lkml.org 
[lkml]   [2017]   [Oct]   [30]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
Subject[PATCH] locking/lockdep: Revise Documentation/locking/crossrelease.txt
Date
I'm afraid the revision is not perfect yet. Of course, the document can
have got much better english by others than me.

But,

I think I should enhance it as much as I can, before they can help it
starting with a better one.

In addition, I removed verboseness as much as possible.

----->8-----
From c7795104ca6ac6dd9f7fd944aee23a2011a6d3a2 Mon Sep 17 00:00:00 2001
From: Byungchul Park <byungchul.park@lge.com>
Date: Mon, 30 Oct 2017 14:51:26 +0900
Subject: [PATCH] locking/lockdep: Revise
Documentation/locking/crossrelease.txt

The document should've been written with a better readability. Revise it
to enhance its readability.

Signed-off-by: Byungchul Park <byungchul.park@lge.com>
---
Documentation/locking/crossrelease.txt | 388 +++++++++++++++------------------
1 file changed, 173 insertions(+), 215 deletions(-)

diff --git a/Documentation/locking/crossrelease.txt b/Documentation/locking/crossrelease.txt
index bdf1423..26c31d7 100644
--- a/Documentation/locking/crossrelease.txt
+++ b/Documentation/locking/crossrelease.txt
@@ -12,10 +12,10 @@ Contents:

(*) Limitation

- - Limit lockdep
+ - Limiting lockdep
- Pros from the limitation
- Cons from the limitation
- - Relax the limitation
+ - Relaxing the limitation

(*) Crossrelease

@@ -32,7 +32,7 @@ Contents:
- Avoid duplication
- Lockless for hot paths

- (*) APPENDIX A: What lockdep does to work aggresively
+ (*) APPENDIX A: How to add dependencies aggresively

(*) APPENDIX B: How to avoid adding false dependencies

@@ -51,36 +51,30 @@ also impossible due to the same reason.

For example:

- A context going to trigger event C is waiting for event A to happen.
- A context going to trigger event A is waiting for event B to happen.
- A context going to trigger event B is waiting for event C to happen.
+ A context going to trigger event C is waiting for event A.
+ A context going to trigger event A is waiting for event B.
+ A context going to trigger event B is waiting for event C.

-A deadlock occurs when these three wait operations run at the same time,
-because event C cannot be triggered if event A does not happen, which in
-turn cannot be triggered if event B does not happen, which in turn
-cannot be triggered if event C does not happen. After all, no event can
-be triggered since any of them never meets its condition to wake up.
+A deadlock occurs when the three waiters run at the same time, because
+event C cannot be triggered if event A does not happen, which in turn
+cannot be triggered if event B does not happen, which in turn cannot be
+triggered if event C does not happen. After all, no event can be
+triggered since any of them never meets its condition to wake up.

-A dependency might exist between two waiters and a deadlock might happen
-due to an incorrect releationship between dependencies. Thus, we must
-define what a dependency is first. A dependency exists between them if:
+A dependency exists between two waiters, and a deadlock happens due to
+an incorrect releationship between dependencies. Thus, we must define
+what a dependency is first. A dependency exists between waiters if:

1. There are two waiters waiting for each event at a given time.
2. The only way to wake up each waiter is to trigger its event.
3. Whether one can be woken up depends on whether the other can.

-Each wait in the example creates its dependency like:
+Each waiter in the example creates its dependency like:

Event C depends on event A.
Event A depends on event B.
Event B depends on event C.

- NOTE: Precisely speaking, a dependency is one between whether a
- waiter for an event can be woken up and whether another waiter for
- another event can be woken up. However from now on, we will describe
- a dependency as if it's one between an event and another event for
- simplicity.
-
And they form circular dependencies like:

-> C -> A -> B -
@@ -101,19 +95,18 @@ Circular dependencies cause a deadlock.
How lockdep works
-----------------

-Lockdep tries to detect a deadlock by checking dependencies created by
-lock operations, acquire and release. Waiting for a lock corresponds to
-waiting for an event, and releasing a lock corresponds to triggering an
-event in the previous section.
+Lockdep tries to detect a deadlock by checking circular dependencies
+created by lock operations, acquire and release, which are wait and
+event respectively.

