<div dir="ltr">This is so interesting. Thanks for letting me know. I'm changing my settings</div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Thu, Aug 27, 2020 at 2:18 AM Gedare Bloom <<a href="mailto:gedare@rtems.org">gedare@rtems.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">The subject line comes through fine, but your mail reader is picking<br>
up some in-reply-to field probably. I think you can change this with<br>
some threading options in either git-format-patch or git-send-email.<br>
<br>
On Wed, Aug 26, 2020 at 2:39 PM Richi Dubey <<a href="mailto:richidubey@gmail.com" target="_blank">richidubey@gmail.com</a>> wrote:<br>
><br>
> I used this command:<br>
><br>
> richi@YouAreAmazing:~/quick-start/src/rtems$ git send-email --to=<a href="mailto:devel@rtems.org" target="_blank">devel@rtems.org</a> v4-0001-Pre-Release-Strong-APA.patch<br>
> v4-0001-Pre-Release-Strong-APA.patch<br>
> (mbox) Adding cc: Richi Dubey <<a href="mailto:richidubey@gmail.com" target="_blank">richidubey@gmail.com</a>> from line 'From: Richi Dubey <<a href="mailto:richidubey@gmail.com" target="_blank">richidubey@gmail.com</a>>'<br>
><br>
> From: Richi Dubey <<a href="mailto:richidubey@gmail.com" target="_blank">richidubey@gmail.com</a>><br>
> To: <a href="mailto:devel@rtems.org" target="_blank">devel@rtems.org</a><br>
> Cc: Richi Dubey <<a href="mailto:richidubey@gmail.com" target="_blank">richidubey@gmail.com</a>><br>
> Subject: [PATCH v4] Pre-Release Strong-APA<br>
> Date: Thu, 27 Aug 2020 02:03:33 +0530<br>
> Message-Id: <<a href="mailto:20200826203333.8491-1-richidubey@gmail.com" target="_blank">20200826203333.8491-1-richidubey@gmail.com</a>><br>
> X-Mailer: git-send-email 2.17.1<br>
><br>
> Send this email? ([y]es|[n]o|[q]uit|[a]ll): y<br>
> OK. Log says:<br>
> Server: <a href="http://smtp.gmail.com" rel="noreferrer" target="_blank">smtp.gmail.com</a><br>
> MAIL FROM:<<a href="mailto:richidubey@gmail.com" target="_blank">richidubey@gmail.com</a>><br>
> RCPT TO:<<a href="mailto:devel@rtems.org" target="_blank">devel@rtems.org</a>><br>
> RCPT TO:<<a href="mailto:richidubey@gmail.com" target="_blank">richidubey@gmail.com</a>><br>
> From: Richi Dubey <<a href="mailto:richidubey@gmail.com" target="_blank">richidubey@gmail.com</a>><br>
> To: <a href="mailto:devel@rtems.org" target="_blank">devel@rtems.org</a><br>
> Cc: Richi Dubey <<a href="mailto:richidubey@gmail.com" target="_blank">richidubey@gmail.com</a>><br>
> Subject: [PATCH v4] Pre-Release Strong-APA<br>
> Date: Thu, 27 Aug 2020 02:03:33 +0530<br>
> Message-Id: <<a href="mailto:20200826203333.8491-1-richidubey@gmail.com" target="_blank">20200826203333.8491-1-richidubey@gmail.com</a>><br>
> X-Mailer: git-send-email 2.17.1<br>
><br>
> Result: 250<br>
><br>
> On Thu, Aug 27, 2020 at 2:07 AM Richi Dubey <<a href="mailto:richidubey@gmail.com" target="_blank">richidubey@gmail.com</a>> wrote:<br>
>><br>
>> I've been trying to send the v4 but the patch keeps coming in v3 mail thread :(.<br>
>><br>
>> What am I missing?<br>
>><br>
>> On Thu, Aug 27, 2020 at 2:03 AM Richi Dubey <<a href="mailto:richidubey@gmail.com" target="_blank">richidubey@gmail.com</a>> wrote:<br>
>>><br>
>>> ---<br>
>>>  cpukit/include/rtems/scheduler.h              |   6 +-<br>
>>>  .../include/rtems/score/schedulerstrongapa.h  | 152 ++-<br>
>>>  cpukit/score/src/schedulerstrongapa.c         | 901 ++++++++++++++----<br>
>>>  3 files changed, 828 insertions(+), 231 deletions(-)<br>
>>><br>
>>> diff --git a/cpukit/include/rtems/scheduler.h b/cpukit/include/rtems/scheduler.h<br>
>>> index 955a83cfb4..6a05c2798a 100644<br>
>>> --- a/cpukit/include/rtems/scheduler.h<br>
>>> +++ b/cpukit/include/rtems/scheduler.h<br>
>>> @@ -257,16 +257,14 @@<br>
>>>    #define RTEMS_SCHEDULER_STRONG_APA( name, prio_count ) \<br>
>>>      static struct { \<br>
>>>        Scheduler_strong_APA_Context Base; \<br>
>>> -      Chain_Control                Ready[ ( prio_count ) ]; \<br>
>>> +      Scheduler_strong_APA_CPU CPU[ CONFIGURE_MAXIMUM_PROCESSORS ]; \<br>
>>>      } SCHEDULER_STRONG_APA_CONTEXT_NAME( name )<br>
>>><br>
>>>    #define RTEMS_SCHEDULER_TABLE_STRONG_APA( name, obj_name ) \<br>
>>>      { \<br>
>>>        &SCHEDULER_STRONG_APA_CONTEXT_NAME( name ).Base.Base.Base, \<br>
>>>        SCHEDULER_STRONG_APA_ENTRY_POINTS, \<br>
>>> -      RTEMS_ARRAY_SIZE( \<br>
>>> -        SCHEDULER_STRONG_APA_CONTEXT_NAME( name ).Ready \<br>
>>> -      ) - 1, \<br>
>>> +      SCHEDULER_STRONG_APA_MAXIMUM_PRIORITY, \<br>
>>>        ( obj_name ) \<br>
>>>        SCHEDULER_CONTROL_IS_NON_PREEMPT_MODE_SUPPORTED( false ) \<br>
>>>      }<br>
>>> diff --git a/cpukit/include/rtems/score/schedulerstrongapa.h b/cpukit/include/rtems/score/schedulerstrongapa.h<br>
>>> index 0ac28cb439..cfd79e932a 100644<br>
>>> --- a/cpukit/include/rtems/score/schedulerstrongapa.h<br>
>>> +++ b/cpukit/include/rtems/score/schedulerstrongapa.h<br>
>>> @@ -6,31 +6,47 @@<br>
>>>   * @brief Strong APA Scheduler API<br>
>>>   */<br>
>>><br>
>>> -/*<br>
>>> - * Copyright (c) 2013, 2018 embedded brains GmbH.  All rights reserved.<br>
>>> +/* SPDX-License-Identifier: BSD-2-Clause<br>
>>>   *<br>
>>> - *  embedded brains GmbH<br>
>>> - *  Dornierstr. 4<br>
>>> - *  82178 Puchheim<br>
>>> - *  Germany<br>
>>> - *  <<a href="mailto:rtems@embedded-brains.de" target="_blank">rtems@embedded-brains.de</a>><br>
>>> + * Copyright (C) 2020 Richi Dubey<br>
>>> + * Copyright (c) 2013, 2018 embedded brains GmbH<br>
>>>   *<br>
>>> - * The license and distribution terms for this file may be<br>
>>> - * found in the file LICENSE in this distribution or at<br>
>>> - * <a href="http://www.rtems.org/license/LICENSE" rel="noreferrer" target="_blank">http://www.rtems.org/license/LICENSE</a>.<br>
>>> + * Redistribution and use in source and binary forms, with or without<br>
>>> + * modification, are permitted provided that the following conditions<br>
>>> + * are met:<br>
>>> + * 1. Redistributions of source code must retain the above copyright<br>
>>> + *    notice, this list of conditions and the following disclaimer.<br>
>>> + * 2. Redistributions in binary form must reproduce the above copyright<br>
>>> + *    notice, this list of conditions and the following disclaimer in the<br>
>>> + *    documentation and/or other materials provided with the distribution.<br>
>>> + *<br>
>>> + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"<br>
>>> + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE<br>
>>> + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE<br>
>>> + * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE<br>
>>> + * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR<br>
>>> + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF<br>
>>> + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS<br>
>>> + * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN<br>
>>> + * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)<br>
>>> + * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE<br>
>>> + * POSSIBILITY OF SUCH DAMAGE.<br>
>>>   */<br>
>>><br>
>>>  #ifndef _RTEMS_SCORE_SCHEDULERSTRONGAPA_H<br>
>>>  #define _RTEMS_SCORE_SCHEDULERSTRONGAPA_H<br>
>>><br>
>>>  #include <rtems/score/scheduler.h><br>
