kernel-ark/drivers/md/raid5.c

4736 lines
132 KiB
C
Raw Normal View History

/*
* raid5.c : Multiple Devices driver for Linux
* Copyright (C) 1996, 1997 Ingo Molnar, Miguel de Icaza, Gadi Oxman
* Copyright (C) 1999, 2000 Ingo Molnar
* Copyright (C) 2002, 2003 H. Peter Anvin
*
* RAID-4/5/6 management functions.
* Thanks to Penguin Computing for making the RAID-6 development possible
* by donating a test server!
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2, or (at your option)
* any later version.
*
* You should have received a copy of the GNU General Public License
* (for example /usr/src/linux/COPYING); if not, write to the Free
* Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
/*
* BITMAP UNPLUGGING:
*
* The sequencing for updating the bitmap reliably is a little
* subtle (and I got it wrong the first time) so it deserves some
* explanation.
*
* We group bitmap updates into batches. Each batch has a number.
* We may write out several batches at once, but that isn't very important.
* conf->bm_write is the number of the last batch successfully written.
* conf->bm_flush is the number of the last batch that was closed to
* new additions.
* When we discover that we will need to write to any block in a stripe
* (in add_stripe_bio) we update the in-memory bitmap and record in sh->bm_seq
* the number of the batch it will be in. This is bm_flush+1.
* When we are ready to do a write, if that batch hasn't been written yet,
* we plug the array and queue the stripe for later.
* When an unplug happens, we increment bm_flush, thus closing the current
* batch.
* When we notice that bm_flush > bm_write, we write out all pending updates
* to the bitmap, and advance bm_write to where bm_flush was.
* This may occasionally write a bit out twice, but is sure never to
* miss any bits.
*/
#include <linux/module.h>
#include <linux/slab.h>
#include <linux/highmem.h>
#include <linux/bitops.h>
#include <linux/kthread.h>
#include <asm/atomic.h>
#include "raid6.h"
#include <linux/raid/bitmap.h>
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
#include <linux/async_tx.h>
/*
* Stripe cache
*/
#define NR_STRIPES 256
#define STRIPE_SIZE PAGE_SIZE
#define STRIPE_SHIFT (PAGE_SHIFT - 9)
#define STRIPE_SECTORS (STRIPE_SIZE>>9)
#define IO_THRESHOLD 1
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
#define BYPASS_THRESHOLD 1
#define NR_HASH (PAGE_SIZE / sizeof(struct hlist_head))
#define HASH_MASK (NR_HASH - 1)
#define stripe_hash(conf, sect) (&((conf)->stripe_hashtbl[((sect) >> STRIPE_SHIFT) & HASH_MASK]))
/* bio's attached to a stripe+device for I/O are linked together in bi_sector
* order without overlap. There may be several bio's per stripe+device, and
* a bio could span several devices.
* When walking this list for a particular stripe+device, we must never proceed
* beyond a bio that extends past this device, as the next bio might no longer
* be valid.
* This macro is used to determine the 'next' bio in the list, given the sector
* of the current stripe+device
*/
#define r5_next_bio(bio, sect) ( ( (bio)->bi_sector + ((bio)->bi_size>>9) < sect + STRIPE_SECTORS) ? (bio)->bi_next : NULL)
/*
* The following can be used to debug the driver
*/
#define RAID5_PARANOIA 1
#if RAID5_PARANOIA && defined(CONFIG_SMP)
# define CHECK_DEVLOCK() assert_spin_locked(&conf->device_lock)
#else
# define CHECK_DEVLOCK()
#endif
#ifdef DEBUG
#define inline
#define __inline__
#endif
#define printk_rl(args...) ((void) (printk_ratelimit() && printk(args)))
#if !RAID6_USE_EMPTY_ZERO_PAGE
/* In .bss so it's zeroed */
const char raid6_empty_zero_page[PAGE_SIZE] __attribute__((aligned(256)));
#endif
static inline int raid6_next_disk(int disk, int raid_disks)
{
disk++;
return (disk < raid_disks) ? disk : 0;
}
static void return_io(struct bio *return_bi)
{
struct bio *bi = return_bi;
while (bi) {
return_bi = bi->bi_next;
bi->bi_next = NULL;
bi->bi_size = 0;
bio_endio(bi, 0);
bi = return_bi;
}
}
static void print_raid5_conf (raid5_conf_t *conf);
static int stripe_operations_active(struct stripe_head *sh)
{
return sh->check_state || sh->reconstruct_state ||
test_bit(STRIPE_BIOFILL_RUN, &sh->state) ||
test_bit(STRIPE_COMPUTE_RUN, &sh->state);
}
static void __release_stripe(raid5_conf_t *conf, struct stripe_head *sh)
{
if (atomic_dec_and_test(&sh->count)) {
BUG_ON(!list_empty(&sh->lru));
BUG_ON(atomic_read(&conf->active_stripes)==0);
if (test_bit(STRIPE_HANDLE, &sh->state)) {
if (test_bit(STRIPE_DELAYED, &sh->state)) {
list_add_tail(&sh->lru, &conf->delayed_list);
blk_plug_device(conf->mddev->queue);
} else if (test_bit(STRIPE_BIT_DELAY, &sh->state) &&
sh->bm_seq - conf->seq_write > 0) {
list_add_tail(&sh->lru, &conf->bitmap_list);
blk_plug_device(conf->mddev->queue);
} else {
clear_bit(STRIPE_BIT_DELAY, &sh->state);
list_add_tail(&sh->lru, &conf->handle_list);
}
md_wakeup_thread(conf->mddev->thread);
} else {
BUG_ON(stripe_operations_active(sh));
if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
atomic_dec(&conf->preread_active_stripes);
if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD)
md_wakeup_thread(conf->mddev->thread);
}
atomic_dec(&conf->active_stripes);
if (!test_bit(STRIPE_EXPANDING, &sh->state)) {
list_add_tail(&sh->lru, &conf->inactive_list);
wake_up(&conf->wait_for_stripe);
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
if (conf->retry_read_aligned)
md_wakeup_thread(conf->mddev->thread);
}
}
}
}
static void release_stripe(struct stripe_head *sh)
{
raid5_conf_t *conf = sh->raid_conf;
unsigned long flags;
spin_lock_irqsave(&conf->device_lock, flags);
__release_stripe(conf, sh);
spin_unlock_irqrestore(&conf->device_lock, flags);
}
static inline void remove_hash(struct stripe_head *sh)
{
pr_debug("remove_hash(), stripe %llu\n",
(unsigned long long)sh->sector);
hlist_del_init(&sh->hash);
}
static inline void insert_hash(raid5_conf_t *conf, struct stripe_head *sh)
{
struct hlist_head *hp = stripe_hash(conf, sh->sector);
pr_debug("insert_hash(), stripe %llu\n",
(unsigned long long)sh->sector);
CHECK_DEVLOCK();
hlist_add_head(&sh->hash, hp);
}
/* find an idle stripe, make sure it is unhashed, and return it. */
static struct stripe_head *get_free_stripe(raid5_conf_t *conf)
{
struct stripe_head *sh = NULL;
struct list_head *first;
CHECK_DEVLOCK();
if (list_empty(&conf->inactive_list))
goto out;
first = conf->inactive_list.next;
sh = list_entry(first, struct stripe_head, lru);
list_del_init(first);
remove_hash(sh);
atomic_inc(&conf->active_stripes);
out:
return sh;
}
static void shrink_buffers(struct stripe_head *sh, int num)
{
struct page *p;
int i;
for (i=0; i<num ; i++) {
p = sh->dev[i].page;
if (!p)
continue;
sh->dev[i].page = NULL;
put_page(p);
}
}
static int grow_buffers(struct stripe_head *sh, int num)
{
int i;
for (i=0; i<num; i++) {
struct page *page;
if (!(page = alloc_page(GFP_KERNEL))) {
return 1;
}
sh->dev[i].page = page;
}
return 0;
}
static void raid5_build_block (struct stripe_head *sh, int i);
static void init_stripe(struct stripe_head *sh, sector_t sector, int pd_idx, int disks)
{
raid5_conf_t *conf = sh->raid_conf;
int i;
BUG_ON(atomic_read(&sh->count) != 0);
BUG_ON(test_bit(STRIPE_HANDLE, &sh->state));
BUG_ON(stripe_operations_active(sh));
CHECK_DEVLOCK();
pr_debug("init_stripe called, stripe %llu\n",
(unsigned long long)sh->sector);
remove_hash(sh);
sh->sector = sector;
sh->pd_idx = pd_idx;
sh->state = 0;
sh->disks = disks;
for (i = sh->disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if (dev->toread || dev->read || dev->towrite || dev->written ||
test_bit(R5_LOCKED, &dev->flags)) {
printk(KERN_ERR "sector=%llx i=%d %p %p %p %p %d\n",
(unsigned long long)sh->sector, i, dev->toread,
dev->read, dev->towrite, dev->written,
test_bit(R5_LOCKED, &dev->flags));
BUG();
}
dev->flags = 0;
raid5_build_block(sh, i);
}
insert_hash(conf, sh);
}
static struct stripe_head *__find_stripe(raid5_conf_t *conf, sector_t sector, int disks)
{
struct stripe_head *sh;
struct hlist_node *hn;
CHECK_DEVLOCK();
pr_debug("__find_stripe, sector %llu\n", (unsigned long long)sector);
hlist_for_each_entry(sh, hn, stripe_hash(conf, sector), hash)
if (sh->sector == sector && sh->disks == disks)
return sh;
pr_debug("__stripe %llu not in cache\n", (unsigned long long)sector);
return NULL;
}
static void unplug_slaves(mddev_t *mddev);
static void raid5_unplug_device(struct request_queue *q);
static struct stripe_head *get_active_stripe(raid5_conf_t *conf, sector_t sector, int disks,
int pd_idx, int noblock)
{
struct stripe_head *sh;
pr_debug("get_stripe, sector %llu\n", (unsigned long long)sector);
spin_lock_irq(&conf->device_lock);
do {
wait_event_lock_irq(conf->wait_for_stripe,
conf->quiesce == 0,
conf->device_lock, /* nothing */);
sh = __find_stripe(conf, sector, disks);
if (!sh) {
if (!conf->inactive_blocked)
sh = get_free_stripe(conf);
if (noblock && sh == NULL)
break;
if (!sh) {
conf->inactive_blocked = 1;
wait_event_lock_irq(conf->wait_for_stripe,
!list_empty(&conf->inactive_list) &&
(atomic_read(&conf->active_stripes)
< (conf->max_nr_stripes *3/4)
|| !conf->inactive_blocked),
conf->device_lock,
raid5_unplug_device(conf->mddev->queue)
);
conf->inactive_blocked = 0;
} else
init_stripe(sh, sector, pd_idx, disks);
} else {
if (atomic_read(&sh->count)) {
BUG_ON(!list_empty(&sh->lru));
} else {
if (!test_bit(STRIPE_HANDLE, &sh->state))
atomic_inc(&conf->active_stripes);
if (list_empty(&sh->lru) &&
!test_bit(STRIPE_EXPANDING, &sh->state))
BUG();
list_del_init(&sh->lru);
}
}
} while (sh == NULL);
if (sh)
atomic_inc(&sh->count);
spin_unlock_irq(&conf->device_lock);
return sh;
}
static void
raid5_end_read_request(struct bio *bi, int error);
static void
raid5_end_write_request(struct bio *bi, int error);
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
static void ops_run_io(struct stripe_head *sh, struct stripe_head_state *s)
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
{
raid5_conf_t *conf = sh->raid_conf;
int i, disks = sh->disks;
might_sleep();
for (i = disks; i--; ) {
int rw;
struct bio *bi;
mdk_rdev_t *rdev;
if (test_and_clear_bit(R5_Wantwrite, &sh->dev[i].flags))
rw = WRITE;
else if (test_and_clear_bit(R5_Wantread, &sh->dev[i].flags))
rw = READ;
else
continue;
bi = &sh->dev[i].req;
bi->bi_rw = rw;
if (rw == WRITE)
bi->bi_end_io = raid5_end_write_request;
else
bi->bi_end_io = raid5_end_read_request;
rcu_read_lock();
rdev = rcu_dereference(conf->disks[i].rdev);
if (rdev && test_bit(Faulty, &rdev->flags))
rdev = NULL;
if (rdev)
atomic_inc(&rdev->nr_pending);
rcu_read_unlock();
if (rdev) {
if (s->syncing || s->expanding || s->expanded)
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
md_sync_acct(rdev->bdev, STRIPE_SECTORS);
set_bit(STRIPE_IO_STARTED, &sh->state);
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
bi->bi_bdev = rdev->bdev;
pr_debug("%s: for %llu schedule op %ld on disc %d\n",
__func__, (unsigned long long)sh->sector,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
bi->bi_rw, i);
atomic_inc(&sh->count);
bi->bi_sector = sh->sector + rdev->data_offset;
bi->bi_flags = 1 << BIO_UPTODATE;
bi->bi_vcnt = 1;
bi->bi_max_vecs = 1;
bi->bi_idx = 0;
bi->bi_io_vec = &sh->dev[i].vec;
bi->bi_io_vec[0].bv_len = STRIPE_SIZE;
bi->bi_io_vec[0].bv_offset = 0;
bi->bi_size = STRIPE_SIZE;
bi->bi_next = NULL;
if (rw == WRITE &&
test_bit(R5_ReWrite, &sh->dev[i].flags))
atomic_add(STRIPE_SECTORS,
&rdev->corrected_errors);
generic_make_request(bi);
} else {
if (rw == WRITE)
set_bit(STRIPE_DEGRADED, &sh->state);
pr_debug("skip op %ld on disc %d for sector %llu\n",
bi->bi_rw, i, (unsigned long long)sh->sector);
clear_bit(R5_LOCKED, &sh->dev[i].flags);
set_bit(STRIPE_HANDLE, &sh->state);
}
}
}
static struct dma_async_tx_descriptor *
async_copy_data(int frombio, struct bio *bio, struct page *page,
sector_t sector, struct dma_async_tx_descriptor *tx)
{
struct bio_vec *bvl;
struct page *bio_page;
int i;
int page_offset;
if (bio->bi_sector >= sector)
page_offset = (signed)(bio->bi_sector - sector) * 512;
else
page_offset = (signed)(sector - bio->bi_sector) * -512;
bio_for_each_segment(bvl, bio, i) {
int len = bio_iovec_idx(bio, i)->bv_len;
int clen;
int b_offset = 0;
if (page_offset < 0) {
b_offset = -page_offset;
page_offset += b_offset;
len -= b_offset;
}
if (len > 0 && page_offset + len > STRIPE_SIZE)
clen = STRIPE_SIZE - page_offset;
else
clen = len;
if (clen > 0) {
b_offset += bio_iovec_idx(bio, i)->bv_offset;
bio_page = bio_iovec_idx(bio, i)->bv_page;
if (frombio)
tx = async_memcpy(page, bio_page, page_offset,
b_offset, clen,
ASYNC_TX_DEP_ACK,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
tx, NULL, NULL);
else
tx = async_memcpy(bio_page, page, b_offset,
page_offset, clen,
ASYNC_TX_DEP_ACK,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
tx, NULL, NULL);
}
if (clen < len) /* hit end of page */
break;
page_offset += len;
}
return tx;
}
static void ops_complete_biofill(void *stripe_head_ref)
{
struct stripe_head *sh = stripe_head_ref;
struct bio *return_bi = NULL;
raid5_conf_t *conf = sh->raid_conf;
int i;
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
pr_debug("%s: stripe %llu\n", __func__,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
(unsigned long long)sh->sector);
/* clear completed biofills */
spin_lock_irq(&conf->device_lock);
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
for (i = sh->disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
/* acknowledge completion of a biofill operation */
/* and check if we need to reply to a read request,
* new R5_Wantfill requests are held off until
* !STRIPE_BIOFILL_RUN
*/
if (test_and_clear_bit(R5_Wantfill, &dev->flags)) {
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
struct bio *rbi, *rbi2;
BUG_ON(!dev->read);
rbi = dev->read;
dev->read = NULL;
while (rbi && rbi->bi_sector <
dev->sector + STRIPE_SECTORS) {
rbi2 = r5_next_bio(rbi, dev->sector);
if (--rbi->bi_phys_segments == 0) {
rbi->bi_next = return_bi;
return_bi = rbi;
}
rbi = rbi2;
}
}
}
spin_unlock_irq(&conf->device_lock);
clear_bit(STRIPE_BIOFILL_RUN, &sh->state);
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
return_io(return_bi);
set_bit(STRIPE_HANDLE, &sh->state);
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
release_stripe(sh);
}
static void ops_run_biofill(struct stripe_head *sh)
{
struct dma_async_tx_descriptor *tx = NULL;
raid5_conf_t *conf = sh->raid_conf;
int i;
pr_debug("%s: stripe %llu\n", __func__,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
(unsigned long long)sh->sector);
for (i = sh->disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if (test_bit(R5_Wantfill, &dev->flags)) {
struct bio *rbi;
spin_lock_irq(&conf->device_lock);
dev->read = rbi = dev->toread;
dev->toread = NULL;
spin_unlock_irq(&conf->device_lock);
while (rbi && rbi->bi_sector <
dev->sector + STRIPE_SECTORS) {
tx = async_copy_data(0, rbi, dev->page,
dev->sector, tx);
rbi = r5_next_bio(rbi, dev->sector);
}
}
}
atomic_inc(&sh->count);
async_trigger_callback(ASYNC_TX_DEP_ACK | ASYNC_TX_ACK, tx,
ops_complete_biofill, sh);
}
static void ops_complete_compute5(void *stripe_head_ref)
{
struct stripe_head *sh = stripe_head_ref;
int target = sh->ops.target;
struct r5dev *tgt = &sh->dev[target];
pr_debug("%s: stripe %llu\n", __func__,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
(unsigned long long)sh->sector);
set_bit(R5_UPTODATE, &tgt->flags);
BUG_ON(!test_bit(R5_Wantcompute, &tgt->flags));
clear_bit(R5_Wantcompute, &tgt->flags);
clear_bit(STRIPE_COMPUTE_RUN, &sh->state);
if (sh->check_state == check_state_compute_run)
sh->check_state = check_state_compute_result;
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
set_bit(STRIPE_HANDLE, &sh->state);
release_stripe(sh);
}
static struct dma_async_tx_descriptor *ops_run_compute5(struct stripe_head *sh)
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
{
/* kernel stack size limits the total number of disks */
int disks = sh->disks;
struct page *xor_srcs[disks];
int target = sh->ops.target;
struct r5dev *tgt = &sh->dev[target];
struct page *xor_dest = tgt->page;
int count = 0;
struct dma_async_tx_descriptor *tx;
int i;
pr_debug("%s: stripe %llu block: %d\n",
__func__, (unsigned long long)sh->sector, target);
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
BUG_ON(!test_bit(R5_Wantcompute, &tgt->flags));
for (i = disks; i--; )
if (i != target)
xor_srcs[count++] = sh->dev[i].page;
atomic_inc(&sh->count);
if (unlikely(count == 1))
tx = async_memcpy(xor_dest, xor_srcs[0], 0, 0, STRIPE_SIZE,
0, NULL, ops_complete_compute5, sh);
else
tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE,
ASYNC_TX_XOR_ZERO_DST, NULL,
ops_complete_compute5, sh);
return tx;
}
static void ops_complete_prexor(void *stripe_head_ref)
{
struct stripe_head *sh = stripe_head_ref;
pr_debug("%s: stripe %llu\n", __func__,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
(unsigned long long)sh->sector);
}
static struct dma_async_tx_descriptor *
ops_run_prexor(struct stripe_head *sh, struct dma_async_tx_descriptor *tx)
{
/* kernel stack size limits the total number of disks */
int disks = sh->disks;
struct page *xor_srcs[disks];
int count = 0, pd_idx = sh->pd_idx, i;
/* existing parity data subtracted */
struct page *xor_dest = xor_srcs[count++] = sh->dev[pd_idx].page;
pr_debug("%s: stripe %llu\n", __func__,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
(unsigned long long)sh->sector);
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
/* Only process blocks that are known to be uptodate */
if (test_bit(R5_Wantdrain, &dev->flags))
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
xor_srcs[count++] = dev->page;
}
tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE,
ASYNC_TX_DEP_ACK | ASYNC_TX_XOR_DROP_DST, tx,
ops_complete_prexor, sh);
return tx;
}
static struct dma_async_tx_descriptor *
ops_run_biodrain(struct stripe_head *sh, struct dma_async_tx_descriptor *tx)
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
{
int disks = sh->disks;
int i;
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
pr_debug("%s: stripe %llu\n", __func__,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
(unsigned long long)sh->sector);
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
struct bio *chosen;
if (test_and_clear_bit(R5_Wantdrain, &dev->flags)) {
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
struct bio *wbi;
spin_lock(&sh->lock);
chosen = dev->towrite;
dev->towrite = NULL;
BUG_ON(dev->written);
wbi = dev->written = chosen;
spin_unlock(&sh->lock);
while (wbi && wbi->bi_sector <
dev->sector + STRIPE_SECTORS) {
tx = async_copy_data(1, wbi, dev->page,
dev->sector, tx);
wbi = r5_next_bio(wbi, dev->sector);
}
}
}
return tx;
}
static void ops_complete_postxor(void *stripe_head_ref)
{
struct stripe_head *sh = stripe_head_ref;
int disks = sh->disks, i, pd_idx = sh->pd_idx;
pr_debug("%s: stripe %llu\n", __func__,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
(unsigned long long)sh->sector);
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if (dev->written || i == pd_idx)
set_bit(R5_UPTODATE, &dev->flags);
}
if (sh->reconstruct_state == reconstruct_state_drain_run)
sh->reconstruct_state = reconstruct_state_drain_result;
else if (sh->reconstruct_state == reconstruct_state_prexor_drain_run)
sh->reconstruct_state = reconstruct_state_prexor_drain_result;
else {
BUG_ON(sh->reconstruct_state != reconstruct_state_run);
sh->reconstruct_state = reconstruct_state_result;
}
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
set_bit(STRIPE_HANDLE, &sh->state);
release_stripe(sh);
}
static void
ops_run_postxor(struct stripe_head *sh, struct dma_async_tx_descriptor *tx)
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
{
/* kernel stack size limits the total number of disks */
int disks = sh->disks;
struct page *xor_srcs[disks];
int count = 0, pd_idx = sh->pd_idx, i;
struct page *xor_dest;
int prexor = 0;
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
unsigned long flags;
pr_debug("%s: stripe %llu\n", __func__,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
(unsigned long long)sh->sector);
/* check if prexor is active which means only process blocks
* that are part of a read-modify-write (written)
*/
if (sh->reconstruct_state == reconstruct_state_prexor_drain_run) {
prexor = 1;
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
xor_dest = xor_srcs[count++] = sh->dev[pd_idx].page;
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if (dev->written)
xor_srcs[count++] = dev->page;
}
} else {
xor_dest = sh->dev[pd_idx].page;
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if (i != pd_idx)
xor_srcs[count++] = dev->page;
}
}
/* 1/ if we prexor'd then the dest is reused as a source
* 2/ if we did not prexor then we are redoing the parity
* set ASYNC_TX_XOR_DROP_DST and ASYNC_TX_XOR_ZERO_DST
* for the synchronous xor case
*/
flags = ASYNC_TX_DEP_ACK | ASYNC_TX_ACK |
(prexor ? ASYNC_TX_XOR_DROP_DST : ASYNC_TX_XOR_ZERO_DST);
atomic_inc(&sh->count);
if (unlikely(count == 1)) {
flags &= ~(ASYNC_TX_XOR_DROP_DST | ASYNC_TX_XOR_ZERO_DST);
tx = async_memcpy(xor_dest, xor_srcs[0], 0, 0, STRIPE_SIZE,
flags, tx, ops_complete_postxor, sh);
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
} else
tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE,
flags, tx, ops_complete_postxor, sh);
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
}
static void ops_complete_check(void *stripe_head_ref)
{
struct stripe_head *sh = stripe_head_ref;
pr_debug("%s: stripe %llu\n", __func__,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
(unsigned long long)sh->sector);
sh->check_state = check_state_check_result;
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
set_bit(STRIPE_HANDLE, &sh->state);
release_stripe(sh);
}
static void ops_run_check(struct stripe_head *sh)
{
/* kernel stack size limits the total number of disks */
int disks = sh->disks;
struct page *xor_srcs[disks];
struct dma_async_tx_descriptor *tx;
int count = 0, pd_idx = sh->pd_idx, i;
struct page *xor_dest = xor_srcs[count++] = sh->dev[pd_idx].page;
pr_debug("%s: stripe %llu\n", __func__,
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
(unsigned long long)sh->sector);
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if (i != pd_idx)
xor_srcs[count++] = dev->page;
}
tx = async_xor_zero_sum(xor_dest, xor_srcs, 0, count, STRIPE_SIZE,
&sh->ops.zero_sum_result, 0, NULL, NULL, NULL);
atomic_inc(&sh->count);
tx = async_trigger_callback(ASYNC_TX_DEP_ACK | ASYNC_TX_ACK, tx,
ops_complete_check, sh);
}
static void raid5_run_ops(struct stripe_head *sh, unsigned long ops_request)
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
{
int overlap_clear = 0, i, disks = sh->disks;
struct dma_async_tx_descriptor *tx = NULL;
if (test_bit(STRIPE_OP_BIOFILL, &ops_request)) {
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
ops_run_biofill(sh);
overlap_clear++;
}
if (test_bit(STRIPE_OP_COMPUTE_BLK, &ops_request)) {
tx = ops_run_compute5(sh);
/* terminate the chain if postxor is not set to be run */
if (tx && !test_bit(STRIPE_OP_POSTXOR, &ops_request))
async_tx_ack(tx);
}
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
if (test_bit(STRIPE_OP_PREXOR, &ops_request))
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
tx = ops_run_prexor(sh, tx);
if (test_bit(STRIPE_OP_BIODRAIN, &ops_request)) {
tx = ops_run_biodrain(sh, tx);
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
overlap_clear++;
}
if (test_bit(STRIPE_OP_POSTXOR, &ops_request))
ops_run_postxor(sh, tx);
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
if (test_bit(STRIPE_OP_CHECK, &ops_request))
md: raid5_run_ops - run stripe operations outside sh->lock When the raid acceleration work was proposed, Neil laid out the following attack plan: 1/ move the xor and copy operations outside spin_lock(&sh->lock) 2/ find/implement an asynchronous offload api The raid5_run_ops routine uses the asynchronous offload api (async_tx) and the stripe_operations member of a stripe_head to carry out xor+copy operations asynchronously, outside the lock. To perform operations outside the lock a new set of state flags is needed to track new requests, in-flight requests, and completed requests. In this new model handle_stripe is tasked with scanning the stripe_head for work, updating the stripe_operations structure, and finally dropping the lock and calling raid5_run_ops for processing. The following flags outline the requests that handle_stripe can make of raid5_run_ops: STRIPE_OP_BIOFILL - copy data into request buffers to satisfy a read request STRIPE_OP_COMPUTE_BLK - generate a missing block in the cache from the other blocks STRIPE_OP_PREXOR - subtract existing data as part of the read-modify-write process STRIPE_OP_BIODRAIN - copy data out of request buffers to satisfy a write request STRIPE_OP_POSTXOR - recalculate parity for new data that has entered the cache STRIPE_OP_CHECK - verify that the parity is correct STRIPE_OP_IO - submit i/o to the member disks (note this was already performed outside the stripe lock, but it made sense to add it as an operation type The flow is: 1/ handle_stripe sets STRIPE_OP_* in sh->ops.pending 2/ raid5_run_ops reads sh->ops.pending, sets sh->ops.ack, and submits the operation to the async_tx api 3/ async_tx triggers the completion callback routine to set sh->ops.complete and release the stripe 4/ handle_stripe runs again to finish the operation and optionally submit new operations that were previously blocked Note this patch just defines raid5_run_ops, subsequent commits (one per major operation type) modify handle_stripe to take advantage of this routine. Changelog: * removed ops_complete_biodrain in favor of ops_complete_postxor and ops_complete_write. * removed the raid5_run_ops workqueue * call bi_end_io for reads in ops_complete_biofill, saves a call to handle_stripe * explicitly handle the 2-disk raid5 case (xor becomes memcpy), Neil Brown * fix race between async engines and bi_end_io call for reads, Neil Brown * remove unnecessary spin_lock from ops_complete_biofill * remove test_and_set/test_and_clear BUG_ONs, Neil Brown * remove explicit interrupt handling for channel switching, this feature was absorbed (i.e. it is now implicit) by the async_tx api * use return_io in ops_complete_biofill Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 20:52:30 +00:00
ops_run_check(sh);
if (overlap_clear)
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if (test_and_clear_bit(R5_Overlap, &dev->flags))
wake_up(&sh->raid_conf->wait_for_overlap);
}
}
static int grow_one_stripe(raid5_conf_t *conf)
{
struct stripe_head *sh;
sh = kmem_cache_alloc(conf->slab_cache, GFP_KERNEL);
if (!sh)
return 0;
memset(sh, 0, sizeof(*sh) + (conf->raid_disks-1)*sizeof(struct r5dev));
sh->raid_conf = conf;
spin_lock_init(&sh->lock);
if (grow_buffers(sh, conf->raid_disks)) {
shrink_buffers(sh, conf->raid_disks);
kmem_cache_free(conf->slab_cache, sh);
return 0;
}
sh->disks = conf->raid_disks;
/* we just created an active stripe so... */
atomic_set(&sh->count, 1);
atomic_inc(&conf->active_stripes);
INIT_LIST_HEAD(&sh->lru);
release_stripe(sh);
return 1;
}
static int grow_stripes(raid5_conf_t *conf, int num)
{
struct kmem_cache *sc;
int devs = conf->raid_disks;
sprintf(conf->cache_name[0], "raid5-%s", mdname(conf->mddev));
sprintf(conf->cache_name[1], "raid5-%s-alt", mdname(conf->mddev));
conf->active_name = 0;
sc = kmem_cache_create(conf->cache_name[conf->active_name],
sizeof(struct stripe_head)+(devs-1)*sizeof(struct r5dev),
0, 0, NULL);
if (!sc)
return 1;
conf->slab_cache = sc;
conf->pool_size = devs;
while (num--)
if (!grow_one_stripe(conf))
return 1;
return 0;
}
#ifdef CONFIG_MD_RAID5_RESHAPE
static int resize_stripes(raid5_conf_t *conf, int newsize)
{
/* Make all the stripes able to hold 'newsize' devices.
* New slots in each stripe get 'page' set to a new page.
*
* This happens in stages:
* 1/ create a new kmem_cache and allocate the required number of
* stripe_heads.
* 2/ gather all the old stripe_heads and tranfer the pages across
* to the new stripe_heads. This will have the side effect of
* freezing the array as once all stripe_heads have been collected,
* no IO will be possible. Old stripe heads are freed once their
* pages have been transferred over, and the old kmem_cache is
* freed when all stripes are done.
* 3/ reallocate conf->disks to be suitable bigger. If this fails,
* we simple return a failre status - no need to clean anything up.
* 4/ allocate new pages for the new slots in the new stripe_heads.
* If this fails, we don't bother trying the shrink the
* stripe_heads down again, we just leave them as they are.
* As each stripe_head is processed the new one is released into
* active service.
*
* Once step2 is started, we cannot afford to wait for a write,
* so we use GFP_NOIO allocations.
*/
struct stripe_head *osh, *nsh;
LIST_HEAD(newstripes);
struct disk_info *ndisks;
int err;
struct kmem_cache *sc;
int i;
if (newsize <= conf->pool_size)
return 0; /* never bother to shrink */
err = md_allow_write(conf->mddev);
if (err)
return err;
/* Step 1 */
sc = kmem_cache_create(conf->cache_name[1-conf->active_name],
sizeof(struct stripe_head)+(newsize-1)*sizeof(struct r5dev),
0, 0, NULL);
if (!sc)
return -ENOMEM;
for (i = conf->max_nr_stripes; i; i--) {
nsh = kmem_cache_alloc(sc, GFP_KERNEL);
if (!nsh)
break;
memset(nsh, 0, sizeof(*nsh) + (newsize-1)*sizeof(struct r5dev));
nsh->raid_conf = conf;
spin_lock_init(&nsh->lock);
list_add(&nsh->lru, &newstripes);
}
if (i) {
/* didn't get enough, give up */
while (!list_empty(&newstripes)) {
nsh = list_entry(newstripes.next, struct stripe_head, lru);
list_del(&nsh->lru);
kmem_cache_free(sc, nsh);
}
kmem_cache_destroy(sc);
return -ENOMEM;
}
/* Step 2 - Must use GFP_NOIO now.
* OK, we have enough stripes, start collecting inactive
* stripes and copying them over
*/
list_for_each_entry(nsh, &newstripes, lru) {
spin_lock_irq(&conf->device_lock);
wait_event_lock_irq(conf->wait_for_stripe,
!list_empty(&conf->inactive_list),
conf->device_lock,
unplug_slaves(conf->mddev)
);
osh = get_free_stripe(conf);
spin_unlock_irq(&conf->device_lock);
atomic_set(&nsh->count, 1);
for(i=0; i<conf->pool_size; i++)
nsh->dev[i].page = osh->dev[i].page;
for( ; i<newsize; i++)
nsh->dev[i].page = NULL;
kmem_cache_free(conf->slab_cache, osh);
}
kmem_cache_destroy(conf->slab_cache);
/* Step 3.
* At this point, we are holding all the stripes so the array
* is completely stalled, so now is a good time to resize
* conf->disks.
*/
ndisks = kzalloc(newsize * sizeof(struct disk_info), GFP_NOIO);
if (ndisks) {
for (i=0; i<conf->raid_disks; i++)
ndisks[i] = conf->disks[i];
kfree(conf->disks);
conf->disks = ndisks;
} else
err = -ENOMEM;
/* Step 4, return new stripes to service */
while(!list_empty(&newstripes)) {
nsh = list_entry(newstripes.next, struct stripe_head, lru);
list_del_init(&nsh->lru);
for (i=conf->raid_disks; i < newsize; i++)
if (nsh->dev[i].page == NULL) {
struct page *p = alloc_page(GFP_NOIO);
nsh->dev[i].page = p;
if (!p)
err = -ENOMEM;
}
release_stripe(nsh);
}
/* critical section pass, GFP_NOIO no longer needed */
conf->slab_cache = sc;
conf->active_name = 1-conf->active_name;
conf->pool_size = newsize;
return err;
}
#endif
static int drop_one_stripe(raid5_conf_t *conf)
{
struct stripe_head *sh;
spin_lock_irq(&conf->device_lock);
sh = get_free_stripe(conf);
spin_unlock_irq(&conf->device_lock);
if (!sh)
return 0;
BUG_ON(atomic_read(&sh->count));
shrink_buffers(sh, conf->pool_size);
kmem_cache_free(conf->slab_cache, sh);
atomic_dec(&conf->active_stripes);
return 1;
}
static void shrink_stripes(raid5_conf_t *conf)
{
while (drop_one_stripe(conf))
;
if (conf->slab_cache)
kmem_cache_destroy(conf->slab_cache);
conf->slab_cache = NULL;
}
static void raid5_end_read_request(struct bio * bi, int error)
{
struct stripe_head *sh = bi->bi_private;
raid5_conf_t *conf = sh->raid_conf;
int disks = sh->disks, i;
int uptodate = test_bit(BIO_UPTODATE, &bi->bi_flags);
char b[BDEVNAME_SIZE];
mdk_rdev_t *rdev;
for (i=0 ; i<disks; i++)
if (bi == &sh->dev[i].req)
break;
pr_debug("end_read_request %llu/%d, count: %d, uptodate %d.\n",
(unsigned long long)sh->sector, i, atomic_read(&sh->count),
uptodate);
if (i == disks) {
BUG();
return;
}
if (uptodate) {
set_bit(R5_UPTODATE, &sh->dev[i].flags);
if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
rdev = conf->disks[i].rdev;
printk_rl(KERN_INFO "raid5:%s: read error corrected"
" (%lu sectors at %llu on %s)\n",
mdname(conf->mddev), STRIPE_SECTORS,
(unsigned long long)(sh->sector
+ rdev->data_offset),
bdevname(rdev->bdev, b));
clear_bit(R5_ReadError, &sh->dev[i].flags);
clear_bit(R5_ReWrite, &sh->dev[i].flags);
}
if (atomic_read(&conf->disks[i].rdev->read_errors))
atomic_set(&conf->disks[i].rdev->read_errors, 0);
} else {
const char *bdn = bdevname(conf->disks[i].rdev->bdev, b);
int retry = 0;
rdev = conf->disks[i].rdev;
clear_bit(R5_UPTODATE, &sh->dev[i].flags);
atomic_inc(&rdev->read_errors);
if (conf->mddev->degraded)
printk_rl(KERN_WARNING
"raid5:%s: read error not correctable "
"(sector %llu on %s).\n",
mdname(conf->mddev),
(unsigned long long)(sh->sector
+ rdev->data_offset),
bdn);
else if (test_bit(R5_ReWrite, &sh->dev[i].flags))
/* Oh, no!!! */
printk_rl(KERN_WARNING
"raid5:%s: read error NOT corrected!! "
"(sector %llu on %s).\n",
mdname(conf->mddev),
(unsigned long long)(sh->sector
+ rdev->data_offset),
bdn);
else if (atomic_read(&rdev->read_errors)
> conf->max_nr_stripes)
printk(KERN_WARNING
"raid5:%s: Too many read errors, failing device %s.\n",
mdname(conf->mddev), bdn);
else
retry = 1;
if (retry)
set_bit(R5_ReadError, &sh->dev[i].flags);
else {
clear_bit(R5_ReadError, &sh->dev[i].flags);
clear_bit(R5_ReWrite, &sh->dev[i].flags);
md_error(conf->mddev, rdev);
}
}
rdev_dec_pending(conf->disks[i].rdev, conf->mddev);
clear_bit(R5_LOCKED, &sh->dev[i].flags);
set_bit(STRIPE_HANDLE, &sh->state);
release_stripe(sh);
}
static void raid5_end_write_request (struct bio *bi, int error)
{
struct stripe_head *sh = bi->bi_private;
raid5_conf_t *conf = sh->raid_conf;
int disks = sh->disks, i;
int uptodate = test_bit(BIO_UPTODATE, &bi->bi_flags);
for (i=0 ; i<disks; i++)
if (bi == &sh->dev[i].req)
break;
pr_debug("end_write_request %llu/%d, count %d, uptodate: %d.\n",
(unsigned long long)sh->sector, i, atomic_read(&sh->count),
uptodate);
if (i == disks) {
BUG();
return;
}
if (!uptodate)
md_error(conf->mddev, conf->disks[i].rdev);
rdev_dec_pending(conf->disks[i].rdev, conf->mddev);
clear_bit(R5_LOCKED, &sh->dev[i].flags);
set_bit(STRIPE_HANDLE, &sh->state);
release_stripe(sh);
}
static sector_t compute_blocknr(struct stripe_head *sh, int i);
static void raid5_build_block (struct stripe_head *sh, int i)
{
struct r5dev *dev = &sh->dev[i];
bio_init(&dev->req);
dev->req.bi_io_vec = &dev->vec;
dev->req.bi_vcnt++;
dev->req.bi_max_vecs++;
dev->vec.bv_page = dev->page;
dev->vec.bv_len = STRIPE_SIZE;
dev->vec.bv_offset = 0;
dev->req.bi_sector = sh->sector;
dev->req.bi_private = sh;
dev->flags = 0;
dev->sector = compute_blocknr(sh, i);
}
static void error(mddev_t *mddev, mdk_rdev_t *rdev)
{
char b[BDEVNAME_SIZE];
raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
pr_debug("raid5: error called\n");
if (!test_bit(Faulty, &rdev->flags)) {
set_bit(MD_CHANGE_DEVS, &mddev->flags);
if (test_and_clear_bit(In_sync, &rdev->flags)) {
unsigned long flags;
spin_lock_irqsave(&conf->device_lock, flags);
mddev->degraded++;
spin_unlock_irqrestore(&conf->device_lock, flags);
/*
* if recovery was running, make sure it aborts.
*/
md: restart recovery cleanly after device failure. When we get any IO error during a recovery (rebuilding a spare), we abort the recovery and restart it. For RAID6 (and multi-drive RAID1) it may not be best to restart at the beginning: when multiple failures can be tolerated, the recovery may be able to continue and re-doing all that has already been done doesn't make sense. We already have the infrastructure to record where a recovery is up to and restart from there, but it is not being used properly. This is because: - We sometimes abort with MD_RECOVERY_ERR rather than just MD_RECOVERY_INTR, which causes the recovery not be be checkpointed. - We remove spares and then re-added them which loses important state information. The distinction between MD_RECOVERY_ERR and MD_RECOVERY_INTR really isn't needed. If there is an error, the relevant drive will be marked as Faulty, and that is enough to ensure correct handling of the error. So we first remove MD_RECOVERY_ERR, changing some of the uses of it to MD_RECOVERY_INTR. Then we cause the attempt to remove a non-faulty device from an array to fail (unless recovery is impossible as the array is too degraded). Then when remove_and_add_spares attempts to remove the devices on which recovery can continue, it will fail, they will remain in place, and recovery will continue on them as desired. Issue: If we are halfway through rebuilding a spare and another drive fails, and a new spare is immediately available, do we want to: 1/ complete the current rebuild, then go back and rebuild the new spare or 2/ restart the rebuild from the start and rebuild both devices in parallel. Both options can be argued for. The code currently takes option 2 as a/ this requires least code change b/ this results in a minimally-degraded array in minimal time. Cc: "Eivind Sarto" <ivan@kasenna.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-05-23 20:04:39 +00:00
set_bit(MD_RECOVERY_INTR, &mddev->recovery);
}
set_bit(Faulty, &rdev->flags);
printk (KERN_ALERT
"raid5: Disk failure on %s, disabling device.\n"
"raid5: Operation continuing on %d devices.\n",
bdevname(rdev->bdev,b), conf->raid_disks - mddev->degraded);
}
}
/*
* Input: a 'big' sector number,
* Output: index of the data and parity disk, and the sector # in them.
*/
static sector_t raid5_compute_sector(sector_t r_sector, unsigned int raid_disks,
unsigned int data_disks, unsigned int * dd_idx,
unsigned int * pd_idx, raid5_conf_t *conf)
{
long stripe;
unsigned long chunk_number;
unsigned int chunk_offset;
sector_t new_sector;
int sectors_per_chunk = conf->chunk_size >> 9;
/* First compute the information on this sector */
/*
* Compute the chunk number and the sector offset inside the chunk
*/
chunk_offset = sector_div(r_sector, sectors_per_chunk);
chunk_number = r_sector;
BUG_ON(r_sector != chunk_number);
/*
* Compute the stripe number
*/
stripe = chunk_number / data_disks;
/*
* Compute the data disk and parity disk indexes inside the stripe
*/
*dd_idx = chunk_number % data_disks;
/*
* Select the parity disk based on the user selected algorithm.
*/
switch(conf->level) {
case 4:
*pd_idx = data_disks;
break;
case 5:
switch (conf->algorithm) {
case ALGORITHM_LEFT_ASYMMETRIC:
*pd_idx = data_disks - stripe % raid_disks;
if (*dd_idx >= *pd_idx)
(*dd_idx)++;
break;
case ALGORITHM_RIGHT_ASYMMETRIC:
*pd_idx = stripe % raid_disks;
if (*dd_idx >= *pd_idx)
(*dd_idx)++;
break;
case ALGORITHM_LEFT_SYMMETRIC:
*pd_idx = data_disks - stripe % raid_disks;
*dd_idx = (*pd_idx + 1 + *dd_idx) % raid_disks;
break;
case ALGORITHM_RIGHT_SYMMETRIC:
*pd_idx = stripe % raid_disks;
*dd_idx = (*pd_idx + 1 + *dd_idx) % raid_disks;
break;
default:
printk(KERN_ERR "raid5: unsupported algorithm %d\n",
conf->algorithm);
}
break;
case 6:
/**** FIX THIS ****/
switch (conf->algorithm) {
case ALGORITHM_LEFT_ASYMMETRIC:
*pd_idx = raid_disks - 1 - (stripe % raid_disks);
if (*pd_idx == raid_disks-1)
(*dd_idx)++; /* Q D D D P */
else if (*dd_idx >= *pd_idx)
(*dd_idx) += 2; /* D D P Q D */
break;
case ALGORITHM_RIGHT_ASYMMETRIC:
*pd_idx = stripe % raid_disks;
if (*pd_idx == raid_disks-1)
(*dd_idx)++; /* Q D D D P */
else if (*dd_idx >= *pd_idx)
(*dd_idx) += 2; /* D D P Q D */
break;
case ALGORITHM_LEFT_SYMMETRIC:
*pd_idx = raid_disks - 1 - (stripe % raid_disks);
*dd_idx = (*pd_idx + 2 + *dd_idx) % raid_disks;
break;
case ALGORITHM_RIGHT_SYMMETRIC:
*pd_idx = stripe % raid_disks;
*dd_idx = (*pd_idx + 2 + *dd_idx) % raid_disks;
break;
default:
printk (KERN_CRIT "raid6: unsupported algorithm %d\n",
conf->algorithm);
}
break;
}
/*
* Finally, compute the new sector number
*/
new_sector = (sector_t)stripe * sectors_per_chunk + chunk_offset;
return new_sector;
}
static sector_t compute_blocknr(struct stripe_head *sh, int i)
{
raid5_conf_t *conf = sh->raid_conf;
int raid_disks = sh->disks;
int data_disks = raid_disks - conf->max_degraded;
sector_t new_sector = sh->sector, check;
int sectors_per_chunk = conf->chunk_size >> 9;
sector_t stripe;
int chunk_offset;
int chunk_number, dummy1, dummy2, dd_idx = i;
sector_t r_sector;
chunk_offset = sector_div(new_sector, sectors_per_chunk);
stripe = new_sector;
BUG_ON(new_sector != stripe);
if (i == sh->pd_idx)
return 0;
switch(conf->level) {
case 4: break;
case 5:
switch (conf->algorithm) {
case ALGORITHM_LEFT_ASYMMETRIC:
case ALGORITHM_RIGHT_ASYMMETRIC:
if (i > sh->pd_idx)
i--;
break;
case ALGORITHM_LEFT_SYMMETRIC:
case ALGORITHM_RIGHT_SYMMETRIC:
if (i < sh->pd_idx)
i += raid_disks;
i -= (sh->pd_idx + 1);
break;
default:
printk(KERN_ERR "raid5: unsupported algorithm %d\n",
conf->algorithm);
}
break;
case 6:
if (i == raid6_next_disk(sh->pd_idx, raid_disks))
return 0; /* It is the Q disk */
switch (conf->algorithm) {
case ALGORITHM_LEFT_ASYMMETRIC:
case ALGORITHM_RIGHT_ASYMMETRIC:
if (sh->pd_idx == raid_disks-1)
i--; /* Q D D D P */
else if (i > sh->pd_idx)
i -= 2; /* D D P Q D */
break;
case ALGORITHM_LEFT_SYMMETRIC:
case ALGORITHM_RIGHT_SYMMETRIC:
if (sh->pd_idx == raid_disks-1)
i--; /* Q D D D P */
else {
/* D D P Q D */
if (i < sh->pd_idx)
i += raid_disks;
i -= (sh->pd_idx + 2);
}
break;
default:
printk (KERN_CRIT "raid6: unsupported algorithm %d\n",
conf->algorithm);
}
break;
}
chunk_number = stripe * data_disks + i;
r_sector = (sector_t)chunk_number * sectors_per_chunk + chunk_offset;
check = raid5_compute_sector (r_sector, raid_disks, data_disks, &dummy1, &dummy2, conf);
if (check != sh->sector || dummy1 != dd_idx || dummy2 != sh->pd_idx) {
printk(KERN_ERR "compute_blocknr: map not correct\n");
return 0;
}
return r_sector;
}
/*
* Copy data between a page in the stripe cache, and one or more bion
* The page could align with the middle of the bio, or there could be
* several bion, each with several bio_vecs, which cover part of the page
* Multiple bion are linked together on bi_next. There may be extras
* at the end of this list. We ignore them.
