> The timespec check is used in different parts internally, I wonder > + it can assume that abstime is not NULL. > +_pthread_cond_clockwait (pthread_cond_t *cond, pthread_mutex_t *mutex, > + return _pthread_cond_wait_common (cond, mutex, clockid, abstime) > + clockid_t clockid = (flags & _PTHREAD_COND_CLOCK_MONOTONIC_MASK) > + unsigned int flags = atomic_load_relaxed (&cond->_data._wrefs) > + /* Relaxed MO is suffice because clock ID bit is only modified

> - return _pthread_cond_wait_common (cond, mutex, abstime) > it can assume that abstime is not NULL. > -664,10 +671,40 _pthread_cond_timedwait (pthread_cond_t *cond, pthread_mutex_t *mutex, > diff -git a/nptl/pthread_cond_wait.c b/nptl/pthread_cond_wait.c > 13 files changed, 314 insertions(+), 14 deletions(-) > sysdeps/unix/sysv/linux/x86_64/64/libpthread.abilist | 1 +. > sysdeps/nptl/pthread-functions.h | 4 +. > * manual/threads.texi: Document pthread_cond_clockwait. > sysdeps/unix/sysv/linux/x86_64/64/libpthread.abilist: Add > sysdeps/unix/sysv/linux/i386/libpthread.abilist, > * sysdeps/unix/sysv/linux/arm/libpthread.abilist, > * nptl/tst-cond27.c: Add test similar to tst-cond5.c, but using struct > * ntpl/tst-cond26.c: New test for passing unsupported and invalid clocks to > combinations of CLOCK_MONOTONIC and CLOCK_REALTIME. > CLOCK_USE_ATTR_CLOCK for existing tests, and add new tests using all

> to call pthread_cond_timedwait or pthread_cond_clockwait. > by using a special magic CLOCK_USE_ATTR_CLOCK value to determine whether > * nptl/tst-cond11.c (run_test): Support testing pthread_cond_clockwait too > with weak alias from pthread_cond_clockwait. (_pthread_cond_clockwait): Add new function > (_pthread_cond_timedwait): Read clock from flags and pass it to > (_pthread_cond_wait): Pass unused clockid parameter. > Use that value rather than reading the clock from the flags. > parameter and comment describing why we don't need to check its value. > * nptl/pthread_cond_wait.c (_pthread_cond_wait_common): Add clockid

> * sysdeps/nptl/pthread-functions.h: Likewise. > _USE_GNU while it's still only proposed for POSIX.) > * nptl/forward.c: Add _pthread_cond_clockwait (not sure if it should be > * nptl/Versions (GLIBC_2.30): Add pthread_cond_clockwait > * nptl/Makefile: Add tst-cond26 and tst-cond27 > proposed as pthread_cond_timedwaitonclock_np, but The Austin Group > CLOCK_MONOTONIC and returns EINVAL if any other clock is specified. > which behaves just like pthread_cond_timedwait except it always measures > int pthread_cond_clockwait (pthread_cond_t *cond, On Tuesday 05 March 2019 at 13:45:13 -0300, Adhemerval Zanella wrote: Re: nptl: Add POSIX-proposed pthread_cond_clockwait - Mike Crowe public inbox for help / color / mirror / Atom feed From: Mike Crowe Ĭc: Re: nptl: Add POSIX-proposed pthread_cond_clockwaitĭate: Sat, 20:23:00 -0000

Mac snip