o Runnable threads are now maintained in priority queues. The

    implementation requires two things:

      1.) The priority queues must be protected during insertion
          and removal of threads.  Since the kernel scheduler
          must modify the priority queues, a spinlock for
          protection cannot be used.   The functions
          _thread_kern_sched_defer() and _thread_kern_sched_undefer()
          were added to {un}defer kernel scheduler activation.

      2.) A thread (active) priority change can be performed only
          when the thread is removed from the priority queue.  The
          implementation uses a threads active priority when
          inserting it into the queue.

    A by-product is that thread switches are much faster.  A
    separate queue is used for waiting and/or blocked threads,
    and it is searched at most 2 times in the kernel scheduler
    when there are active threads.  It should be possible to
    reduce this to once by combining polling of threads waiting
    on I/O with the loop that looks for timed out threads and
    the minimum timeout value.

  o Functions to defer kernel scheduler activation were added.  These
    are _thread_kern_sched_defer() and _thread_kern_sched_undefer()
    and may be called recursively.  These routines do not block the
    scheduling signal, but latch its occurrence.  The signal handler
    will not call the kernel scheduler when the running thread has
    deferred scheduling, but it will be called when running thread
    undefers scheduling.

  o Added support for _POSIX_THREAD_PRIORITY_SCHEDULING.  All the
    POSIX routines required by this should now be implemented.
    One note, SCHED_OTHER, SCHED_FIFO, and SCHED_RR are required
    to be defined by including pthread.h.  These defines are currently
    in sched.h.  I modified pthread.h to include sched.h but don't
    know if this is the proper thing to do.

  o Added support for priority protection and inheritence mutexes.
    This allows definition of _POSIX_THREAD_PRIO_PROTECT and
    _POSIX_THREAD_PRIO_INHERIT.

  o Added additional error checks required by POSIX for mutexes and
    condition variables.

  o Provided a wrapper for sigpending which is marked as a hidden
    syscall.

  o Added a non-portable function as a debugging aid to allow an
    application to monitor thread context switches.  An application
    can install a routine that gets called everytime a thread
    (explicitly created by the application) gets context switched.
    The routine gets passed the pthread IDs of the threads that are
    being switched in and out.  I found this useful, but we can
    get rid of it if you want.

Submitted by: Dan Eischen <eischen@vigrid.com>

1 files changed, 71 insertions, 36 deletions

diff --git a/include/pthread.h b/include/pthread.h
index dc8ebf08b172..d5881827d900 100644
--- a/include/pthread.h
+++ b/include/pthread.h
@@ -42,6 +42,7 @@
 #include <sys/time.h>
 #include <sys/signal.h>
 #include <limits.h>
+#include <sched.h>
 
 /*
  * Run-time invariant values:
@@ -60,9 +61,9 @@
 #define _POSIX_THREADS
 #define _POSIX_THREAD_ATTR_STACKADDR
 #define _POSIX_THREAD_ATTR_STACKSIZE
-/* #define _POSIX_THREAD_PRIORITY_SCHEDULING */
-/* #define _POSIX_THREAD_PRIO_INHERIT   */
-/* #define _POSIX_THREAD_PRIO_PROTECT   */
+#define _POSIX_THREAD_PRIORITY_SCHEDULING
+#define _POSIX_THREAD_PRIO_INHERIT
+#define _POSIX_THREAD_PRIO_PROTECT
 /* #define _POSIX_THREAD_PROCESS_SHARED */
 #define _POSIX_THREAD_SAFE_FUNCTIONS
 
@@ -99,7 +100,6 @@ struct pthread_mutex_attr;
 struct pthread_once;
 struct pthread_rwlock;
 struct pthread_rwlockattr;
-struct sched_param;
 
 /*
  * Primitive system data type definitions required by P1003.1c.
@@ -163,15 +163,33 @@ struct pthread_once {
 #define pthread_attr_default        NULL
 #endif
 