>>> -#include <rtems/score/schedulerpriority.h><br>
>>>  #include <rtems/score/schedulersmp.h><br>
>>> +#include <rtems/score/percpu.h><br>
>>><br>
>>>  #ifdef __cplusplus<br>
>>>  extern "C" {<br>
>>>  #endif /* __cplusplus */<br>
>>><br>
>>> +#define STRONG_SCHEDULER_NODE_OF_CHAIN( node ) \<br>
>>> +  RTEMS_CONTAINER_OF( node, Scheduler_strong_APA_Node, Ready_node )<br>
>>> +<br>
>>>  /**<br>
>>>   * @defgroup RTEMSScoreSchedulerStrongAPA Strong APA Scheduler<br>
>>>   *<br>
>>> @@ -38,42 +54,96 @@ extern "C" {<br>
>>>   *<br>
>>>   * @brief Strong APA Scheduler<br>
>>>   *<br>
>>> - * This is an implementation of the global fixed priority scheduler (G-FP).  It<br>
>>> - * uses one ready chain per priority to ensure constant time insert operations.<br>
>>> - * The scheduled chain uses linear insert operations and has at most processor<br>
>>> - * count entries.  Since the processor and priority count are constants all<br>
>>> - * scheduler operations complete in a bounded execution time.<br>
>>> - *<br>
>>> - * The the_thread preempt mode will be ignored.<br>
>>> + * This is an implementation of the Strong APA scheduler defined by<br>
>>> + * Cerqueira et al. in Linux's Processor Affinity API, Refined:<br>
>>> + * Shifting Real-Time Tasks Towards Higher Schedulability.<br>
>>>   *<br>
>>> + * The scheduled and ready nodes are accessed via the<br>
>>> + * Scheduler_strong_APA_Context::Ready which helps in backtracking when a<br>
>>> + * node which is executing on a CPU gets blocked. New node is allocated to<br>
>>> + * the cpu by checking all the executing nodes in the affinity set of the<br>
>>> + * node and the subsequent nodes executing on the processors in its<br>
>>> + * affinity set.<br>
>>>   * @{<br>
>>>   */<br>
>>><br>
>>>  /**<br>
>>> - * @brief Scheduler context specialization for Strong APA<br>
>>> - * schedulers.<br>
>>> + * @brief Scheduler node specialization for Strong APA schedulers.<br>
>>>   */<br>
>>>  typedef struct {<br>
>>> -  Scheduler_SMP_Context    Base;<br>
>>> -  Priority_bit_map_Control Bit_map;<br>
>>> -  Chain_Control            Ready[ RTEMS_ZERO_LENGTH_ARRAY ];<br>
>>> -} Scheduler_strong_APA_Context;<br>
>>> +  /**<br>
>>> +   * @brief SMP scheduler node.<br>
>>> +   */<br>
>>> +  Scheduler_SMP_Node Base;<br>
>>> +<br>
>>> + /**<br>
>>> +   * @brief Chain node for Scheduler_strong_APA_Context::Ready.<br>
>>> +   */<br>
>>> +  Chain_Node Ready_node;<br>
>>> +<br>
>>> +  /**<br>
>>> +   * @brief CPU that this node would preempt in the backtracking part of<br>
>>> +   * _Scheduler_strong_APA_Get_highest_ready and<br>
>>> +   * _Scheduler_strong_APA_Do_Enqueue.<br>
>>> +   */<br>
>>> +  Per_CPU_Control *cpu_to_preempt;<br>
>>> +<br>
>>> +  /**<br>
>>> +   * @brief The associated affinity set of this node.<br>
>>> +   */<br>
>>> +  Processor_mask Affinity;<br>
>>> +} Scheduler_strong_APA_Node;<br>
>>> +<br>
>>><br>
>>>  /**<br>
>>> - * @brief Scheduler node specialization for Strong APA<br>
>>> - * schedulers.<br>
>>> + * @brief CPU related variables and a CPU_Control to implement BFS.<br>
>>> + */<br>
>>> +typedef struct<br>
>>> +{<br>
>>> +   /**<br>
>>> +   * @brief CPU in a queue.<br>
>>> +   */<br>
>>> +  Per_CPU_Control *cpu;<br>
>>> +<br>
>>> +  /**<br>
>>> +   * @brief The node that would preempt this CPU.<br>
>>> +   */<br>
>>> +  Scheduler_Node *preempting_node;<br>
>>> +<br>
>>> +   /**<br>
>>> +   * @brief Whether or not this cpu has been added to the queue<br>
>>> +   * (visited in BFS).<br>
>>> +   */<br>
>>> +  bool visited;<br>
>>> +<br>
>>> +  /**<br>
>>> +   * @brief The node currently executing on this cpu.<br>
>>> +   */<br>
>>> +   Scheduler_Node *executing;<br>
>>> +} Scheduler_strong_APA_CPU;<br>
>>> +<br>
>>> + /**<br>
>>> + * @brief Scheduler context and node definition for Strong APA scheduler.<br>
>>>   */<br>
>>>  typedef struct {<br>
>>> + /**<br>
>>> +   * @brief @see Scheduler_SMP_Context.<br>
>>> +   */<br>
>>> +  Scheduler_SMP_Context Base;<br>
>>> +<br>
>>>    /**<br>
>>> -   * @brief SMP scheduler node.<br>
>>> +   * @brief Chain of all the ready and scheduled nodes present in<br>
>>> +   * the Strong APA scheduler.<br>
>>>     */<br>
>>> -  Scheduler_SMP_Node Base;<br>
>>> +  Chain_Control Ready;<br>
>>><br>
>>>    /**<br>
>>> -   * @brief The associated ready queue of this node.<br>
>>> +   * @brief Struct with important variables for each cpu.<br>
>>>     */<br>
>>> -  Scheduler_priority_Ready_queue Ready_queue;<br>
>>> -} Scheduler_strong_APA_Node;<br>
>>> +  Scheduler_strong_APA_CPU CPU[ RTEMS_ZERO_LENGTH_ARRAY ];<br>
>>> +} Scheduler_strong_APA_Context;<br>
>>> +<br>
>>> +#define SCHEDULER_STRONG_APA_MAXIMUM_PRIORITY 255<br>
>>><br>
>>>  /**<br>
>>>   * @brief Entry points for the Strong APA Scheduler.<br>
>>> @@ -100,8 +170,8 @@ typedef struct {<br>
>>>      _Scheduler_default_Release_job, \<br>
>>>      _Scheduler_default_Cancel_job, \<br>
>>>      _Scheduler_default_Tick, \<br>
>>> -    _Scheduler_SMP_Start_idle \<br>
>>> -    SCHEDULER_OPERATION_DEFAULT_GET_SET_AFFINITY \<br>
>>> +    _Scheduler_SMP_Start_idle, \<br>
>>> +    _Scheduler_strong_APA_Set_affinity \<br>
>>>    }<br>
>>><br>
>>>  /**<br>
>>> @@ -168,7 +238,7 @@ void _Scheduler_strong_APA_Update_priority(<br>
>>>  /**<br>
>>>   * @brief Asks for help.<br>
>>>   *<br>
>>> - * @param  scheduler The scheduler control instance.<br>
>>> + * @param scheduler The scheduler control instance.<br>
>>>   * @param the_thread The thread that asks for help.<br>
>>>   * @param node The node of @a the_thread.<br>
>>>   *<br>
>>> @@ -246,6 +316,20 @@ void _Scheduler_strong_APA_Yield(<br>
>>>    Scheduler_Node          *node<br>
>>>  );<br>
>>><br>
>>> +/**<br>
>>> + * @brief Sets the affinity .<br>
>>> + *<br>
>>> + * @param scheduler The scheduler control instance.<br>
>>> + * @param the_thread The thread to yield.<br>
>>> + * @param[in, out] node The node of @a the_thread.<br>
>>> + */<br>
>>> +bool _Scheduler_strong_APA_Set_affinity(<br>
>>> +  const Scheduler_Control *scheduler,<br>
>>> +  Thread_Control          *thread,<br>
>>> +  Scheduler_Node          *node_base,<br>
>>> +  const Processor_mask    *affinity<br>
>>> +);<br>
>>> +<br>
>>>  /** @} */<br>
>>><br>
>>>  #ifdef __cplusplus<br>
>>> diff --git a/cpukit/score/src/schedulerstrongapa.c b/cpukit/score/src/schedulerstrongapa.c<br>
>>> index 924cd86412..429bbacece 100644<br>
>>> --- a/cpukit/score/src/schedulerstrongapa.c<br>
>>> +++ b/cpukit/score/src/schedulerstrongapa.c<br>
>>> @@ -6,18 +6,31 @@<br>
>>>   * @brief Strong APA Scheduler Implementation<br>
>>>   */<br>
>>><br>
>>> -/*<br>
>>> - * Copyright (c) 2013, 2016 embedded brains GmbH.  All rights reserved.<br>
>>> +/* SPDX-License-Identifier: BSD-2-Clause<br>
>>> + *<br>
>>> + * Copyright (C) 2020 Richi Dubey<br>
>>> + * Copyright (c) 2013, 2018 embedded brains GmbH<br>
>>>   *<br>
>>> - *  embedded brains GmbH<br>
>>> - *  Dornierstr. 4<br>
>>> - *  82178 Puchheim<br>
>>> - *  Germany<br>
>>> - *  <<a href="mailto:rtems@embedded-brains.de" target="_blank">rtems@embedded-brains.de</a>><br>
>>> + * Redistribution and use in source and binary forms, with or without<br>
