#include <linux/slab.h>
#include <linux/delay.h>
#include <linux/blkdev.h>
+#include <linux/module.h>
#include <linux/seq_file.h>
#include <linux/ratelimit.h>
#include "md.h"
*/
#define NR_RAID10_BIOS 256
-static void allow_barrier(conf_t *conf);
-static void lower_barrier(conf_t *conf);
+/* When there are this many requests queue to be written by
+ * the raid10 thread, we become 'congested' to provide back-pressure
+ * for writeback.
+ */
+static int max_queued_requests = 1024;
+
+static void allow_barrier(struct r10conf *conf);
+static void lower_barrier(struct r10conf *conf);
static void * r10bio_pool_alloc(gfp_t gfp_flags, void *data)
{
- conf_t *conf = data;
- int size = offsetof(struct r10bio_s, devs[conf->copies]);
+ struct r10conf *conf = data;
+ int size = offsetof(struct r10bio, devs[conf->copies]);
/* allocate a r10bio with room for raid_disks entries in the bios array */
return kzalloc(size, gfp_flags);
*/
static void * r10buf_pool_alloc(gfp_t gfp_flags, void *data)
{
- conf_t *conf = data;
+ struct r10conf *conf = data;
struct page *page;
- r10bio_t *r10_bio;
+ struct r10bio *r10_bio;
struct bio *bio;
int i, j;
int nalloc;
static void r10buf_pool_free(void *__r10_bio, void *data)
{
int i;
- conf_t *conf = data;
- r10bio_t *r10bio = __r10_bio;
+ struct r10conf *conf = data;
+ struct r10bio *r10bio = __r10_bio;
int j;
for (j=0; j < conf->copies; j++) {
r10bio_pool_free(r10bio, conf);
}
-static void put_all_bios(conf_t *conf, r10bio_t *r10_bio)
+static void put_all_bios(struct r10conf *conf, struct r10bio *r10_bio)
{
int i;
}
}
-static void free_r10bio(r10bio_t *r10_bio)
+static void free_r10bio(struct r10bio *r10_bio)
{
- conf_t *conf = r10_bio->mddev->private;
+ struct r10conf *conf = r10_bio->mddev->private;
put_all_bios(conf, r10_bio);
mempool_free(r10_bio, conf->r10bio_pool);
}
-static void put_buf(r10bio_t *r10_bio)
+static void put_buf(struct r10bio *r10_bio)
{
- conf_t *conf = r10_bio->mddev->private;
+ struct r10conf *conf = r10_bio->mddev->private;
mempool_free(r10_bio, conf->r10buf_pool);
lower_barrier(conf);
}
-static void reschedule_retry(r10bio_t *r10_bio)
+static void reschedule_retry(struct r10bio *r10_bio)
{
unsigned long flags;
- mddev_t *mddev = r10_bio->mddev;
- conf_t *conf = mddev->private;
+ struct mddev *mddev = r10_bio->mddev;
+ struct r10conf *conf = mddev->private;
spin_lock_irqsave(&conf->device_lock, flags);
list_add(&r10_bio->retry_list, &conf->retry_list);
* operation and are ready to return a success/failure code to the buffer
* cache layer.
*/
-static void raid_end_bio_io(r10bio_t *r10_bio)
+static void raid_end_bio_io(struct r10bio *r10_bio)
{
struct bio *bio = r10_bio->master_bio;
int done;
- conf_t *conf = r10_bio->mddev->private;
+ struct r10conf *conf = r10_bio->mddev->private;
if (bio->bi_phys_segments) {
unsigned long flags;
/*
* Update disk head position estimator based on IRQ completion info.
