Changeset 06496fb in rtems


Ignore:
Timestamp:
Sep 4, 2015, 11:09:17 AM (4 years ago)
Author:
Sebastian Huber <sebastian.huber@…>
Branches:
master
Children:
cee82df
Parents:
41b5c915
Message:

smp: Documentation

Location:
doc/user
Files:
3 edited

Legend:

Unmodified
Added
Removed
  • doc/user/conf.t

    r41b5c915 r06496fb  
    42044204
    42054205@c
    4206 @c === Configuring Clustered/Partitioned Schedulers ===
    4207 @c
    4208 @subsection Configuring Clustered/Partitioned Schedulers
    4209 
    4210 Clustered/partitioned scheduling helps to control the worst-case latencies in
    4211 the system.  The goal is to reduce the amount of shared state in the system and
    4212 thus prevention of lock contention.  Modern multi-processor systems tend to
    4213 have several layers of data and instruction caches.  With clustered/partitioned
    4214 scheduling it is possible to honour the cache topology of a system and thus
    4215 avoid expensive cache synchronization traffic.
     4206@c === Configuring Clustered Schedulers ===
     4207@c
     4208@subsection Configuring Clustered Schedulers
     4209
     4210Clustered scheduling helps to control the worst-case latencies in a
     4211multi-processor system.  The goal is to reduce the amount of shared state in
     4212the system and thus prevention of lock contention.  Modern multi-processor
     4213systems tend to have several layers of data and instruction caches.  With
     4214clustered scheduling it is possible to honour the cache topology of a system
     4215and thus avoid expensive cache synchronization traffic.
    42164216
    42174217We have clustered scheduling in case the set of processors of a system is
    42184218partitioned into non-empty pairwise-disjoint subsets.  These subsets are called
    42194219clusters.  Clusters with a cardinality of one are partitions.  Each cluster is
    4220 owned by exactly one scheduler instance.  In order to use clustered/partitioned
     4220owned by exactly one scheduler instance.  In order to use clustered
    42214221scheduling the application designer has to answer two questions.
    42224222
    42234223@enumerate
    4224 @item How is the set of processors partitioned into clusters/partitions?
    4225 @item Which scheduler is used for which cluster/partition?
     4224@item How is the set of processors partitioned into clusters?
     4225@item Which scheduler is used for which cluster?
    42264226@end enumerate
    42274227
  • doc/user/glossary.texi

    r41b5c915 r06496fb  
    9393We have clustered scheduling in case the set of processors of a system is
    9494partitioned into non-empty pairwise disjoint subsets.  These subsets are called
    95 @dfn{clusters}.  Each cluster is owned by exactly one scheduler instance.
     95@dfn{clusters}.  Clusters with a cardinality of one are partitions.  Each
     96cluster is owned by exactly one scheduler instance.
    9697
    9798@item coalesce
     
    648649at most one scheduler instance.  The processor to scheduler instance assignment
    649650is determined at application configuration time.  @xref{Configuring a System
    650 Configuring Clustered/Partitioned Schedulers}.
     651Configuring Clustered Schedulers}.
    651652
    652653@item segments
  • doc/user/smp.t

    r41b5c915 r06496fb  
    590590
    591591Identifies a scheduler by its name.  The scheduler name is determined by the
    592 scheduler configuration.  @xref{Configuring a System Configuring
    593 Clustered/Partitioned Schedulers}.
     592scheduler configuration.  @xref{Configuring a System Configuring Clustered
     593Schedulers}.
    594594
    595595@subheading NOTES:
Note: See TracChangeset for help on using the changeset viewer.