Merge branch 'generic-dispatch' of git://brick.kernel.dk/data/git/linux-2.6-block

This commit is contained in:
Linus Torvalds 2005-10-28 08:53:49 -07:00
commit 28d721e24c
9 changed files with 466 additions and 854 deletions

View File

@ -906,9 +906,20 @@ Aside:
4. The I/O scheduler 4. The I/O scheduler
I/O schedulers are now per queue. They should be runtime switchable and modular I/O scheduler, a.k.a. elevator, is implemented in two layers. Generic dispatch
but aren't yet. Jens has most bits to do this, but the sysfs implementation is queue and specific I/O schedulers. Unless stated otherwise, elevator is used
missing. to refer to both parts and I/O scheduler to specific I/O schedulers.
Block layer implements generic dispatch queue in ll_rw_blk.c and elevator.c.
The generic dispatch queue is responsible for properly ordering barrier
requests, requeueing, handling non-fs requests and all other subtleties.
Specific I/O schedulers are responsible for ordering normal filesystem
requests. They can also choose to delay certain requests to improve
throughput or whatever purpose. As the plural form indicates, there are
multiple I/O schedulers. They can be built as modules but at least one should
be built inside the kernel. Each queue can choose different one and can also
change to another one dynamically.
A block layer call to the i/o scheduler follows the convention elv_xxx(). This A block layer call to the i/o scheduler follows the convention elv_xxx(). This
calls elevator_xxx_fn in the elevator switch (drivers/block/elevator.c). Oh, calls elevator_xxx_fn in the elevator switch (drivers/block/elevator.c). Oh,
@ -921,44 +932,36 @@ keeping work.
The functions an elevator may implement are: (* are mandatory) The functions an elevator may implement are: (* are mandatory)
elevator_merge_fn called to query requests for merge with a bio elevator_merge_fn called to query requests for merge with a bio
elevator_merge_req_fn " " " with another request elevator_merge_req_fn called when two requests get merged. the one
which gets merged into the other one will be
never seen by I/O scheduler again. IOW, after
being merged, the request is gone.
elevator_merged_fn called when a request in the scheduler has been elevator_merged_fn called when a request in the scheduler has been
involved in a merge. It is used in the deadline involved in a merge. It is used in the deadline
scheduler for example, to reposition the request scheduler for example, to reposition the request
if its sorting order has changed. if its sorting order has changed.
*elevator_next_req_fn returns the next scheduled request, or NULL elevator_dispatch_fn fills the dispatch queue with ready requests.
if there are none (or none are ready). I/O schedulers are free to postpone requests by
not filling the dispatch queue unless @force
is non-zero. Once dispatched, I/O schedulers
are not allowed to manipulate the requests -
they belong to generic dispatch queue.
*elevator_add_req_fn called to add a new request into the scheduler elevator_add_req_fn called to add a new request into the scheduler
elevator_queue_empty_fn returns true if the merge queue is empty. elevator_queue_empty_fn returns true if the merge queue is empty.
Drivers shouldn't use this, but rather check Drivers shouldn't use this, but rather check
if elv_next_request is NULL (without losing the if elv_next_request is NULL (without losing the
request if one exists!) request if one exists!)
elevator_remove_req_fn This is called when a driver claims ownership of
the target request - it now belongs to the
driver. It must not be modified or merged.
Drivers must not lose the request! A subsequent
call of elevator_next_req_fn must return the
_next_ request.
elevator_requeue_req_fn called to add a request to the scheduler. This
is used when the request has alrnadebeen
returned by elv_next_request, but hasn't
completed. If this is not implemented then
elevator_add_req_fn is called instead.
elevator_former_req_fn elevator_former_req_fn
elevator_latter_req_fn These return the request before or after the elevator_latter_req_fn These return the request before or after the
one specified in disk sort order. Used by the one specified in disk sort order. Used by the
block layer to find merge possibilities. block layer to find merge possibilities.
elevator_completed_req_fn called when a request is completed. This might elevator_completed_req_fn called when a request is completed.
come about due to being merged with another or
when the device completes the request.
elevator_may_queue_fn returns true if the scheduler wants to allow the elevator_may_queue_fn returns true if the scheduler wants to allow the
current context to queue a new request even if current context to queue a new request even if
@ -967,13 +970,33 @@ elevator_may_queue_fn returns true if the scheduler wants to allow the
elevator_set_req_fn elevator_set_req_fn
elevator_put_req_fn Must be used to allocate and free any elevator elevator_put_req_fn Must be used to allocate and free any elevator
specific storate for a request. specific storage for a request.
elevator_activate_req_fn Called when device driver first sees a request.
I/O schedulers can use this callback to
determine when actual execution of a request
starts.
elevator_deactivate_req_fn Called when device driver decides to delay
a request by requeueing it.
elevator_init_fn elevator_init_fn
elevator_exit_fn Allocate and free any elevator specific storage elevator_exit_fn Allocate and free any elevator specific storage
for a queue. for a queue.
4.2 I/O scheduler implementation 4.2 Request flows seen by I/O schedulers
All requests seens by I/O schedulers strictly follow one of the following three
flows.
set_req_fn ->
i. add_req_fn -> (merged_fn ->)* -> dispatch_fn -> activate_req_fn ->
(deactivate_req_fn -> activate_req_fn ->)* -> completed_req_fn
ii. add_req_fn -> (merged_fn ->)* -> merge_req_fn
iii. [none]
-> put_req_fn
4.3 I/O scheduler implementation
The generic i/o scheduler algorithm attempts to sort/merge/batch requests for The generic i/o scheduler algorithm attempts to sort/merge/batch requests for
optimal disk scan and request servicing performance (based on generic optimal disk scan and request servicing performance (based on generic
principles and device capabilities), optimized for: principles and device capabilities), optimized for:
@ -993,18 +1016,7 @@ request in sort order to prevent binary tree lookups.
This arrangement is not a generic block layer characteristic however, so This arrangement is not a generic block layer characteristic however, so
elevators may implement queues as they please. elevators may implement queues as they please.
ii. Last merge hint ii. Merge hash
The last merge hint is part of the generic queue layer. I/O schedulers must do
some management on it. For the most part, the most important thing is to make
sure q->last_merge is cleared (set to NULL) when the request on it is no longer
a candidate for merging (for example if it has been sent to the driver).
The last merge performed is cached as a hint for the subsequent request. If
sequential data is being submitted, the hint is used to perform merges without
any scanning. This is not sufficient when there are multiple processes doing
I/O though, so a "merge hash" is used by some schedulers.
iii. Merge hash
AS and deadline use a hash table indexed by the last sector of a request. This AS and deadline use a hash table indexed by the last sector of a request. This
enables merging code to quickly look up "back merge" candidates, even when enables merging code to quickly look up "back merge" candidates, even when
multiple I/O streams are being performed at once on one disk. multiple I/O streams are being performed at once on one disk.
@ -1013,29 +1025,8 @@ multiple I/O streams are being performed at once on one disk.
are far less common than "back merges" due to the nature of most I/O patterns. are far less common than "back merges" due to the nature of most I/O patterns.
Front merges are handled by the binary trees in AS and deadline schedulers. Front merges are handled by the binary trees in AS and deadline schedulers.
iv. Handling barrier cases iii. Plugging the queue to batch requests in anticipation of opportunities for
A request with flags REQ_HARDBARRIER or REQ_SOFTBARRIER must not be ordered merge/sort optimizations
around. That is, they must be processed after all older requests, and before
any newer ones. This includes merges!
In AS and deadline schedulers, barriers have the effect of flushing the reorder
queue. The performance cost of this will vary from nothing to a lot depending
on i/o patterns and device characteristics. Obviously they won't improve
performance, so their use should be kept to a minimum.
v. Handling insertion position directives
A request may be inserted with a position directive. The directives are one of
ELEVATOR_INSERT_BACK, ELEVATOR_INSERT_FRONT, ELEVATOR_INSERT_SORT.
ELEVATOR_INSERT_SORT is a general directive for non-barrier requests.
ELEVATOR_INSERT_BACK is used to insert a barrier to the back of the queue.
ELEVATOR_INSERT_FRONT is used to insert a barrier to the front of the queue, and
overrides the ordering requested by any previous barriers. In practice this is
harmless and required, because it is used for SCSI requeueing. This does not
require flushing the reorder queue, so does not impose a performance penalty.
vi. Plugging the queue to batch requests in anticipation of opportunities for
merge/sort optimizations
This is just the same as in 2.4 so far, though per-device unplugging This is just the same as in 2.4 so far, though per-device unplugging
support is anticipated for 2.5. Also with a priority-based i/o scheduler, support is anticipated for 2.5. Also with a priority-based i/o scheduler,
@ -1069,7 +1060,7 @@ Aside:
blk_kick_queue() to unplug a specific queue (right away ?) blk_kick_queue() to unplug a specific queue (right away ?)
or optionally, all queues, is in the plan. or optionally, all queues, is in the plan.
4.3 I/O contexts 4.4 I/O contexts
I/O contexts provide a dynamically allocated per process data area. They may I/O contexts provide a dynamically allocated per process data area. They may
be used in I/O schedulers, and in the block layer (could be used for IO statis, be used in I/O schedulers, and in the block layer (could be used for IO statis,
priorities for example). See *io_context in drivers/block/ll_rw_blk.c, and priorities for example). See *io_context in drivers/block/ll_rw_blk.c, and

