[linux.git] / Documentation / RCU / rcu.txt

RCU Concepts


The basic idea behind RCU (read-copy update) is to split destructive
operations into two parts, one that prevents anyone from seeing the data
item being destroyed, and one that actually carries out the destruction.
A "grace period" must elapse between the two parts, and this grace period
must be long enough that any readers accessing the item being deleted have
since dropped their references.  For example, an RCU-protected deletion
from a linked list would first remove the item from the list, wait for
a grace period to elapse, then free the element.  See the listRCU.txt
file for more information on using RCU with linked lists.


Frequently Asked Questions

o	Why would anyone want to use RCU?

	The advantage of RCU's two-part approach is that RCU readers need
	not acquire any locks, perform any atomic instructions, write to
	shared memory, or (on CPUs other than Alpha) execute any memory
	barriers.  The fact that these operations are quite expensive
	on modern CPUs is what gives RCU its performance advantages
	in read-mostly situations.  The fact that RCU readers need not
	acquire locks can also greatly simplify deadlock-avoidance code.

o	How can the updater tell when a grace period has completed
	if the RCU readers give no indication when they are done?

	Just as with spinlocks, RCU readers are not permitted to
	block, switch to user-mode execution, or enter the idle loop.
	Therefore, as soon as a CPU is seen passing through any of these
	three states, we know that that CPU has exited any previous RCU
	read-side critical sections.  So, if we remove an item from a
	linked list, and then wait until all CPUs have switched context,
	executed in user mode, or executed in the idle loop, we can
	safely free up that item.

	Preemptible variants of RCU (CONFIG_TREE_PREEMPT_RCU) get the
	same effect, but require that the readers manipulate CPU-local
	counters.  These counters allow limited types of blocking within
	RCU read-side critical sections.  SRCU also uses CPU-local
	counters, and permits general blocking within RCU read-side
	critical sections.  These variants of RCU detect grace periods
	by sampling these counters.

o	If I am running on a uniprocessor kernel, which can only do one
	thing at a time, why should I wait for a grace period?

	See the UP.txt file in this directory.

o	How can I see where RCU is currently used in the Linux kernel?

	Search for "rcu_read_lock", "rcu_read_unlock", "call_rcu",
	"rcu_read_lock_bh", "rcu_read_unlock_bh", "call_rcu_bh",
	"srcu_read_lock", "srcu_read_unlock", "synchronize_rcu",
	"synchronize_net", "synchronize_srcu", and the other RCU
	primitives.  Or grab one of the cscope databases from:

	http://www.rdrop.com/users/paulmck/RCU/linuxusage/rculocktab.html

o	What guidelines should I follow when writing code that uses RCU?

	See the checklist.txt file in this directory.

o	Why the name "RCU"?

	"RCU" stands for "read-copy update".  The file listRCU.txt has
	more information on where this name came from, search for
	"read-copy update" to find it.

o	I hear that RCU is patented?  What is with that?

	Yes, it is.  There are several known patents related to RCU,
	search for the string "Patent" in RTFP.txt to find them.
	Of these, one was allowed to lapse by the assignee, and the
	others have been contributed to the Linux kernel under GPL.
	There are now also LGPL implementations of user-level RCU
	available (http://lttng.org/?q=node/18).

o	I hear that RCU needs work in order to support realtime kernels?

	This work is largely completed.  Realtime-friendly RCU can be
	enabled via the CONFIG_TREE_PREEMPT_RCU kernel configuration
	parameter.  However, work is in progress for enabling priority
	boosting of preempted RCU read-side critical sections.	This is
	needed if you have CPU-bound realtime threads.

o	Where can I find more information on RCU?

	See the RTFP.txt file in this directory.
	Or point your browser at http://www.rdrop.com/users/paulmck/RCU/.

o	What are all these files in this directory?