*/
static void copy_data(int frombio, struct bio *bio,
struct page *page,
sector_t sector)
{
char *pa = page_address(page);
struct bio_vec *bvl;
int i;
int page_offset;
if (bio->bi_sector >= sector)
page_offset = (signed)(bio->bi_sector - sector) * 512;
else
page_offset = (signed)(sector - bio->bi_sector) * -512;
bio_for_each_segment(bvl, bio, i) {
int len = bio_iovec_idx(bio,i)->bv_len;
int clen;
int b_offset = 0;
if (page_offset < 0) {
b_offset = -page_offset;
page_offset += b_offset;
len -= b_offset;
}
if (len > 0 && page_offset + len > STRIPE_SIZE)
clen = STRIPE_SIZE - page_offset;
else clen = len;
if (clen > 0) {
char *ba = __bio_kmap_atomic(bio, i, KM_USER0);
if (frombio)
memcpy(pa+page_offset, ba+b_offset, clen);
else
memcpy(ba+b_offset, pa+page_offset, clen);
__bio_kunmap_atomic(ba, KM_USER0);
}
if (clen < len) /* hit end of page */
break;
page_offset += len;
}
}
async_tx: add the async_tx api The async_tx api provides methods for describing a chain of asynchronous bulk memory transfers/transforms with support for inter-transactional dependencies. It is implemented as a dmaengine client that smooths over the details of different hardware offload engine implementations. Code that is written to the api can optimize for asynchronous operation and the api will fit the chain of operations to the available offload resources. I imagine that any piece of ADMA hardware would register with the 'async_*' subsystem, and a call to async_X would be routed as appropriate, or be run in-line. - Neil Brown async_tx exploits the capabilities of struct dma_async_tx_descriptor to provide an api of the following general format: struct dma_async_tx_descriptor * async_<operation>(..., struct dma_async_tx_descriptor *depend_tx, dma_async_tx_callback cb_fn, void *cb_param) { struct dma_chan *chan = async_tx_find_channel(depend_tx, <operation>); struct dma_device *device = chan ? chan->device : NULL; int int_en = cb_fn ? 1 : 0; struct dma_async_tx_descriptor *tx = device ? device->device_prep_dma_<operation>(chan, len, int_en) : NULL; if (tx) { /* run <operation> asynchronously */ ... tx->tx_set_dest(addr, tx, index); ... tx->tx_set_src(addr, tx, index); ... async_tx_submit(chan, tx, flags, depend_tx, cb_fn, cb_param); } else { /* run <operation> synchronously */ ... <operation> ... async_tx_sync_epilog(flags, depend_tx, cb_fn, cb_param); } return tx; } async_tx_find_channel() returns a capable channel from its pool. The channel pool is organized as a per-cpu array of channel pointers. The async_tx_rebalance() routine is tasked with managing these arrays. In the uniprocessor case async_tx_rebalance() tries to spread responsibility evenly over channels of similar capabilities. For example if there are two copy+xor channels, one will handle copy operations and the other will handle xor. In the SMP case async_tx_rebalance() attempts to spread the operations evenly over the cpus, e.g. cpu0 gets copy channel0 and xor channel0 while cpu1 gets copy channel 1 and xor channel 1. When a dependency is specified async_tx_find_channel defaults to keeping the operation on the same channel. A xor->copy->xor chain will stay on one channel if it supports both operation types, otherwise the transaction will transition between a copy and a xor resource. Currently the raid5 implementation in the MD raid456 driver has been converted to the async_tx api. A driver for the offload engines on the Intel Xscale series of I/O processors, iop-adma, is provided in a later commit. With the iop-adma driver and async_tx, raid456 is able to offload copy, xor, and xor-zero-sum operations to hardware engines. On iop342 tiobench showed higher throughput for sequential writes (20 - 30% improvement) and sequential reads to a degraded array (40 - 55% improvement). For the other cases performance was roughly equal, +/- a few percentage points. On a x86-smp platform the performance of the async_tx implementation (in synchronous mode) was also +/- a few percentage points of the original implementation. According to 'top' on iop342 CPU utilization drops from ~50% to ~15% during a 'resync' while the speed according to /proc/mdstat doubles from ~25 MB/s to ~50 MB/s. The tiobench command line used for testing was: tiobench --size 2048 --block 4096 --block 131072 --dir /mnt/raid --numruns 5 * iop342 had 1GB of memory available Details: * if CONFIG_DMA_ENGINE=n the asynchronous path is compiled away by making async_tx_find_channel a static inline routine that always returns NULL * when a callback is specified for a given transaction an interrupt will fire at operation completion time and the callback will occur in a tasklet. if the the channel does not support interrupts then a live polling wait will be performed * the api is written as a dmaengine client that requests all available channels * In support of dependencies the api implicitly schedules channel-switch interrupts. The interrupt triggers the cleanup tasklet which causes pending operations to be scheduled on the next channel * Xor engines treat an xor destination address differently than a software xor routine. To the software routine the destination address is an implied source, whereas engines treat it as a write-only destination. This patch modifies the xor_blocks routine to take a an explicit destination address to mirror the hardware. Changelog: * fixed a leftover debug print * don't allow callbacks in async_interrupt_cond * fixed xor_block changes * fixed usage of ASYNC_TX_XOR_DROP_DEST * drop dma mapping methods, suggested by Chris Leech * printk warning fixups from Andrew Morton * don't use inline in C files, Adrian Bunk * select the API when MD is enabled * BUG_ON xor source counts <= 1 * implicitly handle hardware concerns like channel switching and interrupts, Neil Brown * remove the per operation type list, and distribute operation capabilities evenly amongst the available channels * simplify async_tx_find_channel to optimize the fast path * introduce the channel_table_initialized flag to prevent early calls to the api * reorganize the code to mimic crypto * include mm.h as not all archs include it in dma-mapping.h * make the Kconfig options non-user visible, Adrian Bunk * move async_tx under crypto since it is meant as 'core' functionality, and the two may share algorithms in the future * move large inline functions into c files * checkpatch.pl fixes * gpl v2 only correction Cc: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 18:10:44 +00:00
#define check_xor() do { \
if (count == MAX_XOR_BLOCKS) { \
xor_blocks(count, STRIPE_SIZE, dest, ptr);\
count = 0; \
} \
} while(0)
static void compute_parity6(struct stripe_head *sh, int method)
{
raid6_conf_t *conf = sh->raid_conf;
int i, pd_idx = sh->pd_idx, qd_idx, d0_idx, disks = sh->disks, count;
struct bio *chosen;
/**** FIX THIS: This could be very bad if disks is close to 256 ****/
void *ptrs[disks];
qd_idx = raid6_next_disk(pd_idx, disks);
d0_idx = raid6_next_disk(qd_idx, disks);
pr_debug("compute_parity, stripe %llu, method %d\n",
(unsigned long long)sh->sector, method);
switch(method) {
case READ_MODIFY_WRITE:
BUG(); /* READ_MODIFY_WRITE N/A for RAID-6 */
case RECONSTRUCT_WRITE:
for (i= disks; i-- ;)
if ( i != pd_idx && i != qd_idx && sh->dev[i].towrite ) {
chosen = sh->dev[i].towrite;
sh->dev[i].towrite = NULL;
if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
wake_up(&conf->wait_for_overlap);
BUG_ON(sh->dev[i].written);
sh->dev[i].written = chosen;
}
break;
case CHECK_PARITY:
BUG(); /* Not implemented yet */
}
for (i = disks; i--;)
if (sh->dev[i].written) {
sector_t sector = sh->dev[i].sector;
struct bio *wbi = sh->dev[i].written;
while (wbi && wbi->bi_sector < sector + STRIPE_SECTORS) {
copy_data(1, wbi, sh->dev[i].page, sector);
wbi = r5_next_bio(wbi, sector);
}
set_bit(R5_LOCKED, &sh->dev[i].flags);
set_bit(R5_UPTODATE, &sh->dev[i].flags);
}
// switch(method) {
// case RECONSTRUCT_WRITE:
// case CHECK_PARITY:
// case UPDATE_PARITY:
/* Note that unlike RAID-5, the ordering of the disks matters greatly. */
/* FIX: Is this ordering of drives even remotely optimal? */
count = 0;
i = d0_idx;
do {
ptrs[count++] = page_address(sh->dev[i].page);
if (count <= disks-2 && !test_bit(R5_UPTODATE, &sh->dev[i].flags))
printk("block %d/%d not uptodate on parity calc\n", i,count);
i = raid6_next_disk(i, disks);
} while ( i != d0_idx );
// break;
// }
raid6_call.gen_syndrome(disks, STRIPE_SIZE, ptrs);
switch(method) {
case RECONSTRUCT_WRITE:
set_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
set_bit(R5_UPTODATE, &sh->dev[qd_idx].flags);
set_bit(R5_LOCKED, &sh->dev[pd_idx].flags);
set_bit(R5_LOCKED, &sh->dev[qd_idx].flags);
break;
case UPDATE_PARITY:
set_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
set_bit(R5_UPTODATE, &sh->dev[qd_idx].flags);
break;
}
}
/* Compute one missing block */
static void compute_block_1(struct stripe_head *sh, int dd_idx, int nozero)
{
int i, count, disks = sh->disks;
async_tx: add the async_tx api The async_tx api provides methods for describing a chain of asynchronous bulk memory transfers/transforms with support for inter-transactional dependencies. It is implemented as a dmaengine client that smooths over the details of different hardware offload engine implementations. Code that is written to the api can optimize for asynchronous operation and the api will fit the chain of operations to the available offload resources. I imagine that any piece of ADMA hardware would register with the 'async_*' subsystem, and a call to async_X would be routed as appropriate, or be run in-line. - Neil Brown async_tx exploits the capabilities of struct dma_async_tx_descriptor to provide an api of the following general format: struct dma_async_tx_descriptor * async_<operation>(..., struct dma_async_tx_descriptor *depend_tx, dma_async_tx_callback cb_fn, void *cb_param) { struct dma_chan *chan = async_tx_find_channel(depend_tx, <operation>); struct dma_device *device = chan ? chan->device : NULL; int int_en = cb_fn ? 1 : 0; struct dma_async_tx_descriptor *tx = device ? device->device_prep_dma_<operation>(chan, len, int_en) : NULL; if (tx) { /* run <operation> asynchronously */ ... tx->tx_set_dest(addr, tx, index); ... tx->tx_set_src(addr, tx, index); ... async_tx_submit(chan, tx, flags, depend_tx, cb_fn, cb_param); } else { /* run <operation> synchronously */ ... <operation> ... async_tx_sync_epilog(flags, depend_tx, cb_fn, cb_param); } return tx; } async_tx_find_channel() returns a capable channel from its pool. The channel pool is organized as a per-cpu array of channel pointers. The async_tx_rebalance() routine is tasked with managing these arrays. In the uniprocessor case async_tx_rebalance() tries to spread responsibility evenly over channels of similar capabilities. For example if there are two copy+xor channels, one will handle copy operations and the other will handle xor. In the SMP case async_tx_rebalance() attempts to spread the operations evenly over the cpus, e.g. cpu0 gets copy channel0 and xor channel0 while cpu1 gets copy channel 1 and xor channel 1. When a dependency is specified async_tx_find_channel defaults to keeping the operation on the same channel. A xor->copy->xor chain will stay on one channel if it supports both operation types, otherwise the transaction will transition between a copy and a xor resource. Currently the raid5 implementation in the MD raid456 driver has been converted to the async_tx api. A driver for the offload engines on the Intel Xscale series of I/O processors, iop-adma, is provided in a later commit. With the iop-adma driver and async_tx, raid456 is able to offload copy, xor, and xor-zero-sum operations to hardware engines. On iop342 tiobench showed higher throughput for sequential writes (20 - 30% improvement) and sequential reads to a degraded array (40 - 55% improvement). For the other cases performance was roughly equal, +/- a few percentage points. On a x86-smp platform the performance of the async_tx implementation (in synchronous mode) was also +/- a few percentage points of the original implementation. According to 'top' on iop342 CPU utilization drops from ~50% to ~15% during a 'resync' while the speed according to /proc/mdstat doubles from ~25 MB/s to ~50 MB/s. The tiobench command line used for testing was: tiobench --size 2048 --block 4096 --block 131072 --dir /mnt/raid --numruns 5 * iop342 had 1GB of memory available Details: * if CONFIG_DMA_ENGINE=n the asynchronous path is compiled away by making async_tx_find_channel a static inline routine that always returns NULL * when a callback is specified for a given transaction an interrupt will fire at operation completion time and the callback will occur in a tasklet. if the the channel does not support interrupts then a live polling wait will be performed * the api is written as a dmaengine client that requests all available channels * In support of dependencies the api implicitly schedules channel-switch interrupts. The interrupt triggers the cleanup tasklet which causes pending operations to be scheduled on the next channel * Xor engines treat an xor destination address differently than a software xor routine. To the software routine the destination address is an implied source, whereas engines treat it as a write-only destination. This patch modifies the xor_blocks routine to take a an explicit destination address to mirror the hardware. Changelog: * fixed a leftover debug print * don't allow callbacks in async_interrupt_cond * fixed xor_block changes * fixed usage of ASYNC_TX_XOR_DROP_DEST * drop dma mapping methods, suggested by Chris Leech * printk warning fixups from Andrew Morton * don't use inline in C files, Adrian Bunk * select the API when MD is enabled * BUG_ON xor source counts <= 1 * implicitly handle hardware concerns like channel switching and interrupts, Neil Brown * remove the per operation type list, and distribute operation capabilities evenly amongst the available channels * simplify async_tx_find_channel to optimize the fast path * introduce the channel_table_initialized flag to prevent early calls to the api * reorganize the code to mimic crypto * include mm.h as not all archs include it in dma-mapping.h * make the Kconfig options non-user visible, Adrian Bunk * move async_tx under crypto since it is meant as 'core' functionality, and the two may share algorithms in the future * move large inline functions into c files * checkpatch.pl fixes * gpl v2 only correction Cc: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 18:10:44 +00:00
void *ptr[MAX_XOR_BLOCKS], *dest, *p;
int pd_idx = sh->pd_idx;
int qd_idx = raid6_next_disk(pd_idx, disks);
pr_debug("compute_block_1, stripe %llu, idx %d\n",
(unsigned long long)sh->sector, dd_idx);
if ( dd_idx == qd_idx ) {
/* We're actually computing the Q drive */
compute_parity6(sh, UPDATE_PARITY);
} else {
async_tx: add the async_tx api The async_tx api provides methods for describing a chain of asynchronous bulk memory transfers/transforms with support for inter-transactional dependencies. It is implemented as a dmaengine client that smooths over the details of different hardware offload engine implementations. Code that is written to the api can optimize for asynchronous operation and the api will fit the chain of operations to the available offload resources. I imagine that any piece of ADMA hardware would register with the 'async_*' subsystem, and a call to async_X would be routed as appropriate, or be run in-line. - Neil Brown async_tx exploits the capabilities of struct dma_async_tx_descriptor to provide an api of the following general format: struct dma_async_tx_descriptor * async_<operation>(..., struct dma_async_tx_descriptor *depend_tx, dma_async_tx_callback cb_fn, void *cb_param) { struct dma_chan *chan = async_tx_find_channel(depend_tx, <operation>); struct dma_device *device = chan ? chan->device : NULL; int int_en = cb_fn ? 1 : 0; struct dma_async_tx_descriptor *tx = device ? device->device_prep_dma_<operation>(chan, len, int_en) : NULL; if (tx) { /* run <operation> asynchronously */ ... tx->tx_set_dest(addr, tx, index); ... tx->tx_set_src(addr, tx, index); ... async_tx_submit(chan, tx, flags, depend_tx, cb_fn, cb_param); } else { /* run <operation> synchronously */ ... <operation> ... async_tx_sync_epilog(flags, depend_tx, cb_fn, cb_param); } return tx; } async_tx_find_channel() returns a capable channel from its pool. The channel pool is organized as a per-cpu array of channel pointers. The async_tx_rebalance() routine is tasked with managing these arrays. In the uniprocessor case async_tx_rebalance() tries to spread responsibility evenly over channels of similar capabilities. For example if there are two copy+xor channels, one will handle copy operations and the other will handle xor. In the SMP case async_tx_rebalance() attempts to spread the operations evenly over the cpus, e.g. cpu0 gets copy channel0 and xor channel0 while cpu1 gets copy channel 1 and xor channel 1. When a dependency is specified async_tx_find_channel defaults to keeping the operation on the same channel. A xor->copy->xor chain will stay on one channel if it supports both operation types, otherwise the transaction will transition between a copy and a xor resource. Currently the raid5 implementation in the MD raid456 driver has been converted to the async_tx api. A driver for the offload engines on the Intel Xscale series of I/O processors, iop-adma, is provided in a later commit. With the iop-adma driver and async_tx, raid456 is able to offload copy, xor, and xor-zero-sum operations to hardware engines. On iop342 tiobench showed higher throughput for sequential writes (20 - 30% improvement) and sequential reads to a degraded array (40 - 55% improvement). For the other cases performance was roughly equal, +/- a few percentage points. On a x86-smp platform the performance of the async_tx implementation (in synchronous mode) was also +/- a few percentage points of the original implementation. According to 'top' on iop342 CPU utilization drops from ~50% to ~15% during a 'resync' while the speed according to /proc/mdstat doubles from ~25 MB/s to ~50 MB/s. The tiobench command line used for testing was: tiobench --size 2048 --block 4096 --block 131072 --dir /mnt/raid --numruns 5 * iop342 had 1GB of memory available Details: * if CONFIG_DMA_ENGINE=n the asynchronous path is compiled away by making async_tx_find_channel a static inline routine that always returns NULL * when a callback is specified for a given transaction an interrupt will fire at operation completion time and the callback will occur in a tasklet. if the the channel does not support interrupts then a live polling wait will be performed * the api is written as a dmaengine client that requests all available channels * In support of dependencies the api implicitly schedules channel-switch interrupts. The interrupt triggers the cleanup tasklet which causes pending operations to be scheduled on the next channel * Xor engines treat an xor destination address differently than a software xor routine. To the software routine the destination address is an implied source, whereas engines treat it as a write-only destination. This patch modifies the xor_blocks routine to take a an explicit destination address to mirror the hardware. Changelog: * fixed a leftover debug print * don't allow callbacks in async_interrupt_cond * fixed xor_block changes * fixed usage of ASYNC_TX_XOR_DROP_DEST * drop dma mapping methods, suggested by Chris Leech * printk warning fixups from Andrew Morton * don't use inline in C files, Adrian Bunk * select the API when MD is enabled * BUG_ON xor source counts <= 1 * implicitly handle hardware concerns like channel switching and interrupts, Neil Brown * remove the per operation type list, and distribute operation capabilities evenly amongst the available channels * simplify async_tx_find_channel to optimize the fast path * introduce the channel_table_initialized flag to prevent early calls to the api * reorganize the code to mimic crypto * include mm.h as not all archs include it in dma-mapping.h * make the Kconfig options non-user visible, Adrian Bunk * move async_tx under crypto since it is meant as 'core' functionality, and the two may share algorithms in the future * move large inline functions into c files * checkpatch.pl fixes * gpl v2 only correction Cc: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 18:10:44 +00:00
dest = page_address(sh->dev[dd_idx].page);
if (!nozero) memset(dest, 0, STRIPE_SIZE);
count = 0;
for (i = disks ; i--; ) {
if (i == dd_idx || i == qd_idx)
continue;
p = page_address(sh->dev[i].page);
if (test_bit(R5_UPTODATE, &sh->dev[i].flags))
ptr[count++] = p;
else
printk("compute_block() %d, stripe %llu, %d"
" not present\n", dd_idx,
(unsigned long long)sh->sector, i);
check_xor();
}