>>> + * modification, are permitted provided that the following conditions<br>
>>> + * are met:<br>
>>> + * 1. Redistributions of source code must retain the above copyright<br>
>>> + *    notice, this list of conditions and the following disclaimer.<br>
>>> + * 2. Redistributions in binary form must reproduce the above copyright<br>
>>> + *    notice, this list of conditions and the following disclaimer in the<br>
>>> + *    documentation and/or other materials provided with the distribution.<br>
>>>   *<br>
>>> - * The license and distribution terms for this file may be<br>
>>> - * found in the file LICENSE in this distribution or at<br>
>>> - * <a href="http://www.rtems.org/license/LICENSE" rel="noreferrer" target="_blank">http://www.rtems.org/license/LICENSE</a>.<br>
>>> + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"<br>
>>> + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE<br>
>>> + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE<br>
>>> + * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE<br>
>>> + * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR<br>
>>> + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF<br>
>>> + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS<br>
>>> + * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN<br>
>>> + * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)<br>
>>> + * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE<br>
>>> + * POSSIBILITY OF SUCH DAMAGE.<br>
>>>   */<br>
>>><br>
>>>  #ifdef HAVE_CONFIG_H<br>
>>> @@ -25,301 +38,754 @@<br>
>>>  #endif<br>
>>><br>
>>>  #include <rtems/score/schedulerstrongapa.h><br>
>>> -#include <rtems/score/schedulerpriorityimpl.h><br>
>>>  #include <rtems/score/schedulersmpimpl.h><br>
>>> +#include <rtems/score/assert.h><br>
>>><br>
>>> -static Scheduler_strong_APA_Context *_Scheduler_strong_APA_Get_self(<br>
>>> -  Scheduler_Context *context<br>
>>> -)<br>
>>> +static inline Scheduler_strong_APA_Context *<br>
>>> +_Scheduler_strong_APA_Get_context( const Scheduler_Control *scheduler )<br>
>>> +{<br>
>>> +  return (Scheduler_strong_APA_Context *) _Scheduler_Get_context( scheduler );<br>
>>> +}<br>
>>> +<br>
>>> +static inline Scheduler_strong_APA_Context *<br>
>>> +_Scheduler_strong_APA_Get_self( Scheduler_Context *context )<br>
>>>  {<br>
>>>    return (Scheduler_strong_APA_Context *) context;<br>
>>>  }<br>
>>><br>
>>> -static Scheduler_strong_APA_Node *<br>
>>> +static inline Scheduler_strong_APA_Node *<br>
>>>  _Scheduler_strong_APA_Node_downcast( Scheduler_Node *node )<br>
>>>  {<br>
>>>    return (Scheduler_strong_APA_Node *) node;<br>
>>>  }<br>
>>><br>
>>> -static void _Scheduler_strong_APA_Move_from_scheduled_to_ready(<br>
>>> +static inline void _Scheduler_strong_APA_Do_update(<br>
>>>    Scheduler_Context *context,<br>
>>> -  Scheduler_Node    *scheduled_to_ready<br>
>>> +  Scheduler_Node    *node,<br>
>>> +  Priority_Control   new_priority<br>
>>>  )<br>
>>>  {<br>
>>> -  Scheduler_strong_APA_Context *self =<br>
>>> -    _Scheduler_strong_APA_Get_self( context );<br>
>>> -  Scheduler_strong_APA_Node *node =<br>
>>> -    _Scheduler_strong_APA_Node_downcast( scheduled_to_ready );<br>
>>> -<br>
>>> -  _Chain_Extract_unprotected( &node->Base.Base.Node.Chain );<br>
>>> -  _Scheduler_priority_Ready_queue_enqueue_first(<br>
>>> -    &node->Base.Base.Node.Chain,<br>
>>> -    &node->Ready_queue,<br>
>>> -    &self->Bit_map<br>
>>> -  );<br>
>>> +  Scheduler_SMP_Node *smp_node;<br>
>>> +  (void) context;<br>
>>> +<br>
>>> +  smp_node = _Scheduler_SMP_Node_downcast( node );<br>
>>> +  _Scheduler_SMP_Node_update_priority( smp_node, new_priority );<br>
>>>  }<br>
>>><br>
>>> -static void _Scheduler_strong_APA_Move_from_ready_to_scheduled(<br>
>>> +/*<br>
>>> + * Returns true if the Strong APA scheduler has ready nodes<br>
>>> + * available for scheduling.<br>
>>> + */<br>
>>> +static inline bool<br>
>>> +    _Scheduler_strong_APA_Has_ready( Scheduler_Context *context )<br>
>>> +{<br>
>>> +  Scheduler_strong_APA_Context *self;<br>
>>> +  const Chain_Node             *tail;<br>
>>> +  Chain_Node                   *next;<br>
>>> +  Scheduler_strong_APA_Node    *node;<br>
>>> +<br>
>>> +  self = _Scheduler_strong_APA_Get_self( context );<br>
>>> +  tail = _Chain_Immutable_tail( &self->Ready );<br>
>>> +  next = _Chain_First( &self->Ready );<br>
>>> +<br>
>>> +  while ( next != tail ) {<br>
>>> +    node = (Scheduler_strong_APA_Node *) STRONG_SCHEDULER_NODE_OF_CHAIN( next );<br>
>>> +<br>
>>> +    if (<br>
>>> +      _Scheduler_SMP_Node_state( &node->Base.Base ) ==<br>
>>> +      SCHEDULER_SMP_NODE_READY<br>
>>> +    ) {<br>
>>> +      return true;<br>
>>> +    }<br>
>>> +<br>
>>> +    next = _Chain_Next( next );<br>
>>> +  }<br>
>>> +<br>
>>> +  return false;<br>
>>> +}<br>
>>> +<br>
>>> +static inline void _Scheduler_strong_APA_Allocate_processor(<br>
>>>    Scheduler_Context *context,<br>
>>> -  Scheduler_Node    *ready_to_scheduled<br>
>>> +  Scheduler_Node    *scheduled_base,<br>
>>> +  Scheduler_Node    *victim_base,<br>
>>> +  Per_CPU_Control   *victim_cpu<br>
>>>  )<br>
>>>  {<br>
>>> +  Scheduler_strong_APA_Node    *scheduled;<br>
>>>    Scheduler_strong_APA_Context *self;<br>
>>> -  Scheduler_strong_APA_Node    *node;<br>
>>> -  Priority_Control              insert_priority;<br>
>>><br>
>>> +  (void) victim_base;<br>
>>> +<br>
>>> +  scheduled = _Scheduler_strong_APA_Node_downcast( scheduled_base );<br>
>>>    self = _Scheduler_strong_APA_Get_self( context );<br>
>>> -  node = _Scheduler_strong_APA_Node_downcast( ready_to_scheduled );<br>
>>><br>
>>> -  _Scheduler_priority_Ready_queue_extract(<br>
>>> -    &node->Base.Base.Node.Chain,<br>
>>> -    &node->Ready_queue,<br>
>>> -    &self->Bit_map<br>
>>> +  self->CPU[ _Per_CPU_Get_index( victim_cpu ) ].executing = scheduled_base;<br>
>>> +<br>
>>> +  _Scheduler_SMP_Allocate_processor_exact(<br>
>>> +    context,<br>
>>> +    &(scheduled->Base.Base),<br>
>>> +    NULL,<br>
>>> +    victim_cpu<br>
>>>    );<br>
>>> -  insert_priority = _Scheduler_SMP_Node_priority( &node->Base.Base );<br>
>>> +}<br>
>>> +<br>
>>> +/*<br>
>>> + * Finds and returns the highest ready node present by accessing the<br>
>>> + * _Strong_APA_Context->CPU with front and rear values.<br>
>>> + */<br>
>>> +<br>
>>> +static inline Scheduler_Node * _Scheduler_strong_APA_Find_highest_ready(<br>
>>> +  Scheduler_strong_APA_Context *self,<br>
>>> +  uint32_t                      front,<br>
>>> +  uint32_t                      rear<br>
>>> +)<br>
>>> +{<br>
>>> +  Scheduler_Node              *highest_ready;<br>
>>> +  Scheduler_strong_APA_CPU    *CPU;<br>
>>> +  const Chain_Node            *tail;<br>
>>> +  Chain_Node                  *next;<br>
>>> +  uint32_t                     index_assigned_cpu;<br>
>>> +  Scheduler_strong_APA_Node   *node;<br>
>>> +  Priority_Control             min_priority_num;<br>
>>> +  Priority_Control             curr_priority;<br>
>>> +  Per_CPU_Control             *assigned_cpu;<br>
>>> +  Scheduler_SMP_Node_state     curr_state;<br>
>>> +  Per_CPU_Control             *curr_CPU;<br>
>>> +  bool                         first_task;<br>
>>> +<br>
>>> +  CPU = self->CPU;<br>
>>> +   /*<br>
