New operation for kref to help avoid locks
Subsystems (2)
PATCH 1/1 New operation for kref to help avoid locks
Greg,
This is the patch for krefs that we talked about. If you don't like it but like the docs, feel free just to take the documentation and cut out the stuff at the end about the new operation.
Thanks,
-Corey
Add a routine to kref that allows the kref_put() routine to be unserialized even when the get routine attempts to kref_get() an object without first holding a valid reference to it. This is useful in situations where this happens multiple times without freeing the object, as it will avoid having to do a lock/semaphore except on the final kref_put().
This also adds some kref documentation to the Documentation directory.
include/linux/kref.h +6 -0
@@ -28,5 +28,11 @@void kref_get (struct kref *kref);void kref_put (struct kref *kref,void (*release) (struct kref *kref)); +/* Get a kref if the count is not already zero. This can be used to +avoid a lock in the routine that calls kref_put(). Returns 1 if +successful or zero if the count was already zero. See +Documentation/kref.txt for details on how to use this. */ +int kref_get_with_check (struct kref *kref); +#endif /* __KERNEL__ */ #endif /* _KREF_H_ */
lib/kref.c +15 -0
@@ -52,6 +52,21 @@release (kref); } +/** +* kref_get_with_check - increment refcount if the refcount is not already 0. +* @kref: object. +*/ +int kref_get_with_check (struct kref *kref) +{ +atomic_inc (&kref->refcount); +if (atomic_read (&kref->refcount) ==1 ) { +atomic_dec (&kref->refcount); +return 0 ; + } +return 1 ; +} +EXPORT_SYMBOL (kref_init);EXPORT_SYMBOL (kref_get);EXPORT_SYMBOL (kref_put); +EXPORT_SYMBOL (kref_get_with_check);
Documentation/kref.txt new file +274 -0
274 lines not on this page · raw message
Quoted 5 lines · On Sat, 26 Feb 2005 09:55:41 -0600 Corey Minyard wrote:
Greg,
This is the patch for krefs that we talked about. If you don't like it but like the docs, feel free just to take the documentation and cut out the stuff at the end about the new operation.
See below for comments to the patch.
Quoted 46 lines
Thanks,
-Corey
[kref_checked.diff text/plain (10576 bytes)]
Add a routine to kref that allows thekref_put()routine to be unserialized even when the get routine attempts tokref_get()an object without first holding a valid reference to it. This is useful in situations where this happens multiple times without freeing the object, as it will avoid having to do a lock/semaphore except on the finalkref_put().This also adds some kref documentation to the Documentation directory.
Signed-off-by Corey Minyard <[email protected]>Index:
linux-2.6.11-rc4/include/linux/kref.h
===================================================================include/linux/kref.h +6
-0@@ -28,5 +28,11 @@void kref_get (struct kref *kref);void kref_put (struct kref *kref,void (*release) (struct kref *kref)); +/* Get a kref if the count is not already zero. This can be used to +avoid a lock in the routine that calls kref_put(). Returns 1 if +successful or zero if the count was already zero. See +Documentation/kref.txt for details on how to use this. */ +int kref_get_with_check (struct kref *kref); +#endif /* __KERNEL__ */ #endif /* _KREF_H_ */ Index:
linux-2.6.11-rc4/lib/kref.c
===================================================================lib/kref.c +3
-0@@ -52,6 +52,21 @@release (kref); } +/** +* kref_get_with_check - increment refcount if the refcount is not already 0. +* @kref: object.
