blob: 1d50cf0c905ef28d62d6e5eda83c6415721504a6 [file] [log] [blame]
Review Checklist for RCU Patches
This document contains a checklist for producing and reviewing patches
that make use of RCU. Violating any of the rules listed below will
result in the same sorts of problems that leaving out a locking primitive
would cause. This list is based on experiences reviewing such patches
over a rather long period of time, but improvements are always welcome!
0. Is RCU being applied to a read-mostly situation? If the data
structure is updated more than about 10% of the time, then
you should strongly consider some other approach, unless
detailed performance measurements show that RCU is nonetheless
the right tool for the job.
The other exception would be where performance is not an issue,
and RCU provides a simpler implementation. An example of this
situation is the dynamic NMI code in the Linux 2.6 kernel,
at least on architectures where NMIs are rare.
1. Does the update code have proper mutual exclusion?
RCU does allow -readers- to run (almost) naked, but -writers- must
still use some sort of mutual exclusion, such as:
a. locking,
b. atomic operations, or
c. restricting updates to a single task.
If you choose #b, be prepared to describe how you have handled
memory barriers on weakly ordered machines (pretty much all of
them -- even x86 allows reads to be reordered), and be prepared
to explain why this added complexity is worthwhile. If you
choose #c, be prepared to explain how this single task does not
become a major bottleneck on big multiprocessor machines (for
example, if the task is updating information relating to itself
that other tasks can read, there by definition can be no
2. Do the RCU read-side critical sections make proper use of
rcu_read_lock() and friends? These primitives are needed
to suppress preemption (or bottom halves, in the case of
rcu_read_lock_bh()) in the read-side critical sections,
and are also an excellent aid to readability.
As a rough rule of thumb, any dereference of an RCU-protected
pointer must be covered by rcu_read_lock() or rcu_read_lock_bh()
or by the appropriate update-side lock.
3. Does the update code tolerate concurrent accesses?
The whole point of RCU is to permit readers to run without
any locks or atomic operations. This means that readers will
be running while updates are in progress. There are a number
of ways to handle this concurrency, depending on the situation:
a. Make updates appear atomic to readers. For example,
pointer updates to properly aligned fields will appear
atomic, as will individual atomic primitives. Operations
performed under a lock and sequences of multiple atomic
primitives will -not- appear to be atomic.
This is almost always the best approach.
b. Carefully order the updates and the reads so that
readers see valid data at all phases of the update.
This is often more difficult than it sounds, especially
given modern CPUs' tendency to reorder memory references.
One must usually liberally sprinkle memory barriers
(smp_wmb(), smp_rmb(), smp_mb()) through the code,
making it difficult to understand and to test.
It is usually better to group the changing data into
a separate structure, so that the change may be made
to appear atomic by updating a pointer to reference
a new structure containing updated values.
4. Weakly ordered CPUs pose special challenges. Almost all CPUs
are weakly ordered -- even i386 CPUs allow reads to be reordered.
RCU code must take all of the following measures to prevent
memory-corruption problems:
a. Readers must maintain proper ordering of their memory
accesses. The rcu_dereference() primitive ensures that
the CPU picks up the pointer before it picks up the data
that the pointer points to. This really is necessary
on Alpha CPUs. If you don't believe me, see:
The rcu_dereference() primitive is also an excellent
documentation aid, letting the person reading the code
know exactly which pointers are protected by RCU.
The rcu_dereference() primitive is used by the various
"_rcu()" list-traversal primitives, such as the
list_for_each_entry_rcu(). Note that it is perfectly
legal (if redundant) for update-side code to use
rcu_dereference() and the "_rcu()" list-traversal
primitives. This is particularly useful in code
that is common to readers and updaters.
b. If the list macros are being used, the list_add_tail_rcu()
and list_add_rcu() primitives must be used in order
to prevent weakly ordered machines from misordering
structure initialization and pointer planting.
Similarly, if the hlist macros are being used, the
hlist_add_head_rcu() primitive is required.
c. If the list macros are being used, the list_del_rcu()
primitive must be used to keep list_del()'s pointer
poisoning from inflicting toxic effects on concurrent
readers. Similarly, if the hlist macros are being used,
the hlist_del_rcu() primitive is required.