+#define PTHREAD_PRIO_NONE	0
+#ifdef _POSIX_THREAD_PRIO_PROTECT
+#define PTHREAD_PRIO_INHERIT	1
+#define PTHREAD_PRIO_PROTECT	2
+#endif
+
+/*
+ * Mutex types (Single UNIX Specification, Version 2, 1997).
+ *
+ * Note that a mutex attribute with one of the following types:
+ *
+ *	PTHREAD_MUTEX_NORMAL
+ *	PTHREAD_MUTEX_RECURSIVE
+ *      MUTEX_TYPE_FAST (deprecated)
+ *	MUTEX_TYPE_COUNTING_FAST (deprecated)
+ *
+ * will deviate from POSIX specified semantics.
+ */
 enum pthread_mutextype {
-	PTHREAD_MUTEX_DEFAULT		= 1,
-	PTHREAD_MUTEX_RECURSIVE		= 2,
-	PTHREAD_MUTEX_NORMAL		= 3,
-	PTHREAD_MUTEX_ERRORCHECK	= 4,
+	PTHREAD_MUTEX_ERRORCHECK	= 1,	/* Default POSIX mutex */
+	PTHREAD_MUTEX_RECURSIVE		= 2,	/* Recursive mutex */
+	PTHREAD_MUTEX_NORMAL		= 3,	/* No error checking */
 	MUTEX_TYPE_MAX
 };
 
-#define MUTEX_TYPE_FAST			PTHREAD_MUTEX_DEFAULT
+#define PTHREAD_MUTEX_DEFAULT		PTHREAD_MUTEX_ERRORCHECK
+#define MUTEX_TYPE_FAST			PTHREAD_MUTEX_NORMAL
 #define MUTEX_TYPE_COUNTING_FAST	PTHREAD_MUTEX_RECURSIVE
 