In short, lockdep does:

1. Detect a new dependency.
- 2. Add the dependency into a global graph.
+ 2. Add the dependency into the global graph.
3. Check if that makes dependencies circular.
- 4. Report a deadlock or its possibility if so.
+ 4. Report a deadlock if so.

-For example, consider a graph built by lockdep that looks like:
+For example, the graph has been built like:

A -> B -
\
@@ -155,33 +148,29 @@ how lockdep works.

CONCLUSION

-Lockdep detects a deadlock or its possibility by checking if circular
-dependencies were created after adding each new dependency.
+Lockdep detects a deadlock by checking if circular dependencies were
+created after adding each new dependency.


==========
Limitation
==========

-Limit lockdep
--------------
+Limiting lockdep
+----------------

-Limiting lockdep to work on only typical locks e.g. spin locks and
-mutexes, which are released within the acquire context, the
-implementation becomes simple but its capacity for detection becomes
-limited. Let's check pros and cons in next section.
+Limiting lockdep to work with only typical locks e.g. spin locks and
+mutexes, the implementation becomes simple because it ensures that
+they are released within the acquire context, but its capacity for
+detection becomes limited. Let's check pros and cons in the next two
+sections.


Pros from the limitation
------------------------

-Given the limitation, when acquiring a lock, locks in a held_locks
-cannot be released if the context cannot acquire it so has to wait to
-acquire it, which means all waiters for the locks in the held_locks are
-stuck. It's an exact case to create dependencies between each lock in
-the held_locks and the lock to acquire.
-
-For example:
+Given the limitation, a dependency can be identified at the time
+acquiring a lock. For example:

CONTEXT X
---------
@@ -192,25 +181,37 @@ For example:

where A and B are different lock classes.

-When acquiring lock A, the held_locks of CONTEXT X is empty thus no
-dependency is added. But when acquiring lock B, lockdep detects and adds
-a new dependency 'A -> B' between lock A in the held_locks and lock B.
-They can be simply added whenever acquiring each lock.
+Remind what a dependency is. A dependency exists if:

-And data required by lockdep exists in a local structure, held_locks
-embedded in task_struct. Forcing to access the data within the context,
-lockdep can avoid racy problems without explicit locks while handling
-the local data.
+ 1. There are two waiters waiting for each event at a given time.
+ 2. The only way to wake up each waiter is to trigger its event.
+ 3. Whether one can be woken up depends on whether the other can.
+
+Therefore, a dependency 'A -> B' exists since:
+
+ 1. A waiter for A and a waiter for B might exist when acquiring B.
+ 2. Only way to wake up each is to release what it waits for.
+ 3. Whether the waiter for A can be woken up depends on whether the
+ other can. In other words, CONTEXT X cannot release A if it stucks
+ in the acquisition of B.

-Lastly, lockdep only needs to keep locks currently being held, to build
-a dependency graph. However, relaxing the limitation, it needs to keep
-even locks already released, because a decision whether they created
-dependencies might be long-deferred.
+And the decision easily can be made at the time acquiring B.
+
+Furthermore, data used to create a dependency can be kept in local
+structure, task_struct, since we've already limited lockdep to handle
+local locks. Thus, by forcing to access the data within the context,
+lockdep can avoid racy problems without explicit protection.
+
+Lastly, lockdep only needs to keep locks currently being held. However,
+relaxing the limitation, it needs to keep even locks already released,
+because the decision, whether they create dependencies, should be
+long-deferred.

To sum up, we can expect several advantages from the limitation:

1. Lockdep can easily identify a dependency when acquiring a lock.
- 2. Races are avoidable while accessing local locks in a held_locks.
+ 2. Races are avoidable without explicit protection, while accessing
+ local locks .
3. Lockdep only needs to keep locks currently being held.

CONCLUSION
@@ -221,9 +222,9 @@ Given the limitation, the implementation becomes simple and efficient.
Cons from the limitation
------------------------

-Given the limitation, lockdep is applicable only to typical locks. For
-example, page locks for page access or completions for synchronization
-cannot work with lockdep.
+Given the limitation, however, lockdep is applicable only to typical
+locks. For example, page locks for page access or completions for
+synchronization cannot work with lockdep.

Can we detect deadlocks below, under the limitation?

@@ -261,19 +262,19 @@ No, we cannot.