lib/kref.c line 67
Quoted 10 lines
lib/kref.c +10
-0+*/ +int kref_get_with_check (struct kref *kref) +{ +atomic_inc (&kref->refcount); +if (atomic_read (&kref->refcount) ==1 ) { +atomic_dec (&kref->refcount); +return 0 ; + } +return 1 ; +}
Consider this situation:
initially refcount is 1
CPU1 does kref_get_with_check() CPU2 does kref_put()
atomic_inc(&kref->refcount)
(increments refcount to 2)
atomic_dec_and_test(&kref->refcount) (decrements refcount to 1)
(kref_put() does nothing more because the refcount did not reach zero)
atomic_read(&kref->refcount)
(finds refcount == 1 and decides
that the entry was being deleted)
atomic_dec(&kref->refcount); (decrements refcount to 0)
(then proceeds thinking that the
entry was about to be finalized,
thus creating a leak)
Because of this, separate atomic_inc+atomic_read sequence is unsafe - you need atomic_inc_return() here instead:
int kref_get_with_check(struct kref *kref)
{
if (atomic_inc_return(&kref->refcount) == 1) {
atomic_dec(&kref->refcount);
return 0;
}
return 1;
}Quoted 71 lines
lib/kref.c +2
-0+EXPORT_SYMBOL (kref_init);EXPORT_SYMBOL (kref_get);EXPORT_SYMBOL (kref_put); +EXPORT_SYMBOL (kref_get_with_check);Index:
linux-2.6.11-rc4/Documentation/kref.txt
===================================================================Documentation/kref.txt new file +61
-0@@ -0,0 +1,274 @@ +krefs allow you to add reference counters to your objects. If you +have objects that are used in multiple places and passed around, and +you don't have refcounts, your code is almost certainly broken. If +you want refcounts, krefs are the way to go. + +To use a kref, add a one to your data structures like: + +struct my_data +{ + . + . + struct kref refcount; + . + . +}; + +The kref can occur anywhere within the data structure. + +You must initialize the kref after you allocate it. To do this, call +kref init as so: + + struct my_data *data; + + data = kmalloc(sizeof(*data), GFP_KERNEL); + if (!data) + return -ENOMEM; + kref_init(&data->refcount); + +This sets the refcount in the kref to 1. + +Once you have a refcount, you must follow the following rules: + +1) If you make a non-temporary copy of a pointer, especially if + it can be passed to another thread of execution, you must + increment the refcount with kref_get() before passing it off: + kref_get(&data->refcount); + If you already have a valid pointer to a kref-ed structure (the + refcount cannot go to zero) you may do this without a lock. + +2) When you are done with a pointer, you must call kref_put(): + kref_put(&data->refcount, data_release); + If this is the last reference to the pointer, the release + routine will be called. If the code never tries to get + a valid pointer to a kref-ed structure without already + holding a valid pointer, it is safe to do this without + a lock. + +3) If the code attempts to gain a reference to a kref-ed structure + without already holding a valid pointer, it must serialize access + where a kref_put() cannot occur during the kref_get(), and the + structure must remain valid during the kref_get(). + +For example, if you allocate some data and then pass it to another +thread to process: + +void data_release(struct kref *ref) +{ + struct my_data *data = container_of(ref, struct my_data, refcount); + kfree(data); +} +
Documentation/kref.txt line 71
Quoted 10 lines
Documentation/kref.txt new file +10
-0+void more_data_handling(void *cb_data) +{ + struct my_data *data = cb_data; + . + . do stuff with data here + . + kref_put(data, data_release); +} + +void my_data_handler(void)
Should be int - you return something from it below.
Documentation/kref.txt line 82
Quoted 11 lines
Documentation/kref.txt new file +11
-0+{ + int rv = 0; + struct my_data *data; + data = kmalloc(sizeof(*data), GFP_KERNEL); + if (!data) + return -ENOMEM; + kref_init(&data->refcount); + + kref_get(&data->refcount); + rv = kthread_run(more_data_handling, data, "more_data_handling"); + if (rv) {
This code is wrong - kthread_run() returns a "struct task_struct *" (which might be ERR_PTR(-ENOMEM) if it fails).