*/
-static inline void update_head_pos(int slot, r10bio_t *r10_bio)
+static inline void update_head_pos(int slot, struct r10bio *r10_bio)
{
- conf_t *conf = r10_bio->mddev->private;
+ struct r10conf *conf = r10_bio->mddev->private;
conf->mirrors[r10_bio->devs[slot].devnum].head_position =
r10_bio->devs[slot].addr + (r10_bio->sectors);
/*
* Find the disk number which triggered given bio
*/
-static int find_bio_disk(conf_t *conf, r10bio_t *r10_bio,
+static int find_bio_disk(struct r10conf *conf, struct r10bio *r10_bio,
struct bio *bio, int *slotp)
{
int slot;
static void raid10_end_read_request(struct bio *bio, int error)
{
int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
- r10bio_t *r10_bio = bio->bi_private;
+ struct r10bio *r10_bio = bio->bi_private;
int slot, dev;
- conf_t *conf = r10_bio->mddev->private;
+ struct r10conf *conf = r10_bio->mddev->private;
slot = r10_bio->read_slot;
}
}
-static void close_write(r10bio_t *r10_bio)
+static void close_write(struct r10bio *r10_bio)
{
/* clear the bitmap if all writes complete successfully */
bitmap_endwrite(r10_bio->mddev->bitmap, r10_bio->sector,
md_write_end(r10_bio->mddev);
}
-static void one_write_done(r10bio_t *r10_bio)
+static void one_write_done(struct r10bio *r10_bio)
{
if (atomic_dec_and_test(&r10_bio->remaining)) {
if (test_bit(R10BIO_WriteError, &r10_bio->state))
static void raid10_end_write_request(struct bio *bio, int error)
{
int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
- r10bio_t *r10_bio = bio->bi_private;
+ struct r10bio *r10_bio = bio->bi_private;
int dev;
int dec_rdev = 1;
- conf_t *conf = r10_bio->mddev->private;
+ struct r10conf *conf = r10_bio->mddev->private;
int slot;
dev = find_bio_disk(conf, r10_bio, bio, &slot);
sector_t first_bad;
int bad_sectors;
- set_bit(R10BIO_Uptodate, &r10_bio->state);
+ /*
+ * Do not set R10BIO_Uptodate if the current device is
+ * rebuilding or Faulty. This is because we cannot use
+ * such device for properly reading the data back (we could
+ * potentially use it, if the current write would have felt
+ * before rdev->recovery_offset, but for simplicity we don't
+ * check this here.
+ */
+ if (test_bit(In_sync, &conf->mirrors[dev].rdev->flags) &&
+ !test_bit(Faulty, &conf->mirrors[dev].rdev->flags))
+ set_bit(R10BIO_Uptodate, &r10_bio->state);
/* Maybe we can clear some bad blocks. */
if (is_badblock(conf->mirrors[dev].rdev,
* sector offset to a virtual address
*/
-static void raid10_find_phys(conf_t *conf, r10bio_t *r10bio)
+static void raid10_find_phys(struct r10conf *conf, struct r10bio *r10bio)
{
int n,f;
sector_t sector;
BUG_ON(slot != conf->copies);
}
-static sector_t raid10_find_virt(conf_t *conf, sector_t sector, int dev)
+static sector_t raid10_find_virt(struct r10conf *conf, sector_t sector, int dev)
{
sector_t offset, chunk, vchunk;
struct bvec_merge_data *bvm,
struct bio_vec *biovec)
{
- mddev_t *mddev = q->queuedata;
+ struct mddev *mddev = q->queuedata;
sector_t sector = bvm->bi_sector + get_start_sect(bvm->bi_bdev);
int max;
unsigned int chunk_sectors = mddev->chunk_sectors;
* FIXME: possibly should rethink readbalancing and do it differently
* depending on near_copies / far_copies geometry.
