wq_has_sleeper(9) [centos man page]

WQ_HAS_SLEEPER(9)						 Linux Networking						 WQ_HAS_SLEEPER(9)

NAME

       wq_has_sleeper - check if there are any waiting processes

SYNOPSIS

       bool wq_has_sleeper(struct socket_wq * wq);

ARGUMENTS

       wq
	   struct socket_wq

DESCRIPTION

       Returns true if socket_wq has waiting processes

       The purpose of the wq_has_sleeper and sock_poll_wait is to wrap the memory barrier call. They were added due to the race found within the
       tcp code.

CONSIDER FOLLOWING TCP CODE PATHS

       CPU1 CPU2

       sys_select receive packet ... ... __add_wait_queue update tp->rcv_nxt ... ... tp->rcv_nxt check sock_def_readable ... { schedule
       rcu_read_lock; wq = rcu_dereference(sk->sk_wq); if (wq && waitqueue_active(wq->wait)) wake_up_interruptible(wq->wait) ... }

       The race for tcp fires when the __add_wait_queue changes done by CPU1 stay in its cache, and so does the tp->rcv_nxt update on CPU2 side.
       The CPU1 could then endup calling schedule and sleep forever if there are no more data on the socket.

COPYRIGHT

Kernel Hackers Manual 3.10					     June 2014							 WQ_HAS_SLEEPER(9)

ptsematest(8)															     ptsematest(8)

NAME

       ptsematest - Start two threads and measure the latency of interprocess communication with POSIX mutex.

SYNTAX

       ptsematest [-a|-a PROC] [-b USEC] [-d DIST] [-i INTV] [-l loops] [-p PRIO] [-t|-t NUM]

DESCRIPTION

       The  program  ptsematest  starts two threads that are synchronized via pthread_mutex_unlock()/pthread_mutex_lock() and measures the latency
       between releasing and getting the lock.

OPTIONS

       -a, --affinity[=PROC]
	      Run on procesor number PROC. If PROC is not specified, run on current processor.

       -b, --breaktrace=USEC
	      Send break trace command when latency > USEC. This is a debugging option to control the latency tracer in  the  realtime	preemption
	      patch.  It is useful to track down unexpected large latencies of a system.

       -d, --distance=DIST
	      Set the distance of thread intervals in microseconds (default is 500 us). When  cylictest is called with the -t option and more than
	      one thread is created, then this distance value is added to the interval of the threads: Interval(thread N) = Interval(thread N-1) +
	      DIST

       -i, --interval=INTV
	      Set  the	base  interval of the thread(s) in microseconds (default is 1000 us). This sets the interval of the first thread. See also
	      -d.

       -l, --loops=LOOPS
	      Set the number of loops. The default is 0 (endless). This option is useful for automated tests with a given number of  test  cycles.
	      ptsematest is stopped once the number of timer intervals has been reached.

       -p, --prio=PRIO
	      Set the priority of the process.

       -t, --threads[=NUM]
	      Set  the number of test threads (default is 1, if this option is not given). If NUM is specified, create NUM test threads. If NUM is
	      not specifed, NUM is set to the number of available CPUs.

EXAMPLES

       The following example was running on a 4-way processor:

       # ptsematest -a -t -p99 -i100 -d25 -l1000000
       #0: ID8672, P99, CPU0, I100; #1: ID8673, P99, CPU0, Cycles 1000000
       #2: ID8674, P98, CPU1, I125; #3: ID8675, P98, CPU1, Cycles 811035
       #4: ID8676, P97, CPU2, I150; #5: ID8677, P97, CPU2, Cycles 668130
       #6: ID8678, P96, CPU3, I175; #7: ID8679, P96, CPU3, Cycles 589423
       #1 -> #0, Min	1, Cur	  1, Avg    2, Max   11
       #3 -> #2, Min	1, Cur	  2, Avg    2, Max   13
       #5 -> #4, Min	1, Cur	  4, Avg    3, Max   12
       #7 -> #6, Min	1, Cur	  4, Avg    2, Max   12

AUTHORS

       Carsten Emde <C.Emde@osadl.org>

SEE ALSO

       pthread_mutex_lock(3p), pthread_mutex_unlock(3p)

									0.1							     ptsematest(8)