async_tx: add the async_tx api The async_tx api provides methods for describing a chain of asynchronous bulk memory transfers/transforms with support for inter-transactional dependencies. It is implemented as a dmaengine client that smooths over the details of different hardware offload engine implementations. Code that is written to the api can optimize for asynchronous operation and the api will fit the chain of operations to the available offload resources. I imagine that any piece of ADMA hardware would register with the 'async_*' subsystem, and a call to async_X would be routed as appropriate, or be run in-line. - Neil Brown async_tx exploits the capabilities of struct dma_async_tx_descriptor to provide an api of the following general format: struct dma_async_tx_descriptor * async_<operation>(..., struct dma_async_tx_descriptor *depend_tx, dma_async_tx_callback cb_fn, void *cb_param) { struct dma_chan *chan = async_tx_find_channel(depend_tx, <operation>); struct dma_device *device = chan ? chan->device : NULL; int int_en = cb_fn ? 1 : 0; struct dma_async_tx_descriptor *tx = device ? device->device_prep_dma_<operation>(chan, len, int_en) : NULL; if (tx) { /* run <operation> asynchronously */ ... tx->tx_set_dest(addr, tx, index); ... tx->tx_set_src(addr, tx, index); ... async_tx_submit(chan, tx, flags, depend_tx, cb_fn, cb_param); } else { /* run <operation> synchronously */ ... <operation> ... async_tx_sync_epilog(flags, depend_tx, cb_fn, cb_param); } return tx; } async_tx_find_channel() returns a capable channel from its pool. The channel pool is organized as a per-cpu array of channel pointers. The async_tx_rebalance() routine is tasked with managing these arrays. In the uniprocessor case async_tx_rebalance() tries to spread responsibility evenly over channels of similar capabilities. For example if there are two copy+xor channels, one will handle copy operations and the other will handle xor. In the SMP case async_tx_rebalance() attempts to spread the operations evenly over the cpus, e.g. cpu0 gets copy channel0 and xor channel0 while cpu1 gets copy channel 1 and xor channel 1. When a dependency is specified async_tx_find_channel defaults to keeping the operation on the same channel. A xor->copy->xor chain will stay on one channel if it supports both operation types, otherwise the transaction will transition between a copy and a xor resource. Currently the raid5 implementation in the MD raid456 driver has been converted to the async_tx api. A driver for the offload engines on the Intel Xscale series of I/O processors, iop-adma, is provided in a later commit. With the iop-adma driver and async_tx, raid456 is able to offload copy, xor, and xor-zero-sum operations to hardware engines. On iop342 tiobench showed higher throughput for sequential writes (20 - 30% improvement) and sequential reads to a degraded array (40 - 55% improvement). For the other cases performance was roughly equal, +/- a few percentage points. On a x86-smp platform the performance of the async_tx implementation (in synchronous mode) was also +/- a few percentage points of the original implementation. According to 'top' on iop342 CPU utilization drops from ~50% to ~15% during a 'resync' while the speed according to /proc/mdstat doubles from ~25 MB/s to ~50 MB/s. The tiobench command line used for testing was: tiobench --size 2048 --block 4096 --block 131072 --dir /mnt/raid --numruns 5 * iop342 had 1GB of memory available Details: * if CONFIG_DMA_ENGINE=n the asynchronous path is compiled away by making async_tx_find_channel a static inline routine that always returns NULL * when a callback is specified for a given transaction an interrupt will fire at operation completion time and the callback will occur in a tasklet. if the the channel does not support interrupts then a live polling wait will be performed * the api is written as a dmaengine client that requests all available channels * In support of dependencies the api implicitly schedules channel-switch interrupts. The interrupt triggers the cleanup tasklet which causes pending operations to be scheduled on the next channel * Xor engines treat an xor destination address differently than a software xor routine. To the software routine the destination address is an implied source, whereas engines treat it as a write-only destination. This patch modifies the xor_blocks routine to take a an explicit destination address to mirror the hardware. Changelog: * fixed a leftover debug print * don't allow callbacks in async_interrupt_cond * fixed xor_block changes * fixed usage of ASYNC_TX_XOR_DROP_DEST * drop dma mapping methods, suggested by Chris Leech * printk warning fixups from Andrew Morton * don't use inline in C files, Adrian Bunk * select the API when MD is enabled * BUG_ON xor source counts <= 1 * implicitly handle hardware concerns like channel switching and interrupts, Neil Brown * remove the per operation type list, and distribute operation capabilities evenly amongst the available channels * simplify async_tx_find_channel to optimize the fast path * introduce the channel_table_initialized flag to prevent early calls to the api * reorganize the code to mimic crypto * include mm.h as not all archs include it in dma-mapping.h * make the Kconfig options non-user visible, Adrian Bunk * move async_tx under crypto since it is meant as 'core' functionality, and the two may share algorithms in the future * move large inline functions into c files * checkpatch.pl fixes * gpl v2 only correction Cc: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Acked-By: NeilBrown <neilb@suse.de>
2007-01-02 18:10:44 +00:00
if (count)
xor_blocks(count, STRIPE_SIZE, dest, ptr);
if (!nozero) set_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
else clear_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
}
}
/* Compute two missing blocks */
static void compute_block_2(struct stripe_head *sh, int dd_idx1, int dd_idx2)
{
int i, count, disks = sh->disks;
int pd_idx = sh->pd_idx;
int qd_idx = raid6_next_disk(pd_idx, disks);
int d0_idx = raid6_next_disk(qd_idx, disks);
int faila, failb;
/* faila and failb are disk numbers relative to d0_idx */
/* pd_idx become disks-2 and qd_idx become disks-1 */
faila = (dd_idx1 < d0_idx) ? dd_idx1+(disks-d0_idx) : dd_idx1-d0_idx;
failb = (dd_idx2 < d0_idx) ? dd_idx2+(disks-d0_idx) : dd_idx2-d0_idx;
BUG_ON(faila == failb);
if ( failb < faila ) { int tmp = faila; faila = failb; failb = tmp; }
pr_debug("compute_block_2, stripe %llu, idx %d,%d (%d,%d)\n",
(unsigned long long)sh->sector, dd_idx1, dd_idx2, faila, failb);
if ( failb == disks-1 ) {
/* Q disk is one of the missing disks */
if ( faila == disks-2 ) {
/* Missing P+Q, just recompute */
compute_parity6(sh, UPDATE_PARITY);
return;
} else {
/* We're missing D+Q; recompute D from P */
compute_block_1(sh, (dd_idx1 == qd_idx) ? dd_idx2 : dd_idx1, 0);
compute_parity6(sh, UPDATE_PARITY); /* Is this necessary? */
return;
}
}
/* We're missing D+P or D+D; build pointer table */
{
/**** FIX THIS: This could be very bad if disks is close to 256 ****/
void *ptrs[disks];
count = 0;
i = d0_idx;
do {
ptrs[count++] = page_address(sh->dev[i].page);
i = raid6_next_disk(i, disks);
if (i != dd_idx1 && i != dd_idx2 &&
!test_bit(R5_UPTODATE, &sh->dev[i].flags))
printk("compute_2 with missing block %d/%d\n", count, i);
} while ( i != d0_idx );
if ( failb == disks-2 ) {
/* We're missing D+P. */
raid6_datap_recov(disks, STRIPE_SIZE, faila, ptrs);
} else {
/* We're missing D+D. */
raid6_2data_recov(disks, STRIPE_SIZE, faila, failb, ptrs);
}
/* Both the above update both missing blocks */
set_bit(R5_UPTODATE, &sh->dev[dd_idx1].flags);
set_bit(R5_UPTODATE, &sh->dev[dd_idx2].flags);
}
}
static void
schedule_reconstruction5(struct stripe_head *sh, struct stripe_head_state *s,
int rcw, int expand)
{
int i, pd_idx = sh->pd_idx, disks = sh->disks;
if (rcw) {
/* if we are not expanding this is a proper write request, and
* there will be bios with new data to be drained into the
* stripe cache
*/
if (!expand) {
sh->reconstruct_state = reconstruct_state_drain_run;
set_bit(STRIPE_OP_BIODRAIN, &s->ops_request);
} else
sh->reconstruct_state = reconstruct_state_run;
set_bit(STRIPE_OP_POSTXOR, &s->ops_request);
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if (dev->towrite) {
set_bit(R5_LOCKED, &dev->flags);
set_bit(R5_Wantdrain, &dev->flags);
if (!expand)
clear_bit(R5_UPTODATE, &dev->flags);
s->locked++;
}
}
if (s->locked + 1 == disks)
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
if (!test_and_set_bit(STRIPE_FULL_WRITE, &sh->state))
atomic_inc(&sh->raid_conf->pending_full_writes);
} else {
BUG_ON(!(test_bit(R5_UPTODATE, &sh->dev[pd_idx].flags) ||
test_bit(R5_Wantcompute, &sh->dev[pd_idx].flags)));
sh->reconstruct_state = reconstruct_state_prexor_drain_run;
set_bit(STRIPE_OP_PREXOR, &s->ops_request);
set_bit(STRIPE_OP_BIODRAIN, &s->ops_request);
set_bit(STRIPE_OP_POSTXOR, &s->ops_request);
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if (i == pd_idx)
continue;
if (dev->towrite &&
(test_bit(R5_UPTODATE, &dev->flags) ||
test_bit(R5_Wantcompute, &dev->flags))) {
set_bit(R5_Wantdrain, &dev->flags);
set_bit(R5_LOCKED, &dev->flags);
clear_bit(R5_UPTODATE, &dev->flags);
s->locked++;
}
}
}
/* keep the parity disk locked while asynchronous operations
* are in flight
*/
set_bit(R5_LOCKED, &sh->dev[pd_idx].flags);
clear_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
s->locked++;
pr_debug("%s: stripe %llu locked: %d ops_request: %lx\n",
__func__, (unsigned long long)sh->sector,
s->locked, s->ops_request);
}
/*
* Each stripe/dev can have one or more bion attached.
* toread/towrite point to the first in a chain.
* The bi_next chain must be in order.
*/
static int add_stripe_bio(struct stripe_head *sh, struct bio *bi, int dd_idx, int forwrite)
{
struct bio **bip;
raid5_conf_t *conf = sh->raid_conf;
int firstwrite=0;
pr_debug("adding bh b#%llu to stripe s#%llu\n",
(unsigned long long)bi->bi_sector,
(unsigned long long)sh->sector);
spin_lock(&sh->lock);
spin_lock_irq(&conf->device_lock);
if (forwrite) {
bip = &sh->dev[dd_idx].towrite;
if (*bip == NULL && sh->dev[dd_idx].written == NULL)
firstwrite = 1;
} else
bip = &sh->dev[dd_idx].toread;
while (*bip && (*bip)->bi_sector < bi->bi_sector) {
if ((*bip)->bi_sector + ((*bip)->bi_size >> 9) > bi->bi_sector)
goto overlap;
bip = & (*bip)->bi_next;
}
if (*bip && (*bip)->bi_sector < bi->bi_sector + ((bi->bi_size)>>9))
goto overlap;
BUG_ON(*bip && bi->bi_next && (*bip) != bi->bi_next);
if (*bip)
bi->bi_next = *bip;
*bip = bi;
bi->bi_phys_segments ++;
spin_unlock_irq(&conf->device_lock);
spin_unlock(&sh->lock);
pr_debug("added bi b#%llu to stripe s#%llu, disk %d.\n",
(unsigned long long)bi->bi_sector,
(unsigned long long)sh->sector, dd_idx);
if (conf->mddev->bitmap && firstwrite) {
bitmap_startwrite(conf->mddev->bitmap, sh->sector,
STRIPE_SECTORS, 0);
sh->bm_seq = conf->seq_flush+1;
set_bit(STRIPE_BIT_DELAY, &sh->state);
}
if (forwrite) {
/* check if page is covered */
sector_t sector = sh->dev[dd_idx].sector;
for (bi=sh->dev[dd_idx].towrite;
sector < sh->dev[dd_idx].sector + STRIPE_SECTORS &&
bi && bi->bi_sector <= sector;
bi = r5_next_bio(bi, sh->dev[dd_idx].sector)) {
if (bi->bi_sector + (bi->bi_size>>9) >= sector)
sector = bi->bi_sector + (bi->bi_size>>9);
}
if (sector >= sh->dev[dd_idx].sector + STRIPE_SECTORS)
set_bit(R5_OVERWRITE, &sh->dev[dd_idx].flags);
}
return 1;
overlap:
set_bit(R5_Overlap, &sh->dev[dd_idx].flags);
spin_unlock_irq(&conf->device_lock);
spin_unlock(&sh->lock);
return 0;
}
static void end_reshape(raid5_conf_t *conf);
static int page_is_zero(struct page *p)
{
char *a = page_address(p);
return ((*(u32*)a) == 0 &&
memcmp(a, a+4, STRIPE_SIZE-4)==0);
}
static int stripe_to_pdidx(sector_t stripe, raid5_conf_t *conf, int disks)
{
int sectors_per_chunk = conf->chunk_size >> 9;
int pd_idx, dd_idx;
int chunk_offset = sector_div(stripe, sectors_per_chunk);
raid5_compute_sector(stripe * (disks - conf->max_degraded)
*sectors_per_chunk + chunk_offset,
disks, disks - conf->max_degraded,
&dd_idx, &pd_idx, conf);
return pd_idx;
}
static void
handle_failed_stripe(raid5_conf_t *conf, struct stripe_head *sh,
struct stripe_head_state *s, int disks,
struct bio **return_bi)
{
int i;
for (i = disks; i--; ) {
struct bio *bi;
int bitmap_end = 0;
if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
mdk_rdev_t *rdev;
rcu_read_lock();
rdev = rcu_dereference(conf->disks[i].rdev);
if (rdev && test_bit(In_sync, &rdev->flags))
/* multiple read failures in one stripe */
md_error(conf->mddev, rdev);
rcu_read_unlock();
}
spin_lock_irq(&conf->device_lock);
/* fail all writes first */
bi = sh->dev[i].towrite;
sh->dev[i].towrite = NULL;
if (bi) {
s->to_write--;
bitmap_end = 1;
}
if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
wake_up(&conf->wait_for_overlap);
while (bi && bi->bi_sector <
sh->dev[i].sector + STRIPE_SECTORS) {
struct bio *nextbi = r5_next_bio(bi, sh->dev[i].sector);
clear_bit(BIO_UPTODATE, &bi->bi_flags);
if (--bi->bi_phys_segments == 0) {
md_write_end(conf->mddev);
bi->bi_next = *return_bi;
*return_bi = bi;
}
bi = nextbi;
}
/* and fail all 'written' */
bi = sh->dev[i].written;
sh->dev[i].written = NULL;
if (bi) bitmap_end = 1;
while (bi && bi->bi_sector <
sh->dev[i].sector + STRIPE_SECTORS) {
struct bio *bi2 = r5_next_bio(bi, sh->dev[i].sector);
clear_bit(BIO_UPTODATE, &bi->bi_flags);
if (--bi->bi_phys_segments == 0) {
md_write_end(conf->mddev);
bi->bi_next = *return_bi;
*return_bi = bi;
}
bi = bi2;
}
/* fail any reads if this device is non-operational and
* the data has not reached the cache yet.
*/
if (!test_bit(R5_Wantfill, &sh->dev[i].flags) &&
(!test_bit(R5_Insync, &sh->dev[i].flags) ||
test_bit(R5_ReadError, &sh->dev[i].flags))) {
bi = sh->dev[i].toread;
sh->dev[i].toread = NULL;
if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
wake_up(&conf->wait_for_overlap);
if (bi) s->to_read--;
while (bi && bi->bi_sector <
sh->dev[i].sector + STRIPE_SECTORS) {
struct bio *nextbi =
r5_next_bio(bi, sh->dev[i].sector);
clear_bit(BIO_UPTODATE, &bi->bi_flags);
if (--bi->bi_phys_segments == 0) {
bi->bi_next = *return_bi;
*return_bi = bi;
}
bi = nextbi;
}
}
spin_unlock_irq(&conf->device_lock);
if (bitmap_end)
bitmap_endwrite(conf->mddev->bitmap, sh->sector,
STRIPE_SECTORS, 0, 0);
}
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
if (test_and_clear_bit(STRIPE_FULL_WRITE, &sh->state))
if (atomic_dec_and_test(&conf->pending_full_writes))
md_wakeup_thread(conf->mddev->thread);
}
/* fetch_block5 - checks the given member device to see if its data needs
* to be read or computed to satisfy a request.
*
* Returns 1 when no more member devices need to be checked, otherwise returns
* 0 to tell the loop in handle_stripe_fill5 to continue
*/
static int fetch_block5(struct stripe_head *sh, struct stripe_head_state *s,
int disk_idx, int disks)
{
struct r5dev *dev = &sh->dev[disk_idx];
struct r5dev *failed_dev = &sh->dev[s->failed_num];
/* is the data in this block needed, and can we get it? */
if (!test_bit(R5_LOCKED, &dev->flags) &&
!test_bit(R5_UPTODATE, &dev->flags) &&
(dev->toread ||
(dev->towrite && !test_bit(R5_OVERWRITE, &dev->flags)) ||
s->syncing || s->expanding ||
(s->failed &&
(failed_dev->toread ||
(failed_dev->towrite &&
!test_bit(R5_OVERWRITE, &failed_dev->flags)))))) {
/* We would like to get this block, possibly by computing it,
* otherwise read it if the backing disk is insync
*/
if ((s->uptodate == disks - 1) &&
(s->failed && disk_idx == s->failed_num)) {
set_bit(STRIPE_COMPUTE_RUN, &sh->state);
set_bit(STRIPE_OP_COMPUTE_BLK, &s->ops_request);
set_bit(R5_Wantcompute, &dev->flags);
sh->ops.target = disk_idx;
s->req_compute = 1;
/* Careful: from this point on 'uptodate' is in the eye
* of raid5_run_ops which services 'compute' operations
* before writes. R5_Wantcompute flags a block that will
* be R5_UPTODATE by the time it is needed for a
* subsequent operation.
*/
s->uptodate++;
return 1; /* uptodate + compute == disks */
} else if (test_bit(R5_Insync, &dev->flags)) {
set_bit(R5_LOCKED, &dev->flags);
set_bit(R5_Wantread, &dev->flags);
s->locked++;
pr_debug("Reading block %d (sync=%d)\n", disk_idx,
s->syncing);
}
}
return 0;
}
/**
* handle_stripe_fill5 - read or compute data to satisfy pending requests.
*/
static void handle_stripe_fill5(struct stripe_head *sh,
struct stripe_head_state *s, int disks)
{
int i;
/* look for blocks to read/compute, skip this if a compute
* is already in flight, or if the stripe contents are in the
* midst of changing due to a write
*/
if (!test_bit(STRIPE_COMPUTE_RUN, &sh->state) && !sh->check_state &&
!sh->reconstruct_state)
for (i = disks; i--; )
if (fetch_block5(sh, s, i, disks))
break;
set_bit(STRIPE_HANDLE, &sh->state);
}
static void handle_stripe_fill6(struct stripe_head *sh,
struct stripe_head_state *s, struct r6_state *r6s,
int disks)
{
int i;
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if (!test_bit(R5_LOCKED, &dev->flags) &&
!test_bit(R5_UPTODATE, &dev->flags) &&
(dev->toread || (dev->towrite &&
!test_bit(R5_OVERWRITE, &dev->flags)) ||
s->syncing || s->expanding ||
(s->failed >= 1 &&
(sh->dev[r6s->failed_num[0]].toread ||
s->to_write)) ||
(s->failed >= 2 &&
(sh->dev[r6s->failed_num[1]].toread ||
s->to_write)))) {
/* we would like to get this block, possibly
* by computing it, but we might not be able to
*/
if ((s->uptodate == disks - 1) &&
(s->failed && (i == r6s->failed_num[0] ||
i == r6s->failed_num[1]))) {
pr_debug("Computing stripe %llu block %d\n",
(unsigned long long)sh->sector, i);
compute_block_1(sh, i, 0);
s->uptodate++;
} else if ( s->uptodate == disks-2 && s->failed >= 2 ) {
/* Computing 2-failure is *very* expensive; only
* do it if failed >= 2
*/
int other;
for (other = disks; other--; ) {
if (other == i)
continue;
if (!test_bit(R5_UPTODATE,
&sh->dev[other].flags))
break;
}
BUG_ON(other < 0);
pr_debug("Computing stripe %llu blocks %d,%d\n",
(unsigned long long)sh->sector,
i, other);
compute_block_2(sh, i, other);
s->uptodate += 2;
} else if (test_bit(R5_Insync, &dev->flags)) {
set_bit(R5_LOCKED, &dev->flags);
set_bit(R5_Wantread, &dev->flags);
s->locked++;
pr_debug("Reading block %d (sync=%d)\n",
i, s->syncing);
}
}
}
set_bit(STRIPE_HANDLE, &sh->state);
}
/* handle_stripe_clean_event
* any written block on an uptodate or failed drive can be returned.
* Note that if we 'wrote' to a failed drive, it will be UPTODATE, but
* never LOCKED, so we don't need to test 'failed' directly.
*/
static void handle_stripe_clean_event(raid5_conf_t *conf,
struct stripe_head *sh, int disks, struct bio **return_bi)
{
int i;
struct r5dev *dev;
for (i = disks; i--; )
if (sh->dev[i].written) {
dev = &sh->dev[i];
if (!test_bit(R5_LOCKED, &dev->flags) &&
test_bit(R5_UPTODATE, &dev->flags)) {
/* We can return any write requests */
struct bio *wbi, *wbi2;
int bitmap_end = 0;
pr_debug("Return write for disc %d\n", i);
spin_lock_irq(&conf->device_lock);
wbi = dev->written;
dev->written = NULL;
while (wbi && wbi->bi_sector <
dev->sector + STRIPE_SECTORS) {
wbi2 = r5_next_bio(wbi, dev->sector);
if (--wbi->bi_phys_segments == 0) {
md_write_end(conf->mddev);
wbi->bi_next = *return_bi;
*return_bi = wbi;
}
wbi = wbi2;
}
if (dev->towrite == NULL)
bitmap_end = 1;
spin_unlock_irq(&conf->device_lock);
if (bitmap_end)
bitmap_endwrite(conf->mddev->bitmap,
sh->sector,
STRIPE_SECTORS,
!test_bit(STRIPE_DEGRADED, &sh->state),
0);
}
}
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
if (test_and_clear_bit(STRIPE_FULL_WRITE, &sh->state))
if (atomic_dec_and_test(&conf->pending_full_writes))
md_wakeup_thread(conf->mddev->thread);
}
static void handle_stripe_dirtying5(raid5_conf_t *conf,
struct stripe_head *sh, struct stripe_head_state *s, int disks)
{
int rmw = 0, rcw = 0, i;
for (i = disks; i--; ) {
/* would I have to read this buffer for read_modify_write */
struct r5dev *dev = &sh->dev[i];
if ((dev->towrite || i == sh->pd_idx) &&
!test_bit(R5_LOCKED, &dev->flags) &&
!(test_bit(R5_UPTODATE, &dev->flags) ||
test_bit(R5_Wantcompute, &dev->flags))) {
if (test_bit(R5_Insync, &dev->flags))
rmw++;
else
rmw += 2*disks; /* cannot read it */
}
/* Would I have to read this buffer for reconstruct_write */
if (!test_bit(R5_OVERWRITE, &dev->flags) && i != sh->pd_idx &&
!test_bit(R5_LOCKED, &dev->flags) &&
!(test_bit(R5_UPTODATE, &dev->flags) ||
test_bit(R5_Wantcompute, &dev->flags))) {
if (test_bit(R5_Insync, &dev->flags)) rcw++;
else
rcw += 2*disks;
}
}
pr_debug("for sector %llu, rmw=%d rcw=%d\n",
(unsigned long long)sh->sector, rmw, rcw);
set_bit(STRIPE_HANDLE, &sh->state);
if (rmw < rcw && rmw > 0)
/* prefer read-modify-write, but need to get some data */
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if ((dev->towrite || i == sh->pd_idx) &&
!test_bit(R5_LOCKED, &dev->flags) &&
!(test_bit(R5_UPTODATE, &dev->flags) ||
test_bit(R5_Wantcompute, &dev->flags)) &&
test_bit(R5_Insync, &dev->flags)) {
if (
test_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
pr_debug("Read_old block "
"%d for r-m-w\n", i);
set_bit(R5_LOCKED, &dev->flags);
set_bit(R5_Wantread, &dev->flags);
s->locked++;
} else {
set_bit(STRIPE_DELAYED, &sh->state);
set_bit(STRIPE_HANDLE, &sh->state);
}
}
}
if (rcw <= rmw && rcw > 0)
/* want reconstruct write, but need to get some data */
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if (!test_bit(R5_OVERWRITE, &dev->flags) &&
i != sh->pd_idx &&
!test_bit(R5_LOCKED, &dev->flags) &&
!(test_bit(R5_UPTODATE, &dev->flags) ||
test_bit(R5_Wantcompute, &dev->flags)) &&
test_bit(R5_Insync, &dev->flags)) {
if (
test_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
pr_debug("Read_old block "
"%d for Reconstruct\n", i);
set_bit(R5_LOCKED, &dev->flags);
set_bit(R5_Wantread, &dev->flags);
s->locked++;
} else {
set_bit(STRIPE_DELAYED, &sh->state);
set_bit(STRIPE_HANDLE, &sh->state);
}
}
}
/* now if nothing is locked, and if we have enough data,
* we can start a write request
*/
/* since handle_stripe can be called at any time we need to handle the
* case where a compute block operation has been submitted and then a
* subsequent call wants to start a write request. raid5_run_ops only
* handles the case where compute block and postxor are requested
* simultaneously. If this is not the case then new writes need to be
* held off until the compute completes.