*/
-static int read_balance(conf_t *conf, r10bio_t *r10_bio, int *max_sectors)
+static int read_balance(struct r10conf *conf, struct r10bio *r10_bio, int *max_sectors)
{
const sector_t this_sector = r10_bio->sector;
int disk, slot;
int sectors = r10_bio->sectors;
int best_good_sectors;
sector_t new_distance, best_dist;
- mdk_rdev_t *rdev;
+ struct md_rdev *rdev;
int do_balance;
int best_slot;
static int raid10_congested(void *data, int bits)
{
- mddev_t *mddev = data;
- conf_t *conf = mddev->private;
+ struct mddev *mddev = data;
+ struct r10conf *conf = mddev->private;
int i, ret = 0;
+ if ((bits & (1 << BDI_async_congested)) &&
+ conf->pending_count >= max_queued_requests)
+ return 1;
+
if (mddev_congested(mddev, bits))
return 1;
rcu_read_lock();
for (i = 0; i < conf->raid_disks && ret == 0; i++) {
- mdk_rdev_t *rdev = rcu_dereference(conf->mirrors[i].rdev);
+ struct md_rdev *rdev = rcu_dereference(conf->mirrors[i].rdev);
if (rdev && !test_bit(Faulty, &rdev->flags)) {
struct request_queue *q = bdev_get_queue(rdev->bdev);
return ret;
}
-static void flush_pending_writes(conf_t *conf)
+static void flush_pending_writes(struct r10conf *conf)
{
/* Any writes that have been queued but are awaiting
* bitmap updates get flushed here.
if (conf->pending_bio_list.head) {
struct bio *bio;
bio = bio_list_get(&conf->pending_bio_list);
+ conf->pending_count = 0;
spin_unlock_irq(&conf->device_lock);
/* flush any pending bitmap writes to disk
* before proceeding w/ I/O */
bitmap_unplug(conf->mddev->bitmap);
+ wake_up(&conf->wait_barrier);
while (bio) { /* submit pending writes */
struct bio *next = bio->bi_next;
* lower_barrier when the particular background IO completes.
*/
-static void raise_barrier(conf_t *conf, int force)
+static void raise_barrier(struct r10conf *conf, int force)
{
BUG_ON(force && !conf->barrier);
spin_lock_irq(&conf->resync_lock);
spin_unlock_irq(&conf->resync_lock);
}
-static void lower_barrier(conf_t *conf)
+static void lower_barrier(struct r10conf *conf)
{
unsigned long flags;
spin_lock_irqsave(&conf->resync_lock, flags);
wake_up(&conf->wait_barrier);
}
-static void wait_barrier(conf_t *conf)
+static void wait_barrier(struct r10conf *conf)
{
spin_lock_irq(&conf->resync_lock);
if (conf->barrier) {
conf->nr_waiting++;
- wait_event_lock_irq(conf->wait_barrier, !conf->barrier,
+ /* Wait for the barrier to drop.
+ * However if there are already pending
+ * requests (preventing the barrier from
+ * rising completely), and the
+ * pre-process bio queue isn't empty,
+ * then don't wait, as we need to empty
+ * that queue to get the nr_pending
+ * count down.
+ */
+ wait_event_lock_irq(conf->wait_barrier,
+ !conf->barrier ||
+ (conf->nr_pending &&
+ current->bio_list &&
+ !bio_list_empty(current->bio_list)),
conf->resync_lock,
- );
+ );
conf->nr_waiting--;
}
conf->nr_pending++;
spin_unlock_irq(&conf->resync_lock);
}
-static void allow_barrier(conf_t *conf)
+static void allow_barrier(struct r10conf *conf)
{
unsigned long flags;
spin_lock_irqsave(&conf->resync_lock, flags);
wake_up(&conf->wait_barrier);
}
-static void freeze_array(conf_t *conf)
+static void freeze_array(struct r10conf *conf)
{
/* stop syncio and normal IO and wait for everything to
* go quiet.
