[bug] gcc-3.4-20050422 compiling glibc-2.3.5 internal compiler error in libm-test.c:ctanh_test()

[Clemens Koller]

Hi, Andrew!

 > Not really, I would now stop for a minute and check your memory
 > and your hardware since this is not reproducible.

Hmm, okay. It's an embedded system without any case and
proper shielding - so some hicups might be possible.

I will see, I am doing several compile tests like that
to verify/learn more about building toolchains...

Thanks anyway,

Clemens


Andrew Pinski wrote:
> 
> On Apr 27, 2005, at 10:19 AM, Clemens Koller wrote:
> 
>> ...and it's not reproducable yet.
>> On a second try the compile was fine and test-idouble just works...
>>
>> ~/newbuild/glibc-2.3.5-build/math$ ./test-idouble
>> testing double (inline functions)
>>
>> Test suite completed:
>>   2562 test cases plus 2337 tests for exception flags executed.
>>   All tests passed successfully.
>>
>> Uargh... great, isn't it?
> 
> 
> Not really, I would now stop for a minute and check your memory
> and your hardware since this is not reproducible.
> 
> -- Pinski
> 
>