[PATCH 2/2] time: Add 64-bit time_t support for ftime

[Adhemerval Zanella]

On 27/10/2020 14:38, Joseph Myers wrote:
> The build of the 64-bit ftime appears broken for Hurd:
> 
> ftime.c: In function '__ftime64':
> ftime.c:27:3: error: implicit declaration of function '__clock_gettime64' [-Werror=implicit-function-declaration]
>    27 |   __clock_gettime64 (CLOCK_REALTIME, &ts);
>       |   ^~~~~~~~~~~~~~~~~
> ftime.c:27:22: error: 'CLOCK_REALTIME' undeclared (first use in this function)
>    27 |   __clock_gettime64 (CLOCK_REALTIME, &ts);
>       |                      ^~~~~~~~~~~~~~
> ftime.c:27:22: note: each undeclared identifier is reported only once for each function it appears in
> ftime.c: In function 'ftime':
> ftime.c:43:9: error: implicit declaration of function 'in_time_t_range'; did you mean 'in_ino_t_range'? [-Werror=implicit-function-declaration]
>    43 |   if (! in_time_t_range (tb64.time))
>       |         ^~~~~~~~~~~~~~~
>       |         in_ino_t_range
> 
> https://sourceware.org/pipermail/libc-testresults/2020q4/006962.html
> 

I think we will need to make the y2038 ftime support Linux specific
for now, I will fix it.