2001-07-01 Ross Johnson <rpj@setup1.ise.canberra.edu.au>

        Contributed by  - Alexander Terekhov.

        * condvar.c: Fixed lost signal bug reported by Timur Aydin
        (taydin@snet.net).
        [RPJ (me) didn't translate the original algorithm
        correctly.]
        * semaphore.c: Added sem_post_multiple; this is a useful
        routine, but it doesn't appear to be standard. For now it's
        not an exported function.

1 files changed, 18 insertions, 221 deletions

diff --git a/sched.c b/sched.c
index 6f558ab..bc8aca0 100644
--- a/sched.c
+++ b/sched.c
@@ -91,11 +91,7 @@ pthread_attr_setschedparam(pthread_attr_t *attr,
 
   priority = param->sched_priority;
 
-  /*
-   * Validate priority level. Don't check the
-   * return values of the function calls because
-   * we're sure they will succeed.
-   */
+  /* Validate priority level. */
   if (priority < sched_get_priority_min(SCHED_OTHER) ||
       priority > sched_get_priority_max(SCHED_OTHER))
     {
@@ -109,7 +105,7 @@ pthread_attr_setschedparam(pthread_attr_t *attr,
 
 int 
 pthread_attr_getschedparam(const pthread_attr_t *attr,
-			       struct sched_param *param)
+                           struct sched_param *param)
 {
   if (is_attr(attr) != 0 || param == NULL)
     {
@@ -177,13 +173,9 @@ pthread_setschedparam(pthread_t thread, int policy,
       return ENOTSUP;
     }
 
-  /*
-   * Validate priority level. Don't check the
-   * return values of the function calls because
-   * we're sure they will succeed.
-   */
-  if (param->sched_priority < sched_get_priority_min(SCHED_OTHER) ||
-      param->sched_priority > sched_get_priority_max(SCHED_OTHER))
+  /* Validate priority level. */
+  if (param->sched_priority < sched_get_priority_min(policy) ||
+      param->sched_priority > sched_get_priority_max(policy))
     {
       return EINVAL;
     }
@@ -253,14 +245,7 @@ sched_get_priority_max(int policy)
 {
   if (policy < SCHED_MIN || policy > SCHED_MAX)
     {
-      errno = EINVAL;
-      return -1;
-    }
-
-  if (policy != SCHED_OTHER)
-    {
-      errno = ENOTSUP;
-      return -1;
+      return EINVAL;
     }
 
 #if (THREAD_PRIORITY_LOWEST > THREAD_PRIORITY_NORMAL)
@@ -278,14 +263,7 @@ sched_get_priority_min(int policy)
 {
   if (policy < SCHED_MIN || policy > SCHED_MAX)
     {
-      errno = EINVAL;
-      return -1;
-    }
-
-  if (policy != SCHED_OTHER)
-    {
-      errno = ENOTSUP;
-      return -1;
+      return EINVAL;
     }
 
 #if (THREAD_PRIORITY_LOWEST > THREAD_PRIORITY_NORMAL)
@@ -299,175 +277,31 @@ sched_get_priority_min(int policy)
 