Quoted 73 lines
Documentation/kref.txt new file +73
-0+ kref_put(&data->refcount); + goto out; + } + + . + . do stuff with data here + . + out: + kref_put(data, data_release); + return rv; +} + +This way, it doesn't matter what order the two threads handle the +data, the put handles knowing when the data is free and releasing it. +The kref_get() does not require a lock, since we already have a valid +pointer that we own a refcount for. The put needs no lock because +nothing tries to get the data without already holding a pointer. + +Note that the "before" in rule 1 is very important. You should never +do something like: + + rv = kthread_run(more_data_handling, data, "more_data_handling"); + if (rv) + goto out; + else + /* BAD BAD BAD - get is after the handoff */ + kref_get(&data->refcount); + +Don't assume you know what you are doing and use the above construct. +First of all, you may not know what you are doing. Second, you may +know what you are doing (there are some situations where locking is +involved where the above may be legal) but someone else who doesn't +know what they are doing may change the code or copy the code. It's +bad style. Don't do it. + +There are some situations where you can optimize the gets and puts. +For instance, if you are done with an object and enqueuing it for +something else or passing it off to something else, there is no reason +to do a get then a put: + + /* Silly extra get and put */ + kref_get(&obj->ref); + enqueue(obj); + kref_put(&obj->ref, obj_cleanup); + +Just do the enqueue. A comment about this is always welcome: + + enqueue(obj); + /* We are done with obj, so we pass our refcount off + to the queue. DON'T TOUCH obj AFTER HERE! */ + +The last rule (rule 3) is the nastiest one to handle. Say, for +instance, you have a list of items that are each kref-ed, and you wish +to get the first one. You can't just pull the first item off the list +and kref_get() it. That violates rule 3 because you are not already +holding a valid pointer. You must add locks or semaphores. For +instance: + +static DECLARE_MUTEX(sem); +static LIST_HEAD(q); +struct my_data +{ + struct kref refcount; + struct list_head link; +}; + +static struct my_data *get_entry() +{ + struct my_data *entry = NULL; + down(&sem); + if (!list_empty(&q)) { + entry = container_of(q.next, struct my_q_entry, link); + kref_get(&entry->refcount);
Documentation/kref.txt line 165
Quoted 10 lines
Documentation/kref.txt new file +10
-0+ } + up(&sem); + return entry; +} + +static void put_entry(struct my_data *entry) +{ + down(&sem); + kref_put(&entry->refcount); + list_del(&entry->link);
1) kref_put(&entry->refcount, release_entry);
2) Wrong order for kref_put() and list_del() - should be opposite.
3) In this case you could safely up(&sem) after list_del(&entry->link), because get_entry() cannot get to this entry anymore, and kref_put() will ensure that only one invocation will actually call the release function. Probably the example where the call to list_del() is inside the release_entry callback (which was omitted here) will illustrate the rule 3 better.
static void release_entry(struct kref *ref)
{
struct my_data *entry = container_of(ref, struct my_data, refcount);
list_del(&entry->link);
kfree(entry);
}static void put_entry(struct my_data *entry)
{
down(&sem);
kref_put(&entry->refcount, release_entry);
up(&sem);
}Quoted 39 lines
Documentation/kref.txt new file +39
-0+ up(&sem); +} + + +There is one exception to rule 3. kref_get_with_check() can be used +to avoid a lock/semaphore in your routines that call kref_put(), except +in the case that release is called. It is useful in situations where +you might be trying to get an object that might have just gone free, +as in our list example above. + +For instance, suppose you have a queue of items to process in your +driver. The head of the queue is the current item being worked on. +You driver has timers, interrupts, and user context threads that may +all attempt to claim the head of the queue. + +So you define three routines, get_working_entry(), put_entry(), and +entry_completed(), one to get the current queue entry that is being +worked on, one to tell the system you are done working on that +entry for now, and one to tell that the entry is completed and +should be dequeued. Here is some example code: + +static LIST_HEAD(q); +DEFINE_SPINLOCK(q_lock); + +struct my_q_entry +{ + struct list_head link; + struct kref refcount; + char data[10]; + void (*done)(struct my_q_entry *); +}; + +struct my_q_entry *get_working_entry(void) +{ + unsigned long flags; + struct my_q_entry *entry = NULL; + + spin_lock_irqsave(&q_lock, flags); + retry:
Documentation/kref.txt line 214
Quoted 10 lines
Documentation/kref.txt new file +10
-0+ if (!list_empty(&q)) { + entry = container_of(q.next, struct my_q_entry, link); + if (!kref_get_with_check(&entry->refcount)) { + /* Entry's refcount has been deleted, but it + * has not yet been removed from the list. We + * raced with put_entry and are between the + * kref_put() decrementing the count and the + * release routine claiming the lock. Just + * remove it and retry. */ + list_del(&entry->link);
This must be list_del_init(&entry->link) - otherwise release_entry() will oops on its list_del() call.