View File

@ -98,7 +98,6 @@ struct as_data {
struct as_rq *next_arq[2]; /* next in sort order */ struct as_rq *next_arq[2]; /* next in sort order */
sector_t last_sector[2]; /* last REQ_SYNC & REQ_ASYNC sectors */ sector_t last_sector[2]; /* last REQ_SYNC & REQ_ASYNC sectors */
struct list_head *dispatch; /* driver dispatch queue */
struct list_head *hash; /* request hash */ struct list_head *hash; /* request hash */
unsigned long exit_prob; /* probability a task will exit while unsigned long exit_prob; /* probability a task will exit while
@ -239,6 +238,25 @@ static struct io_context *as_get_io_context(void)
return ioc; return ioc;
} }
static void as_put_io_context(struct as_rq *arq)
{
struct as_io_context *aic;
if (unlikely(!arq->io_context))
return;
aic = arq->io_context->aic;
if (arq->is_sync == REQ_SYNC && aic) {
spin_lock(&aic->lock);
set_bit(AS_TASK_IORUNNING, &aic->state);
aic->last_end_request = jiffies;
spin_unlock(&aic->lock);
}
put_io_context(arq->io_context);
}
/* /*
* the back merge hash support functions * the back merge hash support functions
*/ */
@ -261,14 +279,6 @@ static inline void as_del_arq_hash(struct as_rq *arq)
__as_del_arq_hash(arq); __as_del_arq_hash(arq);
} }
static void as_remove_merge_hints(request_queue_t *q, struct as_rq *arq)
{
as_del_arq_hash(arq);
if (q->last_merge == arq->request)
q->last_merge = NULL;
}
static void as_add_arq_hash(struct as_data *ad, struct as_rq *arq) static void as_add_arq_hash(struct as_data *ad, struct as_rq *arq)
{ {
struct request *rq = arq->request; struct request *rq = arq->request;
@ -312,7 +322,7 @@ static struct request *as_find_arq_hash(struct as_data *ad, sector_t offset)
BUG_ON(!arq->on_hash); BUG_ON(!arq->on_hash);
if (!rq_mergeable(__rq)) { if (!rq_mergeable(__rq)) {
as_remove_merge_hints(ad->q, arq); as_del_arq_hash(arq);
continue; continue;
} }
@ -950,23 +960,12 @@ static void as_completed_request(request_queue_t *q, struct request *rq)
WARN_ON(!list_empty(&rq->queuelist)); WARN_ON(!list_empty(&rq->queuelist));
if (arq->state == AS_RQ_PRESCHED) {
WARN_ON(arq->io_context);
goto out;
}
if (arq->state == AS_RQ_MERGED)
goto out_ioc;
if (arq->state != AS_RQ_REMOVED) { if (arq->state != AS_RQ_REMOVED) {
printk("arq->state %d\n", arq->state); printk("arq->state %d\n", arq->state);
WARN_ON(1); WARN_ON(1);
goto out; goto out;
} }
if (!blk_fs_request(rq))
goto out;
if (ad->changed_batch && ad->nr_dispatched == 1) { if (ad->changed_batch && ad->nr_dispatched == 1) {
kblockd_schedule_work(&ad->antic_work); kblockd_schedule_work(&ad->antic_work);
ad->changed_batch = 0; ad->changed_batch = 0;
@ -1001,21 +1000,7 @@ static void as_completed_request(request_queue_t *q, struct request *rq)
} }
} }
out_ioc: as_put_io_context(arq);
if (!arq->io_context)
goto out;
if (arq->is_sync == REQ_SYNC) {
struct as_io_context *aic = arq->io_context->aic;
if (aic) {
spin_lock(&aic->lock);
set_bit(AS_TASK_IORUNNING, &aic->state);
aic->last_end_request = jiffies;
spin_unlock(&aic->lock);
}
}
put_io_context(arq->io_context);
out: out:
arq->state = AS_RQ_POSTSCHED; arq->state = AS_RQ_POSTSCHED;
} }
@ -1047,72 +1032,10 @@ static void as_remove_queued_request(request_queue_t *q, struct request *rq)
ad->next_arq[data_dir] = as_find_next_arq(ad, arq); ad->next_arq[data_dir] = as_find_next_arq(ad, arq);
list_del_init(&arq->fifo); list_del_init(&arq->fifo);
as_remove_merge_hints(q, arq); as_del_arq_hash(arq);
as_del_arq_rb(ad, arq); as_del_arq_rb(ad, arq);
} }
/*
* as_remove_dispatched_request is called to remove a request which has gone
* to the dispatch list.
*/
static void as_remove_dispatched_request(request_queue_t *q, struct request *rq)
{
struct as_rq *arq = RQ_DATA(rq);
struct as_io_context *aic;
if (!arq) {
WARN_ON(1);
return;
}
WARN_ON(arq->state != AS_RQ_DISPATCHED);
WARN_ON(ON_RB(&arq->rb_node));
if (arq->io_context && arq->io_context->aic) {
aic = arq->io_context->aic;
if (aic) {
WARN_ON(!atomic_read(&aic->nr_dispatched));
atomic_dec(&aic->nr_dispatched);
}
}
}
/*
* as_remove_request is called when a driver has finished with a request.
* This should be only called for dispatched requests, but for some reason
* a POWER4 box running hwscan it does not.
*/
static void as_remove_request(request_queue_t *q, struct request *rq)
{
struct as_rq *arq = RQ_DATA(rq);
if (unlikely(arq->state == AS_RQ_NEW))
goto out;
if (ON_RB(&arq->rb_node)) {
if (arq->state != AS_RQ_QUEUED) {
printk("arq->state %d\n", arq->state);
WARN_ON(1);
goto out;
}
/*
* We'll lose the aliased request(s) here. I don't think this
* will ever happen, but if it does, hopefully someone will
* report it.
*/
WARN_ON(!list_empty(&rq->queuelist));
as_remove_queued_request(q, rq);
} else {
if (arq->state != AS_RQ_DISPATCHED) {
printk("arq->state %d\n", arq->state);
WARN_ON(1);
goto out;
}
as_remove_dispatched_request(q, rq);
}
out:
arq->state = AS_RQ_REMOVED;
}
/* /*
* as_fifo_expired returns 0 if there are no expired reads on the fifo, * as_fifo_expired returns 0 if there are no expired reads on the fifo,
* 1 otherwise. It is ratelimited so that we only perform the check once per * 1 otherwise. It is ratelimited so that we only perform the check once per
@ -1165,7 +1088,6 @@ static inline int as_batch_expired(struct as_data *ad)
static void as_move_to_dispatch(struct as_data *ad, struct as_rq *arq) static void as_move_to_dispatch(struct as_data *ad, struct as_rq *arq)
{ {
struct request *rq = arq->request; struct request *rq = arq->request;
struct list_head *insert;
const int data_dir = arq->is_sync; const int data_dir = arq->is_sync;
BUG_ON(!ON_RB(&arq->rb_node)); BUG_ON(!ON_RB(&arq->rb_node));
@ -1198,13 +1120,13 @@ static void as_move_to_dispatch(struct as_data *ad, struct as_rq *arq)
/* /*
* take it off the sort and fifo list, add to dispatch queue * take it off the sort and fifo list, add to dispatch queue
*/ */
insert = ad->dispatch->prev;
while (!list_empty(&rq->queuelist)) { while (!list_empty(&rq->queuelist)) {
struct request *__rq = list_entry_rq(rq->queuelist.next); struct request *__rq = list_entry_rq(rq->queuelist.next);
struct as_rq *__arq = RQ_DATA(__rq); struct as_rq *__arq = RQ_DATA(__rq);
list_move_tail(&__rq->queuelist, ad->dispatch); list_del(&__rq->queuelist);
elv_dispatch_add_tail(ad->q, __rq);
if (__arq->io_context && __arq->io_context->aic) if (__arq->io_context && __arq->io_context->aic)
atomic_inc(&__arq->io_context->aic->nr_dispatched); atomic_inc(&__arq->io_context->aic->nr_dispatched);
@ -1218,7 +1140,8 @@ static void as_move_to_dispatch(struct as_data *ad, struct as_rq *arq)
as_remove_queued_request(ad->q, rq); as_remove_queued_request(ad->q, rq);
WARN_ON(arq->state != AS_RQ_QUEUED); WARN_ON(arq->state != AS_RQ_QUEUED);
list_add(&rq->queuelist, insert); elv_dispatch_sort(ad->q, rq);
arq->state = AS_RQ_DISPATCHED; arq->state = AS_RQ_DISPATCHED;
if (arq->io_context && arq->io_context->aic) if (arq->io_context && arq->io_context->aic)
atomic_inc(&arq->io_context->aic->nr_dispatched); atomic_inc(&arq->io_context->aic->nr_dispatched);
@ -1230,12 +1153,42 @@ static void as_move_to_dispatch(struct as_data *ad, struct as_rq *arq)
* read/write expire, batch expire, etc, and moves it to the dispatch * read/write expire, batch expire, etc, and moves it to the dispatch
* queue. Returns 1 if a request was found, 0 otherwise. * queue. Returns 1 if a request was found, 0 otherwise.
*/ */
static int as_dispatch_request(struct as_data *ad) static int as_dispatch_request(request_queue_t *q, int force)
{ {
struct as_data *ad = q->elevator->elevator_data;
struct as_rq *arq; struct as_rq *arq;
const int reads = !list_empty(&ad->fifo_list[REQ_SYNC]); const int reads = !list_empty(&ad->fifo_list[REQ_SYNC]);
const int writes = !list_empty(&ad->fifo_list[REQ_ASYNC]); const int writes = !list_empty(&ad->fifo_list[REQ_ASYNC]);
if (unlikely(force)) {
/*
* Forced dispatch, accounting is useless. Reset
* accounting states and dump fifo_lists. Note that
* batch_data_dir is reset to REQ_SYNC to avoid
* screwing write batch accounting as write batch
* accounting occurs on W->R transition.
*/
int dispatched = 0;
ad->batch_data_dir = REQ_SYNC;
ad->changed_batch = 0;
ad->new_batch = 0;
while (ad->next_arq[REQ_SYNC]) {
as_move_to_dispatch(ad, ad->next_arq[REQ_SYNC]);
dispatched++;
}
ad->last_check_fifo[REQ_SYNC] = jiffies;
while (ad->next_arq[REQ_ASYNC]) {
as_move_to_dispatch(ad, ad->next_arq[REQ_ASYNC]);
dispatched++;
}
ad->last_check_fifo[REQ_ASYNC] = jiffies;
return dispatched;
}
/* Signal that the write batch was uncontended, so we can't time it */ /* Signal that the write batch was uncontended, so we can't time it */
if (ad->batch_data_dir == REQ_ASYNC && !reads) { if (ad->batch_data_dir == REQ_ASYNC && !reads) {
if (ad->current_write_count == 0 || !writes) if (ad->current_write_count == 0 || !writes)
@ -1359,20 +1312,6 @@ static int as_dispatch_request(struct as_data *ad)
return 1; return 1;
} }
static struct request *as_next_request(request_queue_t *q)
{
