Re: [RFC v3 04/23] sysdeps/clock_gettime: Use clock_gettime64 if avaliable

[Lukasz Majewski <lukma at denx dot de>]

Hi Alistair,

> Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
> ---
>  nptl/pthread_mutex_timedlock.c          |  7 ++++++
>  sysdeps/unix/sysv/linux/clock_gettime.c | 30
> +++++++++++++++++++++++++ 2 files changed, 37 insertions(+)
> 
> diff --git a/nptl/pthread_mutex_timedlock.c
> b/nptl/pthread_mutex_timedlock.c index 52c258e33d..8d9cae7d87 100644
> --- a/nptl/pthread_mutex_timedlock.c
> +++ b/nptl/pthread_mutex_timedlock.c
> @@ -402,10 +402,17 @@ __pthread_mutex_clocklock_common
> (pthread_mutex_t *mutex, /* Delay the thread until the timeout is
> reached. Then return ETIMEDOUT.  */
>  		    struct timespec reltime;
> +#ifdef __NR_clock_gettime64
> +		    struct __timespec64 now;
> +
> +		    INTERNAL_SYSCALL (clock_gettime64, __err, 2,
> CLOCK_REALTIME,
> +				      &now);
> +#else
>  		    struct timespec now;
>  
>  		    INTERNAL_SYSCALL (clock_gettime, __err, 2,
> clockid, &now);
> +#endif
>  		    reltime.tv_sec = abstime->tv_sec - now.tv_sec;
>  		    reltime.tv_nsec = abstime->tv_nsec - now.tv_nsec;
>  		    if (reltime.tv_nsec < 0)
> diff --git a/sysdeps/unix/sysv/linux/clock_gettime.c
> b/sysdeps/unix/sysv/linux/clock_gettime.c index
> 5fc47fb7dc..4832069c34 100644 ---
> a/sysdeps/unix/sysv/linux/clock_gettime.c +++
> b/sysdeps/unix/sysv/linux/clock_gettime.c @@ -27,10 +27,40 @@
>  #include <sysdep-vdso.h>
>  
>  /* Get current value of CLOCK and store it in TP.  */
> +
> +#if __WORDSIZE == 32
> +int
> +__clock_gettime (clockid_t clock_id, struct timespec *tp)
> +{
> +   int ret;
> +
> +#ifdef __NR_clock_gettime64
> +  struct __timespec64 tp64;
> +  ret = INLINE_VSYSCALL (clock_gettime64, 2, clock_id, &tp64);
> +
> +  tp->tv_sec = tp64.tv_sec;
> +  tp->tv_nsec = tp64.tv_nsec;
> +
> +  if (! in_time_t_range (tp->tv_sec))
> +    {
> +      __set_errno (EOVERFLOW);
> +      return -1;
> +    }
> +#endif
> +
> +#ifndef __ASSUME_TIME64_SYSCALLS

I think that the context of using the __ASSUME_TIME64_SYSCALLS doesn't
comply with the semantics which was proposed in its introduction patch
[1].

In short:

This define means that the system is supporting 64 bit time, not that
it has (can use) introduced in Linux 5.1 64 bit time related syscalls
(i.e. clock_gettime64, clock_settime64, clock_nanosleep64).

Maybe you could consider using the "conversion" approach as proposed
for __clock_settime64 conversion [2]?

The version in [2] is more "developer/validator" friendly (as suggested
by Arnd) as it uses for __TIMESIZE != 64 archs the 32 bit clock_settime.

The version from [3] is the one recommended by Joseph (which does the
32 to 64 bit conversion and calls clock_settime64).



The full code for a reference with clock_settime conversion and Y2038
support [4].

> +  ret = INLINE_VSYSCALL (clock_gettime, 2, clock_id, tp);
> +#endif
> +
> +  return ret;
> +}
> +#else
>  int
>  __clock_gettime (clockid_t clock_id, struct timespec *tp)
>  {
>    return INLINE_VSYSCALL (clock_gettime, 2, clock_id, tp);
>  }
> +#endif
> +
>  weak_alias (__clock_gettime, clock_gettime)
>  libc_hidden_def (__clock_gettime)

Note:

[1] -
https://github.com/lmajewski/y2038_glibc/commit/1fdbc6002101a78a8a6a076bbb642b3082c2225d

[2] -
https://github.com/lmajewski/y2038_glibc/commit/69f842a8519ca13ed11fab0ff1bcc6fa1a524192

[3] -
https://github.com/lmajewski/y2038_glibc/commit/fa0f5ff6c942beca383daeff3d48829829ace5b1

[4] -

https://github.com/lmajewski/y2038_glibc/commits/Y2038-2.29-glibc-__clock-internal-struct-timespec-v6

Best regards,

Lukasz Majewski

--

DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-59 Fax: (+49)-8142-66989-80 Email: lukma@denx.de

Attachment: pgpNQ64uKO1Cd.pgp
Description: OpenPGP digital signature