CONCLUSION

-Given the limitation, lockdep cannot detect a deadlock or its
-possibility caused by page locks or completions.
+Given the limitation, lockdep cannot detect a deadlock caused by page
+locks or completions.


-Relax the limitation
---------------------
+Relaxing the limitation
+-----------------------

Under the limitation, things to create dependencies are limited to
typical locks. However, synchronization primitives like page locks and
-completions, which are allowed to be released in any context, also
-create dependencies and can cause a deadlock. So lockdep should track
-these locks to do a better job. We have to relax the limitation for
-these locks to work with lockdep.
+completions allowed to be released in any context, also create
+dependencies and can cause a deadlock. So lockdep should track these
+locks to do a better job. We have to relax the limitation for these
+locks to work with lockdep.

Detecting dependencies is very important for lockdep to work because
adding a dependency means adding an opportunity to check whether it
@@ -281,34 +282,33 @@ causes a deadlock. The more lockdep adds dependencies, the more it
thoroughly works. Thus Lockdep has to do its best to detect and add as
many true dependencies into a graph as possible.

-For example, considering only typical locks, lockdep builds a graph like:
+For example:

- A -> B -
- \
- -> E
- /
- C -> D -
+ CONTEXT X CONTEXT Y
+ --------- ---------
+ acquire A
+ acquire B /* A dependency 'A -> B' exists */
+ release B
+ release A held by Y

- where A, B,..., E are different lock classes.
+ where A and B are different lock classes.

-On the other hand, under the relaxation, additional dependencies might
-be created and added. Assuming additional 'FX -> C' and 'E -> GX' are
-added thanks to the relaxation, the graph will be:
+In the case relaxing the limitation, a depedency 'A -> B' exists since:

- A -> B -
- \
- -> E -> GX
- /
- FX -> C -> D -
+ 1. A waiter for A and a waiter for B might exist when acquiring B.
+ 2. Only way to wake up each is to release what it waits for.
+ 3. Whether the waiter for A can be woken up depends on whether the
+ other can. In other words, CONTEXT X cannot release A if it stucks
+ in the acquisition of B.

- where A, B,..., E, FX and GX are different lock classes, and a suffix
- 'X' is added on non-typical locks.
+Considering only typical locks, lockdep would add no dependency. However,
+relaxing the limitation, a real dependency 'A -> B' can be added. Then,
+we have got more chances to check circular dependencies.

-The latter graph gives us more chances to check circular dependencies
-than the former. However, it might suffer performance degradation since
-relaxing the limitation, with which design and implementation of lockdep
-can be efficient, might introduce inefficiency inevitably. So lockdep
-should provide two options, strong detection and efficient detection.
+However, it might suffer performance degradation since relaxing the
+limitation, with which design and implementation of lockdep can be
+efficient, might introduce inefficiency inevitably. So lockdep should
+provide two options i.e. strong detection and efficient detection.

Choosing efficient detection:

@@ -333,12 +333,10 @@ Crossrelease
Introduce crossrelease
----------------------

-In order to allow lockdep to handle additional dependencies by what
-might be released in any context, namely 'crosslock', we have to be able
-to identify those created by crosslocks. The proposed 'crossrelease'
-feature provoides a way to do that.
+To allow lockdep to handle the additional dependencies created by namely
+'crosslock', we propose 'crossrelease' feature.

-Crossrelease feature has to do:
+Crossrelease feature does:

1. Identify dependencies created by crosslocks.
2. Add the dependencies into a dependency graph.
@@ -349,21 +347,16 @@ crossrelease feature has to do is to correctly identify and add true
dependencies into the global graph.

A dependency e.g. 'A -> B' can be identified only in the A's release
-context because a decision required to identify the dependency can be
-made only in the release context. That is to decide whether A can be
-released so that a waiter for A can be woken up. It cannot be made in
-other than the A's release context.
-
-It's no matter for typical locks because each acquire context is same as
-its release context, thus lockdep can decide whether a lock can be
-released in the acquire context. However for crosslocks, lockdep cannot
-make the decision in the acquire context but has to wait until the
-release context is identified.
-
-Therefore, deadlocks by crosslocks cannot be detected just when it
-happens, because those cannot be identified until the crosslocks are
-released. However, deadlock possibilities can be detected and it's very
-worth. See 'APPENDIX A' section to check why.
+context, because a decision required to identify the dependency can be
+made only in the release context. In other words, the decision whether A
+can be released and wake up waiters for it, cannot be made in other than
+the A's release context.
+
+Of course, it's no matter for typical locks because each acquire context
+is the same as its release context, thus lockdep can decide whether a
+lock can be released, in the acquire context. However for crosslocks,
+lockdep cannot make the decision in the acquire context but has to wait
+until the release context is identified.