struct as_data *ad = q->elevator->elevator_data;
struct request *rq = NULL;
/*
* if there are still requests on the dispatch queue, grab the first
*/
if (!list_empty(ad->dispatch) || as_dispatch_request(ad))
rq = list_entry_rq(ad->dispatch->next);
return rq;
}
/* /*
* Add arq to a list behind alias * Add arq to a list behind alias
*/ */
@ -1404,17 +1343,25 @@ as_add_aliased_request(struct as_data *ad, struct as_rq *arq, struct as_rq *alia
/* /*
* Don't want to have to handle merges. * Don't want to have to handle merges.
*/ */
as_remove_merge_hints(ad->q, arq); as_del_arq_hash(arq);
} }
/* /*
* add arq to rbtree and fifo * add arq to rbtree and fifo
*/ */
static void as_add_request(struct as_data *ad, struct as_rq *arq) static void as_add_request(request_queue_t *q, struct request *rq)
{ {
struct as_data *ad = q->elevator->elevator_data;
struct as_rq *arq = RQ_DATA(rq);
struct as_rq *alias; struct as_rq *alias;
int data_dir; int data_dir;
if (arq->state != AS_RQ_PRESCHED) {
printk("arq->state: %d\n", arq->state);
WARN_ON(1);
}
arq->state = AS_RQ_NEW;
if (rq_data_dir(arq->request) == READ if (rq_data_dir(arq->request) == READ
|| current->flags&PF_SYNCWRITE) || current->flags&PF_SYNCWRITE)
arq->is_sync = 1; arq->is_sync = 1;
@ -1437,12 +1384,8 @@ static void as_add_request(struct as_data *ad, struct as_rq *arq)
arq->expires = jiffies + ad->fifo_expire[data_dir]; arq->expires = jiffies + ad->fifo_expire[data_dir];
list_add_tail(&arq->fifo, &ad->fifo_list[data_dir]); list_add_tail(&arq->fifo, &ad->fifo_list[data_dir]);
if (rq_mergeable(arq->request)) { if (rq_mergeable(arq->request))
as_add_arq_hash(ad, arq); as_add_arq_hash(ad, arq);
if (!ad->q->last_merge)
ad->q->last_merge = arq->request;
}
as_update_arq(ad, arq); /* keep state machine up to date */ as_update_arq(ad, arq); /* keep state machine up to date */
} else { } else {
@ -1463,96 +1406,24 @@ static void as_add_request(struct as_data *ad, struct as_rq *arq)
arq->state = AS_RQ_QUEUED; arq->state = AS_RQ_QUEUED;
} }
static void as_activate_request(request_queue_t *q, struct request *rq)
{
struct as_rq *arq = RQ_DATA(rq);
WARN_ON(arq->state != AS_RQ_DISPATCHED);
arq->state = AS_RQ_REMOVED;
if (arq->io_context && arq->io_context->aic)
atomic_dec(&arq->io_context->aic->nr_dispatched);
}
static void as_deactivate_request(request_queue_t *q, struct request *rq) static void as_deactivate_request(request_queue_t *q, struct request *rq)
{ {
struct as_data *ad = q->elevator->elevator_data;
struct as_rq *arq = RQ_DATA(rq); struct as_rq *arq = RQ_DATA(rq);
if (arq) { WARN_ON(arq->state != AS_RQ_REMOVED);
if (arq->state == AS_RQ_REMOVED) { arq->state = AS_RQ_DISPATCHED;
arq->state = AS_RQ_DISPATCHED; if (arq->io_context && arq->io_context->aic)
if (arq->io_context && arq->io_context->aic) atomic_inc(&arq->io_context->aic->nr_dispatched);
atomic_inc(&arq->io_context->aic->nr_dispatched);
}
} else
WARN_ON(blk_fs_request(rq)
&& (!(rq->flags & (REQ_HARDBARRIER|REQ_SOFTBARRIER))) );
/* Stop anticipating - let this request get through */
as_antic_stop(ad);
}
/*
* requeue the request. The request has not been completed, nor is it a
* new request, so don't touch accounting.
*/
static void as_requeue_request(request_queue_t *q, struct request *rq)
{
as_deactivate_request(q, rq);
list_add(&rq->queuelist, &q->queue_head);
}
/*
* Account a request that is inserted directly onto the dispatch queue.
* arq->io_context->aic->nr_dispatched should not need to be incremented
* because only new requests should come through here: requeues go through
* our explicit requeue handler.
*/
static void as_account_queued_request(struct as_data *ad, struct request *rq)
{
if (blk_fs_request(rq)) {
struct as_rq *arq = RQ_DATA(rq);
arq->state = AS_RQ_DISPATCHED;
ad->nr_dispatched++;
}
}
static void
as_insert_request(request_queue_t *q, struct request *rq, int where)
{
struct as_data *ad = q->elevator->elevator_data;
struct as_rq *arq = RQ_DATA(rq);
if (arq) {
if (arq->state != AS_RQ_PRESCHED) {
printk("arq->state: %d\n", arq->state);
WARN_ON(1);
}
arq->state = AS_RQ_NEW;
}
/* barriers must flush the reorder queue */
if (unlikely(rq->flags & (REQ_SOFTBARRIER | REQ_HARDBARRIER)
&& where == ELEVATOR_INSERT_SORT)) {
WARN_ON(1);
where = ELEVATOR_INSERT_BACK;
}
switch (where) {
case ELEVATOR_INSERT_BACK:
while (ad->next_arq[REQ_SYNC])
as_move_to_dispatch(ad, ad->next_arq[REQ_SYNC]);
while (ad->next_arq[REQ_ASYNC])
as_move_to_dispatch(ad, ad->next_arq[REQ_ASYNC]);
list_add_tail(&rq->queuelist, ad->dispatch);
as_account_queued_request(ad, rq);
as_antic_stop(ad);
break;
case ELEVATOR_INSERT_FRONT:
list_add(&rq->queuelist, ad->dispatch);
as_account_queued_request(ad, rq);
as_antic_stop(ad);
break;
case ELEVATOR_INSERT_SORT:
BUG_ON(!blk_fs_request(rq));
as_add_request(ad, arq);
break;
default:
BUG();
return;
}
} }
/* /*
@ -1565,12 +1436,8 @@ static int as_queue_empty(request_queue_t *q)
{ {
struct as_data *ad = q->elevator->elevator_data; struct as_data *ad = q->elevator->elevator_data;
if (!list_empty(&ad->fifo_list[REQ_ASYNC]) return list_empty(&ad->fifo_list[REQ_ASYNC])
|| !list_empty(&ad->fifo_list[REQ_SYNC]) && list_empty(&ad->fifo_list[REQ_SYNC]);
|| !list_empty(ad->dispatch))
return 0;
return 1;
} }
static struct request * static struct request *
@ -1607,15 +1474,6 @@ as_merge(request_queue_t *q, struct request **req, struct bio *bio)
struct request *__rq; struct request *__rq;
int ret; int ret;
/*
* try last_merge to avoid going to hash
*/
ret = elv_try_last_merge(q, bio);
if (ret != ELEVATOR_NO_MERGE) {
__rq = q->last_merge;
goto out_insert;
}
/* /*
* see if the merge hash can satisfy a back merge * see if the merge hash can satisfy a back merge
*/ */
@ -1644,9 +1502,6 @@ as_merge(request_queue_t *q, struct request **req, struct bio *bio)
return ELEVATOR_NO_MERGE; return ELEVATOR_NO_MERGE;
out: out:
if (rq_mergeable(__rq))
q->last_merge = __rq;
out_insert:
if (ret) { if (ret) {
if (rq_mergeable(__rq)) if (rq_mergeable(__rq))
as_hot_arq_hash(ad, RQ_DATA(__rq)); as_hot_arq_hash(ad, RQ_DATA(__rq));
@ -1693,9 +1548,6 @@ static void as_merged_request(request_queue_t *q, struct request *req)
* behind the disk head. We currently don't bother adjusting. * behind the disk head. We currently don't bother adjusting.
*/ */
} }
if (arq->on_hash)
q->last_merge = req;
} }
static void static void
@ -1763,6 +1615,7 @@ as_merged_requests(request_queue_t *q, struct request *req,
* kill knowledge of next, this one is a goner * kill knowledge of next, this one is a goner
*/ */
as_remove_queued_request(q, next); as_remove_queued_request(q, next);
as_put_io_context(anext);
anext->state = AS_RQ_MERGED; anext->state = AS_RQ_MERGED;
} }
@ -1782,7 +1635,7 @@ static void as_work_handler(void *data)
unsigned long flags; unsigned long flags;
spin_lock_irqsave(q->queue_lock, flags); spin_lock_irqsave(q->queue_lock, flags);
if (as_next_request(q)) if (!as_queue_empty(q))
q->request_fn(q); q->request_fn(q);
spin_unlock_irqrestore(q->queue_lock, flags); spin_unlock_irqrestore(q->queue_lock, flags);
} }
@ -1797,7 +1650,9 @@ static void as_put_request(request_queue_t *q, struct request *rq)
return; return;
} }
if (arq->state != AS_RQ_POSTSCHED && arq->state != AS_RQ_PRESCHED) { if (unlikely(arq->state != AS_RQ_POSTSCHED &&
arq->state != AS_RQ_PRESCHED &&
arq->state != AS_RQ_MERGED)) {
printk("arq->state %d\n", arq->state); printk("arq->state %d\n", arq->state);
WARN_ON(1); WARN_ON(1);
} }
@ -1907,7 +1762,6 @@ static int as_init_queue(request_queue_t *q, elevator_t *e)
INIT_LIST_HEAD(&ad->fifo_list[REQ_ASYNC]); INIT_LIST_HEAD(&ad->fifo_list[REQ_ASYNC]);
ad->sort_list[REQ_SYNC] = RB_ROOT; ad->sort_list[REQ_SYNC] = RB_ROOT;
ad->sort_list[REQ_ASYNC] = RB_ROOT; ad->sort_list[REQ_ASYNC] = RB_ROOT;
ad->dispatch = &q->queue_head;
ad->fifo_expire[REQ_SYNC] = default_read_expire; ad->fifo_expire[REQ_SYNC] = default_read_expire;
ad->fifo_expire[REQ_ASYNC] = default_write_expire; ad->fifo_expire[REQ_ASYNC] = default_write_expire;
ad->antic_expire = default_antic_expire; ad->antic_expire = default_antic_expire;
@ -2072,10 +1926,9 @@ static struct elevator_type iosched_as = {
.elevator_merge_fn = as_merge, .elevator_merge_fn = as_merge,
.elevator_merged_fn = as_merged_request, .elevator_merged_fn = as_merged_request,
.elevator_merge_req_fn = as_merged_requests, .elevator_merge_req_fn = as_merged_requests,
.elevator_next_req_fn = as_next_request, .elevator_dispatch_fn = as_dispatch_request,
.elevator_add_req_fn = as_insert_request, .elevator_add_req_fn = as_add_request,
.elevator_remove_req_fn = as_remove_request, .elevator_activate_req_fn = as_activate_request,
.elevator_requeue_req_fn = as_requeue_request,
.elevator_deactivate_req_fn = as_deactivate_request, .elevator_deactivate_req_fn = as_deactivate_request,
.elevator_queue_empty_fn = as_queue_empty, .elevator_queue_empty_fn = as_queue_empty,
.elevator_completed_req_fn = as_completed_request, .elevator_completed_req_fn = as_completed_request,

