[PATCH 1/6] server: Use monotonic clock for relative timeouts.

[Piotr Caban]

Signed-off-by: Piotr Caban 
---
Relative timeouts are expected to wait given amount of time even if 
system clock is changed. It's also behaving incorrectly on daylight 
saving related time change.

server/fd.c | 66 +++++++++++++++++++++++++++++++++++++++++++++--------
  1 file changed, 57 insertions(+), 9 deletions(-)

[Alexandre Julliard]

Piotr Caban  writes:
> +static inline timeout_t monotonic_counter(void)
> +{
> +#ifdef __APPLE__
> - static mach_timebase_info_data_t timebase;
> -
> - if (!timebase.denom) mach_timebase_info( &timebase );
>
> +#ifdef HAVE_MACH_CONTINUOUS_TIME
> - if (&mach_continuous_time != NULL)
> -
>    return mach_continuous_time() * timebase.numer / timebase.denom / 1000000;
>
> +#endif
> - return mach_absolute_time() * timebase.numer / timebase.denom / 1000000;
>
> +#elif defined(HAVE_CLOCK_GETTIME)
> - struct timespec ts;
>
> +#ifdef CLOCK_MONOTONIC_RAW
> - if (!clock_gettime( CLOCK_MONOTONIC_RAW, &ts ))
> -
>    return (timeout_t)ts.tv_sec * TICKS_PER_SEC + ts.tv_nsec / 100;
>
> +#endif
> - if (!clock_gettime( CLOCK_MONOTONIC, &ts ))
> -
>    return (timeout_t)ts.tv_sec * TICKS_PER_SEC + ts.tv_nsec / 100;
>
> +#endif
> - return current_time - server_start_time;
>
> +}

We have almost the same function in request.c, I don't think we want two
of these. Also you didn't change the scaling for the Mac case.