 /*
@@ -179,20 +197,10 @@ enum pthread_mutextype {
  */
 __BEGIN_DECLS
 int		pthread_attr_destroy __P((pthread_attr_t *));
-int		pthread_attr_getinheritsched __P((pthread_attr_t *, int *));
-int		pthread_attr_getschedparam __P((pthread_attr_t *,
-			struct sched_param *));
-int		pthread_attr_getschedpolicy __P((pthread_attr_t *, int *));
-int		pthread_attr_getscope __P((pthread_attr_t *, int *));
 int		pthread_attr_getstacksize __P((pthread_attr_t *, size_t *));
 int		pthread_attr_getstackaddr __P((pthread_attr_t *, void **));
 int		pthread_attr_getdetachstate __P((pthread_attr_t *, int *));
 int		pthread_attr_init __P((pthread_attr_t *));
-int		pthread_attr_setinheritsched __P((pthread_attr_t *, int));
-int		pthread_attr_setschedparam __P((pthread_attr_t *,
-			struct sched_param *));
-int		pthread_attr_setschedpolicy __P((pthread_attr_t *, int));
-int		pthread_attr_setscope __P((pthread_attr_t *, int));
 int		pthread_attr_setstacksize __P((pthread_attr_t *, size_t));
 int		pthread_attr_setstackaddr __P((pthread_attr_t *, void *));
 int		pthread_attr_setdetachstate __P((pthread_attr_t *, int));
@@ -201,10 +209,14 @@ void		pthread_cleanup_push __P((void (*routine) (void *),
 			void *routine_arg));
 int		pthread_condattr_destroy __P((pthread_condattr_t *attr));
 int		pthread_condattr_init __P((pthread_condattr_t *attr));
+
+#if defined(_POSIX_THREAD_PROCESS_SHARED)
 int		pthread_condattr_getpshared __P((pthread_condattr_t *attr,
 			int *pshared));
 int		pthread_condattr_setpshared __P((pthread_condattr_t *attr,
 			int pshared));
+#endif
+
 int		pthread_cond_broadcast __P((pthread_cond_t *));
 int		pthread_cond_destroy __P((pthread_cond_t *));
 int		pthread_cond_init __P((pthread_cond_t *,
@@ -224,27 +236,13 @@ int		pthread_key_create __P((pthread_key_t *,
 			void (*routine) (void *)));
 int		pthread_key_delete __P((pthread_key_t));
 int		pthread_kill __P((struct pthread *, int));
-int		pthread_mutexattr_destroy __P((pthread_mutexattr_t *));
-int		pthread_mutexattr_getprioceiling __P((pthread_mutexattr_t *,
-			int *prioceiling));
-int		pthread_mutexattr_getprotocol __P((pthread_mutexattr_t *,
-			int *protocol));
-int		pthread_mutexattr_getpshared __P((pthread_mutexattr_t *,
-			int *pshared));
 int		pthread_mutexattr_init __P((pthread_mutexattr_t *));
-int		pthread_mutexattr_setprioceiling __P((pthread_mutexattr_t *,
-			int prioceiling));
-int		pthread_mutexattr_setprotocol __P((pthread_mutexattr_t *,
-			int protocol));
-int		pthread_mutexattr_setpshared __P((pthread_mutexattr_t *,
-			int pshared));
+int		pthread_mutexattr_destroy __P((pthread_mutexattr_t *));
 int		pthread_mutexattr_settype __P((pthread_mutexattr_t *, int));
 int		pthread_mutex_destroy __P((pthread_mutex_t *));
-int		pthread_mutex_getprioceiling __P((pthread_mutex_t *));
 int		pthread_mutex_init __P((pthread_mutex_t *,
 			const pthread_mutexattr_t *));
 int		pthread_mutex_lock __P((pthread_mutex_t *));
-int		pthread_mutex_setprioceiling __P((pthread_mutex_t *));
 int		pthread_mutex_trylock __P((pthread_mutex_t *));
 int		pthread_mutex_unlock __P((pthread_mutex_t *));
 int		pthread_once __P((pthread_once_t *,
@@ -274,10 +272,47 @@ int		pthread_testcancel __P((void));
 int		pthread_getprio __P((pthread_t));
 int		pthread_setprio __P((pthread_t, int));
 void		pthread_yield __P((void));
-int		pthread_setschedparam __P((pthread_t pthread, int policy,
-			struct sched_param * param));
+
+#if defined(_POSIX_THREAD_PROCESS_SHARED)
+int		pthread_mutexattr_getpshared __P((pthread_mutexattr_t *,
+			int *pshared));
+int		pthread_mutexattr_setpshared __P((pthread_mutexattr_t *,
+			int pshared));
+#endif
+
+#if defined(_POSIX_THREAD_PRIO_PROTECT)
+int		pthread_mutexattr_getprioceiling __P((pthread_mutexattr_t *,
+			int *prioceiling));
+int		pthread_mutexattr_setprioceiling __P((pthread_mutexattr_t *,
+			int prioceiling));
+int		pthread_mutex_getprioceiling __P((pthread_mutex_t *, int *));
+int		pthread_mutex_setprioceiling __P((pthread_mutex_t *, int, int *));
+#endif
+
+#if defined(_POSIX_THREAD_PRIO_PROTECT) || defined (_POSIX_THREAD_PRIO_INHERIT)
+int		pthread_mutexattr_getprotocol __P((pthread_mutexattr_t *,
+			int *protocol));
+int		pthread_mutexattr_setprotocol __P((pthread_mutexattr_t *,
+			int protocol));
+#endif
+
+#if defined(_POSIX_THREAD_PRIORITY_SCHEDULING)
+int		pthread_attr_getinheritsched __P((pthread_attr_t *, int *));
+int		pthread_attr_getschedparam __P((pthread_attr_t *,
+			struct sched_param *));
+int		pthread_attr_getschedpolicy __P((pthread_attr_t *, int *));
+int		pthread_attr_getscope __P((pthread_attr_t *, int *));
+int		pthread_attr_setinheritsched __P((pthread_attr_t *, int));
+int		pthread_attr_setschedparam __P((pthread_attr_t *,
+			struct sched_param *));
+int		pthread_attr_setschedpolicy __P((pthread_attr_t *, int));
+int		pthread_attr_setscope __P((pthread_attr_t *, int));
 int		pthread_getschedparam __P((pthread_t pthread, int *policy,
 			struct sched_param * param));
+int		pthread_setschedparam __P((pthread_t pthread, int policy,
+			struct sched_param * param));
+#endif
+
 int		pthread_attr_setfloatstate __P((pthread_attr_t *, int));
 int		pthread_attr_getfloatstate __P((pthread_attr_t *, int *));
 int		pthread_attr_setcleanup __P((pthread_attr_t *,