spin_unlock_irq(&conf->resync_lock);
}
-static void unfreeze_array(conf_t *conf)
+static void unfreeze_array(struct r10conf *conf)
{
/* reverse the effect of the freeze */
spin_lock_irq(&conf->resync_lock);
spin_unlock_irq(&conf->resync_lock);
}
-static int make_request(mddev_t *mddev, struct bio * bio)
+static void make_request(struct mddev *mddev, struct bio * bio)
{
- conf_t *conf = mddev->private;
- mirror_info_t *mirror;
- r10bio_t *r10_bio;
+ struct r10conf *conf = mddev->private;
+ struct mirror_info *mirror;
+ struct r10bio *r10_bio;
struct bio *read_bio;
int i;
int chunk_sects = conf->chunk_mask + 1;
const unsigned long do_sync = (bio->bi_rw & REQ_SYNC);
const unsigned long do_fua = (bio->bi_rw & REQ_FUA);
unsigned long flags;
- mdk_rdev_t *blocked_rdev;
+ struct md_rdev *blocked_rdev;
int plugged;
int sectors_handled;
int max_sectors;
if (unlikely(bio->bi_rw & REQ_FLUSH)) {
md_flush_request(mddev, bio);
- return 0;
+ return;
}
/* If this request crosses a chunk boundary, we need to
conf->nr_waiting++;
spin_unlock_irq(&conf->resync_lock);
- if (make_request(mddev, &bp->bio1))
- generic_make_request(&bp->bio1);
- if (make_request(mddev, &bp->bio2))
- generic_make_request(&bp->bio2);
+ make_request(mddev, &bp->bio1);
+ make_request(mddev, &bp->bio2);
spin_lock_irq(&conf->resync_lock);
conf->nr_waiting--;
spin_unlock_irq(&conf->resync_lock);
bio_pair_release(bp);
- return 0;
+ return;
bad_map:
printk("md/raid10:%s: make_request bug: can't convert block across chunks"
" or bigger than %dk %llu %d\n", mdname(mddev), chunk_sects/2,
(unsigned long long)bio->bi_sector, bio->bi_size >> 10);
bio_io_error(bio);
- return 0;
+ return;
}
md_write_start(mddev, bio);
slot = r10_bio->read_slot;
if (disk < 0) {
raid_end_bio_io(r10_bio);
- return 0;
+ return;
}
mirror = conf->mirrors + disk;
/* Could not read all from this device, so we will
* need another r10_bio.
*/
- sectors_handled = (r10_bio->sectors + max_sectors
+ sectors_handled = (r10_bio->sector + max_sectors
- bio->bi_sector);
r10_bio->sectors = max_sectors;
spin_lock_irq(&conf->device_lock);
bio->bi_phys_segments = 2;
else
bio->bi_phys_segments++;
- spin_unlock(&conf->device_lock);
+ spin_unlock_irq(&conf->device_lock);
/* Cannot call generic_make_request directly
* as that will be queued in __generic_make_request
* and subsequent mempool_alloc might block
goto read_again;
} else
generic_make_request(read_bio);
- return 0;
+ return;
}
/*
* WRITE:
*/
+ if (conf->pending_count >= max_queued_requests) {
+ md_wakeup_thread(mddev->thread);
+ wait_event(conf->wait_barrier,
+ conf->pending_count < max_queued_requests);
+ }
/* first select target devices under rcu_lock and
* inc refcount on their rdev. Record them by setting
* bios[x] to bio
for (i = 0; i < conf->copies; i++) {
int d = r10_bio->devs[i].devnum;
- mdk_rdev_t *rdev = rcu_dereference(conf->mirrors[d].rdev);
+ struct md_rdev *rdev = rcu_dereference(conf->mirrors[d].rdev);
if (rdev && unlikely(test_bit(Blocked, &rdev->flags))) {
atomic_inc(&rdev->nr_pending);
blocked_rdev = rdev;
atomic_inc(&r10_bio->remaining);
spin_lock_irqsave(&conf->device_lock, flags);
bio_list_add(&conf->pending_bio_list, mbio);
+ conf->pending_count++;
spin_unlock_irqrestore(&conf->device_lock, flags);
}
if (do_sync || !mddev->bitmap || !plugged)
md_wakeup_thread(mddev->thread);
- return 0;
}
-static void status(struct seq_file *seq, mddev_t *mddev)
+static void status(struct seq_file *seq, struct mddev *mddev)
{
- conf_t *conf = mddev->private;
+ struct r10conf *conf = mddev->private;
int i;
if (conf->near_copies < conf->raid_disks)
* Don't consider the device numbered 'ignore'
* as we might be about to remove it.