>>> +    * When the first task accessed has nothing to compare its priority against<br>
>>> +    * So, it is the task with the highest priority witnessed so far.<br>
>>> +   */<br>
>>> +  first_task = true;<br>
>>> +<br>
>>> +  //Assert rear < sizeof(Context->CPU)<br>
>>> +  _Assert( rear < CONFIGURE_MAXIMUM_PROCESSOR );<br>
>>> +<br>
>>> +  while( front <= rear ) {<br>
>>> +    curr_CPU = CPU[ front ].cpu;<br>
>>> +    front = front + 1;<br>
>>> +<br>
>>> +    tail = _Chain_Immutable_tail( &self->Ready );<br>
>>> +    next = _Chain_First( &self->Ready );<br>
>>> +<br>
>>> +    while ( next != tail ) {<br>
>>> +      node = (Scheduler_strong_APA_Node*) STRONG_SCHEDULER_NODE_OF_CHAIN( next );<br>
>>> +      //Check if the curr_CPU is in the affinity set of the node.<br>
>>> +      if (<br>
>>> +        _Processor_mask_Is_set(&node->Affinity, _Per_CPU_Get_index(curr_CPU))<br>
>>> +      ) {<br>
>>> +        curr_state = _Scheduler_SMP_Node_state( &node->Base.Base );<br>
>>> +<br>
>>> +        if ( curr_state == SCHEDULER_SMP_NODE_SCHEDULED ) {<br>
>>> +          assigned_cpu = _Thread_Get_CPU( node->Base.Base.user );<br>
>>> +          index_assigned_cpu =  _Per_CPU_Get_index( assigned_cpu );<br>
>>> +<br>
>>> +          if ( CPU[ index_assigned_cpu ].visited == false ) {<br>
>>> +            rear = rear + 1;<br>
>>> +            CPU[ rear ].cpu = assigned_cpu;<br>
>>> +            CPU[ index_assigned_cpu ].visited = true;<br>
>>> +            /*<br>
>>> +             * The curr CPU of the queue invoked this node to add its CPU<br>
>>> +             * that it is executing on to the queue. So this node might get<br>
>>> +             * preempted because of the invoker curr_CPU and this curr_CPU<br>
>>> +             * is the CPU that node should preempt in case this node<br>
>>> +             * gets preempted.<br>
>>> +             */<br>
>>> +            node->cpu_to_preempt = curr_CPU;<br>
>>> +          }<br>
>>> +        }<br>
>>> +        else if ( curr_state == SCHEDULER_SMP_NODE_READY ) {<br>
>>> +          curr_priority = _Scheduler_Node_get_priority( &node->Base.Base );<br>
>>> +          curr_priority = SCHEDULER_PRIORITY_PURIFY( curr_priority );<br>
>>> +<br>
>>> +          if ( first_task == true || curr_priority < min_priority_num ) {<br>
>>> +            min_priority_num = curr_priority;<br>
>>> +           highest_ready = &node->Base.Base;<br>
>>> +           first_task = false;<br>
>>> +           /*<br>
>>> +            * In case curr_CPU is filter_CPU, we need to store the<br>
>>> +            * cpu_to_preempt value so that we go back to SMP_*<br>
>>> +            * function, rather than preempting the node ourselves.<br>
>>> +            */<br>
>>> +           node->cpu_to_preempt = curr_CPU;<br>
>>> +         }<br>
>>> +        }<br>
>>> +      }<br>
>>> +    next = _Chain_Next( next );<br>
>>> +    }<br>
>>> +  }<br>
>>> +<br>
>>> +  return highest_ready;<br>
>>> +}<br>
>>> +<br>
>>> +static inline void  _Scheduler_strong_APA_Move_from_ready_to_scheduled(<br>
>>> +  Scheduler_Context *context,<br>
>>> +  Scheduler_Node    *ready_to_scheduled<br>
>>> +)<br>
>>> +{<br>
>>> +  Priority_Control insert_priority;<br>
>>> +<br>
>>> +  insert_priority = _Scheduler_SMP_Node_priority( ready_to_scheduled );<br>
>>>    insert_priority = SCHEDULER_PRIORITY_APPEND( insert_priority );<br>
>>> -  _Chain_Insert_ordered_unprotected(<br>
>>> -    &self->Base.Scheduled,<br>
>>> -    &node->Base.Base.Node.Chain,<br>
>>> -    &insert_priority,<br>
>>> -    _Scheduler_SMP_Priority_less_equal<br>
>>> +  _Scheduler_SMP_Insert_scheduled(<br>
>>> +    context,<br>
>>> +    ready_to_scheduled,<br>
>>> +    insert_priority<br>
>>>    );<br>
>>>  }<br>
>>><br>
>>> -static void _Scheduler_strong_APA_Insert_ready(<br>
>>> +/*<br>
>>> + * Implement the BFS Algorithm for task departure to get the highest ready task<br>
>>> + * for a particular CPU, returns the highest ready Scheduler_Node<br>
>>> + * Scheduler_Node filter here pointst to the victim node that is blocked<br>
>>> + * resulting which this function is called.<br>
>>> + */<br>
>>> +static inline Scheduler_Node *_Scheduler_strong_APA_Get_highest_ready(<br>
>>>    Scheduler_Context *context,<br>
>>> -  Scheduler_Node    *node_base,<br>
>>> -  Priority_Control   insert_priority<br>
>>> +  Scheduler_Node    *filter<br>
>>>  )<br>
>>>  {<br>
>>>    Scheduler_strong_APA_Context *self;<br>
>>> +  Per_CPU_Control              *filter_cpu;<br>
>>>    Scheduler_strong_APA_Node    *node;<br>
>>> +  Scheduler_Node               *highest_ready;<br>
>>> +  Scheduler_Node               *curr_node;<br>
>>> +  Scheduler_Node               *next_node;<br>
>>> +  Scheduler_strong_APA_CPU     *CPU;<br>
>>> +  uint32_t                     front;<br>
>>> +  uint32_t                     rear;<br>
>>> +  uint32_t                     cpu_max;<br>
>>> +  uint32_t                             cpu_index;<br>
>>><br>
>>>    self = _Scheduler_strong_APA_Get_self( context );<br>
>>> -  node = _Scheduler_strong_APA_Node_downcast( node_base );<br>
>>> +  //Denotes front and rear of the queue.<br>
>>> +  front = 0;<br>
>>> +  rear = -1;<br>
>>><br>
>>> -  if ( SCHEDULER_PRIORITY_IS_APPEND( insert_priority ) ) {<br>
>>> -    _Scheduler_priority_Ready_queue_enqueue(<br>
>>> -      &node->Base.Base.Node.Chain,<br>
>>> -      &node->Ready_queue,<br>
>>> -      &self->Bit_map<br>
>>> -    );<br>
>>> -  } else {<br>
>>> -    _Scheduler_priority_Ready_queue_enqueue_first(<br>
>>> -      &node->Base.Base.Node.Chain,<br>
>>> -      &node->Ready_queue,<br>
>>> -      &self->Bit_map<br>
>>> -    );<br>
>>> +  filter_cpu = _Thread_Get_CPU( filter->user );<br>
>>> +  CPU = self->CPU;<br>
>>> +  cpu_max = _SMP_Get_processor_maximum();<br>
>>> +<br>
>>> +  for ( cpu_index = 0 ; cpu_index < cpu_max ; ++cpu_index ) {<br>
>>> +    CPU[ cpu_index ].visited = false;<br>
>>>    }<br>
>>> +<br>
>>> +  rear = rear + 1;<br>
>>> +  CPU[ rear ].cpu = filter_cpu;<br>
>>> +  CPU[ _Per_CPU_Get_index( filter_cpu ) ].visited = true;<br>
>>> +<br>
>>> +  highest_ready = _Scheduler_strong_APA_Find_highest_ready(<br>
>>> +                    self,<br>
>>> +                    front,<br>
>>> +                    rear<br>
>>> +                  );<br>
>>> +<br>
>>> +  if ( highest_ready != filter ) {<br>
>>> +    /*<br>
>>> +     * Backtrack on the path from<br>
>>> +     * filter_cpu to highest_ready, shifting along every task.<br>
>>> +     */<br>
>>> +<br>
>>> +    node = _Scheduler_strong_APA_Node_downcast( highest_ready );<br>
>>> +    /*<br>
>>> +     * Highest ready is not just directly reachable from the victim cpu<br>
>>> +     * So there is need of task shifting .<br>
>>> +     */<br>
>>> +    while( node->cpu_to_preempt !=  filter_cpu ){<br>
>>> +      curr_node = &node->Base.Base;<br>
>>> +      next_node = CPU[ _Per_CPU_Get_index( node->cpu_to_preempt ) ].executing;<br>
>>> +<br>
>>> +      _Scheduler_SMP_Preempt(<br>
>>> +        context,<br>
>>> +        curr_node,<br>
>>> +        next_node,<br>
>>> +        _Scheduler_strong_APA_Allocate_processor<br>
>>> +      );<br>
>>> +<br>
>>> +      if( curr_node == highest_ready) {<br>
>>> +        _Scheduler_strong_APA_Move_from_ready_to_scheduled(context, curr_node);<br>