Quoted 14 lines
Documentation/kref.txt new file +14
-0+ entry = NULL; + goto retry; + } + } + spin_lock_irqrestore(&q_lock, flags); + return entry; +} + +void release_entry(struct kref *ref) +{ + unsigned long flags; + struct my_q_entry *entry = container_of(ref, struct my_q_entry, + refcount); +
Documentation/kref.txt line 238
Quoted 10 lines
Documentation/kref.txt new file +10
-0+ spin_lock_irqsave(&q_lock, flags); + list_del(&entry->link); + spin_lock_irqrestore(&q_lock, flags); + entry->done(entry); + + entry = get_working_entry(); + if (entry) { + /* Start next entry here. */ + } + put_entry(entry)
Obviously this call must be inside the if branch, because your example put_entry() implementation does not accept NULL.
Quoted 36 lines
Documentation/kref.txt new file +36
-0+} + +void put_entry(struct my_q_entry *entry) +{ + kref_put(&entry->refcount, release_entry); +} + +void entry_completed(struct my_q_entry *entry) +{ + kref_put(&entry->refcount, release_entry); +} + + +Note that there is no lock around kref_put(), which should avoid a +lock on most put operations, except for the last one. This works +because if kref_put() sets the refcount to zero, one and only one +thread of execution can be doing the kref_get(). If the kref_get() +beats the kref_put(), everything is fine, the get routine has a +valid entry. If the kref_put() beats the kref_get(), the kref +value will be zero and kref_get_with_check() will fail, but the +release_entry() code will block on the spinlock. So the entry will +be dequeued and everything still works. + +This is generally only useful when you expect do have this happen many +times for a single object. If 99% of the time you have one get and +one put, it's probably not worth the complexity. If you generally +have 20 gets and puts, it saves 19 locking operations, which is +probably worth it. + +Other than this, normal kref rules apply. You can do a kref_get() +without a lock if you are holding a valid kref pointer. + + +Corey Minyard <[email protected]> + +A lot of this was lifted from Greg KH's OLS presentation on krefs.
Quoted 4 lines · Sergey Vlasov
On Sat, 26 Feb 2005 09:55:41 -0600 Corey Minyard wrote:
Quoted 10 lines
Quoted 5 lines
Greg,
This is the patch for krefs that we talked about. If you don't like it but like the docs, feel free just to take the documentation and cut out the stuff at the end about the new operation.
See below for comments to the patch.
Sergey, right on all counts. Here's a patch with the fixes you suggest.
Thanks,
-Corey
Add a routine to kref that allows the kref_put() routine to be unserialized even when the get routine attempts to kref_get() an object without first holding a valid reference to it. This is useful in situations where this happens multiple times without freeing the object, as it will avoid having to do a lock/semaphore except on the final kref_put().
