mirror_zfs/module/zfs/vdev_queue.c

408 lines
10 KiB
C
Raw Normal View History

2008-11-20 23:01:55 +03:00
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
2009-02-18 23:51:31 +03:00
* Copyright 2009 Sun Microsystems, Inc. All rights reserved.
2008-11-20 23:01:55 +03:00
* Use is subject to license terms.
*/
#include <sys/zfs_context.h>
#include <sys/vdev_impl.h>
#include <sys/zio.h>
#include <sys/avl.h>
/*
* These tunables are for performance analysis.
*/
/*
* zfs_vdev_max_pending is the maximum number of i/os concurrently
* pending to each device. zfs_vdev_min_pending is the initial number
* of i/os pending to each device (before it starts ramping up to
* max_pending).
*/
int zfs_vdev_max_pending = 10;
2008-11-20 23:01:55 +03:00
int zfs_vdev_min_pending = 4;
/* deadline = pri + ddi_get_lbolt64() >> time_shift) */
2008-11-20 23:01:55 +03:00
int zfs_vdev_time_shift = 6;
/* exponential I/O issue ramp-up rate */
int zfs_vdev_ramp_rate = 2;
/*
2009-08-18 22:43:27 +04:00
* To reduce IOPs, we aggregate small adjacent I/Os into one large I/O.
* For read I/Os, we also aggregate across small adjacency gaps; for writes
* we include spans of optional I/Os to aid aggregation at the disk even when
* they aren't able to help us aggregate at this level.
2008-11-20 23:01:55 +03:00
*/
int zfs_vdev_aggregation_limit = SPA_MAXBLOCKSIZE;
2009-07-03 02:44:48 +04:00
int zfs_vdev_read_gap_limit = 32 << 10;
2009-08-18 22:43:27 +04:00
int zfs_vdev_write_gap_limit = 4 << 10;
2008-11-20 23:01:55 +03:00
/*
* Virtual device vector for disk I/O scheduling.
*/
int
vdev_queue_deadline_compare(const void *x1, const void *x2)
{
const zio_t *z1 = x1;
const zio_t *z2 = x2;
if (z1->io_deadline < z2->io_deadline)
return (-1);
if (z1->io_deadline > z2->io_deadline)
return (1);
if (z1->io_offset < z2->io_offset)
return (-1);
if (z1->io_offset > z2->io_offset)
return (1);
if (z1 < z2)
return (-1);
if (z1 > z2)
return (1);
return (0);
}
int
vdev_queue_offset_compare(const void *x1, const void *x2)
{
const zio_t *z1 = x1;
const zio_t *z2 = x2;
if (z1->io_offset < z2->io_offset)
return (-1);
if (z1->io_offset > z2->io_offset)
return (1);
if (z1 < z2)
return (-1);
if (z1 > z2)
return (1);
return (0);
}
void
vdev_queue_init(vdev_t *vd)
{
vdev_queue_t *vq = &vd->vdev_queue;
mutex_init(&vq->vq_lock, NULL, MUTEX_DEFAULT, NULL);
avl_create(&vq->vq_deadline_tree, vdev_queue_deadline_compare,
sizeof (zio_t), offsetof(struct zio, io_deadline_node));
avl_create(&vq->vq_read_tree, vdev_queue_offset_compare,
sizeof (zio_t), offsetof(struct zio, io_offset_node));
avl_create(&vq->vq_write_tree, vdev_queue_offset_compare,
sizeof (zio_t), offsetof(struct zio, io_offset_node));
avl_create(&vq->vq_pending_tree, vdev_queue_offset_compare,
sizeof (zio_t), offsetof(struct zio, io_offset_node));
}
void
vdev_queue_fini(vdev_t *vd)
{
vdev_queue_t *vq = &vd->vdev_queue;
avl_destroy(&vq->vq_deadline_tree);
