General Numa functions. More...

Functions
errval_t	numa_available (void)
	checks if numa support is available More...

nodeid_t	numa_max_node (void)
	returns the highest node number available on the current system. More...

coreid_t	numa_max_core (void)
	returns the highest ID of the present cores More...

nodeid_t	numa_current_node (void)
	returns the current node the domain is running on More...

nodeid_t	numa_num_possible_nodes (void)
	returns the size of the node mask More...

nodeid_t	numa_num_configured_nodes (void)
	Obtains the number of all memory nodes in the system. More...

struct bitmap *	numa_get_mems_allowed (void)
	obtains the nodes the domain is allowed to allocate memory from More...

coreid_t	numa_num_configured_cpus (void)
	returns the total numberof CPUs in the system More...

coreid_t	numa_num_task_cpus (void)
	returns the number of cpus that the calling domain is allowed to use. More...

nodeid_t	numa_num_task_nodes (void)
	returns the number of nodes on which the calling domain is allowed to allocate memory More...

size_t	numa_node_size (nodeid_t node, uintptr_t *freep)
	obtains the size of a node More...

lpaddr_t	numa_node_base (nodeid_t node)
	obtains the base address of the numa node More...

nodeid_t	numa_preferred (void)
	returns the preferred node of the current task. More...

void	numa_set_preferred (nodeid_t node)
	sets the preferred node for the current task to node More...

errval_t	numa_run_on_node (nodeid_t node)
	runs the current domain on a specific node. More...

errval_t	numa_run_on_node_mask (struct bitmap *nodemask)
	runs the current domain only on nodes specified in nodemask. More...

struct bitmap *	numa_get_run_node_mask (void)
	returns a mask of CPUs on which the current task is allowed to run. More...

void	numa_set_bind_policy (numa_policy_t strict)
	specify the memory bind policy More...

void	numa_set_strict (numa_policy_t strict)
	enable or disable the strict allocation policy More...

uint32_t	numa_distance (nodeid_t from, nodeid_t to)
	reports the distance in the machine topology between two nodes More...

errval_t	numa_sched_getaffinity (domainid_t did, struct bitmap *mask)
	retrieves a bitmask of the cpus on which a domain may run More...

errval_t	numa_sched_setaffinity (domainid_t did, struct bitmap *mask)
	sets a domain's allowed cpu's to those cpu's specified in mask. More...

size_t	numa_pagesize (void)
	returns the page size More...

errval_t	numa_node_to_cpus (nodeid_t node, struct bitmap *mask)
	converts a node number to a bitmask of CPUs More...

nodeid_t	numa_node_of_cpu (coreid_t cpu)
	returns the node that a cpu belongs to More...

coreid_t	numa_num_node_cpus (nodeid_t node)
	gets the number of cores for the given numa node More...

Variables
struct bitmap *	numa_all_nodes_ptr
	bitmask that is allocated by the library with bits representing all nodes on which the calling task may allocate memory.

struct bitmap *	numa_no_nodes_ptr
	points to a bitmask that is allocated by the library and left all zeroes.

struct bitmap *	numa_all_cpus_ptr
	points to a bitmask that is allocated by the library with bits representing all cpus on which the calling task may execute.

struct numa_topology	numa_topology
	data structure representing the numa topology More...

Detailed Description

General Numa functions.

Function Documentation

errval_t numa_available ( void )

checks if numa support is available

Returns: NUMA_ERR_NOT_AVAILABLE value all other functions are undefined SYS_ERR_OK: NUMA functionality is available

this function must be called before any of the other functions of libnuma. during the call to numa_available the library also gets initialized

nodeid_t numa_current_node ( void )

returns the current node the domain is running on

Returns: ID of the current node

uint32_t numa_distance	(	nodeid_t	from,
		nodeid_t	to
	)

reports the distance in the machine topology between two nodes

Parameters

from	source node to measure the distance
to	target node to measure the distance

Returns: distance between two nodes 0 iff cannot be deterimed

The factors are a multiple of 10. A node has distance 10 to itself.

struct bitmap* numa_get_mems_allowed ( void )

obtains the nodes the domain is allowed to allocate memory from

Returns: bitmask representing the allowing nodes

returns the mask of nodes from which the process is allowed to allocate memory in it's current cpuset context.

struct bitmap* numa_get_run_node_mask ( void )

returns a mask of CPUs on which the current task is allowed to run.

Returns: bitmap represening the coreids the domain is allowed to run

coreid_t numa_max_core ( void )

returns the highest ID of the present cores

Returns: the maximum number of cores in the system

nodeid_t numa_max_node ( void )

returns the highest node number available on the current system.