>>> +      }<br>
>>> +<br>
>>> +      node = _Scheduler_strong_APA_Node_downcast( next_node );<br>
>>> +    }<br>
>>> +    /*<br>
>>> +     * To save the last node so that the caller SMP_* function<br>
>>> +     * can do the allocation<br>
>>> +     */<br>
>>> +      curr_node = &node->Base.Base;<br>
>>> +      highest_ready = curr_node;<br>
>>> +    }<br>
>>> +<br>
>>> +  return highest_ready;<br>
>>>  }<br>
>>><br>
>>> -static void _Scheduler_strong_APA_Extract_from_ready(<br>
>>> +/*<br>
>>> + * Checks the lowest scheduled directly reachable task<br>
>>> + */<br>
>>> +static inline Scheduler_Node *_Scheduler_strong_APA_Get_lowest_scheduled(<br>
>>>    Scheduler_Context *context,<br>
>>> -  Scheduler_Node    *the_thread<br>
>>> +  Scheduler_Node    *filter_base<br>
>>>  )<br>
>>>  {<br>
>>> -  Scheduler_strong_APA_Context *self =<br>
>>> -    _Scheduler_strong_APA_Get_self( context );<br>
>>> -  Scheduler_strong_APA_Node *node =<br>
>>> -    _Scheduler_strong_APA_Node_downcast( the_thread );<br>
>>> -<br>
>>> -  _Scheduler_priority_Ready_queue_extract(<br>
>>> -    &node->Base.Base.Node.Chain,<br>
>>> -    &node->Ready_queue,<br>
>>> -    &self->Bit_map<br>
>>> -  );<br>
>>> +  uint32_t                     cpu_max;<br>
>>> +  uint32_t                     cpu_index;<br>
>>> +  Scheduler_Node               *curr_node;<br>
>>> +  Scheduler_Node               *lowest_scheduled;<br>
>>> +  Priority_Control              max_priority_num;<br>
>>> +  Priority_Control              curr_priority;<br>
>>> +  Scheduler_strong_APA_Node    *filter_strong_node;<br>
>>> +  Scheduler_strong_APA_Context *self;<br>
>>> +<br>
>>> +  self = _Scheduler_strong_APA_Get_self( context );<br>
>>> +  lowest_scheduled = NULL; // To remove compiler warning.<br>
>>> +  max_priority_num = 0;    // Max (Lowest) priority encountered so far.<br>
>>> +  filter_strong_node = _Scheduler_strong_APA_Node_downcast( filter_base );<br>
>>> +<br>
>>> +  //lowest_scheduled is NULL if affinty of a node is 0<br>
>>> +  _Assert( !_Processor_mask_Zero( &filter_strong_node->Affinity ) );<br>
>>> +  cpu_max = _SMP_Get_processor_maximum();<br>
>>> +<br>
>>> +  for ( cpu_index = 0 ; cpu_index < cpu_max ; ++cpu_index ) {<br>
>>> +    //Checks if the CPU is in the affinity set of filter_strong_node<br>
>>> +    if ( _Processor_mask_Is_set( &filter_strong_node->Affinity, cpu_index) ) {<br>
>>> +      Per_CPU_Control *cpu = _Per_CPU_Get_by_index( cpu_index );<br>
>>> +<br>
>>> +      if ( _Per_CPU_Is_processor_online( cpu ) ) {<br>
>>> +        curr_node = self->CPU[ _Per_CPU_Get_index( cpu ) ].executing;<br>
>>> +        curr_priority = _Scheduler_Node_get_priority( curr_node );<br>
>>> +        curr_priority = SCHEDULER_PRIORITY_PURIFY( curr_priority );<br>
>>> +<br>
>>> +        if ( curr_priority > max_priority_num ) {<br>
>>> +          lowest_scheduled = curr_node;<br>
>>> +          max_priority_num = curr_priority;<br>
>>> +        }<br>
>>> +      }<br>
>>> +    }<br>
>>> +  }<br>
>>> +<br>
>>> +  _Assert( lowest_scheduled != NULL );<br>
>>> +  return lowest_scheduled;<br>
>>>  }<br>
>>><br>
>>> -static void _Scheduler_strong_APA_Do_update(<br>
>>> +static inline void _Scheduler_strong_APA_Extract_from_scheduled(<br>
>>>    Scheduler_Context *context,<br>
>>> -  Scheduler_Node *node_to_update,<br>
>>> -  Priority_Control new_priority<br>
>>> +  Scheduler_Node    *node_to_extract<br>
>>>  )<br>
>>>  {<br>
>>> -  Scheduler_strong_APA_Context *self =<br>
>>> -    _Scheduler_strong_APA_Get_self( context );<br>
>>> -  Scheduler_strong_APA_Node *node =<br>
>>> -    _Scheduler_strong_APA_Node_downcast( node_to_update );<br>
>>> -<br>
>>> -  _Scheduler_SMP_Node_update_priority( &node->Base, new_priority );<br>
>>> -  _Scheduler_priority_Ready_queue_update(<br>
>>> -    &node->Ready_queue,<br>
>>> -    SCHEDULER_PRIORITY_UNMAP( new_priority ),<br>
>>> -    &self->Bit_map,<br>
>>> -    &self->Ready[ 0 ]<br>
>>> -  );<br>
>>> +  Scheduler_strong_APA_Context *self;<br>
>>> +  Scheduler_strong_APA_Node    *node;<br>
>>> +<br>
>>> +  self = _Scheduler_strong_APA_Get_self( context );<br>
>>> +  node = _Scheduler_strong_APA_Node_downcast( node_to_extract );<br>
>>> +<br>
>>> +  _Scheduler_SMP_Extract_from_scheduled( &self->Base.Base, &node->Base.Base );<br>
>>> +  //Not removing it from Ready since the node could go in the READY state.<br>
>>>  }<br>
>>><br>
>>> -static Scheduler_strong_APA_Context *<br>
>>> -_Scheduler_strong_APA_Get_context( const Scheduler_Control *scheduler )<br>
>>> +static inline void _Scheduler_strong_APA_Extract_from_ready(<br>
>>> +  Scheduler_Context *context,<br>
>>> +  Scheduler_Node    *node_to_extract<br>
>>> +)<br>
>>>  {<br>
>>> -  return (Scheduler_strong_APA_Context *) _Scheduler_Get_context( scheduler );<br>
>>> +  Scheduler_strong_APA_Context *self;<br>
>>> +  Scheduler_strong_APA_Node    *node;<br>
>>> +<br>
>>> +  self = _Scheduler_strong_APA_Get_self( context );<br>
>>> +  node = _Scheduler_strong_APA_Node_downcast( node_to_extract );<br>
>>> +<br>
>>> +  _Assert( !_Chain_Is_empty(self->Ready) );<br>
>>> +  _Assert( !_Chain_Is_node_off_chain( &node->Ready_node ) );<br>
>>> +<br>
>>> +   _Chain_Extract_unprotected( &node->Ready_node );    //Removed from Ready<br>
>>> +   _Chain_Set_off_chain( &node->Ready_node );<br>
>>>  }<br>
>>><br>
>>> -void _Scheduler_strong_APA_Initialize( const Scheduler_Control *scheduler )<br>
>>> +static inline void _Scheduler_strong_APA_Insert_ready(<br>
>>> +  Scheduler_Context *context,<br>
>>> +  Scheduler_Node    *node_base,<br>
>>> +  Priority_Control   insert_priority<br>
>>> +)<br>
>>>  {<br>
>>> -  Scheduler_strong_APA_Context *self =<br>
>>> -    _Scheduler_strong_APA_Get_context( scheduler );<br>
>>> +  Scheduler_strong_APA_Context *self;<br>
>>> +  Scheduler_strong_APA_Node    *node;<br>
>>><br>
>>> -  _Scheduler_SMP_Initialize( &self->Base );<br>
>>> -  _Priority_bit_map_Initialize( &self->Bit_map );<br>
>>> -  _Scheduler_priority_Ready_queue_initialize(<br>
>>> -    &self->Ready[ 0 ],<br>
>>> -    scheduler->maximum_priority<br>
>>> -  );<br>
>>> +  self = _Scheduler_strong_APA_Get_self( context );<br>
>>> +  node = _Scheduler_strong_APA_Node_downcast( node_base );<br>
>>> +<br>
>>> +  if( _Chain_Is_node_off_chain( &node->Ready_node ) ) {<br>
>>> +    _Chain_Append_unprotected( &self->Ready, &node->Ready_node );<br>
>>> +  }<br>
>>>  }<br>
>>><br>
>>> -void _Scheduler_strong_APA_Node_initialize(<br>
>>> -  const Scheduler_Control *scheduler,<br>
>>> -  Scheduler_Node          *node,<br>
>>> -  Thread_Control          *the_thread,<br>
>>> -  Priority_Control         priority<br>
>>> +static inline void _Scheduler_strong_APA_Move_from_scheduled_to_ready(<br>
>>> +  Scheduler_Context *context,<br>
>>> +  Scheduler_Node    *scheduled_to_ready<br>
>>>  )<br>
>>>  {<br>
>>> -  Scheduler_Context            *context;<br>
>>> -  Scheduler_strong_APA_Context *self;<br>
>>> -  Scheduler_strong_APA_Node    *the_node;<br>
>>> +  Priority_Control insert_priority;<br>
>>><br>
>>> -  the_node = _Scheduler_strong_APA_Node_downcast( node );<br>
>>> -  _Scheduler_SMP_Node_initialize(<br>
>>> -    scheduler,<br>