*/
if ((s->req_compute || !test_bit(STRIPE_COMPUTE_RUN, &sh->state)) &&
(s->locked == 0 && (rcw == 0 || rmw == 0) &&
!test_bit(STRIPE_BIT_DELAY, &sh->state)))
schedule_reconstruction5(sh, s, rcw == 0, 0);
}
static void handle_stripe_dirtying6(raid5_conf_t *conf,
struct stripe_head *sh, struct stripe_head_state *s,
struct r6_state *r6s, int disks)
{
int rcw = 0, must_compute = 0, pd_idx = sh->pd_idx, i;
int qd_idx = r6s->qd_idx;
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
/* Would I have to read this buffer for reconstruct_write */
if (!test_bit(R5_OVERWRITE, &dev->flags)
&& i != pd_idx && i != qd_idx
&& (!test_bit(R5_LOCKED, &dev->flags)
) &&
!test_bit(R5_UPTODATE, &dev->flags)) {
if (test_bit(R5_Insync, &dev->flags)) rcw++;
else {
pr_debug("raid6: must_compute: "
"disk %d flags=%#lx\n", i, dev->flags);
must_compute++;
}
}
}
pr_debug("for sector %llu, rcw=%d, must_compute=%d\n",
(unsigned long long)sh->sector, rcw, must_compute);
set_bit(STRIPE_HANDLE, &sh->state);
if (rcw > 0)
/* want reconstruct write, but need to get some data */
for (i = disks; i--; ) {
struct r5dev *dev = &sh->dev[i];
if (!test_bit(R5_OVERWRITE, &dev->flags)
&& !(s->failed == 0 && (i == pd_idx || i == qd_idx))
&& !test_bit(R5_LOCKED, &dev->flags) &&
!test_bit(R5_UPTODATE, &dev->flags) &&
test_bit(R5_Insync, &dev->flags)) {
if (
test_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
pr_debug("Read_old stripe %llu "
"block %d for Reconstruct\n",
(unsigned long long)sh->sector, i);
set_bit(R5_LOCKED, &dev->flags);
set_bit(R5_Wantread, &dev->flags);
s->locked++;
} else {
pr_debug("Request delayed stripe %llu "
"block %d for Reconstruct\n",
(unsigned long long)sh->sector, i);
set_bit(STRIPE_DELAYED, &sh->state);
set_bit(STRIPE_HANDLE, &sh->state);
}
}
}
/* now if nothing is locked, and if we have enough data, we can start a
* write request
*/
if (s->locked == 0 && rcw == 0 &&
!test_bit(STRIPE_BIT_DELAY, &sh->state)) {
if (must_compute > 0) {
/* We have failed blocks and need to compute them */
switch (s->failed) {
case 0:
BUG();
case 1:
compute_block_1(sh, r6s->failed_num[0], 0);
break;
case 2:
compute_block_2(sh, r6s->failed_num[0],
r6s->failed_num[1]);
break;
default: /* This request should have been failed? */
BUG();
}
}
pr_debug("Computing parity for stripe %llu\n",
(unsigned long long)sh->sector);
compute_parity6(sh, RECONSTRUCT_WRITE);
/* now every locked buffer is ready to be written */
for (i = disks; i--; )
if (test_bit(R5_LOCKED, &sh->dev[i].flags)) {
pr_debug("Writing stripe %llu block %d\n",
(unsigned long long)sh->sector, i);
s->locked++;
set_bit(R5_Wantwrite, &sh->dev[i].flags);
}
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
if (s->locked == disks)
if (!test_and_set_bit(STRIPE_FULL_WRITE, &sh->state))
atomic_inc(&conf->pending_full_writes);
/* after a RECONSTRUCT_WRITE, the stripe MUST be in-sync */
set_bit(STRIPE_INSYNC, &sh->state);
if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
atomic_dec(&conf->preread_active_stripes);
if (atomic_read(&conf->preread_active_stripes) <
IO_THRESHOLD)
md_wakeup_thread(conf->mddev->thread);
}
}
}
static void handle_parity_checks5(raid5_conf_t *conf, struct stripe_head *sh,
struct stripe_head_state *s, int disks)
{
struct r5dev *dev = NULL;
set_bit(STRIPE_HANDLE, &sh->state);
switch (sh->check_state) {
case check_state_idle:
/* start a new check operation if there are no failures */
if (s->failed == 0) {
BUG_ON(s->uptodate != disks);
sh->check_state = check_state_run;
set_bit(STRIPE_OP_CHECK, &s->ops_request);
clear_bit(R5_UPTODATE, &sh->dev[sh->pd_idx].flags);
s->uptodate--;
break;
}
dev = &sh->dev[s->failed_num];
/* fall through */
case check_state_compute_result:
sh->check_state = check_state_idle;
if (!dev)
dev = &sh->dev[sh->pd_idx];
/* check that a write has not made the stripe insync */
if (test_bit(STRIPE_INSYNC, &sh->state))
break;
/* either failed parity check, or recovery is happening */
BUG_ON(!test_bit(R5_UPTODATE, &dev->flags));
BUG_ON(s->uptodate != disks);
set_bit(R5_LOCKED, &dev->flags);
s->locked++;
set_bit(R5_Wantwrite, &dev->flags);
clear_bit(STRIPE_DEGRADED, &sh->state);
set_bit(STRIPE_INSYNC, &sh->state);
break;
case check_state_run:
break; /* we will be called again upon completion */
case check_state_check_result:
sh->check_state = check_state_idle;
/* if a failure occurred during the check operation, leave
* STRIPE_INSYNC not set and let the stripe be handled again
*/
if (s->failed)
break;
/* handle a successful check operation, if parity is correct
* we are done. Otherwise update the mismatch count and repair
* parity if !MD_RECOVERY_CHECK
*/
if (sh->ops.zero_sum_result == 0)
/* parity is correct (on disc,
* not in buffer any more)
*/
set_bit(STRIPE_INSYNC, &sh->state);
else {
conf->mddev->resync_mismatches += STRIPE_SECTORS;
if (test_bit(MD_RECOVERY_CHECK, &conf->mddev->recovery))
/* don't try to repair!! */
set_bit(STRIPE_INSYNC, &sh->state);
else {
sh->check_state = check_state_compute_run;
set_bit(STRIPE_COMPUTE_RUN, &sh->state);
set_bit(STRIPE_OP_COMPUTE_BLK, &s->ops_request);
set_bit(R5_Wantcompute,
&sh->dev[sh->pd_idx].flags);
sh->ops.target = sh->pd_idx;
s->uptodate++;
}
}
break;
case check_state_compute_run:
break;
default:
printk(KERN_ERR "%s: unknown check_state: %d sector: %llu\n",
__func__, sh->check_state,
(unsigned long long) sh->sector);
BUG();
}
}
static void handle_parity_checks6(raid5_conf_t *conf, struct stripe_head *sh,
struct stripe_head_state *s,
struct r6_state *r6s, struct page *tmp_page,
int disks)
{
int update_p = 0, update_q = 0;
struct r5dev *dev;
int pd_idx = sh->pd_idx;
int qd_idx = r6s->qd_idx;
set_bit(STRIPE_HANDLE, &sh->state);
BUG_ON(s->failed > 2);
BUG_ON(s->uptodate < disks);
/* Want to check and possibly repair P and Q.
* However there could be one 'failed' device, in which
* case we can only check one of them, possibly using the
* other to generate missing data
*/
/* If !tmp_page, we cannot do the calculations,
* but as we have set STRIPE_HANDLE, we will soon be called
* by stripe_handle with a tmp_page - just wait until then.
*/
if (tmp_page) {
if (s->failed == r6s->q_failed) {
/* The only possible failed device holds 'Q', so it
* makes sense to check P (If anything else were failed,
* we would have used P to recreate it).
*/
compute_block_1(sh, pd_idx, 1);
if (!page_is_zero(sh->dev[pd_idx].page)) {
compute_block_1(sh, pd_idx, 0);
update_p = 1;
}
}
if (!r6s->q_failed && s->failed < 2) {
/* q is not failed, and we didn't use it to generate
* anything, so it makes sense to check it
*/
memcpy(page_address(tmp_page),
page_address(sh->dev[qd_idx].page),
STRIPE_SIZE);
compute_parity6(sh, UPDATE_PARITY);
if (memcmp(page_address(tmp_page),
page_address(sh->dev[qd_idx].page),
STRIPE_SIZE) != 0) {
clear_bit(STRIPE_INSYNC, &sh->state);
update_q = 1;
}
}
if (update_p || update_q) {
conf->mddev->resync_mismatches += STRIPE_SECTORS;
if (test_bit(MD_RECOVERY_CHECK, &conf->mddev->recovery))
/* don't try to repair!! */
update_p = update_q = 0;
}
/* now write out any block on a failed drive,
* or P or Q if they need it
*/
if (s->failed == 2) {
dev = &sh->dev[r6s->failed_num[1]];
s->locked++;
set_bit(R5_LOCKED, &dev->flags);
set_bit(R5_Wantwrite, &dev->flags);
}
if (s->failed >= 1) {
dev = &sh->dev[r6s->failed_num[0]];
s->locked++;
set_bit(R5_LOCKED, &dev->flags);
set_bit(R5_Wantwrite, &dev->flags);
}
if (update_p) {
dev = &sh->dev[pd_idx];
s->locked++;
set_bit(R5_LOCKED, &dev->flags);
set_bit(R5_Wantwrite, &dev->flags);
}
if (update_q) {
dev = &sh->dev[qd_idx];
s->locked++;
set_bit(R5_LOCKED, &dev->flags);
set_bit(R5_Wantwrite, &dev->flags);
}
clear_bit(STRIPE_DEGRADED, &sh->state);
set_bit(STRIPE_INSYNC, &sh->state);
}
}
static void handle_stripe_expansion(raid5_conf_t *conf, struct stripe_head *sh,
struct r6_state *r6s)
{
int i;
/* We have read all the blocks in this stripe and now we need to
* copy some of them into a target stripe for expand.
*/
struct dma_async_tx_descriptor *tx = NULL;
clear_bit(STRIPE_EXPAND_SOURCE, &sh->state);
for (i = 0; i < sh->disks; i++)
if (i != sh->pd_idx && (!r6s || i != r6s->qd_idx)) {
int dd_idx, pd_idx, j;
struct stripe_head *sh2;
sector_t bn = compute_blocknr(sh, i);
sector_t s = raid5_compute_sector(bn, conf->raid_disks,
conf->raid_disks -
conf->max_degraded, &dd_idx,
&pd_idx, conf);
sh2 = get_active_stripe(conf, s, conf->raid_disks,
pd_idx, 1);
if (sh2 == NULL)
/* so far only the early blocks of this stripe
* have been requested. When later blocks
* get requested, we will try again
*/
continue;
if (!test_bit(STRIPE_EXPANDING, &sh2->state) ||
test_bit(R5_Expanded, &sh2->dev[dd_idx].flags)) {
/* must have already done this block */
release_stripe(sh2);
continue;
}
/* place all the copies on one channel */
tx = async_memcpy(sh2->dev[dd_idx].page,
sh->dev[i].page, 0, 0, STRIPE_SIZE,
ASYNC_TX_DEP_ACK, tx, NULL, NULL);
set_bit(R5_Expanded, &sh2->dev[dd_idx].flags);
set_bit(R5_UPTODATE, &sh2->dev[dd_idx].flags);
for (j = 0; j < conf->raid_disks; j++)
if (j != sh2->pd_idx &&
(!r6s || j != raid6_next_disk(sh2->pd_idx,
sh2->disks)) &&
!test_bit(R5_Expanded, &sh2->dev[j].flags))
break;
if (j == conf->raid_disks) {
set_bit(STRIPE_EXPAND_READY, &sh2->state);
set_bit(STRIPE_HANDLE, &sh2->state);
}
release_stripe(sh2);
}
/* done submitting copies, wait for them to complete */
if (tx) {
async_tx_ack(tx);
dma_wait_for_async_tx(tx);
}
}
/*
* handle_stripe - do things to a stripe.
*
* We lock the stripe and then examine the state of various bits
* to see what needs to be done.
* Possible results:
* return some read request which now have data
* return some write requests which are safely on disc
* schedule a read on some buffers
* schedule a write of some buffers
* return confirmation of parity correctness
*
* buffers are taken off read_list or write_list, and bh_cache buffers
* get BH_Lock set before the stripe lock is released.
*
*/
static bool handle_stripe5(struct stripe_head *sh)
{
raid5_conf_t *conf = sh->raid_conf;
int disks = sh->disks, i;
struct bio *return_bi = NULL;
struct stripe_head_state s;
struct r5dev *dev;
mdk_rdev_t *blocked_rdev = NULL;
2008-06-06 05:45:52 +00:00
int prexor;
memset(&s, 0, sizeof(s));
pr_debug("handling stripe %llu, state=%#lx cnt=%d, pd_idx=%d check:%d "
"reconstruct:%d\n", (unsigned long long)sh->sector, sh->state,
atomic_read(&sh->count), sh->pd_idx, sh->check_state,
sh->reconstruct_state);
spin_lock(&sh->lock);
clear_bit(STRIPE_HANDLE, &sh->state);
clear_bit(STRIPE_DELAYED, &sh->state);
s.syncing = test_bit(STRIPE_SYNCING, &sh->state);
s.expanding = test_bit(STRIPE_EXPAND_SOURCE, &sh->state);
s.expanded = test_bit(STRIPE_EXPAND_READY, &sh->state);
/* Now to look around and see what can be done */
rcu_read_lock();
for (i=disks; i--; ) {
mdk_rdev_t *rdev;
struct r5dev *dev = &sh->dev[i];
clear_bit(R5_Insync, &dev->flags);
pr_debug("check %d: state 0x%lx toread %p read %p write %p "
"written %p\n", i, dev->flags, dev->toread, dev->read,
dev->towrite, dev->written);
/* maybe we can request a biofill operation
*
* new wantfill requests are only permitted while
* ops_complete_biofill is guaranteed to be inactive
*/
if (test_bit(R5_UPTODATE, &dev->flags) && dev->toread &&
!test_bit(STRIPE_BIOFILL_RUN, &sh->state))
set_bit(R5_Wantfill, &dev->flags);
/* now count some things */
if (test_bit(R5_LOCKED, &dev->flags)) s.locked++;
if (test_bit(R5_UPTODATE, &dev->flags)) s.uptodate++;
if (test_bit(R5_Wantcompute, &dev->flags)) s.compute++;
if (test_bit(R5_Wantfill, &dev->flags))
s.to_fill++;
else if (dev->toread)
s.to_read++;
if (dev->towrite) {
s.to_write++;
if (!test_bit(R5_OVERWRITE, &dev->flags))
s.non_overwrite++;
}
if (dev->written)
s.written++;
rdev = rcu_dereference(conf->disks[i].rdev);
if (blocked_rdev == NULL &&
rdev && unlikely(test_bit(Blocked, &rdev->flags))) {
blocked_rdev = rdev;
atomic_inc(&rdev->nr_pending);
}
if (!rdev || !test_bit(In_sync, &rdev->flags)) {
/* The ReadError flag will just be confusing now */
clear_bit(R5_ReadError, &dev->flags);
clear_bit(R5_ReWrite, &dev->flags);
}
if (!rdev || !test_bit(In_sync, &rdev->flags)
|| test_bit(R5_ReadError, &dev->flags)) {
s.failed++;
s.failed_num = i;
} else
set_bit(R5_Insync, &dev->flags);
}
rcu_read_unlock();
if (unlikely(blocked_rdev)) {
if (s.syncing || s.expanding || s.expanded ||
s.to_write || s.written) {
set_bit(STRIPE_HANDLE, &sh->state);
goto unlock;
}
/* There is nothing for the blocked_rdev to block */
rdev_dec_pending(blocked_rdev, conf->mddev);
blocked_rdev = NULL;
}
if (s.to_fill && !test_bit(STRIPE_BIOFILL_RUN, &sh->state)) {
set_bit(STRIPE_OP_BIOFILL, &s.ops_request);
set_bit(STRIPE_BIOFILL_RUN, &sh->state);
}
pr_debug("locked=%d uptodate=%d to_read=%d"
" to_write=%d failed=%d failed_num=%d\n",
s.locked, s.uptodate, s.to_read, s.to_write,
s.failed, s.failed_num);
/* check if the array has lost two devices and, if so, some requests might
* need to be failed
*/
if (s.failed > 1 && s.to_read+s.to_write+s.written)
handle_failed_stripe(conf, sh, &s, disks, &return_bi);
if (s.failed > 1 && s.syncing) {
md_done_sync(conf->mddev, STRIPE_SECTORS,0);
clear_bit(STRIPE_SYNCING, &sh->state);
s.syncing = 0;
}
/* might be able to return some write requests if the parity block
* is safe, or on a failed drive
*/
dev = &sh->dev[sh->pd_idx];
if ( s.written &&
((test_bit(R5_Insync, &dev->flags) &&
!test_bit(R5_LOCKED, &dev->flags) &&
test_bit(R5_UPTODATE, &dev->flags)) ||
(s.failed == 1 && s.failed_num == sh->pd_idx)))
handle_stripe_clean_event(conf, sh, disks, &return_bi);
/* Now we might consider reading some blocks, either to check/generate
* parity, or to satisfy requests
* or to load a block that is being partially written.
*/
if (s.to_read || s.non_overwrite ||
(s.syncing && (s.uptodate + s.compute < disks)) || s.expanding)
handle_stripe_fill5(sh, &s, disks);
/* Now we check to see if any write operations have recently
* completed
*/
2008-06-06 05:45:52 +00:00
prexor = 0;
if (sh->reconstruct_state == reconstruct_state_prexor_drain_result)
2008-06-06 05:45:52 +00:00
prexor = 1;
if (sh->reconstruct_state == reconstruct_state_drain_result ||
sh->reconstruct_state == reconstruct_state_prexor_drain_result) {
sh->reconstruct_state = reconstruct_state_idle;
/* All the 'written' buffers and the parity block are ready to
* be written back to disk
*/
BUG_ON(!test_bit(R5_UPTODATE, &sh->dev[sh->pd_idx].flags));
for (i = disks; i--; ) {
dev = &sh->dev[i];
if (test_bit(R5_LOCKED, &dev->flags) &&
(i == sh->pd_idx || dev->written)) {
pr_debug("Writing block %d\n", i);
set_bit(R5_Wantwrite, &dev->flags);
2008-06-06 05:45:52 +00:00
if (prexor)
continue;
if (!test_bit(R5_Insync, &dev->flags) ||
(i == sh->pd_idx && s.failed == 0))
set_bit(STRIPE_INSYNC, &sh->state);
}
}
if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
atomic_dec(&conf->preread_active_stripes);
if (atomic_read(&conf->preread_active_stripes) <
IO_THRESHOLD)
md_wakeup_thread(conf->mddev->thread);
}
}
/* Now to consider new write requests and what else, if anything
* should be read. We do not handle new writes when:
* 1/ A 'write' operation (copy+xor) is already in flight.
* 2/ A 'check' operation is in flight, as it may clobber the parity
* block.
*/
if (s.to_write && !sh->reconstruct_state && !sh->check_state)
handle_stripe_dirtying5(conf, sh, &s, disks);
/* maybe we need to check and possibly fix the parity for this stripe
* Any reads will already have been scheduled, so we just see if enough
* data is available. The parity check is held off while parity
* dependent operations are in flight.