View File

@ -84,7 +84,6 @@ static int cfq_max_depth = 2;
(node)->rb_left = NULL; \ (node)->rb_left = NULL; \
} while (0) } while (0)
#define RB_CLEAR_ROOT(root) ((root)->rb_node = NULL) #define RB_CLEAR_ROOT(root) ((root)->rb_node = NULL)
#define ON_RB(node) ((node)->rb_color != RB_NONE)
#define rb_entry_crq(node) rb_entry((node), struct cfq_rq, rb_node) #define rb_entry_crq(node) rb_entry((node), struct cfq_rq, rb_node)
#define rq_rb_key(rq) (rq)->sector #define rq_rb_key(rq) (rq)->sector
@ -271,10 +270,7 @@ CFQ_CFQQ_FNS(expired);
#undef CFQ_CFQQ_FNS #undef CFQ_CFQQ_FNS
enum cfq_rq_state_flags { enum cfq_rq_state_flags {
CFQ_CRQ_FLAG_in_flight = 0, CFQ_CRQ_FLAG_is_sync = 0,
CFQ_CRQ_FLAG_in_driver,
CFQ_CRQ_FLAG_is_sync,
CFQ_CRQ_FLAG_requeued,
}; };
#define CFQ_CRQ_FNS(name) \ #define CFQ_CRQ_FNS(name) \
@ -291,14 +287,11 @@ static inline int cfq_crq_##name(const struct cfq_rq *crq) \
return (crq->crq_flags & (1 << CFQ_CRQ_FLAG_##name)) != 0; \ return (crq->crq_flags & (1 << CFQ_CRQ_FLAG_##name)) != 0; \
} }
CFQ_CRQ_FNS(in_flight);
CFQ_CRQ_FNS(in_driver);
CFQ_CRQ_FNS(is_sync); CFQ_CRQ_FNS(is_sync);
CFQ_CRQ_FNS(requeued);
#undef CFQ_CRQ_FNS #undef CFQ_CRQ_FNS
static struct cfq_queue *cfq_find_cfq_hash(struct cfq_data *, unsigned int, unsigned short); static struct cfq_queue *cfq_find_cfq_hash(struct cfq_data *, unsigned int, unsigned short);
static void cfq_dispatch_sort(request_queue_t *, struct cfq_rq *); static void cfq_dispatch_insert(request_queue_t *, struct cfq_rq *);
static void cfq_put_cfqd(struct cfq_data *cfqd); static void cfq_put_cfqd(struct cfq_data *cfqd);
#define process_sync(tsk) ((tsk)->flags & PF_SYNCWRITE) #define process_sync(tsk) ((tsk)->flags & PF_SYNCWRITE)
@ -311,14 +304,6 @@ static inline void cfq_del_crq_hash(struct cfq_rq *crq)
hlist_del_init(&crq->hash); hlist_del_init(&crq->hash);
} }
static void cfq_remove_merge_hints(request_queue_t *q, struct cfq_rq *crq)
{
cfq_del_crq_hash(crq);
if (q->last_merge == crq->request)
q->last_merge = NULL;
}
static inline void cfq_add_crq_hash(struct cfq_data *cfqd, struct cfq_rq *crq) static inline void cfq_add_crq_hash(struct cfq_data *cfqd, struct cfq_rq *crq)
{ {
const int hash_idx = CFQ_MHASH_FN(rq_hash_key(crq->request)); const int hash_idx = CFQ_MHASH_FN(rq_hash_key(crq->request));
@ -347,18 +332,13 @@ static struct request *cfq_find_rq_hash(struct cfq_data *cfqd, sector_t offset)
return NULL; return NULL;
} }
static inline int cfq_pending_requests(struct cfq_data *cfqd)
{
return !list_empty(&cfqd->queue->queue_head) || cfqd->busy_queues;
}
/* /*
* scheduler run of queue, if there are requests pending and no one in the * scheduler run of queue, if there are requests pending and no one in the
* driver that will restart queueing * driver that will restart queueing
*/ */
static inline void cfq_schedule_dispatch(struct cfq_data *cfqd) static inline void cfq_schedule_dispatch(struct cfq_data *cfqd)
{ {
if (!cfqd->rq_in_driver && cfq_pending_requests(cfqd)) if (!cfqd->rq_in_driver && cfqd->busy_queues)
kblockd_schedule_work(&cfqd->unplug_work); kblockd_schedule_work(&cfqd->unplug_work);
} }
@ -366,7 +346,7 @@ static int cfq_queue_empty(request_queue_t *q)
{ {
struct cfq_data *cfqd = q->elevator->elevator_data; struct cfq_data *cfqd = q->elevator->elevator_data;
return !cfq_pending_requests(cfqd); return !cfqd->busy_queues;
} }
/* /*
@ -386,11 +366,6 @@ cfq_choose_req(struct cfq_data *cfqd, struct cfq_rq *crq1, struct cfq_rq *crq2)
if (crq2 == NULL) if (crq2 == NULL)
return crq1; return crq1;
if (cfq_crq_requeued(crq1) && !cfq_crq_requeued(crq2))
return crq1;
else if (cfq_crq_requeued(crq2) && !cfq_crq_requeued(crq1))
return crq2;
if (cfq_crq_is_sync(crq1) && !cfq_crq_is_sync(crq2)) if (cfq_crq_is_sync(crq1) && !cfq_crq_is_sync(crq2))
return crq1; return crq1;
else if (cfq_crq_is_sync(crq2) && !cfq_crq_is_sync(crq1)) else if (cfq_crq_is_sync(crq2) && !cfq_crq_is_sync(crq1))
@ -461,10 +436,7 @@ cfq_find_next_crq(struct cfq_data *cfqd, struct cfq_queue *cfqq,
struct cfq_rq *crq_next = NULL, *crq_prev = NULL; struct cfq_rq *crq_next = NULL, *crq_prev = NULL;
struct rb_node *rbnext, *rbprev; struct rb_node *rbnext, *rbprev;
rbnext = NULL; if (!(rbnext = rb_next(&last->rb_node))) {
if (ON_RB(&last->rb_node))
rbnext = rb_next(&last->rb_node);
if (!rbnext) {
rbnext = rb_first(&cfqq->sort_list); rbnext = rb_first(&cfqq->sort_list);
if (rbnext == &last->rb_node) if (rbnext == &last->rb_node)
rbnext = NULL; rbnext = NULL;
@ -545,13 +517,13 @@ static void cfq_resort_rr_list(struct cfq_queue *cfqq, int preempted)
* the pending list according to last request service * the pending list according to last request service
*/ */
static inline void static inline void
cfq_add_cfqq_rr(struct cfq_data *cfqd, struct cfq_queue *cfqq, int requeue) cfq_add_cfqq_rr(struct cfq_data *cfqd, struct cfq_queue *cfqq)
{ {
BUG_ON(cfq_cfqq_on_rr(cfqq)); BUG_ON(cfq_cfqq_on_rr(cfqq));
cfq_mark_cfqq_on_rr(cfqq); cfq_mark_cfqq_on_rr(cfqq);
cfqd->busy_queues++; cfqd->busy_queues++;
cfq_resort_rr_list(cfqq, requeue); cfq_resort_rr_list(cfqq, 0);
} }
static inline void static inline void
@ -571,22 +543,19 @@ cfq_del_cfqq_rr(struct cfq_data *cfqd, struct cfq_queue *cfqq)
static inline void cfq_del_crq_rb(struct cfq_rq *crq) static inline void cfq_del_crq_rb(struct cfq_rq *crq)
{ {
struct cfq_queue *cfqq = crq->cfq_queue; struct cfq_queue *cfqq = crq->cfq_queue;
struct cfq_data *cfqd = cfqq->cfqd;
const int sync = cfq_crq_is_sync(crq);
if (ON_RB(&crq->rb_node)) { BUG_ON(!cfqq->queued[sync]);
struct cfq_data *cfqd = cfqq->cfqd; cfqq->queued[sync]--;
const int sync = cfq_crq_is_sync(crq);
BUG_ON(!cfqq->queued[sync]); cfq_update_next_crq(crq);
cfqq->queued[sync]--;
cfq_update_next_crq(crq); rb_erase(&crq->rb_node, &cfqq->sort_list);
RB_CLEAR_COLOR(&crq->rb_node);
rb_erase(&crq->rb_node, &cfqq->sort_list); if (cfq_cfqq_on_rr(cfqq) && RB_EMPTY(&cfqq->sort_list))
RB_CLEAR_COLOR(&crq->rb_node); cfq_del_cfqq_rr(cfqd, cfqq);
if (cfq_cfqq_on_rr(cfqq) && RB_EMPTY(&cfqq->sort_list))
cfq_del_cfqq_rr(cfqd, cfqq);
}
} }
static struct cfq_rq * static struct cfq_rq *
@ -627,12 +596,12 @@ static void cfq_add_crq_rb(struct cfq_rq *crq)
* if that happens, put the alias on the dispatch list * if that happens, put the alias on the dispatch list
*/ */
while ((__alias = __cfq_add_crq_rb(crq)) != NULL) while ((__alias = __cfq_add_crq_rb(crq)) != NULL)
cfq_dispatch_sort(cfqd->queue, __alias); cfq_dispatch_insert(cfqd->queue, __alias);
rb_insert_color(&crq->rb_node, &cfqq->sort_list); rb_insert_color(&crq->rb_node, &cfqq->sort_list);
if (!cfq_cfqq_on_rr(cfqq)) if (!cfq_cfqq_on_rr(cfqq))
cfq_add_cfqq_rr(cfqd, cfqq, cfq_crq_requeued(crq)); cfq_add_cfqq_rr(cfqd, cfqq);
/* /*
* check if this request is a better next-serve candidate * check if this request is a better next-serve candidate
@ -643,10 +612,8 @@ static void cfq_add_crq_rb(struct cfq_rq *crq)
static inline void static inline void
cfq_reposition_crq_rb(struct cfq_queue *cfqq, struct cfq_rq *crq) cfq_reposition_crq_rb(struct cfq_queue *cfqq, struct cfq_rq *crq)
{ {
if (ON_RB(&crq->rb_node)) { rb_erase(&crq->rb_node, &cfqq->sort_list);
rb_erase(&crq->rb_node, &cfqq->sort_list); cfqq->queued[cfq_crq_is_sync(crq)]--;
cfqq->queued[cfq_crq_is_sync(crq)]--;
}
cfq_add_crq_rb(crq); cfq_add_crq_rb(crq);
} }
@ -676,49 +643,28 @@ static struct request *cfq_find_rq_rb(struct cfq_data *cfqd, sector_t sector)
return NULL; return NULL;
} }
static void cfq_activate_request(request_queue_t *q, struct request *rq)
{
struct cfq_data *cfqd = q->elevator->elevator_data;
cfqd->rq_in_driver++;
}
static void cfq_deactivate_request(request_queue_t *q, struct request *rq) static void cfq_deactivate_request(request_queue_t *q, struct request *rq)
{ {
struct cfq_data *cfqd = q->elevator->elevator_data; struct cfq_data *cfqd = q->elevator->elevator_data;
struct cfq_rq *crq = RQ_DATA(rq);
if (crq) { WARN_ON(!cfqd->rq_in_driver);
struct cfq_queue *cfqq = crq->cfq_queue; cfqd->rq_in_driver--;
if (cfq_crq_in_driver(crq)) {
cfq_clear_crq_in_driver(crq);
WARN_ON(!cfqd->rq_in_driver);
cfqd->rq_in_driver--;
}
if (cfq_crq_in_flight(crq)) {
const int sync = cfq_crq_is_sync(crq);
cfq_clear_crq_in_flight(crq);
WARN_ON(!cfqq->on_dispatch[sync]);
cfqq->on_dispatch[sync]--;
}
cfq_mark_crq_requeued(crq);
}
} }
/* static void cfq_remove_request(struct request *rq)
* make sure the service time gets corrected on reissue of this request
*/
static void cfq_requeue_request(request_queue_t *q, struct request *rq)
{
cfq_deactivate_request(q, rq);
list_add(&rq->queuelist, &q->queue_head);
}
static void cfq_remove_request(request_queue_t *q, struct request *rq)
{ {
struct cfq_rq *crq = RQ_DATA(rq); struct cfq_rq *crq = RQ_DATA(rq);
if (crq) { list_del_init(&rq->queuelist);
list_del_init(&rq->queuelist); cfq_del_crq_rb(crq);
cfq_del_crq_rb(crq); cfq_del_crq_hash(crq);
cfq_remove_merge_hints(q, crq);
}
} }
static int static int
@ -728,12 +674,6 @@ cfq_merge(request_queue_t *q, struct request **req, struct bio *bio)
struct request *__rq; struct request *__rq;
int ret; int ret;
ret = elv_try_last_merge(q, bio);
if (ret != ELEVATOR_NO_MERGE) {
__rq = q->last_merge;
goto out_insert;
}
__rq = cfq_find_rq_hash(cfqd, bio->bi_sector); __rq = cfq_find_rq_hash(cfqd, bio->bi_sector);
if (__rq && elv_rq_merge_ok(__rq, bio)) { if (__rq && elv_rq_merge_ok(__rq, bio)) {
ret = ELEVATOR_BACK_MERGE; ret = ELEVATOR_BACK_MERGE;
@ -748,8 +688,6 @@ cfq_merge(request_queue_t *q, struct request **req, struct bio *bio)
return ELEVATOR_NO_MERGE; return ELEVATOR_NO_MERGE;
out: out:
q->last_merge = __rq;
out_insert:
*req = __rq; *req = __rq;
return ret; return ret;
} }
@ -762,14 +700,12 @@ static void cfq_merged_request(request_queue_t *q, struct request *req)
cfq_del_crq_hash(crq); cfq_del_crq_hash(crq);
cfq_add_crq_hash(cfqd, crq); cfq_add_crq_hash(cfqd, crq);
if (ON_RB(&crq->rb_node) && (rq_rb_key(req) != crq->rb_key)) { if (rq_rb_key(req) != crq->rb_key) {
struct cfq_queue *cfqq = crq->cfq_queue; struct cfq_queue *cfqq = crq->cfq_queue;
cfq_update_next_crq(crq); cfq_update_next_crq(crq);
cfq_reposition_crq_rb(cfqq, crq); cfq_reposition_crq_rb(cfqq, crq);
} }
q->last_merge = req;
} }
static void static void
@ -785,7 +721,7 @@ cfq_merged_requests(request_queue_t *q, struct request *rq,
time_before(next->start_time, rq->start_time)) time_before(next->start_time, rq->start_time))