CONCLUSION

@@ -375,12 +368,12 @@ Introduce commit
----------------

Since crossrelease defers the work adding true dependencies of
-crosslocks until they are actually released, crossrelease has to queue
-all acquisitions which might create dependencies with the crosslocks.
-Then it identifies dependencies using the queued data in batches at a
-proper time. We call it 'commit'.
+crosslocks until they are eventually released, we have to queue all
+acquisitions, which might create dependencies with the crosslocks, so
+that real dependencies can be identified using the queued data in
+batches at a proper time. We call it 'commit'.

-There are four types of dependencies:
+We have four types of dependencies:

1. TT type: 'typical lock A -> typical lock B'

@@ -409,7 +402,7 @@ There are four types of dependencies:

Lockdep can work without commit for typical locks, but commit step is
necessary once crosslocks are involved. Introducing commit, lockdep
-performs three steps. What lockdep does in each step is:
+performs three steps:

1. Acquisition: For typical locks, lockdep does what it originally did
and queues the lock so that CT type dependencies can be checked using
@@ -455,10 +448,8 @@ Crossrelease introduces two main data structures.
How crossrelease works
----------------------

-It's the key of how crossrelease works, to defer necessary works to an
-appropriate point in time and perform in at once at the commit step.
Let's take a look with examples step by step, starting from how lockdep
-works without crossrelease for typical locks.
+works for typical locks, without crossrelease.

acquire A /* Push A onto held_locks */
acquire B /* Push B onto held_locks and add 'A -> B' */
@@ -469,30 +460,20 @@ works without crossrelease for typical locks.

where A, B and C are different lock classes.

- NOTE: This document assumes that readers already understand how
- lockdep works without crossrelease thus omits details. But there's
- one thing to note. Lockdep pretends to pop a lock from held_locks
- when releasing it. But it's subtly different from the original pop
- operation because lockdep allows other than the top to be poped.
-
-In this case, lockdep adds 'the top of held_locks -> the lock to acquire'
-dependency every time acquiring a lock.
-
-After adding 'A -> B', a dependency graph will be:
+When acquiring B, a dependency 'A -> B' is added:

A -> B

where A and B are different lock classes.

-And after adding 'B -> C', the graph will be:
+And when acquiring C, a dependency 'B -> C' is added:

A -> B -> C

where A, B and C are different lock classes.

-Let's performs commit step even for typical locks to add dependencies.
-Of course, commit step is not necessary for them, however, it would work
-well because this is a more general way.
+Now, let's apply the commit step into the example. Of course, the step
+is not necessary for typical locks, but it would work as well like:

acquire A
/*
@@ -520,9 +501,8 @@ well because this is a more general way.

commit C
/*
- * Add 'C -> ?'
- * Answer the following to decide '?'
- * What has been queued since acquire C: Nothing
+ * Add 'C -> ?' after deciding '?'
+ * What has been queued since acquire C (='?'): Nothing
*
* In hist_locks: A, B, C
* In graph: Empty
@@ -532,9 +512,8 @@ well because this is a more general way.

commit B
/*
- * Add 'B -> ?'
- * Answer the following to decide '?'
- * What has been queued since acquire B: C
+ * Add 'B -> ?' after deciding '?'
+ * What has been queued since acquire B (='?'): C
*
* In hist_locks: A, B, C
* In graph: 'B -> C'
@@ -544,9 +523,8 @@ well because this is a more general way.

commit A
/*
- * Add 'A -> ?'
- * Answer the following to decide '?'
- * What has been queued since acquire A: B, C
+ * Add 'A -> ?' after deciding '?'
+ * What has been queued since acquire A (='?'): B, C
*
* In hist_locks: A, B, C
* In graph: 'B -> C', 'A -> B', 'A -> C'
@@ -556,9 +534,8 @@ well because this is a more general way.

where A, B and C are different lock classes.

