Work around GCC 10 warning regression in string/tester.c

[Jeff Law]

On 10/2/19 6:41 AM, Florian Weimer wrote:
> * Joseph Myers:
> 
>> Any comments on either the patch itself or the principle of working
>> around such issues in glibc to allow more effective testing if not
>> fixed quickly in GCC?
> 
> I think we should disable those middle-end warnings for the entire file
> altogether.
When y'all figure out how you want to do this, can you please pull the
fix into Fedora?  I was running the Fedora tester using the latest gcc
snapshot:

> BUILDSTDERR: tester.c: In function 'test_strncat':
> BUILDSTDERR: tester.c:395:10: error: 'strncat' forming offset [50, 98] is out of the bounds [0, 50] of object 'one' with type 'char[50]' [-Werror=array-bounds]
> BUILDSTDERR:   395 |   (void) strncat (one, two, 99);
> BUILDSTDERR:       |          ^~~~~~~~~~~~~~~~~~~~~~
> BUILDSTDERR: tester.c:61:6: note: 'one' declared here
> BUILDSTDERR:    61 | char one[50];
> BUILDSTDERR:       |      ^~~

So this going to affect Fedora when gcc-10 drops in.

jeff