>>> -    &the_node->Base,<br>
>>> -    the_thread,<br>
>>> -    priority<br>
>>> -  );<br>
>>> +  _Scheduler_SMP_Extract_from_scheduled( context, scheduled_to_ready );<br>
>>> +  insert_priority = _Scheduler_SMP_Node_priority( scheduled_to_ready );<br>
>>><br>
>>> -  context = _Scheduler_Get_context( scheduler );<br>
>>> -  self = _Scheduler_strong_APA_Get_self( context );<br>
>>> -  _Scheduler_priority_Ready_queue_update(<br>
>>> -    &the_node->Ready_queue,<br>
>>> -    SCHEDULER_PRIORITY_UNMAP( priority ),<br>
>>> -    &self->Bit_map,<br>
>>> -    &self->Ready[ 0 ]<br>
>>> +  _Scheduler_strong_APA_Insert_ready(<br>
>>> +    context,<br>
>>> +    scheduled_to_ready,<br>
>>> +    insert_priority<br>
>>>    );<br>
>>>  }<br>
>>><br>
>>> -static bool _Scheduler_strong_APA_Has_ready( Scheduler_Context *context )<br>
>>> +static inline Scheduler_Node* _Scheduler_strong_APA_Get_lowest_reachable(<br>
>>> +  Scheduler_strong_APA_Context *self,<br>
>>> +  uint32_t                      front,<br>
>>> +  uint32_t                      rear,<br>
>>> +  Per_CPU_Control             **cpu_to_preempt<br>
>>> +)<br>
>>>  {<br>
>>> -  Scheduler_strong_APA_Context *self =<br>
>>> -    _Scheduler_strong_APA_Get_self( context );<br>
>>> +  Scheduler_Node              *lowest_reachable;<br>
>>> +  Priority_Control             max_priority_num;<br>
>>> +  uint32_t                    cpu_max;<br>
>>> +  uint32_t                    cpu_index;<br>
>>> +  Thread_Control              *curr_thread;<br>
>>> +  Per_CPU_Control             *curr_CPU;<br>
>>> +  Priority_Control             curr_priority;<br>
>>> +  Scheduler_Node              *curr_node;<br>
>>> +  Scheduler_strong_APA_Node   *curr_strong_node; //Current Strong_APA_Node<br>
>>> +  Scheduler_strong_APA_CPU    *CPU;<br>
>>> +<br>
>>> +  max_priority_num = 0; //Max (Lowest) priority encountered so far.<br>
>>> +  CPU = self->CPU;<br>
>>> +  cpu_max = _SMP_Get_processor_maximum();<br>
>>> +<br>
>>> +  while( front <= rear ) {<br>
>>> +    curr_CPU = CPU[ front ].cpu;<br>
>>> +    front = front + 1;<br>
>>> +<br>
>>> +    curr_node = CPU[ _Per_CPU_Get_index( curr_CPU ) ].executing;<br>
>>> +    curr_thread = curr_node->user;<br>
>>> +<br>
>>> +    curr_priority = _Scheduler_Node_get_priority( curr_node );<br>
>>> +    curr_priority = SCHEDULER_PRIORITY_PURIFY( curr_priority );<br>
>>> +<br>
>>> +    curr_strong_node = _Scheduler_strong_APA_Node_downcast( curr_node );<br>
>>> +<br>
>>> +    if ( curr_priority > max_priority_num ) {<br>
>>> +      lowest_reachable = curr_node;<br>
>>> +      max_priority_num = curr_priority;<br>
>>> +      *cpu_to_preempt = curr_CPU;<br>
>>> +    }<br>
>>> +<br>
>>> +    if ( !curr_thread->is_idle ) {<br>
>>> +      for ( cpu_index = 0 ; cpu_index < cpu_max ; ++cpu_index ) {<br>
>>> +        if ( _Processor_mask_Is_set( &curr_strong_node->Affinity, cpu_index ) ) {<br>
>>> +          //Checks if the thread_CPU is in the affinity set of the node<br>
>>> +          Per_CPU_Control *cpu = _Per_CPU_Get_by_index( cpu_index );<br>
>>> +          if ( _Per_CPU_Is_processor_online( cpu ) && CPU[ cpu_index ].visited == false ) {<br>
>>> +            rear = rear + 1;<br>
>>> +            CPU[ rear ].cpu = cpu;<br>
>>> +            CPU[ cpu_index ].visited = true;<br>
>>> +            CPU[ cpu_index ].preempting_node = curr_node;<br>
>>> +          }<br>
>>> +        }<br>
>>> +      }<br>
>>> +    }<br>
>>> +  }<br>
>>><br>
>>> -  return !_Priority_bit_map_Is_empty( &self->Bit_map );<br>
>>> +  return lowest_reachable;<br>
>>>  }<br>
>>><br>
>>> -static Scheduler_Node *_Scheduler_strong_APA_Get_highest_ready(<br>
>>> +static inline bool _Scheduler_strong_APA_Do_enqueue(<br>
>>>    Scheduler_Context *context,<br>
>>> -  Scheduler_Node    *node<br>
>>> +  Scheduler_Node    *lowest_reachable,<br>
>>> +  Scheduler_Node    *node,<br>
>>> +  Priority_Control  insert_priority,<br>
>>> +  Per_CPU_Control  *cpu_to_preempt<br>
>>>  )<br>
>>>  {<br>
>>> -  Scheduler_strong_APA_Context *self =<br>
>>> -    _Scheduler_strong_APA_Get_self( context );<br>
>>> +  bool                          needs_help;<br>
>>> +  Priority_Control              node_priority;<br>
>>> +  Priority_Control              lowest_priority;<br>
>>> +  Scheduler_strong_APA_CPU     *CPU;<br>
>>> +  Scheduler_Node               *curr_node;<br>
>>> +  Scheduler_strong_APA_Node    *curr_strong_node; //Current Strong_APA_Node<br>
>>> +  Per_CPU_Control              *curr_CPU;<br>
>>> +  Scheduler_strong_APA_Context *self;<br>
>>> +  Scheduler_Node               *next_node;<br>
>>> +<br>
>>> +  self = _Scheduler_strong_APA_Get_self( context );<br>
>>> +  CPU = self->CPU;<br>
>>> +<br>
>>> +  node_priority = _Scheduler_Node_get_priority( node );<br>
>>> +  node_priority = SCHEDULER_PRIORITY_PURIFY( node_priority );<br>
>>> +<br>
>>> +  lowest_priority =  _Scheduler_Node_get_priority( lowest_reachable );<br>
>>> +  lowest_priority = SCHEDULER_PRIORITY_PURIFY( lowest_priority );<br>
>>> +<br>
>>> +  if( lowest_priority > node_priority ) {<br>
>>> +    //Backtrack on the path from<br>
>>> +    //_Thread_Get_CPU(lowest_reachable->user) to lowest_reachable, shifting<br>
>>> +    //along every task<br>
>>> +<br>
>>> +    curr_node = CPU[ _Per_CPU_Get_index(cpu_to_preempt) ].preempting_node;<br>
>>> +    curr_strong_node = _Scheduler_strong_APA_Node_downcast( curr_node );<br>
>>> +    curr_strong_node->cpu_to_preempt = cpu_to_preempt;<br>
>>> +<br>
>>> +    //Save which cpu to preempt in cpu_to_preempt value of the node<br>
>>> +    while ( curr_node != node ) {<br>
>>> +      curr_CPU = _Thread_Get_CPU( curr_node->user );<br>
>>> +      curr_node = CPU[ _Per_CPU_Get_index( curr_CPU ) ].preempting_node;<br>
>>> +      curr_strong_node = _Scheduler_strong_APA_Node_downcast( curr_node );<br>
>>> +      curr_strong_node->cpu_to_preempt =  curr_CPU;<br>
>>> +     }<br>
>>> +<br>
>>> +    curr_CPU = curr_strong_node->cpu_to_preempt;<br>
>>> +    next_node = CPU[ _Per_CPU_Get_index( curr_CPU ) ].executing;<br>
>>> +<br>
>>> +    node_priority = _Scheduler_Node_get_priority( curr_node );<br>
>>> +    node_priority = SCHEDULER_PRIORITY_PURIFY( node_priority );<br>
>>> +<br>
>>> +    _Scheduler_SMP_Enqueue_to_scheduled(<br>
>>> +      context,<br>
>>> +      curr_node,<br>
>>> +      node_priority,<br>
>>> +      next_node,<br>
>>> +      _Scheduler_SMP_Insert_scheduled,<br>
>>> +      _Scheduler_strong_APA_Move_from_scheduled_to_ready,<br>
>>> +      _Scheduler_strong_APA_Allocate_processor<br>
>>> +    );<br>
>>><br>
>>> -  (void) node;<br>
>>> +    curr_node = next_node;<br>
>>> +    curr_strong_node = _Scheduler_strong_APA_Node_downcast( curr_node );<br>
>>><br>
>>> -  return (Scheduler_Node *) _Scheduler_priority_Ready_queue_first(<br>
>>> -    &self->Bit_map,<br>
>>> -    &self->Ready[ 0 ]<br>
>>> -  );<br>
>>> +    while( curr_node !=  lowest_reachable ) {<br>
>>> +      curr_CPU = curr_strong_node->cpu_to_preempt;<br>
>>> +      next_node = CPU[ _Per_CPU_Get_index( curr_CPU ) ].executing;<br>
>>> +      //curr_node preempts the next_node;<br>
>>> +      _Scheduler_SMP_Preempt(<br>
>>> +       context,<br>
>>> +       curr_node,<br>
>>> +       next_node,<br>
