
> On 12/04/2014 05:10 PM, Cyd Haselton wrote:
>> On Thu, Dec 4, 2014 at 10:51 AM, Andrew Haley <aph@redhat.com> wrote:
>>> On 12/04/2014 03:41 PM, Cyd Haselton wrote:
>>>>> Right, so that's your problem.  You need to look at that library to
>>>>>> see if there are any versions of mpn_tdiv_qr, perhaps with different
>>>>>> prefixes to the names.
>>>>>>
>>>>>> Andrew.
>>>>>>
>>>> Two follow up questions:
>>>> 1.  If there are differently-prefixed versions in the library, what
>>>> would I do to allow the make to find them and resolve the original
>>>> problem?
>>>> 2.  Shouldn't the build of libgmp.a include these symbols?
>>>
>>> Let's see what's going on first.  Is there anything in that library
>>> which matches mpn_tdiv_qr ?
>>
>> ...and is not prefixed with "U"?  Unfortunately no.  Grepping for any
>> permutation of mpn_tdiv_qr outputs the same results as previously
>> posted.
>
> Okay, so you must now find out why.  What is in that library?
> Why is this not defined?  You must debug the build.
>
> Andrew.
>
>

Back again after parsing through code from the previous errors.  With
a question.

In gcc-4.9.2/mpfr/div.c:324 mpfr_div references mpn_divrem...but the
error thrown by make references gmpn_divrem.
In gcc-4.9.2/mpfr/div.c:284 mpfr_div references mpn_rshift...but the
error thrown by make references gmpn_rshift.'

Same thing occurs with the undefined references in mpfr_add1 from
gcc-4.9.2/mpfr/add1.c

Why is make complaining about those functions when they aren't being
referenced? Are the functions supposed to be transformed in some way?