avl_destroy(&vq->vq_read_tree);
avl_destroy(&vq->vq_write_tree);
avl_destroy(&vq->vq_pending_tree);
mutex_destroy(&vq->vq_lock);
}
static void
vdev_queue_io_add(vdev_queue_t *vq, zio_t *zio)
{
avl_add(&vq->vq_deadline_tree, zio);
avl_add(zio->io_vdev_tree, zio);
}
static void
vdev_queue_io_remove(vdev_queue_t *vq, zio_t *zio)
{
avl_remove(&vq->vq_deadline_tree, zio);
avl_remove(zio->io_vdev_tree, zio);
}
static void
vdev_queue_agg_io_done(zio_t *aio)
{
2009-02-18 23:51:31 +03:00
zio_t *pio;
2008-11-20 23:01:55 +03:00
2009-02-18 23:51:31 +03:00
while ((pio = zio_walk_parents(aio)) != NULL)
2008-11-20 23:01:55 +03:00
if (aio->io_type == ZIO_TYPE_READ)
2009-02-18 23:51:31 +03:00
bcopy((char *)aio->io_data + (pio->io_offset -
aio->io_offset), pio->io_data, pio->io_size);
2008-11-20 23:01:55 +03:00
zio_buf_free(aio->io_data, aio->io_size);
}
2009-07-03 02:44:48 +04:00
/*
* Compute the range spanned by two i/os, which is the endpoint of the last
* (lio->io_offset + lio->io_size) minus start of the first (fio->io_offset).
* Conveniently, the gap between fio and lio is given by -IO_SPAN(lio, fio);
* thus fio and lio are adjacent if and only if IO_SPAN(lio, fio) == 0.
*/
#define IO_SPAN(fio, lio) ((lio)->io_offset + (lio)->io_size - (fio)->io_offset)
#define IO_GAP(fio, lio) (-IO_SPAN(lio, fio))
2008-11-20 23:01:55 +03:00
static zio_t *
vdev_queue_io_to_issue(vdev_queue_t *vq, uint64_t pending_limit)
{
2009-08-18 22:43:27 +04:00
zio_t *fio, *lio, *aio, *dio, *nio, *mio;
2009-02-18 23:51:31 +03:00
avl_tree_t *t;
2009-01-16 00:59:39 +03:00
int flags;
2009-07-03 02:44:48 +04:00
uint64_t maxspan = zfs_vdev_aggregation_limit;
uint64_t maxgap;
2009-08-18 22:43:27 +04:00
int stretch;
2008-11-20 23:01:55 +03:00
2009-08-18 22:43:27 +04:00
again:
2008-11-20 23:01:55 +03:00
ASSERT(MUTEX_HELD(&vq->vq_lock));
if (avl_numnodes(&vq->vq_pending_tree) >= pending_limit ||
avl_numnodes(&vq->vq_deadline_tree) == 0)
return (NULL);
fio = lio = avl_first(&vq->vq_deadline_tree);
2009-02-18 23:51:31 +03:00
t = fio->io_vdev_tree;
2009-01-16 00:59:39 +03:00
flags = fio->io_flags & ZIO_FLAG_AGG_INHERIT;
2009-07-03 02:44:48 +04:00
maxgap = (t == &vq->vq_read_tree) ? zfs_vdev_read_gap_limit : 0;
2009-01-16 00:59:39 +03:00
if (!(flags & ZIO_FLAG_DONT_AGGREGATE)) {
/*
2009-08-18 22:43:27 +04:00
* We can aggregate I/Os that are sufficiently adjacent and of
* the same flavor, as expressed by the AGG_INHERIT flags.
* The latter requirement is necessary so that certain
* attributes of the I/O, such as whether it's a normal I/O
* or a scrub/resilver, can be preserved in the aggregate.
* We can include optional I/Os, but don't allow them
* to begin a range as they add no benefit in that situation.
*/
/*
* We keep track of the last non-optional I/O.
*/
mio = (fio->io_flags & ZIO_FLAG_OPTIONAL) ? NULL : fio;
/*
* Walk backwards through sufficiently contiguous I/Os
* recording the last non-option I/O.
