uma (9)

Leading comments

-
 Copyright (c) 2001 Dag-Erling Coïdan Smørgrav
 All rights reserved.

 Redistribution and use in source and binary forms, with or without
 modification, are permitted provided that the following conditions
 are met:
 1. Redistributions of source code must retain the above copyright
    notice, this list of conditions and the following disclaimer.
 2. Redistributions in binary form must reproduce the above copyright
    notice, this list of conditions and the following disclaimer in the
    ...

(The comments found at the beginning of the groff file "man9/uma.9freebsd".)

NAME

uma_zcreate uma_zalloc uma_zalloc_arg uma_zfree uma_zfree_arg uma_find_refcnt uma_zdestroy uma_zone_set_max, uma_zone_get_max, uma_zone_get_cur, uma_zone_set_warning - zone allocator

SYNOPSIS

In sys/param.h In sys/queue.h In vm/uma.h Ft uma_zone_t Fo uma_zcreate Fa char *name int size Fa uma_ctor ctor uma_dtor dtor uma_init uminit uma_fini fini Fa int align uint16_t flags Fc Ft void * Fn uma_zalloc uma_zone_t zone int flags Ft void * Fn uma_zalloc_arg uma_zone_t zone void *arg int flags Ft void Fn uma_zfree uma_zone_t zone void *item Ft void Fn uma_zfree_arg uma_zone_t zone void *item void *arg Ft uint32_t * Fn uma_find_refcnt uma_zone_t zone void *item Ft void Fn uma_zdestroy uma_zone_t zone Ft int Fn uma_zone_set_max uma_zone_t zone int nitems Ft int Fn uma_zone_get_max uma_zone_t zone Ft int Fn uma_zone_get_cur uma_zone_t zone Ft void Fn uma_zone_set_warning uma_zone_t zone const char *warning In sys/sysctl.h Fn SYSCTL_UMA_MAX parent nbr name access zone descr Fn SYSCTL_ADD_UMA_MAX ctx parent nbr name access zone descr Fn SYSCTL_UMA_CUR parent nbr name access zone descr Fn SYSCTL_ADD_UMA_CUR ctx parent nbr name access zone descr

DESCRIPTION

The zone allocator provides an efficient interface for managing dynamically-sized collections of items of similar size. The zone allocator can work with preallocated zones as well as with runtime-allocated ones, and is therefore available much earlier in the boot process than other memory management routines.

A zone is an extensible collection of items of identical size. The zone allocator keeps track of which items are in use and which are not, and provides functions for allocating items from the zone and for releasing them back (which makes them available for later use).

After the first allocation of an item, it will have been cleared to zeroes, however subsequent allocations will retain the contents as of the last free.

The Fn uma_zcreate function creates a new zone from which items may then be allocated from. The Fa name argument is a text name of the zone for debugging and stats; this memory should not be freed until the zone has been deallocated.

The Fa ctor and Fa dtor arguments are callback functions that are called by the uma subsystem at the time of the call to Fn uma_zalloc and Fn uma_zfree respectively. Their purpose is to provide hooks for initializing or destroying things that need to be done at the time of the allocation or release of a resource. A good usage for the Fa ctor and Fa dtor callbacks might be to adjust a global count of the number of objects allocated.

The Fa uminit and Fa fini arguments are used to optimize the allocation of objects from the zone. They are called by the uma subsystem whenever it needs to allocate or free several items to satisfy requests or memory pressure. A good use for the Fa uminit and Fa fini callbacks might be to initialize and destroy mutexes contained within the object. This would allow one to re-use already initialized mutexes when an object is returned from the uma subsystem's object cache. They are not called on each call to Fn uma_zalloc and Fn uma_zfree but rather in a batch mode on several objects.

The Fa flags argument of the Fn uma_zcreate is a subset of the following flags:

UMA_ZONE_NOFREE

Slabs of the zone are never returned back to VM.

UMA_ZONE_REFCNT

Each item in the zone would have internal reference counter associated with it. See Fn uma_find_refcnt .

UMA_ZONE_NODUMP

Pages belonging to the zone will not be included into mini-dumps.

UMA_ZONE_PCPU

An allocation from zone would have mp_ncpu shadow copies, that are privately assigned to CPUs. A CPU can address its private copy using base allocation address plus multiple of current CPU id and Fn sizeof struct pcpu :

foo_zone = uma_zcreate(..., UMA_ZONE_PCPU);
 ...
foo_base = uma_zalloc(foo_zone, ...);
 ...
critical_enter();
foo_pcpu = (foo_t *)zpcpu_get(foo_base);
/* do something with foo_pcpu */
critical_exit();

UMA_ZONE_OFFPAGE

By default book-keeping of items within a slab is done in the slab page itself. This flag explicitly tells subsystem that book-keeping structure should be allocated separately from special internal zone. This flag requires either UMA_ZONE_VTOSLAB or UMA_ZONE_HASH since subsystem requires a mechanism to find a book-keeping structure to an item beeing freed. The subsystem may choose to prefer offpage book-keeping for certain zones implicitly.