Returns: ID of the max NUMA node

lpaddr_t numa_node_base ( nodeid_t node )

obtains the base address of the numa node

Returns: physical address of the start of the numa node NUMA_NODE_INVALID if the node does not exist

nodeid_t numa_node_of_cpu ( coreid_t cpu )

returns the node that a cpu belongs to

Parameters

cpu	ID of the core

Returns: node ID on SUCCESS NUMA_NODE_INVALID on FAILURE

size_t numa_node_size	(	nodeid_t	node,
		uintptr_t *	freep
	)

obtains the size of a node

Parameters

node	ID of the NUMA node
freep	returns the number of available bytes of the node

Returns: size of the node in bytes

returns the memory size of a node. If the argument freep is not NULL, it used to return the amount of free memory on the node. On error it returns NUMA_NODE_INVALID

errval_t numa_node_to_cpus	(	nodeid_t	node,
		struct bitmap *	mask
	)

converts a node number to a bitmask of CPUs

Parameters

node	the ID of the node
mask	bitmap representing the CPUs of this node

Returns: SYS_ERR_OK on SUCCESS NUMA_ERR_BITMAP_RANGE on FAILURE (too small bitmap)

The user must pass a bitmask structure with a mask buffer long enough to represent all possible cpu's

coreid_t numa_num_configured_cpus ( void )

returns the total numberof CPUs in the system

Returns: total number of CPUs in the system

returns the number of cpus in the system. This count includes any cpus that are currently disabled.

nodeid_t numa_num_configured_nodes ( void )

Obtains the number of all memory nodes in the system.

Returns: number of memory nodes in the system

returns the number of memory nodes in the system. This count includes any nodes that are currently disabled.

coreid_t numa_num_node_cpus ( nodeid_t node )

gets the number of cores for the given numa node

Parameters

node	NUMA node to get the number of cores

Returns: number of cores for the node

nodeid_t numa_num_possible_nodes ( void )

returns the size of the node mask

Returns: size of the node mask

coreid_t numa_num_task_cpus ( void )

returns the number of cpus that the calling domain is allowed to use.

Returns: number of CPUs the domain is allowed to use

nodeid_t numa_num_task_nodes ( void )

returns the number of nodes on which the calling domain is allowed to allocate memory

Returns: number of nodes the domain is allowed to use

size_t numa_pagesize ( void )

returns the page size

Returns: the number of bytes in a page

nodeid_t numa_preferred ( void )

returns the preferred node of the current task.

Returns: node ID where memory is preferably allocated

errval_t numa_run_on_node ( nodeid_t node )

runs the current domain on a specific node.

Parameters

node	ID of the node to run the domain on

Returns: SYS_ERR_OK on SUCCESS errval on FAILURE

Passing -1 permits the kernel to schedule on all nodes again

errval_t numa_run_on_node_mask ( struct bitmap * nodemask )

runs the current domain only on nodes specified in nodemask.

Parameters

nodemask bitmap representing the nodes to run the domain on

Returns: SYS_ERR_OK on SUCCESS errval on FAILURE

errval_t numa_sched_getaffinity	(	domainid_t	did,
		struct bitmap *	mask
	)

retrieves a bitmask of the cpus on which a domain may run

Parameters

did	domain ID
mask	returned bitmask

Returns: SYS_ERR_OK on success errval on FAILURE

errval_t numa_sched_setaffinity	(	domainid_t	did,
		struct bitmap *	mask
	)

sets a domain's allowed cpu's to those cpu's specified in mask.

Parameters

did	domain ID
mask	bitmap representing the CPUs

Returns: SYS_ERR_OK on success errval on FAILURE

void numa_set_bind_policy ( numa_policy_t strict )

specify the memory bind policy

Parameters

strict numa policy to apply

specifies whether calls that bind memory to a specific node should use the preferred policy or a strict policy.

void numa_set_preferred ( nodeid_t node )

sets the preferred node for the current task to node

Parameters

node	ID of the node to set preferred

The system will attempt to allocate memory from the preferred node, but will fall back to other nodes if no memory is available on the the preferred node

Passing a node of -1 argument specifies local allocation

void numa_set_strict ( numa_policy_t strict )

enable or disable the strict allocation policy

Parameters

strict numa policy to apply

s a flag that says whether the functions allocating on specific nodes should use a strict policy. Strict means the allocation will fail if the memory cannot be allocated on the target node.

Variable Documentation

struct numa_topology numa_topology

data structure representing the numa topology

numa interleave mask for allocations

Functions

Variables

Detailed Description

Function Documentation

Variable Documentation