2009-01-16 00:59:39 +03:00
*/
2009-02-18 23:51:31 +03:00
while ((dio = AVL_PREV(t, fio)) != NULL &&
2009-01-16 00:59:39 +03:00
(dio->io_flags & ZIO_FLAG_AGG_INHERIT) == flags &&
2009-08-18 22:43:27 +04:00
IO_SPAN(dio, lio) <= maxspan &&
IO_GAP(dio, fio) <= maxgap) {
2009-01-16 00:59:39 +03:00
fio = dio;
2009-08-18 22:43:27 +04:00
if (mio == NULL && !(fio->io_flags & ZIO_FLAG_OPTIONAL))
mio = fio;
}
/*
* Skip any initial optional I/Os.
*/
while ((fio->io_flags & ZIO_FLAG_OPTIONAL) && fio != lio) {
fio = AVL_NEXT(t, fio);
ASSERT(fio != NULL);
}
2009-07-03 02:44:48 +04:00
2009-08-18 22:43:27 +04:00
/*
* Walk forward through sufficiently contiguous I/Os.
*/
2009-02-18 23:51:31 +03:00
while ((dio = AVL_NEXT(t, lio)) != NULL &&
2009-01-16 00:59:39 +03:00
(dio->io_flags & ZIO_FLAG_AGG_INHERIT) == flags &&
2009-08-18 22:43:27 +04:00
IO_SPAN(fio, dio) <= maxspan &&
IO_GAP(lio, dio) <= maxgap) {
2009-01-16 00:59:39 +03:00
lio = dio;
2009-08-18 22:43:27 +04:00
if (!(lio->io_flags & ZIO_FLAG_OPTIONAL))
mio = lio;
}
/*
* Now that we've established the range of the I/O aggregation
* we must decide what to do with trailing optional I/Os.
* For reads, there's nothing to do. While we are unable to
* aggregate further, it's possible that a trailing optional
* I/O would allow the underlying device to aggregate with
* subsequent I/Os. We must therefore determine if the next
* non-optional I/O is close enough to make aggregation
* worthwhile.
*/
stretch = B_FALSE;
if (t != &vq->vq_read_tree && mio != NULL) {
nio = lio;
while ((dio = AVL_NEXT(t, nio)) != NULL &&
IO_GAP(nio, dio) == 0 &&
IO_GAP(mio, dio) <= zfs_vdev_write_gap_limit) {
nio = dio;
if (!(nio->io_flags & ZIO_FLAG_OPTIONAL)) {
stretch = B_TRUE;
break;
}
}
}
if (stretch) {
/* This may be a no-op. */
VERIFY((dio = AVL_NEXT(t, lio)) != NULL);
dio->io_flags &= ~ZIO_FLAG_OPTIONAL;
} else {
while (lio != mio && lio != fio) {
ASSERT(lio->io_flags & ZIO_FLAG_OPTIONAL);
lio = AVL_PREV(t, lio);
ASSERT(lio != NULL);
}
}
2008-11-20 23:01:55 +03:00
}
if (fio != lio) {
2009-07-03 02:44:48 +04:00
uint64_t size = IO_SPAN(fio, lio);
2008-11-20 23:01:55 +03:00
ASSERT(size <= zfs_vdev_aggregation_limit);
aio = zio_vdev_delegated_io(fio->io_vd, fio->io_offset,
zio_buf_alloc(size), size, fio->io_type, ZIO_PRIORITY_AGG,
2009-01-16 00:59:39 +03:00
flags | ZIO_FLAG_DONT_CACHE | ZIO_FLAG_DONT_QUEUE,
2008-11-20 23:01:55 +03:00
vdev_queue_agg_io_done, NULL);
2009-07-03 02:44:48 +04:00
nio = fio;
do {
dio = nio;
nio = AVL_NEXT(t, dio);
2008-11-20 23:01:55 +03:00
ASSERT(dio->io_type == aio->io_type);
2009-02-18 23:51:31 +03:00
ASSERT(dio->io_vdev_tree == t);