This also adds some kref documentation to the Documentation directory.
include/linux/kref.h +6 -0
@@ -28,5 +28,11 @@void kref_get (struct kref *kref);void kref_put (struct kref *kref,void (*release) (struct kref *kref)); +/* Get a kref if the count is not already zero. This can be used to +avoid a lock in the routine that calls kref_put(). Returns 1 if +successful or zero if the count was already zero. See +Documentation/kref.txt for details on how to use this. */ +int kref_get_with_check (struct kref *kref); +#endif /* __KERNEL__ */ #endif /* _KREF_H_ */
lib/kref.c +14 -0
@@ -52,6 +52,20 @@release (kref); } +/** +* kref_get_with_check - increment refcount if the refcount is not already 0. +* @kref: object. +*/ +int kref_get_with_check (struct kref *kref) +{ +if (atomic_inc_return (&kref->refcount) ==1 ) { +atomic_dec (&kref->refcount); +return 0 ; + } +return 1 ; +} +EXPORT_SYMBOL (kref_init);EXPORT_SYMBOL (kref_get);EXPORT_SYMBOL (kref_put); +EXPORT_SYMBOL (kref_get_with_check);
Documentation/kref.txt new file +283 -0
283 lines not on this page · raw message
Quoted 9 lines · Corey Minyard
Add a routine to kref that allows the
kref_put()routine to be unserialized even when the get routine attempts tokref_get()an object without first holding a valid reference to it. This is useful in situations where this happens multiple times without freeing the object, as it will avoid having to do a lock/semaphore except on the finalkref_put().This also adds some kref documentation to the Documentation directory.
I like the first part of the documentation, that's nice.
But I don't like the new kref_get_with_check() function that you implemented. If you look in the -mm tree, kref_put() now returns if this was the last put on the reference count or not, to help with lists of objects with a kref in it.
Perhaps you can use that to implement what you need instead?
thanks,
greg k-h
Quoted 11 lines · Greg KH
Quoted 9 lines · Corey Minyard
Add a routine to kref that allows the
kref_put()routine to be unserialized even when the get routine attempts tokref_get()an object without first holding a valid reference to it. This is useful in situations where this happens multiple times without freeing the object, as it will avoid having to do a lock/semaphore except on the finalkref_put().This also adds some kref documentation to the Documentation directory.
Quoted 8 lines
I like the first part of the documentation, that's nice.
But I don't like the new
kref_get_with_check()function that you implemented. If you look in the -mm tree,kref_put()now returns if this was the last put on the reference count or not, to help with lists of objects with a kref in it.Perhaps you can use that to implement what you need instead?
note that I'm not convinced the "lockless" implementation actually is faster. It still uses an atomic variable, which is just as expensive as taking a lock normally...
Quoted 19 lines · Arjan van de Ven
Quoted 18 lines · Greg KH
Quoted 9 lines · Corey Minyard
Add a routine to kref that allows the
kref_put()routine to be unserialized even when the get routine attempts tokref_get()an object without first holding a valid reference to it. This is useful in situations where this happens multiple times without freeing the object, as it will avoid having to do a lock/semaphore except on the finalkref_put().This also adds some kref documentation to the Documentation directory.
I like the first part of the documentation, that's nice.
But I don't like the new
kref_get_with_check()function that you implemented. If you look in the -mm tree,kref_put()now returns if this was the last put on the reference count or not, to help with lists of objects with a kref in it.
Quoted 5 lines
Quoted 1 line
Perhaps you can use that to implement what you need instead?
note that I'm not convinced the "lockless" implementation actually is faster. It still uses an atomic variable, which is just as expensive as taking a lock normally...
I have never stated it would be "faster" that I know of, and you still need a lock to protect some of the paths. But that is documented in my 2004 ols paper about kref.
thanks,
greg k-h
Quoted 19 lines · Arjan van de Ven
Quoted 16 lines · Greg KH
Quoted 11 lines · Corey Minyard
Add a routine to kref that allows the
kref_put()routine to be unserialized even when the get routine attempts tokref_get()an object without first holding a valid reference to it. This is useful in situations where this happens multiple times without freeing the object, as it will avoid having to do a lock/semaphore except on the finalkref_put().This also adds some kref documentation to the Documentation directory.
I like the first part of the documentation, that's nice.
Quoted 8 lines
Quoted 8 lines
But I don't like the new
kref_get_with_check()function that you implemented. If you look in the -mm tree,kref_put()now returns if this was the last put on the reference count or not, to help with lists of objects with a kref in it.Perhaps you can use that to implement what you need instead?