	See 00-INDEX for the list.
Commit	Line	Data
1da177e4 LT	1	RCU Concepts
	2
	3
	4	The basic idea behind RCU (read-copy update) is to split destructive
	5	operations into two parts, one that prevents anyone from seeing the data
	6	item being destroyed, and one that actually carries out the destruction.
	7	A "grace period" must elapse between the two parts, and this grace period
	8	must be long enough that any readers accessing the item being deleted have
	9	since dropped their references. For example, an RCU-protected deletion
	10	from a linked list would first remove the item from the list, wait for
	11	a grace period to elapse, then free the element. See the listRCU.txt
	12	file for more information on using RCU with linked lists.
	13
	14
	15	Frequently Asked Questions
	16
	17	o Why would anyone want to use RCU?
	18
	19	The advantage of RCU's two-part approach is that RCU readers need
	20	not acquire any locks, perform any atomic instructions, write to
	21	shared memory, or (on CPUs other than Alpha) execute any memory
	22	barriers. The fact that these operations are quite expensive
	23	on modern CPUs is what gives RCU its performance advantages
	24	in read-mostly situations. The fact that RCU readers need not
	25	acquire locks can also greatly simplify deadlock-avoidance code.
	26
	27	o How can the updater tell when a grace period has completed
	28	if the RCU readers give no indication when they are done?
	29
	30	Just as with spinlocks, RCU readers are not permitted to
	31	block, switch to user-mode execution, or enter the idle loop.
	32	Therefore, as soon as a CPU is seen passing through any of these
	33	three states, we know that that CPU has exited any previous RCU
	34	read-side critical sections. So, if we remove an item from a
	35	linked list, and then wait until all CPUs have switched context,
	36	executed in user mode, or executed in the idle loop, we can
	37	safely free up that item.
	38
6b3ef48a	39	Preemptible variants of RCU (CONFIG_TREE_PREEMPT_RCU) get the
f85d6c71	40	same effect, but require that the readers manipulate CPU-local
9ceae0e2 PM	41	counters. These counters allow limited types of blocking within
	42	RCU read-side critical sections. SRCU also uses CPU-local
	43	counters, and permits general blocking within RCU read-side
	44	critical sections. These variants of RCU detect grace periods
	45	by sampling these counters.
f85d6c71	46
1da177e4 LT	47	o If I am running on a uniprocessor kernel, which can only do one
	48	thing at a time, why should I wait for a grace period?
	49
	50	See the UP.txt file in this directory.
	51
	52	o How can I see where RCU is currently used in the Linux kernel?
	53
a83f1fe2 PM	54	Search for "rcu_read_lock", "rcu_read_unlock", "call_rcu",
a83f1fe2 PM	55	"rcu_read_lock_bh", "rcu_read_unlock_bh", "call_rcu_bh",
621934ee	56	"srcu_read_lock", "srcu_read_unlock", "synchronize_rcu",
f85d6c71 PM	57	"synchronize_net", "synchronize_srcu", and the other RCU
	58	primitives. Or grab one of the cscope databases from:
	59
	60	http://www.rdrop.com/users/paulmck/RCU/linuxusage/rculocktab.html
1da177e4 LT	61
	62	o What guidelines should I follow when writing code that uses RCU?
	63
	64	See the checklist.txt file in this directory.
	65
	66	o Why the name "RCU"?
	67
	68	"RCU" stands for "read-copy update". The file listRCU.txt has
	69	more information on where this name came from, search for
	70	"read-copy update" to find it.
	71
	72	o I hear that RCU is patented? What is with that?
	73
	74	Yes, it is. There are several known patents related to RCU,
	75	search for the string "Patent" in RTFP.txt to find them.
	76	Of these, one was allowed to lapse by the assignee, and the
	77	others have been contributed to the Linux kernel under GPL.
4c54005c PM	78	There are now also LGPL implementations of user-level RCU
4c54005c PM	79	available (http://lttng.org/?q=node/18).
1da177e4	80
dd81eca8 PM	81	o I hear that RCU needs work in order to support realtime kernels?
dd81eca8 PM	82
f85d6c71	83	This work is largely completed. Realtime-friendly RCU can be
6b3ef48a PM	84	enabled via the CONFIG_TREE_PREEMPT_RCU kernel configuration
	85	parameter. However, work is in progress for enabling priority
	86	boosting of preempted RCU read-side critical sections. This is
	87	needed if you have CPU-bound realtime threads.
dd81eca8	88
1da177e4 LT	89	o Where can I find more information on RCU?
	90
	91	See the RTFP.txt file in this directory.
dd81eca8 PM	92	Or point your browser at http://www.rdrop.com/users/paulmck/RCU/.
	93
	94	o What are all these files in this directory?
	95
4c54005c	96	See 00-INDEX for the list.