2009-08-18 22:43:27 +04:00
if (dio->io_flags & ZIO_FLAG_NODATA) {
ASSERT(dio->io_type == ZIO_TYPE_WRITE);
bzero((char *)aio->io_data + (dio->io_offset -
aio->io_offset), dio->io_size);
} else if (dio->io_type == ZIO_TYPE_WRITE) {
2009-02-18 23:51:31 +03:00
bcopy(dio->io_data, (char *)aio->io_data +
(dio->io_offset - aio->io_offset),
dio->io_size);
2009-08-18 22:43:27 +04:00
}
2009-02-18 23:51:31 +03:00
zio_add_child(dio, aio);
2008-11-20 23:01:55 +03:00
vdev_queue_io_remove(vq, dio);
zio_vdev_io_bypass(dio);
2009-02-18 23:51:31 +03:00
zio_execute(dio);
2009-07-03 02:44:48 +04:00
} while (dio != lio);
2008-11-20 23:01:55 +03:00
avl_add(&vq->vq_pending_tree, aio);
return (aio);
}
2009-02-18 23:51:31 +03:00
ASSERT(fio->io_vdev_tree == t);
2008-11-20 23:01:55 +03:00
vdev_queue_io_remove(vq, fio);
2009-08-18 22:43:27 +04:00
/*
* If the I/O is or was optional and therefore has no data, we need to
* simply discard it. We need to drop the vdev queue's lock to avoid a
* deadlock that we could encounter since this I/O will complete
* immediately.
*/
if (fio->io_flags & ZIO_FLAG_NODATA) {
mutex_exit(&vq->vq_lock);
zio_vdev_io_bypass(fio);
zio_execute(fio);
mutex_enter(&vq->vq_lock);
goto again;
}
2008-11-20 23:01:55 +03:00
avl_add(&vq->vq_pending_tree, fio);
return (fio);
}
zio_t *
vdev_queue_io(zio_t *zio)
{
vdev_queue_t *vq = &zio->io_vd->vdev_queue;
zio_t *nio;
ASSERT(zio->io_type == ZIO_TYPE_READ || zio->io_type == ZIO_TYPE_WRITE);
if (zio->io_flags & ZIO_FLAG_DONT_QUEUE)
return (zio);
zio->io_flags |= ZIO_FLAG_DONT_CACHE | ZIO_FLAG_DONT_QUEUE;
if (zio->io_type == ZIO_TYPE_READ)
zio->io_vdev_tree = &vq->vq_read_tree;
else
zio->io_vdev_tree = &vq->vq_write_tree;
mutex_enter(&vq->vq_lock);
zio->io_deadline = (ddi_get_lbolt64() >> zfs_vdev_time_shift) +
zio->io_priority;
2008-11-20 23:01:55 +03:00
vdev_queue_io_add(vq, zio);
nio = vdev_queue_io_to_issue(vq, zfs_vdev_min_pending);
mutex_exit(&vq->vq_lock);
if (nio == NULL)
return (NULL);
if (nio->io_done == vdev_queue_agg_io_done) {
zio_nowait(nio);
return (NULL);
}
return (nio);
}
void
vdev_queue_io_done(zio_t *zio)
{
vdev_queue_t *vq = &zio->io_vd->vdev_queue;
int i;
2008-11-20 23:01:55 +03:00
mutex_enter(&vq->vq_lock);
avl_remove(&vq->vq_pending_tree, zio);
for (i = 0; i < zfs_vdev_ramp_rate; i++) {
zio_t *nio = vdev_queue_io_to_issue(vq, zfs_vdev_max_pending);
2008-11-20 23:01:55 +03:00
if (nio == NULL)
break;
mutex_exit(&vq->vq_lock);
if (nio->io_done == vdev_queue_agg_io_done) {
zio_nowait(nio);
} else {
zio_vdev_io_reissue(nio);
zio_execute(nio);
}
mutex_enter(&vq->vq_lock);
}
mutex_exit(&vq->vq_lock);
}