*/
-static int enough(conf_t *conf, int ignore)
+static int enough(struct r10conf *conf, int ignore)
{
int first = 0;
do {
int n = conf->copies;
int cnt = 0;
+ int this = first;
while (n--) {
- if (conf->mirrors[first].rdev &&
- first != ignore)
+ if (conf->mirrors[this].rdev &&
+ this != ignore)
cnt++;
- first = (first+1) % conf->raid_disks;
+ this = (this+1) % conf->raid_disks;
}
if (cnt == 0)
return 0;
+ first = (first + conf->near_copies) % conf->raid_disks;
} while (first != 0);
return 1;
}
-static void error(mddev_t *mddev, mdk_rdev_t *rdev)
+static void error(struct mddev *mddev, struct md_rdev *rdev)
{
char b[BDEVNAME_SIZE];
- conf_t *conf = mddev->private;
+ struct r10conf *conf = mddev->private;
/*
* If it is not operational, then we have already marked it as dead
spin_lock_irqsave(&conf->device_lock, flags);
mddev->degraded++;
spin_unlock_irqrestore(&conf->device_lock, flags);
- /*
- * if recovery is running, make sure it aborts.
- */
- set_bit(MD_RECOVERY_INTR, &mddev->recovery);
}
+ /*
+ * If recovery is running, make sure it aborts.
+ */
+ set_bit(MD_RECOVERY_INTR, &mddev->recovery);
set_bit(Blocked, &rdev->flags);
set_bit(Faulty, &rdev->flags);
set_bit(MD_CHANGE_DEVS, &mddev->flags);
+ set_bit(MD_CHANGE_PENDING, &mddev->flags);
printk(KERN_ALERT
"md/raid10:%s: Disk failure on %s, disabling device.\n"
"md/raid10:%s: Operation continuing on %d devices.\n",
mdname(mddev), conf->raid_disks - mddev->degraded);
}
-static void print_conf(conf_t *conf)
+static void print_conf(struct r10conf *conf)
{
int i;
- mirror_info_t *tmp;
+ struct mirror_info *tmp;
printk(KERN_DEBUG "RAID10 conf printout:\n");
if (!conf) {
}
}
-static void close_sync(conf_t *conf)
+static void close_sync(struct r10conf *conf)
{
wait_barrier(conf);
allow_barrier(conf);
conf->r10buf_pool = NULL;
}
-static int raid10_spare_active(mddev_t *mddev)
+static int raid10_spare_active(struct mddev *mddev)
{
int i;
- conf_t *conf = mddev->private;
- mirror_info_t *tmp;
+ struct r10conf *conf = mddev->private;
+ struct mirror_info *tmp;
int count = 0;
unsigned long flags;
}
-static int raid10_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
+static int raid10_add_disk(struct mddev *mddev, struct md_rdev *rdev)
{
- conf_t *conf = mddev->private;
+ struct r10conf *conf = mddev->private;
int err = -EEXIST;
int mirror;
int first = 0;
else
mirror = first;
for ( ; mirror <= last ; mirror++) {
- mirror_info_t *p = &conf->mirrors[mirror];
+ struct mirror_info *p = &conf->mirrors[mirror];
if (p->recovery_disabled == mddev->recovery_disabled)
continue;
- if (!p->rdev)
+ if (p->rdev)
continue;
disk_stack_limits(mddev->gendisk, rdev->bdev,
}
p->head_position = 0;
+ p->recovery_disabled = mddev->recovery_disabled - 1;
rdev->raid_disk = mirror;
err = 0;
if (rdev->saved_raid_disk != mirror)
return err;
}
-static int raid10_remove_disk(mddev_t *mddev, int number)
+static int raid10_remove_disk(struct mddev *mddev, int number)
{
- conf_t *conf = mddev->private;
+ struct r10conf *conf = mddev->private;