UMA_ZONE_ZINIT

The zone will have its Ft uma_init method set to internal method that initializes a new allocated slab to all zeros. Do not mistake Ft uma_init method with Ft uma_ctor . A zone with UMA_ZONE_ZINIT flag would not return zeroed memory on every Fn uma_zalloc .

UMA_ZONE_HASH

The zone should use an internal hash table to find slab book-keeping structure where an allocation being freed belongs to.

UMA_ZONE_VTOSLAB

The zone should use special field of Vt vm_page_t to find slab book-keeping structure where an allocation being freed belongs to.

UMA_ZONE_MALLOC

The zone is for the malloc(9) subsystem.

UMA_ZONE_VM

The zone is for the VM subsystem.

To allocate an item from a zone, simply call Fn uma_zalloc with a pointer to that zone and set the Fa flags argument to selected flags as documented in malloc(9). It will return a pointer to an item if successful, or NULL in the rare case where all items in the zone are in use and the allocator is unable to grow the zone and M_NOWAIT is specified.

Items are released back to the zone from which they were allocated by calling Fn uma_zfree with a pointer to the zone and a pointer to the item. If Fa item is NULL then Fn uma_zfree does nothing.

The variations Fn uma_zalloc_arg and Fn uma_zfree_arg allow to specify an argument for the ctor and dtor functions, respectively.

If zone was created with UMA_ZONE_REFCNT flag, then pointer to reference counter for an item can be retrieved with help of the Fn uma_find_refcnt function.

Created zones, which are empty, can be destroyed using Fn uma_zdestroy , freeing all memory that was allocated for the zone. All items allocated from the zone with Fn uma_zalloc must have been freed with Fn uma_zfree before.

The Fn uma_zone_set_max function limits the number of items (and therefore memory) that can be allocated to Fa zone . The Fa nitems argument specifies the requested upper limit number of items. The effective limit is returned to the caller, as it may end up being higher than requested due to the implementation rounding up to ensure all memory pages allocated to the zone are utilised to capacity. The limit applies to the total number of items in the zone, which includes allocated items, free items and free items in the per-cpu caches. On systems with more than one CPU it may not be possible to allocate the specified number of items even when there is no shortage of memory, because all of the remaining free items may be in the caches of the other CPUs when the limit is hit.

The Fn uma_zone_get_max function returns the effective upper limit number of items for a zone.

The Fn uma_zone_get_cur function returns the approximate current occupancy of the zone. The returned value is approximate because appropriate synchronisation to determine an exact value is not performed by the implementation. This ensures low overhead at the expense of potentially stale data being used in the calculation.

The Fn uma_zone_set_warning function sets a warning that will be printed on the system console when the given zone becomes full and fails to allocate an item. The warning will be printed not often than every five minutes. Warnings can be turned off globally by setting the vm.zone_warnings sysctl tunable to 0

The Fn SYSCTL_UMA_MAX parent nbr name access zone descr macro declares a static sysctl oid that exports the effective upper limit number of items for a zone. The Fa zone argument should be a pointer to Vt uma_zone_t . A read of the oid returns value obtained through Fn uma_zone_get_max . A write to the oid sets new value via Fn uma_zone_set_max . The Fn SYSCTL_ADD_UMA_MAX ctx parent nbr name access zone descr macro is provided to create this type of oid dynamically.

The Fn SYSCTL_UMA_CUR parent nbr name access zone descr macro declares a static read only sysctl oid that exports the approximate current occupancy of the zone. The Fa zone argument should be a pointer to Vt uma_zone_t . A read of the oid returns value obtained through Fn uma_zone_get_cur . The Fn SYSCTL_ADD_UMA_CUR ctx parent nbr name zone descr macro is provided to create this type of oid dynamically.

RETURN VALUES

The Fn uma_zalloc function returns a pointer to an item, or NULL if the zone ran out of unused items and M_NOWAIT was specified.

SEE ALSO

malloc(9)

HISTORY

The zone allocator first appeared in Fx 3.0 . It was radically changed in Fx 5.0 to function as a slab allocator.

AUTHORS

An -nosplit The zone allocator was written by An John S. Dyson . The zone allocator was rewritten in large parts by An Jeff Roberson Aq jeff@FreeBSD.org to function as a slab allocator.

This manual page was written by An Dag-Erling Sm/orgrav Aq des@FreeBSD.org . Changes for UMA by An Jeroen Ruigrok van der Werven Aq asmodai@FreeBSD.org .