*/
if (sh->check_state ||
(s.syncing && s.locked == 0 &&
!test_bit(STRIPE_COMPUTE_RUN, &sh->state) &&
!test_bit(STRIPE_INSYNC, &sh->state)))
handle_parity_checks5(conf, sh, &s, disks);
if (s.syncing && s.locked == 0 && test_bit(STRIPE_INSYNC, &sh->state)) {
md_done_sync(conf->mddev, STRIPE_SECTORS,1);
clear_bit(STRIPE_SYNCING, &sh->state);
}
/* If the failed drive is just a ReadError, then we might need to progress
* the repair/check process
*/
if (s.failed == 1 && !conf->mddev->ro &&
test_bit(R5_ReadError, &sh->dev[s.failed_num].flags)
&& !test_bit(R5_LOCKED, &sh->dev[s.failed_num].flags)
&& test_bit(R5_UPTODATE, &sh->dev[s.failed_num].flags)
) {
dev = &sh->dev[s.failed_num];
if (!test_bit(R5_ReWrite, &dev->flags)) {
set_bit(R5_Wantwrite, &dev->flags);
set_bit(R5_ReWrite, &dev->flags);
set_bit(R5_LOCKED, &dev->flags);
s.locked++;
} else {
/* let's read it back */
set_bit(R5_Wantread, &dev->flags);
set_bit(R5_LOCKED, &dev->flags);
s.locked++;
}
}
/* Finish reconstruct operations initiated by the expansion process */
if (sh->reconstruct_state == reconstruct_state_result) {
sh->reconstruct_state = reconstruct_state_idle;
clear_bit(STRIPE_EXPANDING, &sh->state);
for (i = conf->raid_disks; i--; ) {
set_bit(R5_Wantwrite, &sh->dev[i].flags);
set_bit(R5_LOCKED, &sh->dev[i].flags);
s.locked++;
}
}
if (s.expanded && test_bit(STRIPE_EXPANDING, &sh->state) &&
!sh->reconstruct_state) {
/* Need to write out all blocks after computing parity */
sh->disks = conf->raid_disks;
sh->pd_idx = stripe_to_pdidx(sh->sector, conf,
conf->raid_disks);
schedule_reconstruction5(sh, &s, 1, 1);
} else if (s.expanded && !sh->reconstruct_state && s.locked == 0) {
clear_bit(STRIPE_EXPAND_READY, &sh->state);
atomic_dec(&conf->reshape_stripes);
wake_up(&conf->wait_for_overlap);
md_done_sync(conf->mddev, STRIPE_SECTORS, 1);
}
if (s.expanding && s.locked == 0 &&
!test_bit(STRIPE_COMPUTE_RUN, &sh->state))
handle_stripe_expansion(conf, sh, NULL);
unlock:
spin_unlock(&sh->lock);
/* wait for this device to become unblocked */
if (unlikely(blocked_rdev))
md_wait_for_blocked_rdev(blocked_rdev, conf->mddev);
if (s.ops_request)
raid5_run_ops(sh, s.ops_request);
ops_run_io(sh, &s);
return_io(return_bi);
return blocked_rdev == NULL;
}
static bool handle_stripe6(struct stripe_head *sh, struct page *tmp_page)
{
raid6_conf_t *conf = sh->raid_conf;
int disks = sh->disks;
struct bio *return_bi = NULL;
int i, pd_idx = sh->pd_idx;
struct stripe_head_state s;
struct r6_state r6s;
struct r5dev *dev, *pdev, *qdev;
mdk_rdev_t *blocked_rdev = NULL;
r6s.qd_idx = raid6_next_disk(pd_idx, disks);
pr_debug("handling stripe %llu, state=%#lx cnt=%d, "
"pd_idx=%d, qd_idx=%d\n",
(unsigned long long)sh->sector, sh->state,
atomic_read(&sh->count), pd_idx, r6s.qd_idx);
memset(&s, 0, sizeof(s));
spin_lock(&sh->lock);
clear_bit(STRIPE_HANDLE, &sh->state);
clear_bit(STRIPE_DELAYED, &sh->state);
s.syncing = test_bit(STRIPE_SYNCING, &sh->state);
s.expanding = test_bit(STRIPE_EXPAND_SOURCE, &sh->state);
s.expanded = test_bit(STRIPE_EXPAND_READY, &sh->state);
/* Now to look around and see what can be done */
rcu_read_lock();
for (i=disks; i--; ) {
mdk_rdev_t *rdev;
dev = &sh->dev[i];
clear_bit(R5_Insync, &dev->flags);
pr_debug("check %d: state 0x%lx read %p write %p written %p\n",
i, dev->flags, dev->toread, dev->towrite, dev->written);
/* maybe we can reply to a read */
if (test_bit(R5_UPTODATE, &dev->flags) && dev->toread) {
struct bio *rbi, *rbi2;
pr_debug("Return read for disc %d\n", i);
spin_lock_irq(&conf->device_lock);
rbi = dev->toread;
dev->toread = NULL;
if (test_and_clear_bit(R5_Overlap, &dev->flags))
wake_up(&conf->wait_for_overlap);
spin_unlock_irq(&conf->device_lock);
while (rbi && rbi->bi_sector < dev->sector + STRIPE_SECTORS) {
copy_data(0, rbi, dev->page, dev->sector);
rbi2 = r5_next_bio(rbi, dev->sector);
spin_lock_irq(&conf->device_lock);
if (--rbi->bi_phys_segments == 0) {
rbi->bi_next = return_bi;
return_bi = rbi;
}
spin_unlock_irq(&conf->device_lock);
rbi = rbi2;
}
}
/* now count some things */
if (test_bit(R5_LOCKED, &dev->flags)) s.locked++;
if (test_bit(R5_UPTODATE, &dev->flags)) s.uptodate++;
if (dev->toread)
s.to_read++;
if (dev->towrite) {
s.to_write++;
if (!test_bit(R5_OVERWRITE, &dev->flags))
s.non_overwrite++;
}
if (dev->written)
s.written++;
rdev = rcu_dereference(conf->disks[i].rdev);
if (blocked_rdev == NULL &&
rdev && unlikely(test_bit(Blocked, &rdev->flags))) {
blocked_rdev = rdev;
atomic_inc(&rdev->nr_pending);
}
if (!rdev || !test_bit(In_sync, &rdev->flags)) {
/* The ReadError flag will just be confusing now */
clear_bit(R5_ReadError, &dev->flags);
clear_bit(R5_ReWrite, &dev->flags);
}
if (!rdev || !test_bit(In_sync, &rdev->flags)
|| test_bit(R5_ReadError, &dev->flags)) {
if (s.failed < 2)
r6s.failed_num[s.failed] = i;
s.failed++;
} else
set_bit(R5_Insync, &dev->flags);
}
rcu_read_unlock();
if (unlikely(blocked_rdev)) {
if (s.syncing || s.expanding || s.expanded ||
s.to_write || s.written) {
set_bit(STRIPE_HANDLE, &sh->state);
goto unlock;
}
/* There is nothing for the blocked_rdev to block */
rdev_dec_pending(blocked_rdev, conf->mddev);
blocked_rdev = NULL;
}
pr_debug("locked=%d uptodate=%d to_read=%d"
" to_write=%d failed=%d failed_num=%d,%d\n",
s.locked, s.uptodate, s.to_read, s.to_write, s.failed,
r6s.failed_num[0], r6s.failed_num[1]);
/* check if the array has lost >2 devices and, if so, some requests
* might need to be failed
*/
if (s.failed > 2 && s.to_read+s.to_write+s.written)
handle_failed_stripe(conf, sh, &s, disks, &return_bi);
if (s.failed > 2 && s.syncing) {
md_done_sync(conf->mddev, STRIPE_SECTORS,0);
clear_bit(STRIPE_SYNCING, &sh->state);
s.syncing = 0;
}
/*
* might be able to return some write requests if the parity blocks
* are safe, or on a failed drive
*/
pdev = &sh->dev[pd_idx];
r6s.p_failed = (s.failed >= 1 && r6s.failed_num[0] == pd_idx)
|| (s.failed >= 2 && r6s.failed_num[1] == pd_idx);
qdev = &sh->dev[r6s.qd_idx];
r6s.q_failed = (s.failed >= 1 && r6s.failed_num[0] == r6s.qd_idx)
|| (s.failed >= 2 && r6s.failed_num[1] == r6s.qd_idx);
if ( s.written &&
( r6s.p_failed || ((test_bit(R5_Insync, &pdev->flags)
&& !test_bit(R5_LOCKED, &pdev->flags)
&& test_bit(R5_UPTODATE, &pdev->flags)))) &&
( r6s.q_failed || ((test_bit(R5_Insync, &qdev->flags)
&& !test_bit(R5_LOCKED, &qdev->flags)
&& test_bit(R5_UPTODATE, &qdev->flags)))))
handle_stripe_clean_event(conf, sh, disks, &return_bi);
/* Now we might consider reading some blocks, either to check/generate
* parity, or to satisfy requests
* or to load a block that is being partially written.
*/
if (s.to_read || s.non_overwrite || (s.to_write && s.failed) ||
(s.syncing && (s.uptodate < disks)) || s.expanding)
handle_stripe_fill6(sh, &s, &r6s, disks);
/* now to consider writing and what else, if anything should be read */
if (s.to_write)
handle_stripe_dirtying6(conf, sh, &s, &r6s, disks);
/* maybe we need to check and possibly fix the parity for this stripe
* Any reads will already have been scheduled, so we just see if enough
* data is available
*/
if (s.syncing && s.locked == 0 && !test_bit(STRIPE_INSYNC, &sh->state))
handle_parity_checks6(conf, sh, &s, &r6s, tmp_page, disks);
if (s.syncing && s.locked == 0 && test_bit(STRIPE_INSYNC, &sh->state)) {
md_done_sync(conf->mddev, STRIPE_SECTORS,1);
clear_bit(STRIPE_SYNCING, &sh->state);
}
/* If the failed drives are just a ReadError, then we might need
* to progress the repair/check process
*/
if (s.failed <= 2 && !conf->mddev->ro)
for (i = 0; i < s.failed; i++) {
dev = &sh->dev[r6s.failed_num[i]];
if (test_bit(R5_ReadError, &dev->flags)
&& !test_bit(R5_LOCKED, &dev->flags)
&& test_bit(R5_UPTODATE, &dev->flags)
) {
if (!test_bit(R5_ReWrite, &dev->flags)) {
set_bit(R5_Wantwrite, &dev->flags);
set_bit(R5_ReWrite, &dev->flags);
set_bit(R5_LOCKED, &dev->flags);
} else {
/* let's read it back */
set_bit(R5_Wantread, &dev->flags);
set_bit(R5_LOCKED, &dev->flags);
}
}
}
if (s.expanded && test_bit(STRIPE_EXPANDING, &sh->state)) {
/* Need to write out all blocks after computing P&Q */
sh->disks = conf->raid_disks;
sh->pd_idx = stripe_to_pdidx(sh->sector, conf,
conf->raid_disks);
compute_parity6(sh, RECONSTRUCT_WRITE);
for (i = conf->raid_disks ; i-- ; ) {
set_bit(R5_LOCKED, &sh->dev[i].flags);
s.locked++;
set_bit(R5_Wantwrite, &sh->dev[i].flags);
}
clear_bit(STRIPE_EXPANDING, &sh->state);
} else if (s.expanded) {
clear_bit(STRIPE_EXPAND_READY, &sh->state);
atomic_dec(&conf->reshape_stripes);
wake_up(&conf->wait_for_overlap);
md_done_sync(conf->mddev, STRIPE_SECTORS, 1);
}
if (s.expanding && s.locked == 0 &&
!test_bit(STRIPE_COMPUTE_RUN, &sh->state))
handle_stripe_expansion(conf, sh, &r6s);
unlock:
spin_unlock(&sh->lock);
/* wait for this device to become unblocked */
if (unlikely(blocked_rdev))
md_wait_for_blocked_rdev(blocked_rdev, conf->mddev);
ops_run_io(sh, &s);
return_io(return_bi);
return blocked_rdev == NULL;
}
/* returns true if the stripe was handled */
static bool handle_stripe(struct stripe_head *sh, struct page *tmp_page)
{
if (sh->raid_conf->level == 6)
return handle_stripe6(sh, tmp_page);
else
return handle_stripe5(sh);
}
static void raid5_activate_delayed(raid5_conf_t *conf)
{
if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD) {
while (!list_empty(&conf->delayed_list)) {
struct list_head *l = conf->delayed_list.next;
struct stripe_head *sh;
sh = list_entry(l, struct stripe_head, lru);
list_del_init(l);
clear_bit(STRIPE_DELAYED, &sh->state);
if (!test_and_set_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
atomic_inc(&conf->preread_active_stripes);
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
list_add_tail(&sh->lru, &conf->hold_list);
}
} else
blk_plug_device(conf->mddev->queue);
}
static void activate_bit_delay(raid5_conf_t *conf)
{
/* device_lock is held */
struct list_head head;
list_add(&head, &conf->bitmap_list);
list_del_init(&conf->bitmap_list);
while (!list_empty(&head)) {
struct stripe_head *sh = list_entry(head.next, struct stripe_head, lru);
list_del_init(&sh->lru);
atomic_inc(&sh->count);
__release_stripe(conf, sh);
}
}
static void unplug_slaves(mddev_t *mddev)
{
raid5_conf_t *conf = mddev_to_conf(mddev);
int i;
rcu_read_lock();
for (i=0; i<mddev->raid_disks; i++) {
mdk_rdev_t *rdev = rcu_dereference(conf->disks[i].rdev);
if (rdev && !test_bit(Faulty, &rdev->flags) && atomic_read(&rdev->nr_pending)) {
struct request_queue *r_queue = bdev_get_queue(rdev->bdev);
atomic_inc(&rdev->nr_pending);
rcu_read_unlock();
blk_unplug(r_queue);
rdev_dec_pending(rdev, mddev);
rcu_read_lock();
}
}
rcu_read_unlock();
}
static void raid5_unplug_device(struct request_queue *q)
{
mddev_t *mddev = q->queuedata;
raid5_conf_t *conf = mddev_to_conf(mddev);
unsigned long flags;
spin_lock_irqsave(&conf->device_lock, flags);
if (blk_remove_plug(q)) {
conf->seq_flush++;
raid5_activate_delayed(conf);
}
md_wakeup_thread(mddev->thread);
spin_unlock_irqrestore(&conf->device_lock, flags);
unplug_slaves(mddev);
}
static int raid5_congested(void *data, int bits)
{
mddev_t *mddev = data;
raid5_conf_t *conf = mddev_to_conf(mddev);
/* No difference between reads and writes. Just check
* how busy the stripe_cache is
*/
if (conf->inactive_blocked)
return 1;
if (conf->quiesce)
return 1;
if (list_empty_careful(&conf->inactive_list))
return 1;
return 0;
}
/* We want read requests to align with chunks where possible,
* but write requests don't need to.
*/
static int raid5_mergeable_bvec(struct request_queue *q,
struct bvec_merge_data *bvm,
struct bio_vec *biovec)
{
mddev_t *mddev = q->queuedata;
sector_t sector = bvm->bi_sector + get_start_sect(bvm->bi_bdev);
int max;
unsigned int chunk_sectors = mddev->chunk_size >> 9;
unsigned int bio_sectors = bvm->bi_size >> 9;
if ((bvm->bi_rw & 1) == WRITE)
return biovec->bv_len; /* always allow writes to be mergeable */
max = (chunk_sectors - ((sector & (chunk_sectors - 1)) + bio_sectors)) << 9;
if (max < 0) max = 0;
if (max <= biovec->bv_len && bio_sectors == 0)
return biovec->bv_len;
else
return max;
}
static int in_chunk_boundary(mddev_t *mddev, struct bio *bio)
{
sector_t sector = bio->bi_sector + get_start_sect(bio->bi_bdev);
unsigned int chunk_sectors = mddev->chunk_size >> 9;
unsigned int bio_sectors = bio->bi_size >> 9;
return chunk_sectors >=
((sector & (chunk_sectors - 1)) + bio_sectors);
}
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
/*
* add bio to the retry LIFO ( in O(1) ... we are in interrupt )
* later sampled by raid5d.
*/
static void add_bio_to_retry(struct bio *bi,raid5_conf_t *conf)
{
unsigned long flags;
spin_lock_irqsave(&conf->device_lock, flags);
bi->bi_next = conf->retry_read_aligned_list;
conf->retry_read_aligned_list = bi;
spin_unlock_irqrestore(&conf->device_lock, flags);
md_wakeup_thread(conf->mddev->thread);
}
static struct bio *remove_bio_from_retry(raid5_conf_t *conf)
{
struct bio *bi;
bi = conf->retry_read_aligned;
if (bi) {
conf->retry_read_aligned = NULL;
return bi;
}
bi = conf->retry_read_aligned_list;
if(bi) {
conf->retry_read_aligned_list = bi->bi_next;
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
bi->bi_next = NULL;
bi->bi_phys_segments = 1; /* biased count of active stripes */
bi->bi_hw_segments = 0; /* count of processed stripes */
}
return bi;
}
/*
* The "raid5_align_endio" should check if the read succeeded and if it
* did, call bio_endio on the original bio (having bio_put the new bio
* first).
* If the read failed..
*/
static void raid5_align_endio(struct bio *bi, int error)
{
struct bio* raid_bi = bi->bi_private;
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
mddev_t *mddev;
raid5_conf_t *conf;
int uptodate = test_bit(BIO_UPTODATE, &bi->bi_flags);
mdk_rdev_t *rdev;
bio_put(bi);
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
mddev = raid_bi->bi_bdev->bd_disk->queue->queuedata;
conf = mddev_to_conf(mddev);
rdev = (void*)raid_bi->bi_next;
raid_bi->bi_next = NULL;
rdev_dec_pending(rdev, conf->mddev);
if (!error && uptodate) {
bio_endio(raid_bi, 0);
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
if (atomic_dec_and_test(&conf->active_aligned_reads))
wake_up(&conf->wait_for_stripe);
return;
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
}
pr_debug("raid5_align_endio : io error...handing IO for a retry\n");
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
add_bio_to_retry(raid_bi, conf);
}
static int bio_fits_rdev(struct bio *bi)
{
struct request_queue *q = bdev_get_queue(bi->bi_bdev);
if ((bi->bi_size>>9) > q->max_sectors)
return 0;
blk_recount_segments(q, bi);
if (bi->bi_phys_segments > q->max_phys_segments ||
bi->bi_hw_segments > q->max_hw_segments)
return 0;
if (q->merge_bvec_fn)
/* it's too hard to apply the merge_bvec_fn at this stage,
* just just give up
*/
return 0;
return 1;
}
static int chunk_aligned_read(struct request_queue *q, struct bio * raid_bio)
{
mddev_t *mddev = q->queuedata;
raid5_conf_t *conf = mddev_to_conf(mddev);
const unsigned int raid_disks = conf->raid_disks;
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
const unsigned int data_disks = raid_disks - conf->max_degraded;
unsigned int dd_idx, pd_idx;
struct bio* align_bi;
mdk_rdev_t *rdev;
if (!in_chunk_boundary(mddev, raid_bio)) {
pr_debug("chunk_aligned_read : non aligned\n");
return 0;
}
/*
* use bio_clone to make a copy of the bio
*/
align_bi = bio_clone(raid_bio, GFP_NOIO);
if (!align_bi)
return 0;
/*
* set bi_end_io to a new function, and set bi_private to the
* original bio.
*/
align_bi->bi_end_io = raid5_align_endio;
align_bi->bi_private = raid_bio;
/*
* compute position
*/
align_bi->bi_sector = raid5_compute_sector(raid_bio->bi_sector,
raid_disks,
data_disks,
&dd_idx,
&pd_idx,
conf);
rcu_read_lock();
rdev = rcu_dereference(conf->disks[dd_idx].rdev);
if (rdev && test_bit(In_sync, &rdev->flags)) {
atomic_inc(&rdev->nr_pending);
rcu_read_unlock();
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
raid_bio->bi_next = (void*)rdev;
align_bi->bi_bdev = rdev->bdev;
align_bi->bi_flags &= ~(1 << BIO_SEG_VALID);
align_bi->bi_sector += rdev->data_offset;
if (!bio_fits_rdev(align_bi)) {
/* too big in some way */
bio_put(align_bi);
rdev_dec_pending(rdev, mddev);
return 0;
}
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
spin_lock_irq(&conf->device_lock);
wait_event_lock_irq(conf->wait_for_stripe,
conf->quiesce == 0,
conf->device_lock, /* nothing */);
atomic_inc(&conf->active_aligned_reads);
spin_unlock_irq(&conf->device_lock);
generic_make_request(align_bi);
return 1;
} else {
rcu_read_unlock();
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
bio_put(align_bi);
return 0;
}
}
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
/* __get_priority_stripe - get the next stripe to process
*
* Full stripe writes are allowed to pass preread active stripes up until
* the bypass_threshold is exceeded. In general the bypass_count
* increments when the handle_list is handled before the hold_list; however, it
* will not be incremented when STRIPE_IO_STARTED is sampled set signifying a
* stripe with in flight i/o. The bypass_count will be reset when the
* head of the hold_list has changed, i.e. the head was promoted to the
* handle_list.
*/
static struct stripe_head *__get_priority_stripe(raid5_conf_t *conf)
{
struct stripe_head *sh;
pr_debug("%s: handle: %s hold: %s full_writes: %d bypass_count: %d\n",
__func__,
list_empty(&conf->handle_list) ? "empty" : "busy",
list_empty(&conf->hold_list) ? "empty" : "busy",
atomic_read(&conf->pending_full_writes), conf->bypass_count);
if (!list_empty(&conf->handle_list)) {
sh = list_entry(conf->handle_list.next, typeof(*sh), lru);
if (list_empty(&conf->hold_list))
conf->bypass_count = 0;
else if (!test_bit(STRIPE_IO_STARTED, &sh->state)) {
if (conf->hold_list.next == conf->last_hold)
conf->bypass_count++;
else {
conf->last_hold = conf->hold_list.next;
conf->bypass_count -= conf->bypass_threshold;
if (conf->bypass_count < 0)
conf->bypass_count = 0;
}
}
} else if (!list_empty(&conf->hold_list) &&
((conf->bypass_threshold &&
conf->bypass_count > conf->bypass_threshold) ||
atomic_read(&conf->pending_full_writes) == 0)) {
sh = list_entry(conf->hold_list.next,
typeof(*sh), lru);
conf->bypass_count -= conf->bypass_threshold;
if (conf->bypass_count < 0)
conf->bypass_count = 0;
} else
return NULL;
list_del_init(&sh->lru);
atomic_inc(&sh->count);
BUG_ON(atomic_read(&sh->count) != 1);
return sh;
}
static int make_request(struct request_queue *q, struct bio * bi)
{
mddev_t *mddev = q->queuedata;
raid5_conf_t *conf = mddev_to_conf(mddev);
unsigned int dd_idx, pd_idx;
sector_t new_sector;
sector_t logical_sector, last_sector;
struct stripe_head *sh;
const int rw = bio_data_dir(bi);
int remaining;
if (unlikely(bio_barrier(bi))) {
bio_endio(bi, -EOPNOTSUPP);
return 0;
}
md_write_start(mddev, bi);
disk_stat_inc(mddev->gendisk, ios[rw]);
disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bi));
if (rw == READ &&
mddev->reshape_position == MaxSector &&
chunk_aligned_read(q,bi))
return 0;
logical_sector = bi->bi_sector & ~((sector_t)STRIPE_SECTORS-1);
last_sector = bi->bi_sector + (bi->bi_size>>9);
bi->bi_next = NULL;
bi->bi_phys_segments = 1; /* over-loaded to count active stripes */
for (;logical_sector < last_sector; logical_sector += STRIPE_SECTORS) {
DEFINE_WAIT(w);
int disks, data_disks;
retry:
prepare_to_wait(&conf->wait_for_overlap, &w, TASK_UNINTERRUPTIBLE);
if (likely(conf->expand_progress == MaxSector))
disks = conf->raid_disks;
else {
/* spinlock is needed as expand_progress may be
* 64bit on a 32bit platform, and so it might be
* possible to see a half-updated value
* Ofcourse expand_progress could change after
* the lock is dropped, so once we get a reference
* to the stripe that we think it is, we will have
* to check again.
*/
spin_lock_irq(&conf->device_lock);
disks = conf->raid_disks;
if (logical_sector >= conf->expand_progress)
disks = conf->previous_raid_disks;
else {
if (logical_sector >= conf->expand_lo) {
spin_unlock_irq(&conf->device_lock);
schedule();
goto retry;
}
}
spin_unlock_irq(&conf->device_lock);
}
data_disks = disks - conf->max_degraded;
new_sector = raid5_compute_sector(logical_sector, disks, data_disks,
&dd_idx, &pd_idx, conf);
pr_debug("raid5: make_request, sector %llu logical %llu\n",
(unsigned long long)new_sector,
(unsigned long long)logical_sector);
sh = get_active_stripe(conf, new_sector, disks, pd_idx, (bi->bi_rw&RWA_MASK));
if (sh) {
if (unlikely(conf->expand_progress != MaxSector)) {
/* expansion might have moved on while waiting for a
* stripe, so we must do the range check again.
* Expansion could still move past after this
* test, but as we are holding a reference to
* 'sh', we know that if that happens,
* STRIPE_EXPANDING will get set and the expansion
* won't proceed until we finish with the stripe.
*/
int must_retry = 0;
spin_lock_irq(&conf->device_lock);
if (logical_sector < conf->expand_progress &&
disks == conf->previous_raid_disks)
/* mismatch, need to try again */
must_retry = 1;
spin_unlock_irq(&conf->device_lock);
if (must_retry) {
release_stripe(sh);
goto retry;
}
}
/* FIXME what if we get a false positive because these
* are being updated.