int err = 0;
- mdk_rdev_t *rdev;
- mirror_info_t *p = conf->mirrors+ number;
+ struct md_rdev *rdev;
+ struct mirror_info *p = conf->mirrors+ number;
print_conf(conf);
rdev = p->rdev;
static void end_sync_read(struct bio *bio, int error)
{
- r10bio_t *r10_bio = bio->bi_private;
- conf_t *conf = r10_bio->mddev->private;
+ struct r10bio *r10_bio = bio->bi_private;
+ struct r10conf *conf = r10_bio->mddev->private;
int d;
d = find_bio_disk(conf, r10_bio, bio, NULL);
}
}
-static void end_sync_request(r10bio_t *r10_bio)
+static void end_sync_request(struct r10bio *r10_bio)
{
- mddev_t *mddev = r10_bio->mddev;
+ struct mddev *mddev = r10_bio->mddev;
while (atomic_dec_and_test(&r10_bio->remaining)) {
if (r10_bio->master_bio == NULL) {
md_done_sync(mddev, s, 1);
break;
} else {
- r10bio_t *r10_bio2 = (r10bio_t *)r10_bio->master_bio;
+ struct r10bio *r10_bio2 = (struct r10bio *)r10_bio->master_bio;
if (test_bit(R10BIO_MadeGood, &r10_bio->state) ||
test_bit(R10BIO_WriteError, &r10_bio->state))
reschedule_retry(r10_bio);
static void end_sync_write(struct bio *bio, int error)
{
int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
- r10bio_t *r10_bio = bio->bi_private;
- mddev_t *mddev = r10_bio->mddev;
- conf_t *conf = mddev->private;
+ struct r10bio *r10_bio = bio->bi_private;
+ struct mddev *mddev = r10_bio->mddev;
+ struct r10conf *conf = mddev->private;
int d;
sector_t first_bad;
int bad_sectors;
* We check if all blocks are in-sync and only write to blocks that
* aren't in sync
*/
-static void sync_request_write(mddev_t *mddev, r10bio_t *r10_bio)
+static void sync_request_write(struct mddev *mddev, struct r10bio *r10_bio)
{
- conf_t *conf = mddev->private;
+ struct r10conf *conf = mddev->private;
int i, first;
struct bio *tbio, *fbio;
* The second for writing.
*
*/
-static void fix_recovery_read_error(r10bio_t *r10_bio)
+static void fix_recovery_read_error(struct r10bio *r10_bio)
{
/* We got a read error during recovery.
* We repeat the read in smaller page-sized sections.
* If a read fails, record a bad block on both old and
* new devices.
*/
- mddev_t *mddev = r10_bio->mddev;
- conf_t *conf = mddev->private;
+ struct mddev *mddev = r10_bio->mddev;
+ struct r10conf *conf = mddev->private;
struct bio *bio = r10_bio->devs[0].bio;
sector_t sect = 0;
int sectors = r10_bio->sectors;
while (sectors) {
int s = sectors;
- mdk_rdev_t *rdev;
+ struct md_rdev *rdev;
sector_t addr;
int ok;
if (rdev != conf->mirrors[dw].rdev) {
/* need bad block on destination too */
- mdk_rdev_t *rdev2 = conf->mirrors[dw].rdev;
+ struct md_rdev *rdev2 = conf->mirrors[dw].rdev;
addr = r10_bio->devs[1].addr + sect;
ok = rdev_set_badblocks(rdev2, addr, s, 0);
if (!ok) {
}
}
-static void recovery_request_write(mddev_t *mddev, r10bio_t *r10_bio)
+static void recovery_request_write(struct mddev *mddev, struct r10bio *r10_bio)
{
- conf_t *conf = mddev->private;
+ struct r10conf *conf = mddev->private;
int d;
struct bio *wbio;
* since the last recorded read error.