The list_replace_rcu() primitive may be used to
replace an old structure with a new one in an
RCU-protected list.
d. Updates must ensure that initialization of a given
structure happens before pointers to that structure are
publicized. Use the rcu_assign_pointer() primitive
when publicizing a pointer to a structure that can
be traversed by an RCU read-side critical section.
5. If call_rcu(), or a related primitive such as call_rcu_bh(),
is used, the callback function must be written to be called
from softirq context. In particular, it cannot block.
6. Since synchronize_rcu() can block, it cannot be called from
any sort of irq context.
7. If the updater uses call_rcu(), then the corresponding readers
must use rcu_read_lock() and rcu_read_unlock(). If the updater
uses call_rcu_bh(), then the corresponding readers must use
rcu_read_lock_bh() and rcu_read_unlock_bh(). Mixing things up
will result in confusion and broken kernels.
One exception to this rule: rcu_read_lock() and rcu_read_unlock()
may be substituted for rcu_read_lock_bh() and rcu_read_unlock_bh()
in cases where local bottom halves are already known to be
disabled, for example, in irq or softirq context. Commenting
such cases is a must, of course! And the jury is still out on
whether the increased speed is worth it.
8. Although synchronize_rcu() is a bit slower than is call_rcu(),
it usually results in simpler code. So, unless update
performance is critically important or the updaters cannot block,
synchronize_rcu() should be used in preference to call_rcu().
An especially important property of the synchronize_rcu()
primitive is that it automatically self-limits: if grace periods
are delayed for whatever reason, then the synchronize_rcu()
primitive will correspondingly delay updates. In contrast,
code using call_rcu() should explicitly limit update rate in
cases where grace periods are delayed, as failing to do so can
result in excessive realtime latencies or even OOM conditions.
Ways of gaining this self-limiting property when using call_rcu()
a. Keeping a count of the number of data-structure elements
used by the RCU-protected data structure, including those
waiting for a grace period to elapse. Enforce a limit
on this number, stalling updates as needed to allow
previously deferred frees to complete.
Alternatively, limit only the number awaiting deferred
free rather than the total number of elements.
b. Limiting update rate. For example, if updates occur only
once per hour, then no explicit rate limiting is required,
unless your system is already badly broken. The dcache
subsystem takes this approach -- updates are guarded
by a global lock, limiting their rate.
c. Trusted update -- if updates can only be done manually by
superuser or some other trusted user, then it might not
be necessary to automatically limit them. The theory
here is that superuser already has lots of ways to crash
the machine.
d. Use call_rcu_bh() rather than call_rcu(), in order to take
advantage of call_rcu_bh()'s faster grace periods.
e. Periodically invoke synchronize_rcu(), permitting a limited
number of updates per grace period.
9. All RCU list-traversal primitives, which include
list_for_each_rcu(), list_for_each_entry_rcu(),
list_for_each_continue_rcu(), and list_for_each_safe_rcu(),
must be within an RCU read-side critical section. RCU
read-side critical sections are delimited by rcu_read_lock()
and rcu_read_unlock(), or by similar primitives such as
rcu_read_lock_bh() and rcu_read_unlock_bh().
Use of the _rcu() list-traversal primitives outside of an
RCU read-side critical section causes no harm other than
a slight performance degradation on Alpha CPUs. It can
also be quite helpful in reducing code bloat when common
code is shared between readers and updaters.
10. Conversely, if you are in an RCU read-side critical section,
you -must- use the "_rcu()" variants of the list macros.
Failing to do so will break Alpha and confuse people reading
your code.
11. Note that synchronize_rcu() -only- guarantees to wait until
all currently executing rcu_read_lock()-protected RCU read-side
critical sections complete. It does -not- necessarily guarantee
that all currently running interrupts, NMIs, preempt_disable()
code, or idle loops will complete. Therefore, if you do not have
rcu_read_lock()-protected read-side critical sections, do -not-
use synchronize_rcu().
If you want to wait for some of these other things, you might
instead need to use synchronize_irq() or synchronize_sched().
12. Any lock acquired by an RCU callback must be acquired elsewhere
with irq disabled, e.g., via spin_lock_irqsave(). Failing to
disable irq on a given acquisition of that lock will result in
deadlock as soon as the RCU callback happens to interrupt that
acquisition's critical section.