*/
if (logical_sector >= mddev->suspend_lo &&
logical_sector < mddev->suspend_hi) {
release_stripe(sh);
schedule();
goto retry;
}
if (test_bit(STRIPE_EXPANDING, &sh->state) ||
!add_stripe_bio(sh, bi, dd_idx, (bi->bi_rw&RW_MASK))) {
/* Stripe is busy expanding or
* add failed due to overlap. Flush everything
* and wait a while
*/
raid5_unplug_device(mddev->queue);
release_stripe(sh);
schedule();
goto retry;
}
finish_wait(&conf->wait_for_overlap, &w);
set_bit(STRIPE_HANDLE, &sh->state);
clear_bit(STRIPE_DELAYED, &sh->state);
release_stripe(sh);
} else {
/* cannot get stripe for read-ahead, just give-up */
clear_bit(BIO_UPTODATE, &bi->bi_flags);
finish_wait(&conf->wait_for_overlap, &w);
break;
}
}
spin_lock_irq(&conf->device_lock);
remaining = --bi->bi_phys_segments;
spin_unlock_irq(&conf->device_lock);
if (remaining == 0) {
if ( rw == WRITE )
md_write_end(mddev);
bio_endio(bi, 0);
}
return 0;
}
static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped)
{
/* reshaping is quite different to recovery/resync so it is
* handled quite separately ... here.
*
* On each call to sync_request, we gather one chunk worth of
* destination stripes and flag them as expanding.
* Then we find all the source stripes and request reads.
* As the reads complete, handle_stripe will copy the data
* into the destination stripe and release that stripe.
*/
raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
struct stripe_head *sh;
int pd_idx;
sector_t first_sector, last_sector;
int raid_disks = conf->previous_raid_disks;
int data_disks = raid_disks - conf->max_degraded;
int new_data_disks = conf->raid_disks - conf->max_degraded;
int i;
int dd_idx;
sector_t writepos, safepos, gap;
if (sector_nr == 0 &&
conf->expand_progress != 0) {
/* restarting in the middle, skip the initial sectors */
sector_nr = conf->expand_progress;
sector_div(sector_nr, new_data_disks);
*skipped = 1;
return sector_nr;
}
/* we update the metadata when there is more than 3Meg
* in the block range (that is rather arbitrary, should
* probably be time based) or when the data about to be
* copied would over-write the source of the data at
* the front of the range.
* i.e. one new_stripe forward from expand_progress new_maps
* to after where expand_lo old_maps to
*/
writepos = conf->expand_progress +
conf->chunk_size/512*(new_data_disks);
sector_div(writepos, new_data_disks);
safepos = conf->expand_lo;
sector_div(safepos, data_disks);
gap = conf->expand_progress - conf->expand_lo;
if (writepos >= safepos ||
gap > (new_data_disks)*3000*2 /*3Meg*/) {
/* Cannot proceed until we've updated the superblock... */
wait_event(conf->wait_for_overlap,
atomic_read(&conf->reshape_stripes)==0);
mddev->reshape_position = conf->expand_progress;
set_bit(MD_CHANGE_DEVS, &mddev->flags);
md_wakeup_thread(mddev->thread);
wait_event(mddev->sb_wait, mddev->flags == 0 ||
kthread_should_stop());
spin_lock_irq(&conf->device_lock);
conf->expand_lo = mddev->reshape_position;
spin_unlock_irq(&conf->device_lock);
wake_up(&conf->wait_for_overlap);
}
for (i=0; i < conf->chunk_size/512; i+= STRIPE_SECTORS) {
int j;
int skipped = 0;
pd_idx = stripe_to_pdidx(sector_nr+i, conf, conf->raid_disks);
sh = get_active_stripe(conf, sector_nr+i,
conf->raid_disks, pd_idx, 0);
set_bit(STRIPE_EXPANDING, &sh->state);
atomic_inc(&conf->reshape_stripes);
/* If any of this stripe is beyond the end of the old
* array, then we need to zero those blocks
*/
for (j=sh->disks; j--;) {
sector_t s;
if (j == sh->pd_idx)
continue;
if (conf->level == 6 &&
j == raid6_next_disk(sh->pd_idx, sh->disks))
continue;
s = compute_blocknr(sh, j);
if (s < mddev->array_sectors) {
skipped = 1;
continue;
}
memset(page_address(sh->dev[j].page), 0, STRIPE_SIZE);
set_bit(R5_Expanded, &sh->dev[j].flags);
set_bit(R5_UPTODATE, &sh->dev[j].flags);
}
if (!skipped) {
set_bit(STRIPE_EXPAND_READY, &sh->state);
set_bit(STRIPE_HANDLE, &sh->state);
}
release_stripe(sh);
}
spin_lock_irq(&conf->device_lock);
conf->expand_progress = (sector_nr + i) * new_data_disks;
spin_unlock_irq(&conf->device_lock);
/* Ok, those stripe are ready. We can start scheduling
* reads on the source stripes.
* The source stripes are determined by mapping the first and last
* block on the destination stripes.
*/
first_sector =
raid5_compute_sector(sector_nr*(new_data_disks),
raid_disks, data_disks,
&dd_idx, &pd_idx, conf);
last_sector =
raid5_compute_sector((sector_nr+conf->chunk_size/512)
*(new_data_disks) -1,
raid_disks, data_disks,
&dd_idx, &pd_idx, conf);
if (last_sector >= (mddev->size<<1))
last_sector = (mddev->size<<1)-1;
while (first_sector <= last_sector) {
pd_idx = stripe_to_pdidx(first_sector, conf,
conf->previous_raid_disks);
sh = get_active_stripe(conf, first_sector,
conf->previous_raid_disks, pd_idx, 0);
set_bit(STRIPE_EXPAND_SOURCE, &sh->state);
set_bit(STRIPE_HANDLE, &sh->state);
release_stripe(sh);
first_sector += STRIPE_SECTORS;
}
/* If this takes us to the resync_max point where we have to pause,
* then we need to write out the superblock.
*/
sector_nr += conf->chunk_size>>9;
if (sector_nr >= mddev->resync_max) {
/* Cannot proceed until we've updated the superblock... */
wait_event(conf->wait_for_overlap,
atomic_read(&conf->reshape_stripes) == 0);
mddev->reshape_position = conf->expand_progress;
set_bit(MD_CHANGE_DEVS, &mddev->flags);
md_wakeup_thread(mddev->thread);
wait_event(mddev->sb_wait,
!test_bit(MD_CHANGE_DEVS, &mddev->flags)
|| kthread_should_stop());
spin_lock_irq(&conf->device_lock);
conf->expand_lo = mddev->reshape_position;
spin_unlock_irq(&conf->device_lock);
wake_up(&conf->wait_for_overlap);
}
return conf->chunk_size>>9;
}
/* FIXME go_faster isn't used */
static inline sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, int go_faster)
{
raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
struct stripe_head *sh;
int pd_idx;
int raid_disks = conf->raid_disks;
sector_t max_sector = mddev->size << 1;
int sync_blocks;
int still_degraded = 0;
int i;
if (sector_nr >= max_sector) {
/* just being told to finish up .. nothing much to do */
unplug_slaves(mddev);
if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery)) {
end_reshape(conf);
return 0;
}
if (mddev->curr_resync < max_sector) /* aborted */
bitmap_end_sync(mddev->bitmap, mddev->curr_resync,
&sync_blocks, 1);
else /* completed sync */
conf->fullsync = 0;
bitmap_close_sync(mddev->bitmap);
return 0;
}
if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery))
return reshape_request(mddev, sector_nr, skipped);
/* No need to check resync_max as we never do more than one
* stripe, and as resync_max will always be on a chunk boundary,
* if the check in md_do_sync didn't fire, there is no chance
* of overstepping resync_max here
*/
/* if there is too many failed drives and we are trying
* to resync, then assert that we are finished, because there is
* nothing we can do.
*/
if (mddev->degraded >= conf->max_degraded &&
test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
sector_t rv = (mddev->size << 1) - sector_nr;
*skipped = 1;
return rv;
}
if (!bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, 1) &&
!test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery) &&
!conf->fullsync && sync_blocks >= STRIPE_SECTORS) {
/* we can skip this block, and probably more */
sync_blocks /= STRIPE_SECTORS;
*skipped = 1;
return sync_blocks * STRIPE_SECTORS; /* keep things rounded to whole stripes */
}
bitmap_cond_end_sync(mddev->bitmap, sector_nr);
pd_idx = stripe_to_pdidx(sector_nr, conf, raid_disks);
sh = get_active_stripe(conf, sector_nr, raid_disks, pd_idx, 1);
if (sh == NULL) {
sh = get_active_stripe(conf, sector_nr, raid_disks, pd_idx, 0);
/* make sure we don't swamp the stripe cache if someone else
* is trying to get access
*/
schedule_timeout_uninterruptible(1);
}
/* Need to check if array will still be degraded after recovery/resync
* We don't need to check the 'failed' flag as when that gets set,
* recovery aborts.
*/
for (i=0; i<mddev->raid_disks; i++)
if (conf->disks[i].rdev == NULL)
still_degraded = 1;
bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, still_degraded);
spin_lock(&sh->lock);
set_bit(STRIPE_SYNCING, &sh->state);
clear_bit(STRIPE_INSYNC, &sh->state);
spin_unlock(&sh->lock);
/* wait for any blocked device to be handled */
while(unlikely(!handle_stripe(sh, NULL)))
;
release_stripe(sh);
return STRIPE_SECTORS;
}
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
static int retry_aligned_read(raid5_conf_t *conf, struct bio *raid_bio)
{
/* We may not be able to submit a whole bio at once as there
* may not be enough stripe_heads available.
* We cannot pre-allocate enough stripe_heads as we may need
* more than exist in the cache (if we allow ever large chunks).
* So we do one stripe head at a time and record in
* ->bi_hw_segments how many have been done.
*
* We *know* that this entire raid_bio is in one chunk, so
* it will be only one 'dd_idx' and only need one call to raid5_compute_sector.
*/
struct stripe_head *sh;
int dd_idx, pd_idx;
sector_t sector, logical_sector, last_sector;
int scnt = 0;
int remaining;
int handled = 0;
logical_sector = raid_bio->bi_sector & ~((sector_t)STRIPE_SECTORS-1);
sector = raid5_compute_sector( logical_sector,
conf->raid_disks,
conf->raid_disks - conf->max_degraded,
&dd_idx,
&pd_idx,
conf);
last_sector = raid_bio->bi_sector + (raid_bio->bi_size>>9);
for (; logical_sector < last_sector;
logical_sector += STRIPE_SECTORS,
sector += STRIPE_SECTORS,
scnt++) {
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
if (scnt < raid_bio->bi_hw_segments)
/* already done this stripe */
continue;
sh = get_active_stripe(conf, sector, conf->raid_disks, pd_idx, 1);
if (!sh) {
/* failed to get a stripe - must wait */
raid_bio->bi_hw_segments = scnt;
conf->retry_read_aligned = raid_bio;
return handled;
}
set_bit(R5_ReadError, &sh->dev[dd_idx].flags);
if (!add_stripe_bio(sh, raid_bio, dd_idx, 0)) {
release_stripe(sh);
raid_bio->bi_hw_segments = scnt;
conf->retry_read_aligned = raid_bio;
return handled;
}
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
handle_stripe(sh, NULL);
release_stripe(sh);
handled++;
}
spin_lock_irq(&conf->device_lock);
remaining = --raid_bio->bi_phys_segments;
spin_unlock_irq(&conf->device_lock);
if (remaining == 0)
bio_endio(raid_bio, 0);
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
if (atomic_dec_and_test(&conf->active_aligned_reads))
wake_up(&conf->wait_for_stripe);
return handled;
}
/*
* This is our raid5 kernel thread.
*
* We scan the hash table for stripes which can be handled now.
* During the scan, completed stripes are saved for us by the interrupt
* handler, so that they will not have to wait for our next wakeup.
*/
static void raid5d(mddev_t *mddev)
{
struct stripe_head *sh;
raid5_conf_t *conf = mddev_to_conf(mddev);
int handled;
pr_debug("+++ raid5d active\n");
md_check_recovery(mddev);
handled = 0;
spin_lock_irq(&conf->device_lock);
while (1) {
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
struct bio *bio;
if (conf->seq_flush != conf->seq_write) {
int seq = conf->seq_flush;
spin_unlock_irq(&conf->device_lock);
bitmap_unplug(mddev->bitmap);
spin_lock_irq(&conf->device_lock);
conf->seq_write = seq;
activate_bit_delay(conf);
}
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
while ((bio = remove_bio_from_retry(conf))) {
int ok;
spin_unlock_irq(&conf->device_lock);
ok = retry_aligned_read(conf, bio);
spin_lock_irq(&conf->device_lock);
if (!ok)
break;
handled++;
}
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
sh = __get_priority_stripe(conf);
if (!sh)
break;
spin_unlock_irq(&conf->device_lock);
handled++;
handle_stripe(sh, conf->spare_page);
release_stripe(sh);
spin_lock_irq(&conf->device_lock);
}
pr_debug("%d stripes handled\n", handled);
spin_unlock_irq(&conf->device_lock);
async_tx_issue_pending_all();
unplug_slaves(mddev);
pr_debug("--- raid5d inactive\n");
}
static ssize_t
raid5_show_stripe_cache_size(mddev_t *mddev, char *page)
{
raid5_conf_t *conf = mddev_to_conf(mddev);
if (conf)
return sprintf(page, "%d\n", conf->max_nr_stripes);
else
return 0;
}
static ssize_t
raid5_store_stripe_cache_size(mddev_t *mddev, const char *page, size_t len)
{
raid5_conf_t *conf = mddev_to_conf(mddev);
unsigned long new;
int err;
if (len >= PAGE_SIZE)
return -EINVAL;
if (!conf)
return -ENODEV;
if (strict_strtoul(page, 10, &new))
return -EINVAL;
if (new <= 16 || new > 32768)
return -EINVAL;
while (new < conf->max_nr_stripes) {
if (drop_one_stripe(conf))
conf->max_nr_stripes--;
else
break;
}
err = md_allow_write(mddev);
if (err)
return err;
while (new > conf->max_nr_stripes) {
if (grow_one_stripe(conf))
conf->max_nr_stripes++;
else break;
}
return len;
}
static struct md_sysfs_entry
raid5_stripecache_size = __ATTR(stripe_cache_size, S_IRUGO | S_IWUSR,
raid5_show_stripe_cache_size,
raid5_store_stripe_cache_size);
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
static ssize_t
raid5_show_preread_threshold(mddev_t *mddev, char *page)
{
raid5_conf_t *conf = mddev_to_conf(mddev);
if (conf)
return sprintf(page, "%d\n", conf->bypass_threshold);
else
return 0;
}
static ssize_t
raid5_store_preread_threshold(mddev_t *mddev, const char *page, size_t len)
{
raid5_conf_t *conf = mddev_to_conf(mddev);
unsigned long new;
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
if (len >= PAGE_SIZE)
return -EINVAL;
if (!conf)
return -ENODEV;
if (strict_strtoul(page, 10, &new))
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
return -EINVAL;
if (new > conf->max_nr_stripes)
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
return -EINVAL;
conf->bypass_threshold = new;
return len;
}
static struct md_sysfs_entry
raid5_preread_bypass_threshold = __ATTR(preread_bypass_threshold,
S_IRUGO | S_IWUSR,
raid5_show_preread_threshold,
raid5_store_preread_threshold);
static ssize_t
stripe_cache_active_show(mddev_t *mddev, char *page)
{
raid5_conf_t *conf = mddev_to_conf(mddev);
if (conf)
return sprintf(page, "%d\n", atomic_read(&conf->active_stripes));
else
return 0;
}
static struct md_sysfs_entry
raid5_stripecache_active = __ATTR_RO(stripe_cache_active);
static struct attribute *raid5_attrs[] = {
&raid5_stripecache_size.attr,
&raid5_stripecache_active.attr,
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
&raid5_preread_bypass_threshold.attr,
NULL,
};
static struct attribute_group raid5_attrs_group = {
.name = NULL,
.attrs = raid5_attrs,
};
static int run(mddev_t *mddev)
{
raid5_conf_t *conf;
int raid_disk, memory;
mdk_rdev_t *rdev;
struct disk_info *disk;
struct list_head *tmp;
int working_disks = 0;
if (mddev->level != 5 && mddev->level != 4 && mddev->level != 6) {
printk(KERN_ERR "raid5: %s: raid level not set to 4/5/6 (%d)\n",
mdname(mddev), mddev->level);
return -EIO;
}
if (mddev->reshape_position != MaxSector) {
/* Check that we can continue the reshape.
* Currently only disks can change, it must
* increase, and we must be past the point where
* a stripe over-writes itself
*/
sector_t here_new, here_old;
int old_disks;
int max_degraded = (mddev->level == 5 ? 1 : 2);
if (mddev->new_level != mddev->level ||
mddev->new_layout != mddev->layout ||
mddev->new_chunk != mddev->chunk_size) {
printk(KERN_ERR "raid5: %s: unsupported reshape "
"required - aborting.\n",
mdname(mddev));
return -EINVAL;
}
if (mddev->delta_disks <= 0) {
printk(KERN_ERR "raid5: %s: unsupported reshape "
"(reduce disks) required - aborting.\n",
mdname(mddev));
return -EINVAL;
}
old_disks = mddev->raid_disks - mddev->delta_disks;
/* reshape_position must be on a new-stripe boundary, and one
* further up in new geometry must map after here in old
* geometry.
*/
here_new = mddev->reshape_position;
if (sector_div(here_new, (mddev->chunk_size>>9)*
(mddev->raid_disks - max_degraded))) {
printk(KERN_ERR "raid5: reshape_position not "
"on a stripe boundary\n");
return -EINVAL;
}
/* here_new is the stripe we will write to */
here_old = mddev->reshape_position;
sector_div(here_old, (mddev->chunk_size>>9)*
(old_disks-max_degraded));
/* here_old is the first stripe that we might need to read
* from */
if (here_new >= here_old) {
/* Reading from the same stripe as writing to - bad */
printk(KERN_ERR "raid5: reshape_position too early for "
"auto-recovery - aborting.\n");
return -EINVAL;
}
printk(KERN_INFO "raid5: reshape will continue\n");
/* OK, we should be able to continue; */
}
mddev->private = kzalloc(sizeof (raid5_conf_t), GFP_KERNEL);
if ((conf = mddev->private) == NULL)
goto abort;
if (mddev->reshape_position == MaxSector) {
conf->previous_raid_disks = conf->raid_disks = mddev->raid_disks;
} else {
conf->raid_disks = mddev->raid_disks;
conf->previous_raid_disks = mddev->raid_disks - mddev->delta_disks;
}
conf->disks = kzalloc(conf->raid_disks * sizeof(struct disk_info),
GFP_KERNEL);
if (!conf->disks)
goto abort;
conf->mddev = mddev;
if ((conf->stripe_hashtbl = kzalloc(PAGE_SIZE, GFP_KERNEL)) == NULL)
goto abort;
if (mddev->level == 6) {
conf->spare_page = alloc_page(GFP_KERNEL);
if (!conf->spare_page)
goto abort;
}
spin_lock_init(&conf->device_lock);
mddev->queue->queue_lock = &conf->device_lock;
init_waitqueue_head(&conf->wait_for_stripe);
init_waitqueue_head(&conf->wait_for_overlap);
INIT_LIST_HEAD(&conf->handle_list);
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
INIT_LIST_HEAD(&conf->hold_list);
INIT_LIST_HEAD(&conf->delayed_list);
INIT_LIST_HEAD(&conf->bitmap_list);
INIT_LIST_HEAD(&conf->inactive_list);
atomic_set(&conf->active_stripes, 0);
atomic_set(&conf->preread_active_stripes, 0);
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
atomic_set(&conf->active_aligned_reads, 0);
md: introduce get_priority_stripe() to improve raid456 write performance Improve write performance by preventing the delayed_list from dumping all its stripes onto the handle_list in one shot. Delayed stripes are now further delayed by being held on the 'hold_list'. The 'hold_list' is bypassed when: * a STRIPE_IO_STARTED stripe is found at the head of 'handle_list' * 'handle_list' is empty and i/o is being done to satisfy full stripe-width write requests * 'bypass_count' is less than 'bypass_threshold'. By default the threshold is 1, i.e. every other stripe handled is a preread stripe provided the top two conditions are false. Benchmark data: System: 2x Xeon 5150, 4x SATA, mem=1GB Baseline: 2.6.24-rc7 Configuration: mdadm --create /dev/md0 /dev/sd[b-e] -n 4 -l 5 --assume-clean Test1: dd if=/dev/zero of=/dev/md0 bs=1024k count=2048 * patched: +33% (stripe_cache_size = 256), +25% (stripe_cache_size = 512) Test2: tiobench --size 2048 --numruns 5 --block 4096 --block 131072 (XFS) * patched: +13% * patched + preread_bypass_threshold = 0: +37% Changes since v1: * reduce bypass_threshold from (chunk_size / sectors_per_chunk) to (1) and make it configurable. This defaults to fairness and modest performance gains out of the box. Changes since v2: * [neilb@suse.de]: kill STRIPE_PRIO_HI and preread_needed as they are not necessary, the important change was clearing STRIPE_DELAYED in add_stripe_bio and this has been moved out to make_request for the hang fix. * [neilb@suse.de]: simplify get_priority_stripe * [dan.j.williams@intel.com]: reset the bypass_count when ->hold_list is sampled empty (+11%) * [dan.j.williams@intel.com]: decrement the bypass_count at the detection of stripes being naturally promoted off of hold_list +2%. Note, resetting bypass_count instead of decrementing on these events yields +4% but that is probably too aggressive. Changes since v3: * cosmetic fixups Tested-by: James W. Laferriere <babydr@baby-dragons.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 09:15:53 +00:00
conf->bypass_threshold = BYPASS_THRESHOLD;
pr_debug("raid5: run(%s) called.\n", mdname(mddev));
rdev_for_each(rdev, tmp, mddev) {
raid_disk = rdev->raid_disk;
if (raid_disk >= conf->raid_disks
|| raid_disk < 0)
continue;
disk = conf->disks + raid_disk;
disk->rdev = rdev;
if (test_bit(In_sync, &rdev->flags)) {
char b[BDEVNAME_SIZE];
printk(KERN_INFO "raid5: device %s operational as raid"
" disk %d\n", bdevname(rdev->bdev,b),
raid_disk);
working_disks++;
} else
/* Cannot rely on bitmap to complete recovery */
conf->fullsync = 1;
}
/*
* 0 for a fully functional array, 1 or 2 for a degraded array.
