Re: [patch 4/4 v2] mempolicy: update NUMA memory policy documentation

[Randy Dunlap <randy.dunlap@oracle.com>]

David Rientjes wrote:
> Updates Documentation/vm/numa_memory_policy.txt and
> Documentation/filesystems/tmpfs.txt to describe optional mempolicy mode
> flags.
> 
> Cc: Paul Jackson <pj@sgi.com>
> Cc: Christoph Lameter <clameter@sgi.com>
> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
> Cc: Andi Kleen <ak@suse.de>
> Cc: Randy Dunlap <randy.dunlap@oracle.com>
> Signed-off-by: David Rientjes <rientjes@google.com>

Acked-by: Randy Dunlap <randy.dunlap@oracle.com>

Thanks.

> ---
>  Includes fixes to problems identified by Randy Dunlap.
> 
>  Documentation/filesystems/tmpfs.txt     |   11 ++++++++
>  Documentation/vm/numa_memory_policy.txt |   41 +++++++++++++++++++++++++++----
>  2 files changed, 47 insertions(+), 5 deletions(-)
> 
> diff --git a/Documentation/filesystems/tmpfs.txt b/Documentation/filesystems/tmpfs.txt
> --- a/Documentation/filesystems/tmpfs.txt
> +++ b/Documentation/filesystems/tmpfs.txt
> @@ -92,6 +92,17 @@ NodeList format is a comma-separated list of decimal numbers and ranges,
>  a range being two hyphen-separated decimal numbers, the smallest and
>  largest node numbers in the range.  For example, mpol=bind:0-3,5,7,9-15
>  
> +It is possible to specify a static NodeList by appending '=static' to
> +the memory policy mode in the mpol= argument.  This will require that
> +tasks or VMA's restricted to a subset of allowed nodes are only allowed
> +to effect the memory policy over those nodes.  No remapping of the
> +NodeList when the policy is rebound, which is the default behavior, is
> +allowed when '=static' is specified.  For example:
> +
> +mpol=bind=static:NodeList	will only allocate from each node in
> +				the NodeList without remapping the
> +				NodeList if the policy is rebound
> +
>  Note that trying to mount a tmpfs with an mpol option will fail if the
>  running kernel does not support NUMA; and will fail if its nodelist
>  specifies a node which is not online.  If your system relies on that
> diff --git a/Documentation/vm/numa_memory_policy.txt b/Documentation/vm/numa_memory_policy.txt
> --- a/Documentation/vm/numa_memory_policy.txt
> +++ b/Documentation/vm/numa_memory_policy.txt
> @@ -135,9 +135,11 @@ most general to most specific:
>  
>  Components of Memory Policies
>  
> -    A Linux memory policy is a tuple consisting of a "mode" and an optional set
> -    of nodes.  The mode determine the behavior of the policy, while the
> -    optional set of nodes can be viewed as the arguments to the behavior.
> +    A Linux memory policy consists of a "mode", optional mode flags, and an
> +    optional set of nodes.  The mode determines the behavior of the policy,
> +    the optional mode flags determine the behavior of the mode, and the
> +    optional set of nodes can be viewed as the arguments to the policy
> +    behavior.
>  
>     Internally, memory policies are implemented by a reference counted
>     structure, struct mempolicy.  Details of this structure will be discussed
> @@ -145,7 +147,12 @@ Components of Memory Policies
>  
>  	Note:  in some functions AND in the struct mempolicy itself, the mode
>  	is called "policy".  However, to avoid confusion with the policy tuple,
> -	this document will continue to use the term "mode".
> +	this document will continue to use the term "mode".  Since the mode and
> +	optional mode flags are stored in the same struct mempolicy member
> +	(specifically, pol->policy), you must use mpol_mode(pol->policy) to
> +	access only the mode and mpol_flags(pol->policy) to access only the
> +	flags.  Any function with a formal of type enum mempolicy_mode only
> +	refers to the mode.
>  
>     Linux memory policy supports the following 4 behavioral modes:
>  
> @@ -231,6 +238,28 @@ Components of Memory Policies
>  	    the temporary interleaved system default policy works in this
>  	    mode.
>  
> +   Linux memory policy supports the following optional mode flag:
> +
> +	MPOL_F_STATIC_NODES:  This flag specifies that the nodemask passed by
> +	the user should not be remapped if the task or VMA's set of accessible
> +	nodes changes after the memory policy has been defined.
> +
> +	    Without this flag, anytime a mempolicy is rebound because of a
> +	    change in the set of accessible nodes, the node (Preferred) or
> +	    nodemask (Bind, Interleave) is remapped to the new set of
> +	    accessible nodes.  This may result in nodes being used that were
> +	    previously undesired.  With this flag, the policy is either
> +	    effected over the user's specified nodemask or the Default
> +	    behavior is used.
> +
> +	    For example, consider a task that is attached to a cpuset with
> +	    mems 1-3 that sets an Interleave policy over the same set.  If
> +	    the cpuset's mems change to 3-5, the Interleave will now occur
> +	    over nodes 3, 4, and 5.  With this flag, however, since only
> +	    node 3 is accessible from the user's nodemask, the "interleave"
> +	    only occurs over that node.  If no nodes from the user's
> +	    nodemask are now accessible, the Default behavior is used.
> +
>  MEMORY POLICY APIs
>  
>  Linux supports 3 system calls for controlling memory policy.  These APIS
> @@ -251,7 +280,9 @@ Set [Task] Memory Policy:
>  	Set's the calling task's "task/process memory policy" to mode
>  	specified by the 'mode' argument and the set of nodes defined
>  	by 'nmask'.  'nmask' points to a bit mask of node ids containing
> -	at least 'maxnode' ids.
> +	at least 'maxnode' ids.  Optional mode flags may be passed by
> +	combining the 'mode' argument with the flag (for example:
> +	MPOL_INTERLEAVE | MPOL_F_STATIC_NODES).
>  
>  	See the set_mempolicy(2) man page for more details
>  


-- 
~Randy