 
 int
-sched_rr_get_interval(pid_t pid, struct timespec * interval)
-     /*
-      * ------------------------------------------------------
-      * DOCPUBLIC
-      *      This function updates the timespec structure
-      *      referenced by the interval argument to contain
-      *      the current quantum for the process executing
-      *      under the SCHED_RR policy. If a process, running under
-      *      the round-robin scheduling policy, runs without
-      *      blocking or yielding for more than this amount of
-      *      time, it may be preempted by another runnable
-      *      process (at the same priority).
-      *
-      *      If the PID argument is zero, the current execution
-      *      time limit for the calling process is returned.
-      *
-      * PARAMETERS
-      *      pid    Process identifier.
-      *
-      *
-      * DESCRIPTION
-      *      This function updates the timespec structure
-      *      referenced by the interval argument to contain
-      *      the current quantum for the process executing
-      *      under the SCHED_RR policy. If a process, running under
-      *      the round-robin scheduling policy, runs without
-      *      blocking or yielding for more than this amount of
-      *      time, it may be preempted by another runnable
-      *      process (at the same priority).
-      *
-      *      If the PID argument is zero, the current execution
-      *      time limit for the calling process is returned.
-      *
-      * RESULTS
-      *      We don't support SCHED_RR so this routine always fails.
-      *
-      *      NOTE: Since this is part of POSIX 1003.1b
-      *                (realtime extensions), it is defined as returning
-      *                -1 if an error occurs and sets errno to the actual
-      *                error.
-      *
-      *      errno:  ESRCH           Pid doesn't refer to a valid process.
-      *              EPERM           The process referenced by pid does
-      *                              not allow query access.
-      *              ENOTSUP         SCHED_RR is not supported.
-      *
-      * ------------------------------------------------------
-      */
-{
-  /*
-   * Provide other valid side-effects
-   * such as EPERM and ESRCH errors.
-   */
-  if (0 != pid)
-    {
-      DWORD selfPid = GetCurrentProcessId();
-
-      if (pid != selfPid)
-        {
-          HANDLE h = OpenProcess(PROCESS_QUERY_INFORMATION, FALSE, pid);
-
-          if (NULL == h)
-            {
-              errno = (GetLastError() == (0xFF & ERROR_ACCESS_DENIED)
-                       ? EPERM
-                       : ESRCH );
-
-              return -1;
-            }
-
-          (void) CloseHandle(h);
-        }
-    }
-
-  /*
-   * We can't return ENOSYS and stay strictly compliant with the
-   * standard. We don't support round-robin scheduling so
-   * we return ENOTSUP instead. This is consistent with
-   * routines which return ENOTSUP if SCHED_RR is given
-   * as the policy parameter.
-   */
-  errno = ENOTSUP;
-  return -1;
-}
-
-
-int
 sched_setscheduler(pid_t pid, int policy)
-     /*
-      * ------------------------------------------------------
-      * DOCPUBLIC
-      *      This function returns sets the scheduling policy and
-      *      scheduling parameters of the process specified by pid
-      *      to policy and the parameters specified in the
-      *      sched_param structure pointed to by param, respectively.
-      *      The value of the sched_priority member in the sched_param
-      *      structure is any integer within the inclusive priority
-      *      range for the scheduling policy specified by policy.
-      *      If the value of pid is negative, the behaviour of the
-      *      sched_setscheduler() function is unspecified.
-      *
-      * PARAMETERS
-      *      pid        Identifier of the process in which the
-      *                 policy will be set.
-      *      policy     The new policy value.
-      *
-      *
-      * DESCRIPTION
-      *      This function returns sets the scheduling policy and
-      *      scheduling parameters of the process specified by pid
-      *      to policy and the parameters specified in the
-      *      sched_param structure pointed to by param, respectively.
-      *      The value of the sched_priority member in the sched_param
-      *      structure is any integer within the inclusive priority
-      *      range for the scheduling policy specified by policy.
-      *      If the value of pid is negative, the behaviour of the
-      *      sched_setscheduler() function is unspecified.
-      *
-      * RESULTS
-      *      SCHED_OTHER            on success this is the only possible
-      *                             value that can be returned.
-      *
-      *      NOTE: Since this is part of POSIX 1003.1b
-      *                (realtime extensions), it is defined as returning
-      *                -1 if an error occurs and sets errno to the actual
-      *                error.
-      *
-      *      errno:  ESRCH           'pid' doesn't refer to an existing
-      *                              process.
-      *              EPERM           The process referenced by pid does
-      *                              not allow set access.
-      *              EINVAL          'policy' is not a valid policy value.
-      *              ENOSYS          the policy specified is not supported.
-      *
-      * ------------------------------------------------------
-      */
 {
   /*
    * Win32 only has one policy which we call SCHED_OTHER.
    * However, we try to provide other valid side-effects
-   * such as EPERM and ESRCH errors.
+   * such as EPERM and ESRCH errors. Choosing to check
+   * for a valid policy last allows us to get the most value out
+   * of this function.
    */
   if (0 != pid)
     {
-      DWORD selfPid = GetCurrentProcessId();
+      int selfPid = (int) GetCurrentProcessId();
 