list_move(&rq->queuelist, &next->queuelist); list_move(&rq->queuelist, &next->queuelist);
cfq_remove_request(q, next); cfq_remove_request(next);
} }
static inline void static inline void
@ -992,53 +928,15 @@ static int cfq_arm_slice_timer(struct cfq_data *cfqd, struct cfq_queue *cfqq)
return 1; return 1;
} }
/* static void cfq_dispatch_insert(request_queue_t *q, struct cfq_rq *crq)
* we dispatch cfqd->cfq_quantum requests in total from the rr_list queues,
* this function sector sorts the selected request to minimize seeks. we start
* at cfqd->last_sector, not 0.
*/
static void cfq_dispatch_sort(request_queue_t *q, struct cfq_rq *crq)
{ {
struct cfq_data *cfqd = q->elevator->elevator_data; struct cfq_data *cfqd = q->elevator->elevator_data;
struct cfq_queue *cfqq = crq->cfq_queue; struct cfq_queue *cfqq = crq->cfq_queue;
struct list_head *head = &q->queue_head, *entry = head;
struct request *__rq;
sector_t last;
list_del(&crq->request->queuelist);
last = cfqd->last_sector;
list_for_each_entry_reverse(__rq, head, queuelist) {
struct cfq_rq *__crq = RQ_DATA(__rq);
if (blk_barrier_rq(__rq))
break;
if (!blk_fs_request(__rq))
break;
if (cfq_crq_requeued(__crq))
break;
if (__rq->sector <= crq->request->sector)
break;
if (__rq->sector > last && crq->request->sector < last) {
last = crq->request->sector + crq->request->nr_sectors;
break;
}
entry = &__rq->queuelist;
}
cfqd->last_sector = last;
cfqq->next_crq = cfq_find_next_crq(cfqd, cfqq, crq); cfqq->next_crq = cfq_find_next_crq(cfqd, cfqq, crq);
cfq_remove_request(crq->request);
cfq_del_crq_rb(crq);
cfq_remove_merge_hints(q, crq);
cfq_mark_crq_in_flight(crq);
cfq_clear_crq_requeued(crq);
cfqq->on_dispatch[cfq_crq_is_sync(crq)]++; cfqq->on_dispatch[cfq_crq_is_sync(crq)]++;
list_add_tail(&crq->request->queuelist, entry); elv_dispatch_sort(q, crq->request);
} }
/* /*
@ -1159,7 +1057,7 @@ __cfq_dispatch_requests(struct cfq_data *cfqd, struct cfq_queue *cfqq,
/* /*
* finally, insert request into driver dispatch list * finally, insert request into driver dispatch list
*/ */
cfq_dispatch_sort(cfqd->queue, crq); cfq_dispatch_insert(cfqd->queue, crq);
cfqd->dispatch_slice++; cfqd->dispatch_slice++;
dispatched++; dispatched++;
@ -1194,7 +1092,7 @@ __cfq_dispatch_requests(struct cfq_data *cfqd, struct cfq_queue *cfqq,
} }
static int static int
cfq_dispatch_requests(request_queue_t *q, int max_dispatch, int force) cfq_dispatch_requests(request_queue_t *q, int force)
{ {
struct cfq_data *cfqd = q->elevator->elevator_data; struct cfq_data *cfqd = q->elevator->elevator_data;
struct cfq_queue *cfqq; struct cfq_queue *cfqq;
@ -1204,12 +1102,25 @@ cfq_dispatch_requests(request_queue_t *q, int max_dispatch, int force)
cfqq = cfq_select_queue(cfqd, force); cfqq = cfq_select_queue(cfqd, force);
if (cfqq) { if (cfqq) {
int max_dispatch;
/*
* if idle window is disabled, allow queue buildup
*/
if (!cfq_cfqq_idle_window(cfqq) &&
cfqd->rq_in_driver >= cfqd->cfq_max_depth)
return 0;
cfq_clear_cfqq_must_dispatch(cfqq); cfq_clear_cfqq_must_dispatch(cfqq);
cfq_clear_cfqq_wait_request(cfqq); cfq_clear_cfqq_wait_request(cfqq);
del_timer(&cfqd->idle_slice_timer); del_timer(&cfqd->idle_slice_timer);
if (cfq_class_idle(cfqq)) if (!force) {
max_dispatch = 1; max_dispatch = cfqd->cfq_quantum;
if (cfq_class_idle(cfqq))
max_dispatch = 1;
} else
max_dispatch = INT_MAX;
return __cfq_dispatch_requests(cfqd, cfqq, max_dispatch); return __cfq_dispatch_requests(cfqd, cfqq, max_dispatch);
} }
@ -1217,93 +1128,6 @@ cfq_dispatch_requests(request_queue_t *q, int max_dispatch, int force)
return 0; return 0;
} }
static inline void cfq_account_dispatch(struct cfq_rq *crq)
{
struct cfq_queue *cfqq = crq->cfq_queue;
struct cfq_data *cfqd = cfqq->cfqd;
if (unlikely(!blk_fs_request(crq->request)))
return;
/*
* accounted bit is necessary since some drivers will call
* elv_next_request() many times for the same request (eg ide)
*/
if (cfq_crq_in_driver(crq))
return;
cfq_mark_crq_in_driver(crq);
cfqd->rq_in_driver++;
}
static inline void
cfq_account_completion(struct cfq_queue *cfqq, struct cfq_rq *crq)
{
struct cfq_data *cfqd = cfqq->cfqd;
unsigned long now;
if (!cfq_crq_in_driver(crq))
return;
now = jiffies;
WARN_ON(!cfqd->rq_in_driver);
cfqd->rq_in_driver--;
if (!cfq_class_idle(cfqq))
cfqd->last_end_request = now;
if (!cfq_cfqq_dispatched(cfqq)) {
if (cfq_cfqq_on_rr(cfqq)) {
cfqq->service_last = now;
cfq_resort_rr_list(cfqq, 0);
}
if (cfq_cfqq_expired(cfqq)) {
__cfq_slice_expired(cfqd, cfqq, 0);
cfq_schedule_dispatch(cfqd);
}
}
if (cfq_crq_is_sync(crq))
crq->io_context->last_end_request = now;
}
static struct request *cfq_next_request(request_queue_t *q)
{
struct cfq_data *cfqd = q->elevator->elevator_data;
struct request *rq;
if (!list_empty(&q->queue_head)) {
struct cfq_rq *crq;
dispatch:
rq = list_entry_rq(q->queue_head.next);
crq = RQ_DATA(rq);
if (crq) {
struct cfq_queue *cfqq = crq->cfq_queue;
/*
* if idle window is disabled, allow queue buildup
*/
if (!cfq_crq_in_driver(crq) &&
!cfq_cfqq_idle_window(cfqq) &&
!blk_barrier_rq(rq) &&
cfqd->rq_in_driver >= cfqd->cfq_max_depth)
return NULL;
cfq_remove_merge_hints(q, crq);
cfq_account_dispatch(crq);
}
return rq;
}
if (cfq_dispatch_requests(q, cfqd->cfq_quantum, 0))
goto dispatch;
return NULL;
}
/* /*
* task holds one reference to the queue, dropped when task exits. each crq * task holds one reference to the queue, dropped when task exits. each crq
* in-flight on this queue also holds a reference, dropped when crq is freed. * in-flight on this queue also holds a reference, dropped when crq is freed.
@ -1816,8 +1640,9 @@ cfq_crq_enqueued(struct cfq_data *cfqd, struct cfq_queue *cfqq,
} }
} }
static void cfq_enqueue(struct cfq_data *cfqd, struct request *rq) static void cfq_insert_request(request_queue_t *q, struct request *rq)
{ {
struct cfq_data *cfqd = q->elevator->elevator_data;
struct cfq_rq *crq = RQ_DATA(rq); struct cfq_rq *crq = RQ_DATA(rq);
struct cfq_queue *cfqq = crq->cfq_queue; struct cfq_queue *cfqq = crq->cfq_queue;
@ -1827,66 +1652,43 @@ static void cfq_enqueue(struct cfq_data *cfqd, struct request *rq)
list_add_tail(&rq->queuelist, &cfqq->fifo); list_add_tail(&rq->queuelist, &cfqq->fifo);
if (rq_mergeable(rq)) { if (rq_mergeable(rq))
cfq_add_crq_hash(cfqd, crq); cfq_add_crq_hash(cfqd, crq);
if (!cfqd->queue->last_merge)
cfqd->queue->last_merge = rq;
}
cfq_crq_enqueued(cfqd, cfqq, crq); cfq_crq_enqueued(cfqd, cfqq, crq);
} }
static void
cfq_insert_request(request_queue_t *q, struct request *rq, int where)
{
struct cfq_data *cfqd = q->elevator->elevator_data;
switch (where) {
case ELEVATOR_INSERT_BACK:
while (cfq_dispatch_requests(q, INT_MAX, 1))
;
list_add_tail(&rq->queuelist, &q->queue_head);
/*
* If we were idling with pending requests on
* inactive cfqqs, force dispatching will
* remove the idle timer and the queue won't
* be kicked by __make_request() afterward.
* Kick it here.
*/
cfq_schedule_dispatch(cfqd);
break;
case ELEVATOR_INSERT_FRONT:
list_add(&rq->queuelist, &q->queue_head);
break;
case ELEVATOR_INSERT_SORT:
BUG_ON(!blk_fs_request(rq));
cfq_enqueue(cfqd, rq);
break;
default:
printk("%s: bad insert point %d\n", __FUNCTION__,where);
return;
}
}
static void cfq_completed_request(request_queue_t *q, struct request *rq) static void cfq_completed_request(request_queue_t *q, struct request *rq)
{ {
struct cfq_rq *crq = RQ_DATA(rq); struct cfq_rq *crq = RQ_DATA(rq);
struct cfq_queue *cfqq; struct cfq_queue *cfqq = crq->cfq_queue;
struct cfq_data *cfqd = cfqq->cfqd;
const int sync = cfq_crq_is_sync(crq);
unsigned long now;
if (unlikely(!blk_fs_request(rq))) now = jiffies;
return;
cfqq = crq->cfq_queue; WARN_ON(!cfqd->rq_in_driver);
WARN_ON(!cfqq->on_dispatch[sync]);
cfqd->rq_in_driver--;
cfqq->on_dispatch[sync]--;
if (cfq_crq_in_flight(crq)) { if (!cfq_class_idle(cfqq))
const int sync = cfq_crq_is_sync(crq); cfqd->last_end_request = now;
WARN_ON(!cfqq->on_dispatch[sync]); if (!cfq_cfqq_dispatched(cfqq)) {
cfqq->on_dispatch[sync]--; if (cfq_cfqq_on_rr(cfqq)) {
cfqq->service_last = now;
cfq_resort_rr_list(cfqq, 0);
}
if (cfq_cfqq_expired(cfqq)) {
__cfq_slice_expired(cfqd, cfqq, 0);
cfq_schedule_dispatch(cfqd);
}
} }
cfq_account_completion(cfqq, crq); if (cfq_crq_is_sync(crq))
crq->io_context->last_end_request = now;
} }
static struct request * static struct request *
@ -2118,9 +1920,6 @@ cfq_set_request(request_queue_t *q, struct request *rq, struct bio *bio,
INIT_HLIST_NODE(&crq->hash); INIT_HLIST_NODE(&crq->hash);
crq->cfq_queue = cfqq; crq->cfq_queue = cfqq;
crq->io_context = cic; crq->io_context = cic;
cfq_clear_crq_in_flight(crq);
cfq_clear_crq_in_driver(crq);
cfq_clear_crq_requeued(crq);
if (rw == READ || process_sync(tsk)) if (rw == READ || process_sync(tsk))
cfq_mark_crq_is_sync(crq); cfq_mark_crq_is_sync(crq);
@ -2201,7 +2000,7 @@ static void cfq_idle_slice_timer(unsigned long data)
* only expire and reinvoke request handler, if there are * only expire and reinvoke request handler, if there are
* other queues with pending requests * other queues with pending requests
*/ */
if (!cfq_pending_requests(cfqd)) { if (!cfqd->busy_queues) {
cfqd->idle_slice_timer.expires = min(now + cfqd->cfq_slice_idle, cfqq->slice_end); cfqd->idle_slice_timer.expires = min(now + cfqd->cfq_slice_idle, cfqq->slice_end);
add_timer(&cfqd->idle_slice_timer); add_timer(&cfqd->idle_slice_timer);
goto out_cont; goto out_cont;
@ -2576,10 +2375,9 @@ static struct elevator_type iosched_cfq = {
.elevator_merge_fn = cfq_merge, .elevator_merge_fn = cfq_merge,
.elevator_merged_fn = cfq_merged_request, .elevator_merged_fn = cfq_merged_request,
.elevator_merge_req_fn = cfq_merged_requests, .elevator_merge_req_fn = cfq_merged_requests,
.elevator_next_req_fn = cfq_next_request, .elevator_dispatch_fn = cfq_dispatch_requests,
.elevator_add_req_fn = cfq_insert_request, .elevator_add_req_fn = cfq_insert_request,
.elevator_remove_req_fn = cfq_remove_request, .elevator_activate_req_fn = cfq_activate_request,
.elevator_requeue_req_fn = cfq_requeue_request,
.elevator_deactivate_req_fn = cfq_deactivate_request, .elevator_deactivate_req_fn = cfq_deactivate_request,
.elevator_queue_empty_fn = cfq_queue_empty, .elevator_queue_empty_fn = cfq_queue_empty,
.elevator_completed_req_fn = cfq_completed_request, .elevator_completed_req_fn = cfq_completed_request,