-In this case, dependencies are added at the commit step as described.
-
-After commits for A, B and C, the graph will be:
+Dependencies are added at the commit step as described. After commits
+for A, B and C, the graph becomes:

A -> B -> C

@@ -566,15 +543,13 @@ After commits for A, B and C, the graph will be:

NOTE: A dependency 'A -> C' is optimized out.

-We can see the former graph built without commit step is same as the
-latter graph built using commit steps. Of course the former way leads to
-earlier finish for building the graph, which means we can detect a
-deadlock or its possibility sooner. So the former way would be prefered
-when possible. But we cannot avoid using the latter way for crosslocks.
+We can see the former graph is the same as the latter graph. Of course,
+the former way leads to earlier finish for building the graph, which
+means we can detect a deadlock or its possibility sooner. So the former
+way would be prefered when possible. But, we cannot avoid using the
+latter way for crosslocks.

-Let's look at how commit steps work for crosslocks. In this case, the
-commit step is performed only on crosslock AX as real. And it assumes
-that the AX release context is different from the AX acquire context.
+Lastly, let's look at how commit steps work for crosslocks in practice.

BX RELEASE CONTEXT BX ACQUIRE CONTEXT
------------------ ------------------
@@ -658,8 +633,8 @@ that the AX release context is different from the AX acquire context.
*/
commit BX
/*
- * Add 'BX -> ?'
- * What has been queued since acquire BX: C, E
+ * Add 'BX -> ?' after deciding '?'
+ * What has been queued since acquire BX (='?'): C, E
*
* In held_locks: Empty
* In hist_locks: D, E
@@ -687,14 +662,10 @@ that the AX release context is different from the AX acquire context.
where A, BX, C,..., E are different lock classes, and a suffix 'X' is
added on crosslocks.

-Crossrelease considers all acquisitions after acqiuring BX are
-candidates which might create dependencies with BX. True dependencies
-will be determined when identifying the release context of BX. Meanwhile,
-all typical locks are queued so that they can be used at the commit step.
-And then two dependencies 'BX -> C' and 'BX -> E' are added at the
-commit step when identifying the release context.
+All typical locks are queued so that they can be used at the commit step.
+And then two dependencies 'BX -> C' and 'BX -> E' are added at the step.

-The final graph will be, with crossrelease:
+The final graph would be, with crossrelease:

-> C
/
@@ -707,7 +678,7 @@ The final graph will be, with crossrelease:
where A, BX, C,..., E are different lock classes, and a suffix 'X' is
added on crosslocks.

-However, the final graph will be, without crossrelease:
+However, the final graph would be, without crossrelease:

A -> D

@@ -720,7 +691,7 @@ caused by crosslocks.

CONCLUSION

-We checked how crossrelease works with several examples.
+We checked how crossrelease works, with several examples.


=============
@@ -730,7 +701,7 @@ Optimizations
Avoid duplication
-----------------

-Crossrelease feature uses a cache like what lockdep already uses for
+Crossrelease feature uses a cache, like what lockdep already uses for
dependency chains, but this time it's for caching CT type dependencies.
Once that dependency is cached, the same will never be added again.

@@ -738,34 +709,36 @@ Once that dependency is cached, the same will never be added again.
Lockless for hot paths
----------------------

-To keep all locks for later use at the commit step, crossrelease adopts
-a local array embedded in task_struct, which makes access to the data
-lockless by forcing it to happen only within the owner context. It's
-like how lockdep handles held_locks. Lockless implmentation is important
-since typical locks are very frequently acquired and released.
+To keep all locks for later use, crossrelease adopts a local array in
+task_struct, which makes the access lockless by forcing it to happen
+within the owner context. It's like how lockdep handles held_locks.
+Lockless is very important since typical locks are very frequently
+acquired and released.