Yes, that helps a lot. I had actually already implemented something like that :). But that's a different thing than avoiding the lock.
It's just that with the I2C stuff, you may be calling kref_put() 20-30 times for a single operation. That's a lot of lock/unlock operations. But it is wierd, so I understand. Thanks.
Quoted 6 lines
note that I'm not convinced the "lockless" implementation actually is faster. It still uses an atomic variable, which is just as expensive as taking a lock normally...
Just doing an atomic operation is not faster than doing a lock, an atomic operation, then an unlock? Am I missing something?
-Corey
Quoted 3 lines
Just doing an atomic operation is not faster than doing a lock, an atomic operation, then an unlock? Am I missing something?
if the lock and the atomic are on the same cacheline they're the same cost on most modern cpus...
Quoted 8 lines · Arjan van de Ven
Quoted 3 lines
Just doing an atomic operation is not faster than doing a lock, an atomic operation, then an unlock? Am I missing something?
if the lock and the atomic are on the same cacheline they're the same cost on most modern cpus...
Ah, I see. Not likely to ever be the case with this. The lock will likely be with the main data structure (the list, or whatever) and the refcount will be in the individual item in the main data structure (list entry).
-Corey
Quoted 6 lines · Corey Minyard
Quoted 4 lines · Arjan van de Ven
Quoted 2 lines
Just doing an atomic operation is not faster than doing a lock, an atomic operation, then an unlock? Am I missing something?
Quoted 6 lines
Quoted 4 lines
if the lock and the atomic are on the same cacheline they're the same cost on most modern cpus...
Ah, I see. Not likely to ever be the case with this. The lock will likely be with the main data structure (the list, or whatever) and the refcount will be in the individual item in the main data structure (list entry).
Is get_with_check actually going to be useful for anything? It seems like it promotes complex and potentially unsafe schemes.
eg. In your queue example, it would usually be better to have a refcount for being on queue, and entry_completed would remove the entry from the queue and accordingly drop the refcount. The release function would then just free it.
Quoted 13 lines · Nick Piggin
Quoted 11 lines · Corey Minyard
Quoted 9 lines · Arjan van de Ven
Quoted 2 lines
Just doing an atomic operation is not faster than doing a lock, an atomic operation, then an unlock? Am I missing something?
if the lock and the atomic are on the same cacheline they're the same cost on most modern cpus...
Quoted 5 lines
Quoted 2 lines
Ah, I see. Not likely to ever be the case with this. The lock will likely be with the main data structure (the list, or whatever) and the refcount will be in the individual item in the main data structure (list entry).
Is
get_with_checkactually going to be useful for anything? It seems like it promotes complex and potentially unsafe schemes.
It is certainly more complex to use this, and I'm guessing that's why Greg rejected it. Certainly a valid problem.
Quoted 5 lines
eg. In your queue example, it would usually be better to have a refcount for being on queue, and
entry_completedwould remove the entry from the queue and accordingly drop the refcount. The release function would then just free it.
True. But if things picked up entries of the queue and incremented their refcount, then you would need a lock. The same technique would apply. But your example would be the more common one, I would think.
Thanks,
-Corey
Quoted 9 lines · Corey Minyard
Quoted 2 lines · Nick Piggin
Is
get_with_checkactually going to be useful for anything? It seems like it promotes complex and potentially unsafe schemes.It is certainly more complex to use this, and I'm guessing that's why Greg rejected it. Certainly a valid problem.
Quoted 1 line
Quoted 8 lines
Quoted 4 lines
eg. In your queue example, it would usually be better to have a refcount for being on queue, and
entry_completedwould remove the entry from the queue and accordingly drop the refcount. The release function would then just free it.True. But if things picked up entries of the queue and incremented their refcount, then you would need a lock. The same technique would apply. But your example would be the more common one, I would think.
Well, but you take a lock in your system too, to protect the queue (ie. in get_working_entry()).
Nick