View File

@ -50,7 +50,6 @@ struct deadline_data {
* next in sort order. read, write or both are NULL * next in sort order. read, write or both are NULL
*/ */
struct deadline_rq *next_drq[2]; struct deadline_rq *next_drq[2];
struct list_head *dispatch; /* driver dispatch queue */
struct list_head *hash; /* request hash */ struct list_head *hash; /* request hash */
unsigned int batching; /* number of sequential requests made */ unsigned int batching; /* number of sequential requests made */
sector_t last_sector; /* head position */ sector_t last_sector; /* head position */
@ -113,15 +112,6 @@ static inline void deadline_del_drq_hash(struct deadline_rq *drq)
__deadline_del_drq_hash(drq); __deadline_del_drq_hash(drq);
} }
static void
deadline_remove_merge_hints(request_queue_t *q, struct deadline_rq *drq)
{
deadline_del_drq_hash(drq);
if (q->last_merge == drq->request)
q->last_merge = NULL;
}
static inline void static inline void
deadline_add_drq_hash(struct deadline_data *dd, struct deadline_rq *drq) deadline_add_drq_hash(struct deadline_data *dd, struct deadline_rq *drq)
{ {
@ -239,10 +229,9 @@ deadline_del_drq_rb(struct deadline_data *dd, struct deadline_rq *drq)
dd->next_drq[data_dir] = rb_entry_drq(rbnext); dd->next_drq[data_dir] = rb_entry_drq(rbnext);
} }
if (ON_RB(&drq->rb_node)) { BUG_ON(!ON_RB(&drq->rb_node));
rb_erase(&drq->rb_node, DRQ_RB_ROOT(dd, drq)); rb_erase(&drq->rb_node, DRQ_RB_ROOT(dd, drq));
RB_CLEAR(&drq->rb_node); RB_CLEAR(&drq->rb_node);
}
} }
static struct request * static struct request *
@ -286,7 +275,7 @@ deadline_find_first_drq(struct deadline_data *dd, int data_dir)
/* /*
* add drq to rbtree and fifo * add drq to rbtree and fifo
*/ */
static inline void static void
deadline_add_request(struct request_queue *q, struct request *rq) deadline_add_request(struct request_queue *q, struct request *rq)
{ {
struct deadline_data *dd = q->elevator->elevator_data; struct deadline_data *dd = q->elevator->elevator_data;
@ -301,12 +290,8 @@ deadline_add_request(struct request_queue *q, struct request *rq)
drq->expires = jiffies + dd->fifo_expire[data_dir]; drq->expires = jiffies + dd->fifo_expire[data_dir];
list_add_tail(&drq->fifo, &dd->fifo_list[data_dir]); list_add_tail(&drq->fifo, &dd->fifo_list[data_dir]);
if (rq_mergeable(rq)) { if (rq_mergeable(rq))
deadline_add_drq_hash(dd, drq); deadline_add_drq_hash(dd, drq);
if (!q->last_merge)
q->last_merge = rq;
}
} }
/* /*
@ -315,14 +300,11 @@ deadline_add_request(struct request_queue *q, struct request *rq)
static void deadline_remove_request(request_queue_t *q, struct request *rq) static void deadline_remove_request(request_queue_t *q, struct request *rq)
{ {
struct deadline_rq *drq = RQ_DATA(rq); struct deadline_rq *drq = RQ_DATA(rq);
struct deadline_data *dd = q->elevator->elevator_data;
if (drq) { list_del_init(&drq->fifo);
struct deadline_data *dd = q->elevator->elevator_data; deadline_del_drq_rb(dd, drq);
deadline_del_drq_hash(drq);
list_del_init(&drq->fifo);
deadline_remove_merge_hints(q, drq);
deadline_del_drq_rb(dd, drq);
}
} }
static int static int
@ -332,15 +314,6 @@ deadline_merge(request_queue_t *q, struct request **req, struct bio *bio)
struct request *__rq; struct request *__rq;
int ret; int ret;
/*
* try last_merge to avoid going to hash
*/
ret = elv_try_last_merge(q, bio);
if (ret != ELEVATOR_NO_MERGE) {
__rq = q->last_merge;
goto out_insert;
}
/* /*
* see if the merge hash can satisfy a back merge * see if the merge hash can satisfy a back merge
*/ */
@ -373,8 +346,6 @@ deadline_merge(request_queue_t *q, struct request **req, struct bio *bio)
return ELEVATOR_NO_MERGE; return ELEVATOR_NO_MERGE;
out: out:
q->last_merge = __rq;
out_insert:
if (ret) if (ret)
deadline_hot_drq_hash(dd, RQ_DATA(__rq)); deadline_hot_drq_hash(dd, RQ_DATA(__rq));
*req = __rq; *req = __rq;
@ -399,8 +370,6 @@ static void deadline_merged_request(request_queue_t *q, struct request *req)
deadline_del_drq_rb(dd, drq); deadline_del_drq_rb(dd, drq);
deadline_add_drq_rb(dd, drq); deadline_add_drq_rb(dd, drq);
} }
q->last_merge = req;
} }
static void static void
@ -452,7 +421,7 @@ deadline_move_to_dispatch(struct deadline_data *dd, struct deadline_rq *drq)
request_queue_t *q = drq->request->q; request_queue_t *q = drq->request->q;
deadline_remove_request(q, drq->request); deadline_remove_request(q, drq->request);
list_add_tail(&drq->request->queuelist, dd->dispatch); elv_dispatch_add_tail(q, drq->request);
} }
/* /*
@ -502,8 +471,9 @@ static inline int deadline_check_fifo(struct deadline_data *dd, int ddir)
* deadline_dispatch_requests selects the best request according to * deadline_dispatch_requests selects the best request according to
* read/write expire, fifo_batch, etc * read/write expire, fifo_batch, etc
*/ */
static int deadline_dispatch_requests(struct deadline_data *dd) static int deadline_dispatch_requests(request_queue_t *q, int force)
{ {
struct deadline_data *dd = q->elevator->elevator_data;
const int reads = !list_empty(&dd->fifo_list[READ]); const int reads = !list_empty(&dd->fifo_list[READ]);
const int writes = !list_empty(&dd->fifo_list[WRITE]); const int writes = !list_empty(&dd->fifo_list[WRITE]);
struct deadline_rq *drq; struct deadline_rq *drq;
@ -597,65 +567,12 @@ static int deadline_dispatch_requests(struct deadline_data *dd)
return 1; return 1;
} }
static struct request *deadline_next_request(request_queue_t *q)
{
struct deadline_data *dd = q->elevator->elevator_data;
struct request *rq;
/*
* if there are still requests on the dispatch queue, grab the first one
*/
if (!list_empty(dd->dispatch)) {
dispatch:
rq = list_entry_rq(dd->dispatch->next);
return rq;
}
if (deadline_dispatch_requests(dd))
goto dispatch;
return NULL;
}
static void
deadline_insert_request(request_queue_t *q, struct request *rq, int where)
{
struct deadline_data *dd = q->elevator->elevator_data;
/* barriers must flush the reorder queue */
if (unlikely(rq->flags & (REQ_SOFTBARRIER | REQ_HARDBARRIER)
&& where == ELEVATOR_INSERT_SORT))
where = ELEVATOR_INSERT_BACK;
switch (where) {
case ELEVATOR_INSERT_BACK:
while (deadline_dispatch_requests(dd))
;
list_add_tail(&rq->queuelist, dd->dispatch);
break;
case ELEVATOR_INSERT_FRONT:
list_add(&rq->queuelist, dd->dispatch);
break;
case ELEVATOR_INSERT_SORT:
BUG_ON(!blk_fs_request(rq));
deadline_add_request(q, rq);
break;
default:
printk("%s: bad insert point %d\n", __FUNCTION__,where);
return;
}
}
static int deadline_queue_empty(request_queue_t *q) static int deadline_queue_empty(request_queue_t *q)
{ {
struct deadline_data *dd = q->elevator->elevator_data; struct deadline_data *dd = q->elevator->elevator_data;
if (!list_empty(&dd->fifo_list[WRITE]) return list_empty(&dd->fifo_list[WRITE])
|| !list_empty(&dd->fifo_list[READ]) && list_empty(&dd->fifo_list[READ]);
|| !list_empty(dd->dispatch))
return 0;
return 1;
} }
static struct request * static struct request *
@ -733,7 +650,6 @@ static int deadline_init_queue(request_queue_t *q, elevator_t *e)
INIT_LIST_HEAD(&dd->fifo_list[WRITE]); INIT_LIST_HEAD(&dd->fifo_list[WRITE]);
dd->sort_list[READ] = RB_ROOT; dd->sort_list[READ] = RB_ROOT;
dd->sort_list[WRITE] = RB_ROOT; dd->sort_list[WRITE] = RB_ROOT;
dd->dispatch = &q->queue_head;
dd->fifo_expire[READ] = read_expire; dd->fifo_expire[READ] = read_expire;
dd->fifo_expire[WRITE] = write_expire; dd->fifo_expire[WRITE] = write_expire;
dd->writes_starved = writes_starved; dd->writes_starved = writes_starved;
@ -748,10 +664,8 @@ static void deadline_put_request(request_queue_t *q, struct request *rq)
struct deadline_data *dd = q->elevator->elevator_data; struct deadline_data *dd = q->elevator->elevator_data;
struct deadline_rq *drq = RQ_DATA(rq); struct deadline_rq *drq = RQ_DATA(rq);
if (drq) { mempool_free(drq, dd->drq_pool);
mempool_free(drq, dd->drq_pool); rq->elevator_private = NULL;
rq->elevator_private = NULL;
}
} }
static int static int
@ -917,9 +831,8 @@ static struct elevator_type iosched_deadline = {
.elevator_merge_fn = deadline_merge, .elevator_merge_fn = deadline_merge,
.elevator_merged_fn = deadline_merged_request, .elevator_merged_fn = deadline_merged_request,
.elevator_merge_req_fn = deadline_merged_requests, .elevator_merge_req_fn = deadline_merged_requests,
.elevator_next_req_fn = deadline_next_request, .elevator_dispatch_fn = deadline_dispatch_requests,
.elevator_add_req_fn = deadline_insert_request, .elevator_add_req_fn = deadline_add_request,
.elevator_remove_req_fn = deadline_remove_request,
.elevator_queue_empty_fn = deadline_queue_empty, .elevator_queue_empty_fn = deadline_queue_empty,
.elevator_former_req_fn = deadline_former_request, .elevator_former_req_fn = deadline_former_request,
.elevator_latter_req_fn = deadline_latter_request, .elevator_latter_req_fn = deadline_latter_request,