-=================================================
-APPENDIX A: What lockdep does to work aggresively
-=================================================
+===============================================
+APPENDIX A: How to add dependencies aggresively
+===============================================

-A deadlock actually occurs when all wait operations creating circular
+A deadlock actually occurs only when all waiters creating circular
dependencies run at the same time. Even though they don't, a potential
-deadlock exists if the problematic dependencies exist. Thus it's
-meaningful to detect not only an actual deadlock but also its potential
-possibility. The latter is rather valuable. When a deadlock occurs
-actually, we can identify what happens in the system by some means or
-other even without lockdep. However, there's no way to detect possiblity
-without lockdep unless the whole code is parsed in head. It's terrible.
-Lockdep does the both, and crossrelease only focuses on the latter.
+deadlock exists if the problematic dependencies exist.
+
+So we have to detect not only an actual deadlock but also its potential
+possibility. The latter is rather valuable. When a deadlock actually
+occures, we can identify what happens in the system by some means or
+other. However, there's no way to detect possiblity without tools like
+lockdep, unless the whole code is parsed in head. It's terrible. Lockdep
+does the both, and crossrelease focuses on the latter.

Whether or not a deadlock actually occurs depends on several factors.
-For example, what order contexts are switched in is a factor. Assuming
+For example, the order in which contexts are scheduled is a factor. If
circular dependencies exist, a deadlock would occur when contexts are
-switched so that all wait operations creating the dependencies run
-simultaneously. Thus to detect a deadlock possibility even in the case
-that it has not occured yet, lockdep should consider all possible
-combinations of dependencies, trying to:
+scheduled so that all waiters creating the dependencies run at the same
+time.
+
+To detect the possibilities, lockdep should consider all possible
+scenarios aggresively, trying to:

1. Use a global dependency graph.

@@ -789,8 +762,8 @@ combinations of dependencies, trying to:

CONCLUSION

-Lockdep detects not only an actual deadlock but also its possibility,
-and the latter is more valuable.
+Lockdep detects not only an actual deadlock but also its possibility
+aggresively, and the latter is more valuable.


==================================================
@@ -805,24 +778,8 @@ Remind what a dependency is. A dependency exists if:

For example:

- acquire A
- acquire B /* A dependency 'A -> B' exists */
- release B
- release A
-
- where A and B are different lock classes.
-
-A depedency 'A -> B' exists since:
-
- 1. A waiter for A and a waiter for B might exist when acquiring B.
- 2. Only way to wake up each is to release what it waits for.
- 3. Whether the waiter for A can be woken up depends on whether the
- other can. IOW, TASK X cannot release A if it fails to acquire B.
-
-For another example:
-
- TASK X TASK Y
- ------ ------
+ CONTEXT X CONTEXT Y
+ --------- ---------
acquire AX
acquire B /* A dependency 'AX -> B' exists */
release B
@@ -831,18 +788,18 @@ For another example:
where AX and B are different lock classes, and a suffix 'X' is added
on crosslocks.

-Even in this case involving crosslocks, the same rule can be applied. A
-depedency 'AX -> B' exists since:
+A depedency 'AX -> B' exists since:

1. A waiter for AX and a waiter for B might exist when acquiring B.
2. Only way to wake up each is to release what it waits for.
3. Whether the waiter for AX can be woken up depends on whether the
- other can. IOW, TASK X cannot release AX if it fails to acquire B.
+ other can. In other words, CONTEXT X cannot release AX if it
+ stucks in the acquisition of B.

Let's take a look at more complicated example:

- TASK X TASK Y
- ------ ------
+ CONTEXT X CONTEXT Y
+ --------- ---------
acquire B
release B
fork Y
@@ -862,11 +819,12 @@ example. Thus the dependency 'AX -> B' cannot be created.

It would be ideal if the full set of true ones can be considered. But
we can ensure nothing but what actually happened. Relying on what
-actually happens at runtime, we can anyway add only true ones, though
-they might be a subset of true ones. It's similar to how lockdep works
-for typical locks. There might be more true dependencies than what
-lockdep has detected in runtime. Lockdep has no choice but to rely on
-what actually happens. Crossrelease also relies on it.
+actually happens at runtime, we can anyway avoid adding false ones,
+though they might be a subset of true ones.
+
+It's similar to how lockdep works for typical locks. There might be more
+true dependencies than lockdep has detected. Lockdep has no choice but
+to rely on what actually happens. Crossrelease also relies on it.

CONCLUSION

--
1.9.1
\
 
 \ /
  Last update: 2017-10-30 07:19    [W:0.086 / U:0.444 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site