       if (pid != selfPid)
         {
-          HANDLE h = OpenProcess(PROCESS_SET_INFORMATION, FALSE, pid);
+          HANDLE h = OpenProcess(PROCESS_SET_INFORMATION, FALSE, (DWORD) pid);
 
           if (NULL == h)
             {
-              errno = (GetLastError() == (0xFF & ERROR_ACCESS_DENIED)
-                       ? EPERM
-                       : ESRCH );
-
+              errno = (GetLastError() == (0xFF & ERROR_ACCESS_DENIED)) ? EPERM : ESRCH;
               return -1;
             }
-
-          (void) CloseHandle(h);
         }
     }
 
-  if (policy < SCHED_MIN || policy > SCHED_MAX)
-    {
-      errno = EINVAL;
-      return -1;
-    }
-
   if (SCHED_OTHER != policy)
     {
       errno = ENOSYS;
@@ -484,39 +318,6 @@ sched_setscheduler(pid_t pid, int policy)
 
 int
 sched_getscheduler(pid_t pid)
-     /*
-      * ------------------------------------------------------
-      * DOCPUBLIC
-      *      This function returns the scheduling policy of the
-      *      process specified by pid. If the value of pid is
-      *      negative, the behaviour of the sched_getscheduler()
-      *      function is unspecified.
-      *
-      * PARAMETERS
-      *      pid    Process identifier.
-      *
-      *
-      * DESCRIPTION
-      *      This function returns the scheduling policy of the
-      *      process specified by pid. If the value of pid is
-      *      negative, the behaviour of the sched_getscheduler()
-      *      function is unspecified.
-      *
-      * RESULTS
-      *      SCHED_OTHER            on success this is the only possible
-      *                             value that can be returned.
-      *
-      *      NOTE: Since this is part of POSIX 1003.1b
-      *                (realtime extensions), it is defined as returning
-      *                -1 if an error occurs and sets errno to the actual
-      *                error.
-      *
-      *      errno:  ESRCH           Pid doesn't refer to a valid process.
-      *              EPERM           The process referenced by pid does
-      *                              not allow query access.
-      *
-      * ------------------------------------------------------
-      */
 {
   /*
    * Win32 only has one policy which we call SCHED_OTHER.
@@ -525,22 +326,17 @@ sched_getscheduler(pid_t pid)
    */
   if (0 != pid)
     {
-      DWORD selfPid = GetCurrentProcessId();
+      int selfPid = (int) GetCurrentProcessId();
 
       if (pid != selfPid)
         {
-          HANDLE h = OpenProcess(PROCESS_QUERY_INFORMATION, FALSE, pid);
+          HANDLE h = OpenProcess(PROCESS_QUERY_INFORMATION, FALSE, (DWORD) pid);
 
           if (NULL == h)
             {
-              errno = (GetLastError() == (0xFF & ERROR_ACCESS_DENIED)
-                       ? EPERM
-                       : ESRCH );
-
+              errno = (GetLastError() == (0xFF & ERROR_ACCESS_DENIED)) ? EPERM : ESRCH;
               return -1;
             }
-
-          (void) CloseHandle(h);
         }
     }
 
@@ -569,7 +365,8 @@ sched_yield(void)
       *                error.
       *
       * RESULTS
-      *              0               always succeeds
+      *              0               successfully created semaphore,
+      *              ENOSYS          sched_yield not supported,
       *
       * ------------------------------------------------------
       */