>>> +       _Scheduler_strong_APA_Allocate_processor<br>
>>> +      );<br>
>>> +<br>
>>> +      curr_node = next_node;<br>
>>> +      curr_strong_node = _Scheduler_strong_APA_Node_downcast( curr_node );<br>
>>> +    }<br>
>>> +<br>
>>> +    _Scheduler_strong_APA_Move_from_scheduled_to_ready( context, lowest_reachable );<br>
>>> +<br>
>>> +    needs_help = false;<br>
>>> +  } else {<br>
>>> +    needs_help = true;<br>
>>> +  }<br>
>>> +<br>
>>> +  //Add it to Ready chain since it is now either scheduled or just ready.<br>
>>> +  _Scheduler_strong_APA_Insert_ready(context,node,insert_priority);<br>
>>> +<br>
>>> +  return needs_help;<br>
>>>  }<br>
>>><br>
>>> -void _Scheduler_strong_APA_Block(<br>
>>> -  const Scheduler_Control *scheduler,<br>
>>> -  Thread_Control          *the_thread,<br>
>>> -  Scheduler_Node          *node<br>
>>> +/*<br>
>>> + * BFS Algorithm for task arrival<br>
>>> + * Enqueue node either in the scheduled chain or in the ready chain.<br>
>>> + * node is the newly arrived node and is currently not scheduled.<br>
>>> + */<br>
>>> +static inline bool _Scheduler_strong_APA_Enqueue(<br>
>>> +  Scheduler_Context *context,<br>
>>> +  Scheduler_Node    *node,<br>
>>> +  Priority_Control   insert_priority<br>
>>>  )<br>
>>>  {<br>
>>> -  Scheduler_Context *context = _Scheduler_Get_context( scheduler );<br>
>>> +  Scheduler_strong_APA_Context *self;<br>
>>> +  Scheduler_strong_APA_CPU     *CPU;<br>
>>> +  uint32_t                     cpu_max;<br>
>>> +  uint32_t                     cpu_index;<br>
>>> +  Per_CPU_Control              *cpu_to_preempt;<br>
>>> +  Scheduler_Node               *lowest_reachable;<br>
>>> +  Scheduler_strong_APA_Node    *strong_node;<br>
>>><br>
>>> -  _Scheduler_SMP_Block(<br>
>>> +  //Denotes front and rear of the queue<br>
>>> +  uint32_t     front;<br>
>>> +  uint32_t     rear;<br>
>>> +<br>
>>> +  front = 0;<br>
>>> +  rear = -1;<br>
>>> +<br>
>>> +  self = _Scheduler_strong_APA_Get_self( context );<br>
>>> +  strong_node = _Scheduler_strong_APA_Node_downcast( node );<br>
>>> +  cpu_max = _SMP_Get_processor_maximum();<br>
>>> +  CPU = self->CPU;<br>
>>> +<br>
>>> +  for ( cpu_index = 0 ; cpu_index < cpu_max ; ++cpu_index ) {<br>
>>> +    CPU[ cpu_index ].visited = false;<br>
>>> +<br>
>>> +    //Checks if the thread_CPU is in the affinity set of the node<br>
>>> +    if ( _Processor_mask_Is_set( &strong_node->Affinity, cpu_index) ) {<br>
>>> +      Per_CPU_Control *cpu = _Per_CPU_Get_by_index( cpu_index );<br>
>>> +<br>
>>> +      if ( _Per_CPU_Is_processor_online( cpu ) ) {<br>
>>> +        rear = rear + 1;<br>
>>> +        CPU[ rear ].cpu = cpu;<br>
>>> +        CPU[ cpu_index ].visited = true;<br>
>>> +        CPU[ cpu_index ].preempting_node = node;<br>
>>> +      }<br>
>>> +    }<br>
>>> +  }<br>
>>> +<br>
>>> +  //This assert makes sure that there always exist an element in the<br>
>>> +  // Queue when we start the queue traversal.<br>
>>> +  _Assert( !_Processor_mask_Zero( &strong_node->Affinity ) );<br>
>>> +<br>
>>> +  lowest_reachable = _Scheduler_strong_APA_Get_lowest_reachable(<br>
>>> +                       self,<br>
>>> +                       front,<br>
>>> +                       rear,<br>
>>> +                       &cpu_to_preempt<br>
>>> +                     );<br>
>>> +<br>
>>> +  return _Scheduler_strong_APA_Do_enqueue(<br>
>>> +           context,<br>
>>> +           lowest_reachable,<br>
>>> +           node,<br>
>>> +           insert_priority,<br>
>>> +           cpu_to_preempt<br>
>>> +         );<br>
>>> +}<br>
>>> +<br>
>>> +static inline bool _Scheduler_strong_APA_Enqueue_scheduled(<br>
>>> +  Scheduler_Context *context,<br>
>>> +  Scheduler_Node    *node,<br>
>>> +  Priority_Control   insert_priority<br>
>>> +)<br>
>>> +{<br>
>>> +  return _Scheduler_SMP_Enqueue_scheduled(<br>
>>>      context,<br>
>>> -    the_thread,<br>
>>>      node,<br>
>>> -    _Scheduler_SMP_Extract_from_scheduled,<br>
>>> +    insert_priority,<br>
>>> +    _Scheduler_SMP_Priority_less_equal,<br>
>>>      _Scheduler_strong_APA_Extract_from_ready,<br>
>>>      _Scheduler_strong_APA_Get_highest_ready,<br>
>>> +    _Scheduler_strong_APA_Insert_ready,<br>
>>> +    _Scheduler_SMP_Insert_scheduled,<br>
>>>      _Scheduler_strong_APA_Move_from_ready_to_scheduled,<br>
>>> -    _Scheduler_SMP_Allocate_processor_exact<br>
>>> +    _Scheduler_strong_APA_Allocate_processor<br>
>>>    );<br>
>>>  }<br>
>>><br>
>>> -static bool _Scheduler_strong_APA_Enqueue(<br>
>>> +static inline bool _Scheduler_strong_APA_Do_ask_for_help(<br>
>>>    Scheduler_Context *context,<br>
>>> -  Scheduler_Node    *node,<br>
>>> -  Priority_Control   insert_priority<br>
>>> +  Thread_Control    *the_thread,<br>
>>> +  Scheduler_Node    *node<br>
>>>  )<br>
>>>  {<br>
>>> -  return _Scheduler_SMP_Enqueue(<br>
>>> +  return _Scheduler_SMP_Ask_for_help(<br>
>>>      context,<br>
>>> +    the_thread,<br>
>>>      node,<br>
>>> -    insert_priority,<br>
>>>      _Scheduler_SMP_Priority_less_equal,<br>
>>>      _Scheduler_strong_APA_Insert_ready,<br>
>>>      _Scheduler_SMP_Insert_scheduled,<br>
>>>      _Scheduler_strong_APA_Move_from_scheduled_to_ready,<br>
>>> -    _Scheduler_SMP_Get_lowest_scheduled,<br>
>>> -    _Scheduler_SMP_Allocate_processor_exact<br>
>>> +    _Scheduler_strong_APA_Get_lowest_scheduled,<br>
>>> +    _Scheduler_strong_APA_Allocate_processor<br>
>>>    );<br>
>>>  }<br>
>>><br>
>>> -static bool _Scheduler_strong_APA_Enqueue_scheduled(<br>
>>> +static inline void _Scheduler_strong_APA_Register_idle(<br>
>>>    Scheduler_Context *context,<br>
>>> -  Scheduler_Node    *node,<br>
>>> -  Priority_Control  insert_priority<br>
>>> +  Scheduler_Node    *idle_base,<br>
>>> +  Per_CPU_Control   *cpu<br>
>>>  )<br>
>>>  {<br>
>>> -  return _Scheduler_SMP_Enqueue_scheduled(<br>
>>> +  (void) context;<br>
>>> +  (void) idle_base;<br>
>>> +  (void) cpu;<br>
>>> +  /*<br>
>>> +   * We do not maintain a variable to access the scheduled<br>
>>> +   * node for a CPU. So this function does nothing.<br>
>>> +   */<br>
>>> +}<br>
>>> +<br>
>>> +static inline  void  _Scheduler_strong_APA_Do_set_affinity(<br>
>>> +  Scheduler_Context *context,<br>
>>> +  Scheduler_Node    *node_base,<br>
>>> +  void              *arg<br>
>>> +)<br>
>>> +{<br>
>>> +  Scheduler_strong_APA_Node *node;<br>
>>> +<br>
>>> +  node = _Scheduler_strong_APA_Node_downcast( node_base );<br>
>>> +  node->Affinity = *(( const Processor_mask *) arg);<br>
>>> +}<br>
>>> +<br>
>>> +void _Scheduler_strong_APA_Initialize( const Scheduler_Control *scheduler )<br>
>>> +{<br>
>>> +  Scheduler_strong_APA_Context *self =<br>
>>> +      _Scheduler_strong_APA_Get_context( scheduler );<br>
>>> +<br>
>>> +  _Scheduler_SMP_Initialize( &self->Base );<br>
>>> +  _Chain_Initialize_empty( &self->Ready );<br>
>>> +}<br>
>>> +<br>
>>> +void _Scheduler_strong_APA_Yield(<br>
>>> +  const Scheduler_Control *scheduler,<br>
>>> +  Thread_Control          *thread,<br>
>>> +  Scheduler_Node          *node<br>
>>> +)<br>
>>> +{<br>
>>> +  Scheduler_Context *context = _Scheduler_Get_context( scheduler );<br>
>>> +<br>
>>> +  _Scheduler_SMP_Yield(<br>
>>>      context,<br>
>>> +    thread,<br>