*/
mddev->degraded = conf->raid_disks - working_disks;
conf->mddev = mddev;
conf->chunk_size = mddev->chunk_size;
conf->level = mddev->level;
if (conf->level == 6)
conf->max_degraded = 2;
else
conf->max_degraded = 1;
conf->algorithm = mddev->layout;
conf->max_nr_stripes = NR_STRIPES;
conf->expand_progress = mddev->reshape_position;
/* device size must be a multiple of chunk size */
mddev->size &= ~(mddev->chunk_size/1024 -1);
mddev->resync_max_sectors = mddev->size << 1;
if (conf->level == 6 && conf->raid_disks < 4) {
printk(KERN_ERR "raid6: not enough configured devices for %s (%d, minimum 4)\n",
mdname(mddev), conf->raid_disks);
goto abort;
}
if (!conf->chunk_size || conf->chunk_size % 4) {
printk(KERN_ERR "raid5: invalid chunk size %d for %s\n",
conf->chunk_size, mdname(mddev));
goto abort;
}
if (conf->algorithm > ALGORITHM_RIGHT_SYMMETRIC) {
printk(KERN_ERR
"raid5: unsupported parity algorithm %d for %s\n",
conf->algorithm, mdname(mddev));
goto abort;
}
if (mddev->degraded > conf->max_degraded) {
printk(KERN_ERR "raid5: not enough operational devices for %s"
" (%d/%d failed)\n",
mdname(mddev), mddev->degraded, conf->raid_disks);
goto abort;
}
if (mddev->degraded > 0 &&
mddev->recovery_cp != MaxSector) {
if (mddev->ok_start_degraded)
printk(KERN_WARNING
"raid5: starting dirty degraded array: %s"
"- data corruption possible.\n",
mdname(mddev));
else {
printk(KERN_ERR
"raid5: cannot start dirty degraded array for %s\n",
mdname(mddev));
goto abort;
}
}
{
mddev->thread = md_register_thread(raid5d, mddev, "%s_raid5");
if (!mddev->thread) {
printk(KERN_ERR
"raid5: couldn't allocate thread for %s\n",
mdname(mddev));
goto abort;
}
}
memory = conf->max_nr_stripes * (sizeof(struct stripe_head) +
conf->raid_disks * ((sizeof(struct bio) + PAGE_SIZE))) / 1024;
if (grow_stripes(conf, conf->max_nr_stripes)) {
printk(KERN_ERR
"raid5: couldn't allocate %dkB for buffers\n", memory);
shrink_stripes(conf);
md_unregister_thread(mddev->thread);
goto abort;
} else
printk(KERN_INFO "raid5: allocated %dkB for %s\n",
memory, mdname(mddev));
if (mddev->degraded == 0)
printk("raid5: raid level %d set %s active with %d out of %d"
" devices, algorithm %d\n", conf->level, mdname(mddev),
mddev->raid_disks-mddev->degraded, mddev->raid_disks,
conf->algorithm);
else
printk(KERN_ALERT "raid5: raid level %d set %s active with %d"
" out of %d devices, algorithm %d\n", conf->level,
mdname(mddev), mddev->raid_disks - mddev->degraded,
mddev->raid_disks, conf->algorithm);
print_raid5_conf(conf);
if (conf->expand_progress != MaxSector) {
printk("...ok start reshape thread\n");
conf->expand_lo = conf->expand_progress;
atomic_set(&conf->reshape_stripes, 0);
clear_bit(MD_RECOVERY_SYNC, &mddev->recovery);
clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
set_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);
set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
mddev->sync_thread = md_register_thread(md_do_sync, mddev,
"%s_reshape");
}
/* read-ahead size must cover two whole stripes, which is
* 2 * (datadisks) * chunksize where 'n' is the number of raid devices
*/
{
int data_disks = conf->previous_raid_disks - conf->max_degraded;
int stripe = data_disks *
(mddev->chunk_size / PAGE_SIZE);
if (mddev->queue->backing_dev_info.ra_pages < 2 * stripe)
mddev->queue->backing_dev_info.ra_pages = 2 * stripe;
}
/* Ok, everything is just fine now */
if (sysfs_create_group(&mddev->kobj, &raid5_attrs_group))
printk(KERN_WARNING
"raid5: failed to create sysfs attributes for %s\n",
mdname(mddev));
mddev->queue->unplug_fn = raid5_unplug_device;
mddev->queue->backing_dev_info.congested_data = mddev;
mddev->queue->backing_dev_info.congested_fn = raid5_congested;
mddev->array_sectors = 2 * mddev->size * (conf->previous_raid_disks -
conf->max_degraded);
blk_queue_merge_bvec(mddev->queue, raid5_mergeable_bvec);
return 0;
abort:
if (conf) {
print_raid5_conf(conf);
safe_put_page(conf->spare_page);
kfree(conf->disks);
kfree(conf->stripe_hashtbl);
kfree(conf);
}
mddev->private = NULL;
printk(KERN_ALERT "raid5: failed to run raid set %s\n", mdname(mddev));
return -EIO;
}
static int stop(mddev_t *mddev)
{
raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
md_unregister_thread(mddev->thread);
mddev->thread = NULL;
shrink_stripes(conf);
kfree(conf->stripe_hashtbl);
mddev->queue->backing_dev_info.congested_fn = NULL;
blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
sysfs_remove_group(&mddev->kobj, &raid5_attrs_group);
kfree(conf->disks);
kfree(conf);
mddev->private = NULL;
return 0;
}
#ifdef DEBUG
static void print_sh (struct seq_file *seq, struct stripe_head *sh)
{
int i;
seq_printf(seq, "sh %llu, pd_idx %d, state %ld.\n",
(unsigned long long)sh->sector, sh->pd_idx, sh->state);
seq_printf(seq, "sh %llu, count %d.\n",
(unsigned long long)sh->sector, atomic_read(&sh->count));
seq_printf(seq, "sh %llu, ", (unsigned long long)sh->sector);
for (i = 0; i < sh->disks; i++) {
seq_printf(seq, "(cache%d: %p %ld) ",
i, sh->dev[i].page, sh->dev[i].flags);
}
seq_printf(seq, "\n");
}
static void printall (struct seq_file *seq, raid5_conf_t *conf)
{
struct stripe_head *sh;
struct hlist_node *hn;
int i;
spin_lock_irq(&conf->device_lock);
for (i = 0; i < NR_HASH; i++) {
hlist_for_each_entry(sh, hn, &conf->stripe_hashtbl[i], hash) {
if (sh->raid_conf != conf)
continue;
print_sh(seq, sh);
}
}
spin_unlock_irq(&conf->device_lock);
}
#endif
static void status (struct seq_file *seq, mddev_t *mddev)
{
raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
int i;
seq_printf (seq, " level %d, %dk chunk, algorithm %d", mddev->level, mddev->chunk_size >> 10, mddev->layout);
seq_printf (seq, " [%d/%d] [", conf->raid_disks, conf->raid_disks - mddev->degraded);
for (i = 0; i < conf->raid_disks; i++)
seq_printf (seq, "%s",
conf->disks[i].rdev &&
test_bit(In_sync, &conf->disks[i].rdev->flags) ? "U" : "_");
seq_printf (seq, "]");
#ifdef DEBUG
seq_printf (seq, "\n");
printall(seq, conf);
#endif
}
static void print_raid5_conf (raid5_conf_t *conf)
{
int i;
struct disk_info *tmp;
printk("RAID5 conf printout:\n");
if (!conf) {
printk("(conf==NULL)\n");
return;
}
printk(" --- rd:%d wd:%d\n", conf->raid_disks,
conf->raid_disks - conf->mddev->degraded);
for (i = 0; i < conf->raid_disks; i++) {
char b[BDEVNAME_SIZE];
tmp = conf->disks + i;
if (tmp->rdev)
printk(" disk %d, o:%d, dev:%s\n",
i, !test_bit(Faulty, &tmp->rdev->flags),
bdevname(tmp->rdev->bdev,b));
}
}
static int raid5_spare_active(mddev_t *mddev)
{
int i;
raid5_conf_t *conf = mddev->private;
struct disk_info *tmp;
for (i = 0; i < conf->raid_disks; i++) {
tmp = conf->disks + i;
if (tmp->rdev
&& !test_bit(Faulty, &tmp->rdev->flags)
&& !test_and_set_bit(In_sync, &tmp->rdev->flags)) {
unsigned long flags;
spin_lock_irqsave(&conf->device_lock, flags);
mddev->degraded--;
spin_unlock_irqrestore(&conf->device_lock, flags);
}
}
print_raid5_conf(conf);
return 0;
}
static int raid5_remove_disk(mddev_t *mddev, int number)
{
raid5_conf_t *conf = mddev->private;
int err = 0;
mdk_rdev_t *rdev;
struct disk_info *p = conf->disks + number;
print_raid5_conf(conf);
rdev = p->rdev;
if (rdev) {
if (test_bit(In_sync, &rdev->flags) ||
atomic_read(&rdev->nr_pending)) {
err = -EBUSY;
goto abort;
}
md: restart recovery cleanly after device failure. When we get any IO error during a recovery (rebuilding a spare), we abort the recovery and restart it. For RAID6 (and multi-drive RAID1) it may not be best to restart at the beginning: when multiple failures can be tolerated, the recovery may be able to continue and re-doing all that has already been done doesn't make sense. We already have the infrastructure to record where a recovery is up to and restart from there, but it is not being used properly. This is because: - We sometimes abort with MD_RECOVERY_ERR rather than just MD_RECOVERY_INTR, which causes the recovery not be be checkpointed. - We remove spares and then re-added them which loses important state information. The distinction between MD_RECOVERY_ERR and MD_RECOVERY_INTR really isn't needed. If there is an error, the relevant drive will be marked as Faulty, and that is enough to ensure correct handling of the error. So we first remove MD_RECOVERY_ERR, changing some of the uses of it to MD_RECOVERY_INTR. Then we cause the attempt to remove a non-faulty device from an array to fail (unless recovery is impossible as the array is too degraded). Then when remove_and_add_spares attempts to remove the devices on which recovery can continue, it will fail, they will remain in place, and recovery will continue on them as desired. Issue: If we are halfway through rebuilding a spare and another drive fails, and a new spare is immediately available, do we want to: 1/ complete the current rebuild, then go back and rebuild the new spare or 2/ restart the rebuild from the start and rebuild both devices in parallel. Both options can be argued for. The code currently takes option 2 as a/ this requires least code change b/ this results in a minimally-degraded array in minimal time. Cc: "Eivind Sarto" <ivan@kasenna.com> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-05-23 20:04:39 +00:00
/* Only remove non-faulty devices if recovery
* isn't possible.
*/
if (!test_bit(Faulty, &rdev->flags) &&
mddev->degraded <= conf->max_degraded) {
err = -EBUSY;
goto abort;
}
p->rdev = NULL;
synchronize_rcu();
if (atomic_read(&rdev->nr_pending)) {
/* lost the race, try later */
err = -EBUSY;
p->rdev = rdev;
}
}
abort:
print_raid5_conf(conf);
return err;
}
static int raid5_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
{
raid5_conf_t *conf = mddev->private;
int err = -EEXIST;
int disk;
struct disk_info *p;
int first = 0;
int last = conf->raid_disks - 1;
if (mddev->degraded > conf->max_degraded)
/* no point adding a device */
return -EINVAL;
if (rdev->raid_disk >= 0)
first = last = rdev->raid_disk;
/*
* find the disk ... but prefer rdev->saved_raid_disk
* if possible.
*/
if (rdev->saved_raid_disk >= 0 &&
rdev->saved_raid_disk >= first &&
conf->disks[rdev->saved_raid_disk].rdev == NULL)
disk = rdev->saved_raid_disk;
else
disk = first;
for ( ; disk <= last ; disk++)
if ((p=conf->disks + disk)->rdev == NULL) {
clear_bit(In_sync, &rdev->flags);
rdev->raid_disk = disk;
err = 0;
if (rdev->saved_raid_disk != disk)
conf->fullsync = 1;
rcu_assign_pointer(p->rdev, rdev);
break;
}
print_raid5_conf(conf);
return err;
}
static int raid5_resize(mddev_t *mddev, sector_t sectors)
{
/* no resync is happening, and there is enough space
* on all devices, so we can resize.
* We need to make sure resync covers any new space.
* If the array is shrinking we should possibly wait until
* any io in the removed space completes, but it hardly seems
* worth it.
*/
raid5_conf_t *conf = mddev_to_conf(mddev);
sectors &= ~((sector_t)mddev->chunk_size/512 - 1);
mddev->array_sectors = sectors * (mddev->raid_disks
- conf->max_degraded);
set_capacity(mddev->gendisk, mddev->array_sectors);
mddev->changed = 1;
if (sectors/2 > mddev->size && mddev->recovery_cp == MaxSector) {
mddev->recovery_cp = mddev->size << 1;
set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
}
mddev->size = sectors /2;
mddev->resync_max_sectors = sectors;
return 0;
}
#ifdef CONFIG_MD_RAID5_RESHAPE
static int raid5_check_reshape(mddev_t *mddev)
{
raid5_conf_t *conf = mddev_to_conf(mddev);
int err;
if (mddev->delta_disks < 0 ||
mddev->new_level != mddev->level)
return -EINVAL; /* Cannot shrink array or change level yet */
if (mddev->delta_disks == 0)
return 0; /* nothing to do */
if (mddev->bitmap)
/* Cannot grow a bitmap yet */
return -EBUSY;
/* Can only proceed if there are plenty of stripe_heads.
* We need a minimum of one full stripe,, and for sensible progress
* it is best to have about 4 times that.
* If we require 4 times, then the default 256 4K stripe_heads will
* allow for chunk sizes up to 256K, which is probably OK.
* If the chunk size is greater, user-space should request more
* stripe_heads first.
*/
if ((mddev->chunk_size / STRIPE_SIZE) * 4 > conf->max_nr_stripes ||
(mddev->new_chunk / STRIPE_SIZE) * 4 > conf->max_nr_stripes) {
printk(KERN_WARNING "raid5: reshape: not enough stripes. Needed %lu\n",
(mddev->chunk_size / STRIPE_SIZE)*4);
return -ENOSPC;
}
err = resize_stripes(conf, conf->raid_disks + mddev->delta_disks);
if (err)
return err;
if (mddev->degraded > conf->max_degraded)
return -EINVAL;
/* looks like we might be able to manage this */
return 0;
}
static int raid5_start_reshape(mddev_t *mddev)
{
raid5_conf_t *conf = mddev_to_conf(mddev);
mdk_rdev_t *rdev;
struct list_head *rtmp;
int spares = 0;
int added_devices = 0;
unsigned long flags;
if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery))
return -EBUSY;
rdev_for_each(rdev, rtmp, mddev)
if (rdev->raid_disk < 0 &&
!test_bit(Faulty, &rdev->flags))
spares++;
if (spares - mddev->degraded < mddev->delta_disks - conf->max_degraded)
/* Not enough devices even to make a degraded array
* of that size
*/
return -EINVAL;
atomic_set(&conf->reshape_stripes, 0);
spin_lock_irq(&conf->device_lock);
conf->previous_raid_disks = conf->raid_disks;
conf->raid_disks += mddev->delta_disks;
conf->expand_progress = 0;
conf->expand_lo = 0;
spin_unlock_irq(&conf->device_lock);
/* Add some new drives, as many as will fit.
* We know there are enough to make the newly sized array work.
*/
rdev_for_each(rdev, rtmp, mddev)
if (rdev->raid_disk < 0 &&
!test_bit(Faulty, &rdev->flags)) {
if (raid5_add_disk(mddev, rdev) == 0) {
char nm[20];
set_bit(In_sync, &rdev->flags);
added_devices++;
rdev->recovery_offset = 0;
sprintf(nm, "rd%d", rdev->raid_disk);
if (sysfs_create_link(&mddev->kobj,
&rdev->kobj, nm))
printk(KERN_WARNING
"raid5: failed to create "
" link %s for %s\n",
nm, mdname(mddev));
} else
break;
}
spin_lock_irqsave(&conf->device_lock, flags);
mddev->degraded = (conf->raid_disks - conf->previous_raid_disks) - added_devices;
spin_unlock_irqrestore(&conf->device_lock, flags);
mddev->raid_disks = conf->raid_disks;
mddev->reshape_position = 0;
set_bit(MD_CHANGE_DEVS, &mddev->flags);
clear_bit(MD_RECOVERY_SYNC, &mddev->recovery);
clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
set_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);
set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
mddev->sync_thread = md_register_thread(md_do_sync, mddev,
"%s_reshape");
if (!mddev->sync_thread) {
mddev->recovery = 0;
spin_lock_irq(&conf->device_lock);
mddev->raid_disks = conf->raid_disks = conf->previous_raid_disks;
conf->expand_progress = MaxSector;
spin_unlock_irq(&conf->device_lock);
return -EAGAIN;
}
md_wakeup_thread(mddev->sync_thread);
md_new_event(mddev);
return 0;
}
#endif
static void end_reshape(raid5_conf_t *conf)
{
struct block_device *bdev;
if (!test_bit(MD_RECOVERY_INTR, &conf->mddev->recovery)) {
conf->mddev->array_sectors = 2 * conf->mddev->size *
(conf->raid_disks - conf->max_degraded);
set_capacity(conf->mddev->gendisk, conf->mddev->array_sectors);
conf->mddev->changed = 1;
bdev = bdget_disk(conf->mddev->gendisk, 0);
if (bdev) {
mutex_lock(&bdev->bd_inode->i_mutex);
i_size_write(bdev->bd_inode,
(loff_t)conf->mddev->array_sectors << 9);
mutex_unlock(&bdev->bd_inode->i_mutex);
bdput(bdev);
}
spin_lock_irq(&conf->device_lock);
conf->expand_progress = MaxSector;
spin_unlock_irq(&conf->device_lock);
conf->mddev->reshape_position = MaxSector;
/* read-ahead size must cover two whole stripes, which is
* 2 * (datadisks) * chunksize where 'n' is the number of raid devices
*/
{
int data_disks = conf->previous_raid_disks - conf->max_degraded;
int stripe = data_disks *
(conf->mddev->chunk_size / PAGE_SIZE);
if (conf->mddev->queue->backing_dev_info.ra_pages < 2 * stripe)
conf->mddev->queue->backing_dev_info.ra_pages = 2 * stripe;
}
}
}
static void raid5_quiesce(mddev_t *mddev, int state)
{
raid5_conf_t *conf = mddev_to_conf(mddev);
switch(state) {
case 2: /* resume for a suspend */
wake_up(&conf->wait_for_overlap);
break;
case 1: /* stop all writes */
spin_lock_irq(&conf->device_lock);
conf->quiesce = 1;
wait_event_lock_irq(conf->wait_for_stripe,
[PATCH] md: allow reads that have bypassed the cache to be retried on failure If a bypass-the-cache read fails, we simply try again through the cache. If it fails again it will trigger normal recovery precedures. update 1: From: NeilBrown <neilb@suse.de> 1/ chunk_aligned_read and retry_aligned_read assume that data_disks == raid_disks - 1 which is not true for raid6. So when an aligned read request bypasses the cache, we can get the wrong data. 2/ The cloned bio is being used-after-free in raid5_align_endio (to test BIO_UPTODATE). 3/ We forgot to add rdev->data_offset when submitting a bio for aligned-read 4/ clone_bio calls blk_recount_segments and then we change bi_bdev, so we need to invalidate the segment counts. 5/ We don't de-reference the rdev when the read completes. This means we need to record the rdev to so it is still available in the end_io routine. Fortunately bi_next in the original bio is unused at this point so we can stuff it in there. 6/ We leak a cloned bio if the target rdev is not usable. From: NeilBrown <neilb@suse.de> update 2: 1/ When aligned requests fail (read error) they need to be retried via the normal method (stripe cache). As we cannot be sure that we can process a single read in one go (we may not be able to allocate all the stripes needed) we store a bio-being-retried and a list of bioes-that-still-need-to-be-retried. When find a bio that needs to be retried, we should add it to the list, not to single-bio... 2/ We were never incrementing 'scnt' when resubmitting failed aligned requests. [akpm@osdl.org: build fix] Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-10 10:20:47 +00:00
atomic_read(&conf->active_stripes) == 0 &&
atomic_read(&conf->active_aligned_reads) == 0,
conf->device_lock, /* nothing */);
spin_unlock_irq(&conf->device_lock);
break;
case 0: /* re-enable writes */
spin_lock_irq(&conf->device_lock);
conf->quiesce = 0;
wake_up(&conf->wait_for_stripe);
wake_up(&conf->wait_for_overlap);
spin_unlock_irq(&conf->device_lock);
break;
}
}
static struct mdk_personality raid6_personality =
{
.name = "raid6",
.level = 6,
.owner = THIS_MODULE,
.make_request = make_request,
.run = run,
.stop = stop,
.status = status,
.error_handler = error,
.hot_add_disk = raid5_add_disk,
.hot_remove_disk= raid5_remove_disk,
.spare_active = raid5_spare_active,
.sync_request = sync_request,
.resize = raid5_resize,
#ifdef CONFIG_MD_RAID5_RESHAPE
.check_reshape = raid5_check_reshape,
.start_reshape = raid5_start_reshape,
#endif
.quiesce = raid5_quiesce,
};
static struct mdk_personality raid5_personality =
{
.name = "raid5",
.level = 5,
.owner = THIS_MODULE,
.make_request = make_request,
.run = run,
.stop = stop,
.status = status,
.error_handler = error,
.hot_add_disk = raid5_add_disk,
.hot_remove_disk= raid5_remove_disk,
.spare_active = raid5_spare_active,
.sync_request = sync_request,
.resize = raid5_resize,
#ifdef CONFIG_MD_RAID5_RESHAPE
.check_reshape = raid5_check_reshape,
.start_reshape = raid5_start_reshape,
#endif
.quiesce = raid5_quiesce,
};
static struct mdk_personality raid4_personality =
{
.name = "raid4",
.level = 4,
.owner = THIS_MODULE,
.make_request = make_request,
.run = run,
.stop = stop,
.status = status,
.error_handler = error,
.hot_add_disk = raid5_add_disk,
.hot_remove_disk= raid5_remove_disk,
.spare_active = raid5_spare_active,
.sync_request = sync_request,
.resize = raid5_resize,
#ifdef CONFIG_MD_RAID5_RESHAPE
.check_reshape = raid5_check_reshape,
.start_reshape = raid5_start_reshape,
#endif
.quiesce = raid5_quiesce,
};
static int __init raid5_init(void)
{
int e;
e = raid6_select_algo();
if ( e )
return e;
register_md_personality(&raid6_personality);
register_md_personality(&raid5_personality);
register_md_personality(&raid4_personality);
return 0;
}
static void raid5_exit(void)
{
unregister_md_personality(&raid6_personality);
unregister_md_personality(&raid5_personality);
unregister_md_personality(&raid4_personality);
}
module_init(raid5_init);
module_exit(raid5_exit);
MODULE_LICENSE("GPL");
MODULE_ALIAS("md-personality-4"); /* RAID5 */
MODULE_ALIAS("md-raid5");
MODULE_ALIAS("md-raid4");
MODULE_ALIAS("md-level-5");
MODULE_ALIAS("md-level-4");
MODULE_ALIAS("md-personality-8"); /* RAID6 */
MODULE_ALIAS("md-raid6");
MODULE_ALIAS("md-level-6");
/* This used to be two separate modules, they were: */
MODULE_ALIAS("raid5");
MODULE_ALIAS("raid6");