View File

@ -83,15 +83,6 @@ inline int elv_try_merge(struct request *__rq, struct bio *bio)
} }
EXPORT_SYMBOL(elv_try_merge); EXPORT_SYMBOL(elv_try_merge);
inline int elv_try_last_merge(request_queue_t *q, struct bio *bio)
{
if (q->last_merge)
return elv_try_merge(q->last_merge, bio);
return ELEVATOR_NO_MERGE;
}
EXPORT_SYMBOL(elv_try_last_merge);
static struct elevator_type *elevator_find(const char *name) static struct elevator_type *elevator_find(const char *name)
{ {
struct elevator_type *e = NULL; struct elevator_type *e = NULL;
@ -143,6 +134,8 @@ static int elevator_attach(request_queue_t *q, struct elevator_type *e,
INIT_LIST_HEAD(&q->queue_head); INIT_LIST_HEAD(&q->queue_head);
q->last_merge = NULL; q->last_merge = NULL;
q->elevator = eq; q->elevator = eq;
q->end_sector = 0;
q->boundary_rq = NULL;
if (eq->ops->elevator_init_fn) if (eq->ops->elevator_init_fn)
ret = eq->ops->elevator_init_fn(q, eq); ret = eq->ops->elevator_init_fn(q, eq);
@ -225,9 +218,52 @@ void elevator_exit(elevator_t *e)
kfree(e); kfree(e);
} }
/*
* Insert rq into dispatch queue of q. Queue lock must be held on
* entry. If sort != 0, rq is sort-inserted; otherwise, rq will be
* appended to the dispatch queue. To be used by specific elevators.
*/
void elv_dispatch_sort(request_queue_t *q, struct request *rq)
{
sector_t boundary;
struct list_head *entry;
if (q->last_merge == rq)
q->last_merge = NULL;
boundary = q->end_sector;
list_for_each_prev(entry, &q->queue_head) {
struct request *pos = list_entry_rq(entry);
if (pos->flags & (REQ_SOFTBARRIER|REQ_HARDBARRIER|REQ_STARTED))
break;
if (rq->sector >= boundary) {
if (pos->sector < boundary)
continue;
} else {
if (pos->sector >= boundary)
break;
}
if (rq->sector >= pos->sector)
break;
}
list_add(&rq->queuelist, entry);
}
int elv_merge(request_queue_t *q, struct request **req, struct bio *bio) int elv_merge(request_queue_t *q, struct request **req, struct bio *bio)
{ {
elevator_t *e = q->elevator; elevator_t *e = q->elevator;
int ret;
if (q->last_merge) {
ret = elv_try_merge(q->last_merge, bio);
if (ret != ELEVATOR_NO_MERGE) {
*req = q->last_merge;
return ret;
}
}
if (e->ops->elevator_merge_fn) if (e->ops->elevator_merge_fn)
return e->ops->elevator_merge_fn(q, req, bio); return e->ops->elevator_merge_fn(q, req, bio);
@ -241,6 +277,8 @@ void elv_merged_request(request_queue_t *q, struct request *rq)
if (e->ops->elevator_merged_fn) if (e->ops->elevator_merged_fn)
e->ops->elevator_merged_fn(q, rq); e->ops->elevator_merged_fn(q, rq);
q->last_merge = rq;
} }
void elv_merge_requests(request_queue_t *q, struct request *rq, void elv_merge_requests(request_queue_t *q, struct request *rq,
@ -248,20 +286,13 @@ void elv_merge_requests(request_queue_t *q, struct request *rq,
{ {
elevator_t *e = q->elevator; elevator_t *e = q->elevator;
if (q->last_merge == next)
q->last_merge = NULL;
if (e->ops->elevator_merge_req_fn) if (e->ops->elevator_merge_req_fn)
e->ops->elevator_merge_req_fn(q, rq, next); e->ops->elevator_merge_req_fn(q, rq, next);
q->last_merge = rq;
} }
/* void elv_requeue_request(request_queue_t *q, struct request *rq)
* For careful internal use by the block layer. Essentially the same as
* a requeue in that it tells the io scheduler that this request is not
* active in the driver or hardware anymore, but we don't want the request
* added back to the scheduler. Function is not exported.
*/
void elv_deactivate_request(request_queue_t *q, struct request *rq)
{ {
elevator_t *e = q->elevator; elevator_t *e = q->elevator;
@ -269,19 +300,14 @@ void elv_deactivate_request(request_queue_t *q, struct request *rq)
* it already went through dequeue, we need to decrement the * it already went through dequeue, we need to decrement the
* in_flight count again * in_flight count again
*/ */
if (blk_account_rq(rq)) if (blk_account_rq(rq)) {
q->in_flight--; q->in_flight--;
if (blk_sorted_rq(rq) && e->ops->elevator_deactivate_req_fn)
e->ops->elevator_deactivate_req_fn(q, rq);
}
rq->flags &= ~REQ_STARTED; rq->flags &= ~REQ_STARTED;
if (e->ops->elevator_deactivate_req_fn)
e->ops->elevator_deactivate_req_fn(q, rq);
}
void elv_requeue_request(request_queue_t *q, struct request *rq)
{
elv_deactivate_request(q, rq);
/* /*
* if this is the flush, requeue the original instead and drop the flush * if this is the flush, requeue the original instead and drop the flush
*/ */
@ -290,55 +316,91 @@ void elv_requeue_request(request_queue_t *q, struct request *rq)
rq = rq->end_io_data; rq = rq->end_io_data;
} }
/* __elv_add_request(q, rq, ELEVATOR_INSERT_FRONT, 0);
* the request is prepped and may have some resources allocated.
* allowing unprepped requests to pass this one may cause resource
* deadlock. turn on softbarrier.
*/
rq->flags |= REQ_SOFTBARRIER;
/*
* if iosched has an explicit requeue hook, then use that. otherwise
* just put the request at the front of the queue
*/
if (q->elevator->ops->elevator_requeue_req_fn)
q->elevator->ops->elevator_requeue_req_fn(q, rq);
else
__elv_add_request(q, rq, ELEVATOR_INSERT_FRONT, 0);
} }
void __elv_add_request(request_queue_t *q, struct request *rq, int where, void __elv_add_request(request_queue_t *q, struct request *rq, int where,
int plug) int plug)
{ {
/* if (rq->flags & (REQ_SOFTBARRIER | REQ_HARDBARRIER)) {
* barriers implicitly indicate back insertion /*
*/ * barriers implicitly indicate back insertion
if (rq->flags & (REQ_SOFTBARRIER | REQ_HARDBARRIER) && */
where == ELEVATOR_INSERT_SORT) if (where == ELEVATOR_INSERT_SORT)
where = ELEVATOR_INSERT_BACK; where = ELEVATOR_INSERT_BACK;
/*
* this request is scheduling boundary, update end_sector
*/
if (blk_fs_request(rq)) {
q->end_sector = rq_end_sector(rq);
q->boundary_rq = rq;
}
}
if (plug) if (plug)
blk_plug_device(q); blk_plug_device(q);
rq->q = q; rq->q = q;
if (!test_bit(QUEUE_FLAG_DRAIN, &q->queue_flags)) { if (unlikely(test_bit(QUEUE_FLAG_DRAIN, &q->queue_flags))) {
q->elevator->ops->elevator_add_req_fn(q, rq, where);
if (blk_queue_plugged(q)) {
int nrq = q->rq.count[READ] + q->rq.count[WRITE]
- q->in_flight;
if (nrq >= q->unplug_thresh)
__generic_unplug_device(q);
}
} else
/* /*
* if drain is set, store the request "locally". when the drain * if drain is set, store the request "locally". when the drain
* is finished, the requests will be handed ordered to the io * is finished, the requests will be handed ordered to the io
* scheduler * scheduler
*/ */
list_add_tail(&rq->queuelist, &q->drain_list); list_add_tail(&rq->queuelist, &q->drain_list);
return;
}
switch (where) {
case ELEVATOR_INSERT_FRONT:
rq->flags |= REQ_SOFTBARRIER;
list_add(&rq->queuelist, &q->queue_head);
break;
case ELEVATOR_INSERT_BACK:
rq->flags |= REQ_SOFTBARRIER;
while (q->elevator->ops->elevator_dispatch_fn(q, 1))
;
list_add_tail(&rq->queuelist, &q->queue_head);
/*
* We kick the queue here for the following reasons.
* - The elevator might have returned NULL previously
* to delay requests and returned them now. As the
* queue wasn't empty before this request, ll_rw_blk
* won't run the queue on return, resulting in hang.
* - Usually, back inserted requests won't be merged
* with anything. There's no point in delaying queue
* processing.
*/
blk_remove_plug(q);
q->request_fn(q);
break;
case ELEVATOR_INSERT_SORT:
BUG_ON(!blk_fs_request(rq));
rq->flags |= REQ_SORTED;
q->elevator->ops->elevator_add_req_fn(q, rq);
if (q->last_merge == NULL && rq_mergeable(rq))
q->last_merge = rq;
break;
default:
printk(KERN_ERR "%s: bad insertion point %d\n",
__FUNCTION__, where);
BUG();
}
if (blk_queue_plugged(q)) {
int nrq = q->rq.count[READ] + q->rq.count[WRITE]
- q->in_flight;
if (nrq >= q->unplug_thresh)
__generic_unplug_device(q);
}
} }
void elv_add_request(request_queue_t *q, struct request *rq, int where, void elv_add_request(request_queue_t *q, struct request *rq, int where,
@ -353,13 +415,19 @@ void elv_add_request(request_queue_t *q, struct request *rq, int where,
static inline struct request *__elv_next_request(request_queue_t *q) static inline struct request *__elv_next_request(request_queue_t *q)
{ {
struct request *rq = q->elevator->ops->elevator_next_req_fn(q); struct request *rq;
if (unlikely(list_empty(&q->queue_head) &&
!q->elevator->ops->elevator_dispatch_fn(q, 0)))
return NULL;
rq = list_entry_rq(q->queue_head.next);
/* /*
* if this is a barrier write and the device has to issue a * if this is a barrier write and the device has to issue a
* flush sequence to support it, check how far we are * flush sequence to support it, check how far we are
*/ */
if (rq && blk_fs_request(rq) && blk_barrier_rq(rq)) { if (blk_fs_request(rq) && blk_barrier_rq(rq)) {
BUG_ON(q->ordered == QUEUE_ORDERED_NONE); BUG_ON(q->ordered == QUEUE_ORDERED_NONE);
if (q->ordered == QUEUE_ORDERED_FLUSH && if (q->ordered == QUEUE_ORDERED_FLUSH &&
@ -376,15 +444,30 @@ struct request *elv_next_request(request_queue_t *q)
int ret; int ret;
while ((rq = __elv_next_request(q)) != NULL) { while ((rq = __elv_next_request(q)) != NULL) {
/* if (!(rq->flags & REQ_STARTED)) {
* just mark as started even if we don't start it, a request elevator_t *e = q->elevator;
* that has been delayed should not be passed by new incoming
* requests
*/
rq->flags |= REQ_STARTED;
if (rq == q->last_merge) /*
q->last_merge = NULL; * This is the first time the device driver
* sees this request (possibly after
* requeueing). Notify IO scheduler.
*/
if (blk_sorted_rq(rq) &&
e->ops->elevator_activate_req_fn)
e->ops->elevator_activate_req_fn(q, rq);
/*
* just mark as started even if we don't start
* it, a request that has been delayed should
* not be passed by new incoming requests
*/
rq->flags |= REQ_STARTED;
}
if (!q->boundary_rq || q->boundary_rq == rq) {
q->end_sector = rq_end_sector(rq);
q->boundary_rq = NULL;
}
if ((rq->flags & REQ_DONTPREP) || !q->prep_rq_fn) if ((rq->flags & REQ_DONTPREP) || !q->prep_rq_fn)
break; break;
@ -396,9 +479,9 @@ struct request *elv_next_request(request_queue_t *q)
/* /*
* the request may have been (partially) prepped. * the request may have been (partially) prepped.
* we need to keep this request in the front to * we need to keep this request in the front to
* avoid resource deadlock. turn on softbarrier. * avoid resource deadlock. REQ_STARTED will
* prevent other fs requests from passing this one.
*/ */
rq->flags |= REQ_SOFTBARRIER;
rq = NULL; rq = NULL;
break; break;
} else if (ret == BLKPREP_KILL) { } else if (ret == BLKPREP_KILL) {
@ -421,42 +504,32 @@ struct request *elv_next_request(request_queue_t *q)
return rq; return rq;
} }
void elv_remove_request(request_queue_t *q, struct request *rq) void elv_dequeue_request(request_queue_t *q, struct request *rq)
{ {
elevator_t *e = q->elevator; BUG_ON(list_empty(&rq->queuelist));
list_del_init(&rq->queuelist);
/* /*
* the time frame between a request being removed from the lists * the time frame between a request being removed from the lists
* and to it is freed is accounted as io that is in progress at * and to it is freed is accounted as io that is in progress at
* the driver side. note that we only account requests that the * the driver side.
* driver has seen (REQ_STARTED set), to avoid false accounting
* for request-request merges
*/ */
if (blk_account_rq(rq)) if (blk_account_rq(rq))
q->in_flight++; q->in_flight++;
/*
* the main clearing point for q->last_merge is on retrieval of
* request by driver (it calls elv_next_request()), but it _can_
* also happen here if a request is added to the queue but later
* deleted without ever being given to driver (merged with another
* request).
*/
if (rq == q->last_merge)
q->last_merge = NULL;
if (e->ops->elevator_remove_req_fn)
e->ops->elevator_remove_req_fn(q, rq);
} }
int elv_queue_empty(request_queue_t *q) int elv_queue_empty(request_queue_t *q)
{ {
elevator_t *e = q->elevator; elevator_t *e = q->elevator;
if (!list_empty(&q->queue_head))
return 0;
if (e->ops->elevator_queue_empty_fn) if (e->ops->elevator_queue_empty_fn)
return e->ops->elevator_queue_empty_fn(q); return e->ops->elevator_queue_empty_fn(q);
return list_empty(&q->queue_head); return 1;
} }
struct request *elv_latter_request(request_queue_t *q, struct request *rq) struct request *elv_latter_request(request_queue_t *q, struct request *rq)
@ -528,11 +601,11 @@ void elv_completed_request(request_queue_t *q, struct request *rq)
/* /*
* request is released from the driver, io must be done * request is released from the driver, io must be done
*/ */
if (blk_account_rq(rq)) if (blk_account_rq(rq)) {
q->in_flight--; q->in_flight--;
if (blk_sorted_rq(rq) && e->ops->elevator_completed_req_fn)
if (e->ops->elevator_completed_req_fn) e->ops->elevator_completed_req_fn(q, rq);
e->ops->elevator_completed_req_fn(q, rq); }
} }
int elv_register_queue(struct request_queue *q) int elv_register_queue(struct request_queue *q)
@ -705,11 +778,12 @@ ssize_t elv_iosched_show(request_queue_t *q, char *name)
return len; return len;
} }
EXPORT_SYMBOL(elv_dispatch_sort);
EXPORT_SYMBOL(elv_add_request); EXPORT_SYMBOL(elv_add_request);
EXPORT_SYMBOL(__elv_add_request); EXPORT_SYMBOL(__elv_add_request);
EXPORT_SYMBOL(elv_requeue_request); EXPORT_SYMBOL(elv_requeue_request);
EXPORT_SYMBOL(elv_next_request); EXPORT_SYMBOL(elv_next_request);
EXPORT_SYMBOL(elv_remove_request); EXPORT_SYMBOL(elv_dequeue_request);
EXPORT_SYMBOL(elv_queue_empty); EXPORT_SYMBOL(elv_queue_empty);
EXPORT_SYMBOL(elv_completed_request); EXPORT_SYMBOL(elv_completed_request);
EXPORT_SYMBOL(elevator_exit); EXPORT_SYMBOL(elevator_exit);