>>>      node,<br>
>>> -    insert_priority,<br>
>>> -    _Scheduler_SMP_Priority_less_equal,<br>
>>>      _Scheduler_strong_APA_Extract_from_ready,<br>
>>> -    _Scheduler_strong_APA_Get_highest_ready,<br>
>>> -    _Scheduler_strong_APA_Insert_ready,<br>
>>> -    _Scheduler_SMP_Insert_scheduled,<br>
>>> -    _Scheduler_strong_APA_Move_from_ready_to_scheduled,<br>
>>> -    _Scheduler_SMP_Allocate_processor_exact<br>
>>> +    _Scheduler_strong_APA_Enqueue,<br>
>>> +    _Scheduler_strong_APA_Enqueue_scheduled<br>
>>>    );<br>
>>>  }<br>
>>><br>
>>> -void _Scheduler_strong_APA_Unblock(<br>
>>> +void _Scheduler_strong_APA_Block(<br>
>>>    const Scheduler_Control *scheduler,<br>
>>> -  Thread_Control          *the_thread,<br>
>>> +  Thread_Control          *thread,<br>
>>>    Scheduler_Node          *node<br>
>>>  )<br>
>>>  {<br>
>>>    Scheduler_Context *context = _Scheduler_Get_context( scheduler );<br>
>>> -<br>
>>> -  _Scheduler_SMP_Unblock(<br>
>>> +  //The extract from ready automatically removes the node from Ready chain.<br>
>>> +  _Scheduler_SMP_Block(<br>
>>>      context,<br>
>>> -    the_thread,<br>
>>> +    thread,<br>
>>>      node,<br>
>>> -    _Scheduler_strong_APA_Do_update,<br>
>>> -    _Scheduler_strong_APA_Enqueue<br>
>>> +    _Scheduler_strong_APA_Extract_from_scheduled,<br>
>>> +    _Scheduler_strong_APA_Extract_from_ready,<br>
>>> +    _Scheduler_strong_APA_Get_highest_ready,<br>
>>> +    _Scheduler_strong_APA_Move_from_ready_to_scheduled,<br>
>>> +    _Scheduler_strong_APA_Allocate_processor<br>
>>>    );<br>
>>>  }<br>
>>><br>
>>> -static bool _Scheduler_strong_APA_Do_ask_for_help(<br>
>>> -  Scheduler_Context *context,<br>
>>> -  Thread_Control    *the_thread,<br>
>>> -  Scheduler_Node    *node<br>
>>> +void _Scheduler_strong_APA_Unblock(<br>
>>> +  const Scheduler_Control *scheduler,<br>
>>> +  Thread_Control          *thread,<br>
>>> +  Scheduler_Node          *node<br>
>>>  )<br>
>>>  {<br>
>>> -  return _Scheduler_SMP_Ask_for_help(<br>
>>> +  Scheduler_Context *context = _Scheduler_Get_context( scheduler );<br>
>>> +<br>
>>> +  _Scheduler_SMP_Unblock(<br>
>>>      context,<br>
>>> -    the_thread,<br>
>>> +    thread,<br>
>>>      node,<br>
>>> -    _Scheduler_SMP_Priority_less_equal,<br>
>>> -    _Scheduler_strong_APA_Insert_ready,<br>
>>> -    _Scheduler_SMP_Insert_scheduled,<br>
>>> -    _Scheduler_strong_APA_Move_from_scheduled_to_ready,<br>
>>> -    _Scheduler_SMP_Get_lowest_scheduled,<br>
>>> -    _Scheduler_SMP_Allocate_processor_lazy<br>
>>> +    _Scheduler_strong_APA_Do_update,<br>
>>> +    _Scheduler_strong_APA_Enqueue<br>
>>>    );<br>
>>>  }<br>
>>><br>
>>>  void _Scheduler_strong_APA_Update_priority(<br>
>>>    const Scheduler_Control *scheduler,<br>
>>> -  Thread_Control          *the_thread,<br>
>>> +  Thread_Control          *thread,<br>
>>>    Scheduler_Node          *node<br>
>>>  )<br>
>>>  {<br>
>>> @@ -327,7 +793,7 @@ void _Scheduler_strong_APA_Update_priority(<br>
>>><br>
>>>    _Scheduler_SMP_Update_priority(<br>
>>>      context,<br>
>>> -    the_thread,<br>
>>> +    thread,<br>
>>>      node,<br>
>>>      _Scheduler_strong_APA_Extract_from_ready,<br>
>>>      _Scheduler_strong_APA_Do_update,<br>
>>> @@ -345,7 +811,11 @@ bool _Scheduler_strong_APA_Ask_for_help(<br>
>>>  {<br>
>>>    Scheduler_Context *context = _Scheduler_Get_context( scheduler );<br>
>>><br>
>>> -  return _Scheduler_strong_APA_Do_ask_for_help( context, the_thread, node );<br>
>>> +  return _Scheduler_strong_APA_Do_ask_for_help(<br>
>>> +    context,<br>
>>> +    the_thread,<br>
>>> +    node<br>
>>> +  );<br>
>>>  }<br>
>>><br>
>>>  void _Scheduler_strong_APA_Reconsider_help_request(<br>
>>> @@ -381,7 +851,7 @@ void _Scheduler_strong_APA_Withdraw_node(<br>
>>>      _Scheduler_strong_APA_Extract_from_ready,<br>
>>>      _Scheduler_strong_APA_Get_highest_ready,<br>
>>>      _Scheduler_strong_APA_Move_from_ready_to_scheduled,<br>
>>> -    _Scheduler_SMP_Allocate_processor_lazy<br>
>>> +    _Scheduler_strong_APA_Allocate_processor<br>
>>>    );<br>
>>>  }<br>
>>><br>
>>> @@ -416,20 +886,65 @@ Thread_Control *_Scheduler_strong_APA_Remove_processor(<br>
>>>    );<br>
>>>  }<br>
>>><br>
>>> -void _Scheduler_strong_APA_Yield(<br>
>>> +void _Scheduler_strong_APA_Node_initialize(<br>
>>>    const Scheduler_Control *scheduler,<br>
>>> +  Scheduler_Node          *node,<br>
>>>    Thread_Control          *the_thread,<br>
>>> -  Scheduler_Node          *node<br>
>>> +  Priority_Control         priority<br>
>>>  )<br>
>>>  {<br>
>>> -  Scheduler_Context *context = _Scheduler_Get_context( scheduler );<br>
>>> +  Scheduler_SMP_Node *smp_node;<br>
>>> +  Scheduler_strong_APA_Node *strong_node;<br>
>>><br>
>>> -  _Scheduler_SMP_Yield(<br>
>>> -    context,<br>
>>> -    the_thread,<br>
>>> -    node,<br>
>>> -    _Scheduler_strong_APA_Extract_from_ready,<br>
>>> -    _Scheduler_strong_APA_Enqueue,<br>
>>> -    _Scheduler_strong_APA_Enqueue_scheduled<br>
>>> +  smp_node = _Scheduler_SMP_Node_downcast( node );<br>
>>> +  strong_node = _Scheduler_strong_APA_Node_downcast( node );<br>
>>> +<br>
>>> +  _Scheduler_SMP_Node_initialize( scheduler, smp_node, the_thread, priority );<br>
>>> +<br>
>>> +  _Processor_mask_Assign(<br>
>>> +    &strong_node->Affinity,<br>
>>> +   _SMP_Get_online_processors()<br>
>>>    );<br>
>>>  }<br>
>>> +<br>
>>> +bool _Scheduler_strong_APA_Set_affinity(<br>
>>> +  const Scheduler_Control *scheduler,<br>
>>> +  Thread_Control          *thread,<br>
>>> +  Scheduler_Node          *node_base,<br>
>>> +  const Processor_mask    *affinity<br>
>>> +)<br>
>>> +{<br>
>>> +  Scheduler_Context         *context;<br>
>>> +  Scheduler_strong_APA_Node *node;<br>
>>> +  Processor_mask             local_affinity;<br>
>>> +<br>
>>> +  context = _Scheduler_Get_context( scheduler );<br>
>>> +  _Processor_mask_And( &local_affinity, &context->Processors, affinity );<br>
>>> +<br>
>>> +  if ( _Processor_mask_Is_zero( &local_affinity ) ) {<br>
>>> +    return false;<br>
>>> +  }<br>
>>> +<br>
>>> +  node = _Scheduler_strong_APA_Node_downcast( node_base );<br>
>>> +<br>
>>> +  if ( _Processor_mask_Is_equal( &node->Affinity, affinity ) )<br>
>>> +    return true;       //Nothing to do. Return true.<br>
>>> +<br>
>>> + _Processor_mask_Assign( &node->Affinity, &local_affinity );<br>
>>> +<br>
>>> + _Scheduler_SMP_Set_affinity(<br>
>>> +   context,<br>
>>> +   thread,<br>
>>> +   node_base,<br>
>>> +   &local_affinity,<br>
>>> +   _Scheduler_strong_APA_Do_set_affinity,<br>
>>> +   _Scheduler_strong_APA_Extract_from_ready,<br>
>>> +   _Scheduler_strong_APA_Get_highest_ready,<br>
>>> +   _Scheduler_strong_APA_Move_from_ready_to_scheduled,<br>
>>> +   _Scheduler_strong_APA_Enqueue,<br>
>>> +   _Scheduler_strong_APA_Allocate_processor<br>
>>> + );<br>
>>> +<br>
>>> +  return true;<br>
>>> +}<br>
>>> +<br>
>>> --<br>
>>> 2.17.1<br>
>>><br>
> _______________________________________________<br>
> devel mailing list<br>
> <a href="mailto:devel@rtems.org" target="_blank">devel@rtems.org</a><br>
> <a href="http://lists.rtems.org/mailman/listinfo/devel" rel="noreferrer" target="_blank">http://lists.rtems.org/mailman/listinfo/devel</a><br>
</blockquote></div>