View File

@ -353,6 +353,8 @@ static void blk_pre_flush_end_io(struct request *flush_rq)
struct request *rq = flush_rq->end_io_data; struct request *rq = flush_rq->end_io_data;
request_queue_t *q = rq->q; request_queue_t *q = rq->q;
elv_completed_request(q, flush_rq);
rq->flags |= REQ_BAR_PREFLUSH; rq->flags |= REQ_BAR_PREFLUSH;
if (!flush_rq->errors) if (!flush_rq->errors)
@ -369,6 +371,8 @@ static void blk_post_flush_end_io(struct request *flush_rq)
struct request *rq = flush_rq->end_io_data; struct request *rq = flush_rq->end_io_data;
request_queue_t *q = rq->q; request_queue_t *q = rq->q;
elv_completed_request(q, flush_rq);
rq->flags |= REQ_BAR_POSTFLUSH; rq->flags |= REQ_BAR_POSTFLUSH;
q->end_flush_fn(q, flush_rq); q->end_flush_fn(q, flush_rq);
@ -408,8 +412,6 @@ struct request *blk_start_pre_flush(request_queue_t *q, struct request *rq)
if (!list_empty(&rq->queuelist)) if (!list_empty(&rq->queuelist))
blkdev_dequeue_request(rq); blkdev_dequeue_request(rq);
elv_deactivate_request(q, rq);
flush_rq->end_io_data = rq; flush_rq->end_io_data = rq;
flush_rq->end_io = blk_pre_flush_end_io; flush_rq->end_io = blk_pre_flush_end_io;
@ -1040,6 +1042,7 @@ EXPORT_SYMBOL(blk_queue_invalidate_tags);
static char *rq_flags[] = { static char *rq_flags[] = {
"REQ_RW", "REQ_RW",
"REQ_FAILFAST", "REQ_FAILFAST",
"REQ_SORTED",
"REQ_SOFTBARRIER", "REQ_SOFTBARRIER",
"REQ_HARDBARRIER", "REQ_HARDBARRIER",
"REQ_CMD", "REQ_CMD",
@ -2456,6 +2459,8 @@ static void __blk_put_request(request_queue_t *q, struct request *req)
if (unlikely(--req->ref_count)) if (unlikely(--req->ref_count))
return; return;
elv_completed_request(q, req);
req->rq_status = RQ_INACTIVE; req->rq_status = RQ_INACTIVE;
req->rl = NULL; req->rl = NULL;
@ -2466,8 +2471,6 @@ static void __blk_put_request(request_queue_t *q, struct request *req)
if (rl) { if (rl) {
int rw = rq_data_dir(req); int rw = rq_data_dir(req);
elv_completed_request(q, req);
BUG_ON(!list_empty(&req->queuelist)); BUG_ON(!list_empty(&req->queuelist));
blk_free_request(q, req); blk_free_request(q, req);
@ -2477,14 +2480,14 @@ static void __blk_put_request(request_queue_t *q, struct request *req)
void blk_put_request(struct request *req) void blk_put_request(struct request *req)
{ {
/* unsigned long flags;
* if req->rl isn't set, this request didnt originate from the request_queue_t *q = req->q;
* block layer, so it's safe to just disregard it
*/
if (req->rl) {
unsigned long flags;
request_queue_t *q = req->q;
/*
* Gee, IDE calls in w/ NULL q. Fix IDE and remove the
* following if (q) test.
*/
if (q) {
spin_lock_irqsave(q->queue_lock, flags); spin_lock_irqsave(q->queue_lock, flags);
__blk_put_request(q, req); __blk_put_request(q, req);
spin_unlock_irqrestore(q->queue_lock, flags); spin_unlock_irqrestore(q->queue_lock, flags);

View File

@ -7,57 +7,19 @@
#include <linux/module.h> #include <linux/module.h>
#include <linux/init.h> #include <linux/init.h>
/* static void elevator_noop_add_request(request_queue_t *q, struct request *rq)
* See if we can find a request that this buffer can be coalesced with.
*/
static int elevator_noop_merge(request_queue_t *q, struct request **req,
struct bio *bio)
{ {
int ret; elv_dispatch_add_tail(q, rq);
ret = elv_try_last_merge(q, bio);
if (ret != ELEVATOR_NO_MERGE)
*req = q->last_merge;
return ret;
} }
static void elevator_noop_merge_requests(request_queue_t *q, struct request *req, static int elevator_noop_dispatch(request_queue_t *q, int force)
struct request *next)
{ {
list_del_init(&next->queuelist); return 0;
}
static void elevator_noop_add_request(request_queue_t *q, struct request *rq,
int where)
{
if (where == ELEVATOR_INSERT_FRONT)
list_add(&rq->queuelist, &q->queue_head);
else
list_add_tail(&rq->queuelist, &q->queue_head);
/*
* new merges must not precede this barrier
*/
if (rq->flags & REQ_HARDBARRIER)
q->last_merge = NULL;
else if (!q->last_merge)
q->last_merge = rq;
}
static struct request *elevator_noop_next_request(request_queue_t *q)
{
if (!list_empty(&q->queue_head))
return list_entry_rq(q->queue_head.next);
return NULL;
} }
static struct elevator_type elevator_noop = { static struct elevator_type elevator_noop = {
.ops = { .ops = {
.elevator_merge_fn = elevator_noop_merge, .elevator_dispatch_fn = elevator_noop_dispatch,
.elevator_merge_req_fn = elevator_noop_merge_requests,
.elevator_next_req_fn = elevator_noop_next_request,
.elevator_add_req_fn = elevator_noop_add_request, .elevator_add_req_fn = elevator_noop_add_request,
}, },
.elevator_name = "noop", .elevator_name = "noop",

View File

@ -203,6 +203,7 @@ struct request {
enum rq_flag_bits { enum rq_flag_bits {
__REQ_RW, /* not set, read. set, write */ __REQ_RW, /* not set, read. set, write */
__REQ_FAILFAST, /* no low level driver retries */ __REQ_FAILFAST, /* no low level driver retries */
__REQ_SORTED, /* elevator knows about this request */
__REQ_SOFTBARRIER, /* may not be passed by ioscheduler */ __REQ_SOFTBARRIER, /* may not be passed by ioscheduler */
__REQ_HARDBARRIER, /* may not be passed by drive either */ __REQ_HARDBARRIER, /* may not be passed by drive either */
__REQ_CMD, /* is a regular fs rw request */ __REQ_CMD, /* is a regular fs rw request */
@ -235,6 +236,7 @@ enum rq_flag_bits {
#define REQ_RW (1 << __REQ_RW) #define REQ_RW (1 << __REQ_RW)
#define REQ_FAILFAST (1 << __REQ_FAILFAST) #define REQ_FAILFAST (1 << __REQ_FAILFAST)
#define REQ_SORTED (1 << __REQ_SORTED)
#define REQ_SOFTBARRIER (1 << __REQ_SOFTBARRIER) #define REQ_SOFTBARRIER (1 << __REQ_SOFTBARRIER)
#define REQ_HARDBARRIER (1 << __REQ_HARDBARRIER) #define REQ_HARDBARRIER (1 << __REQ_HARDBARRIER)
#define REQ_CMD (1 << __REQ_CMD) #define REQ_CMD (1 << __REQ_CMD)
@ -332,6 +334,12 @@ struct request_queue
prepare_flush_fn *prepare_flush_fn; prepare_flush_fn *prepare_flush_fn;
end_flush_fn *end_flush_fn; end_flush_fn *end_flush_fn;
/*
* Dispatch queue sorting
*/
sector_t end_sector;
struct request *boundary_rq;
/* /*
* Auto-unplugging state * Auto-unplugging state
*/ */
@ -454,6 +462,7 @@ enum {
#define blk_pm_request(rq) \ #define blk_pm_request(rq) \
((rq)->flags & (REQ_PM_SUSPEND | REQ_PM_RESUME)) ((rq)->flags & (REQ_PM_SUSPEND | REQ_PM_RESUME))
#define blk_sorted_rq(rq) ((rq)->flags & REQ_SORTED)
#define blk_barrier_rq(rq) ((rq)->flags & REQ_HARDBARRIER) #define blk_barrier_rq(rq) ((rq)->flags & REQ_HARDBARRIER)
#define blk_barrier_preflush(rq) ((rq)->flags & REQ_BAR_PREFLUSH) #define blk_barrier_preflush(rq) ((rq)->flags & REQ_BAR_PREFLUSH)
#define blk_barrier_postflush(rq) ((rq)->flags & REQ_BAR_POSTFLUSH) #define blk_barrier_postflush(rq) ((rq)->flags & REQ_BAR_POSTFLUSH)
@ -611,12 +620,21 @@ extern void end_request(struct request *req, int uptodate);
static inline void blkdev_dequeue_request(struct request *req) static inline void blkdev_dequeue_request(struct request *req)
{ {
BUG_ON(list_empty(&req->queuelist)); elv_dequeue_request(req->q, req);
}
list_del_init(&req->queuelist); /*
* This should be in elevator.h, but that requires pulling in rq and q
*/
static inline void elv_dispatch_add_tail(struct request_queue *q,
struct request *rq)
{
if (q->last_merge == rq)
q->last_merge = NULL;
if (req->rl) q->end_sector = rq_end_sector(rq);
elv_remove_request(req->q, req); q->boundary_rq = rq;
list_add_tail(&rq->queuelist, &q->queue_head);
} }
/* /*

View File

@ -8,18 +8,17 @@ typedef void (elevator_merge_req_fn) (request_queue_t *, struct request *, struc
typedef void (elevator_merged_fn) (request_queue_t *, struct request *); typedef void (elevator_merged_fn) (request_queue_t *, struct request *);
typedef struct request *(elevator_next_req_fn) (request_queue_t *); typedef int (elevator_dispatch_fn) (request_queue_t *, int);
typedef void (elevator_add_req_fn) (request_queue_t *, struct request *, int); typedef void (elevator_add_req_fn) (request_queue_t *, struct request *);
typedef int (elevator_queue_empty_fn) (request_queue_t *); typedef int (elevator_queue_empty_fn) (request_queue_t *);
typedef void (elevator_remove_req_fn) (request_queue_t *, struct request *);
typedef void (elevator_requeue_req_fn) (request_queue_t *, struct request *);
typedef struct request *(elevator_request_list_fn) (request_queue_t *, struct request *); typedef struct request *(elevator_request_list_fn) (request_queue_t *, struct request *);
typedef void (elevator_completed_req_fn) (request_queue_t *, struct request *); typedef void (elevator_completed_req_fn) (request_queue_t *, struct request *);
typedef int (elevator_may_queue_fn) (request_queue_t *, int, struct bio *); typedef int (elevator_may_queue_fn) (request_queue_t *, int, struct bio *);
typedef int (elevator_set_req_fn) (request_queue_t *, struct request *, struct bio *, gfp_t); typedef int (elevator_set_req_fn) (request_queue_t *, struct request *, struct bio *, gfp_t);
typedef void (elevator_put_req_fn) (request_queue_t *, struct request *); typedef void (elevator_put_req_fn) (request_queue_t *, struct request *);
typedef void (elevator_activate_req_fn) (request_queue_t *, struct request *);
typedef void (elevator_deactivate_req_fn) (request_queue_t *, struct request *); typedef void (elevator_deactivate_req_fn) (request_queue_t *, struct request *);
typedef int (elevator_init_fn) (request_queue_t *, elevator_t *); typedef int (elevator_init_fn) (request_queue_t *, elevator_t *);
@ -31,10 +30,9 @@ struct elevator_ops
elevator_merged_fn *elevator_merged_fn; elevator_merged_fn *elevator_merged_fn;
elevator_merge_req_fn *elevator_merge_req_fn; elevator_merge_req_fn *elevator_merge_req_fn;
elevator_next_req_fn *elevator_next_req_fn; elevator_dispatch_fn *elevator_dispatch_fn;
elevator_add_req_fn *elevator_add_req_fn; elevator_add_req_fn *elevator_add_req_fn;
elevator_remove_req_fn *elevator_remove_req_fn; elevator_activate_req_fn *elevator_activate_req_fn;
elevator_requeue_req_fn *elevator_requeue_req_fn;
elevator_deactivate_req_fn *elevator_deactivate_req_fn; elevator_deactivate_req_fn *elevator_deactivate_req_fn;
elevator_queue_empty_fn *elevator_queue_empty_fn; elevator_queue_empty_fn *elevator_queue_empty_fn;
@ -81,15 +79,15 @@ struct elevator_queue
/* /*
* block elevator interface * block elevator interface
*/ */
extern void elv_dispatch_sort(request_queue_t *, struct request *);
extern void elv_add_request(request_queue_t *, struct request *, int, int); extern void elv_add_request(request_queue_t *, struct request *, int, int);
extern void __elv_add_request(request_queue_t *, struct request *, int, int); extern void __elv_add_request(request_queue_t *, struct request *, int, int);
extern int elv_merge(request_queue_t *, struct request **, struct bio *); extern int elv_merge(request_queue_t *, struct request **, struct bio *);
extern void elv_merge_requests(request_queue_t *, struct request *, extern void elv_merge_requests(request_queue_t *, struct request *,
struct request *); struct request *);
extern void elv_merged_request(request_queue_t *, struct request *); extern void elv_merged_request(request_queue_t *, struct request *);
extern void elv_remove_request(request_queue_t *, struct request *); extern void elv_dequeue_request(request_queue_t *, struct request *);
extern void elv_requeue_request(request_queue_t *, struct request *); extern void elv_requeue_request(request_queue_t *, struct request *);
extern void elv_deactivate_request(request_queue_t *, struct request *);
extern int elv_queue_empty(request_queue_t *); extern int elv_queue_empty(request_queue_t *);
extern struct request *elv_next_request(struct request_queue *q); extern struct request *elv_next_request(struct request_queue *q);
extern struct request *elv_former_request(request_queue_t *, struct request *); extern struct request *elv_former_request(request_queue_t *, struct request *);
@ -142,4 +140,6 @@ enum {
ELV_MQUEUE_MUST, ELV_MQUEUE_MUST,
}; };
#define rq_end_sector(rq) ((rq)->sector + (rq)->nr_sectors)
#endif #endif