*
*/
-static void check_decay_read_errors(mddev_t *mddev, mdk_rdev_t *rdev)
+static void check_decay_read_errors(struct mddev *mddev, struct md_rdev *rdev)
{
struct timespec cur_time_mon;
unsigned long hours_since_last;
atomic_set(&rdev->read_errors, read_errors >> hours_since_last);
}
-static int r10_sync_page_io(mdk_rdev_t *rdev, sector_t sector,
+static int r10_sync_page_io(struct md_rdev *rdev, sector_t sector,
int sectors, struct page *page, int rw)
{
sector_t first_bad;
* 3. Performs writes following reads for array synchronising.
*/
-static void fix_read_error(conf_t *conf, mddev_t *mddev, r10bio_t *r10_bio)
+static void fix_read_error(struct r10conf *conf, struct mddev *mddev, struct r10bio *r10_bio)
{
int sect = 0; /* Offset from r10_bio->sector */
int sectors = r10_bio->sectors;
- mdk_rdev_t*rdev;
+ struct md_rdev*rdev;
int max_read_errors = atomic_read(&mddev->max_corr_read_errors);
int d = r10_bio->devs[r10_bio->read_slot].devnum;
if (r10_sync_page_io(rdev,
r10_bio->devs[sl].addr +
sect,
- s<<9, conf->tmppage, WRITE)
+ s, conf->tmppage, WRITE)
== 0) {
/* Well, this device is dead */
printk(KERN_NOTICE
switch (r10_sync_page_io(rdev,
r10_bio->devs[sl].addr +
sect,
- s<<9, conf->tmppage,
+ s, conf->tmppage,
READ)) {
case 0:
/* Well, this device is dead */
return test_bit(BIO_UPTODATE, &bio->bi_flags);
}
-static int narrow_write_error(r10bio_t *r10_bio, int i)
+static int narrow_write_error(struct r10bio *r10_bio, int i)
{
struct bio *bio = r10_bio->master_bio;
- mddev_t *mddev = r10_bio->mddev;
- conf_t *conf = mddev->private;
- mdk_rdev_t *rdev = conf->mirrors[r10_bio->devs[i].devnum].rdev;
+ struct mddev *mddev = r10_bio->mddev;
+ struct r10conf *conf = mddev->private;
+ struct md_rdev *rdev = conf->mirrors[r10_bio->devs[i].devnum].rdev;
/* bio has the data to be written to slot 'i' where
* we just recently had a write error.
* We repeatedly clone the bio and trim down to one block,
return ok;
}
-static void handle_read_error(mddev_t *mddev, r10bio_t *r10_bio)
+static void handle_read_error(struct mddev *mddev, struct r10bio *r10_bio)
{
int slot = r10_bio->read_slot;
int mirror = r10_bio->devs[slot].devnum;
struct bio *bio;
- conf_t *conf = mddev->private;
- mdk_rdev_t *rdev;
+ struct r10conf *conf = mddev->private;
+ struct md_rdev *rdev;
char b[BDEVNAME_SIZE];
unsigned long do_sync;
int max_sectors;
rdev = conf->mirrors[mirror].rdev;
printk_ratelimited(
KERN_ERR
- "md/raid10:%s: %s: redirecting"
+ "md/raid10:%s: %s: redirecting "
"sector %llu to another mirror\n",
mdname(mddev),
bdevname(rdev->bdev, b),
generic_make_request(bio);
}
-static void handle_write_completed(conf_t *conf, r10bio_t *r10_bio)
+static void handle_write_completed(struct r10conf *conf, struct r10bio *r10_bio)
{
/* Some sort of write request has finished and it
* succeeded in writing where we thought there was a
* a bad block.
*/
int m;
- mdk_rdev_t *rdev;
+ struct md_rdev *rdev;
if (test_bit(R10BIO_IsSync, &r10_bio->state) ||
test_bit(R10BIO_IsRecover, &r10_bio->state)) {
}
put_buf(r10_bio);
} else {
+ bool fail = false;
for (m = 0; m < conf->copies; m++) {
int dev = r10_bio->devs[m].devnum;
struct bio *bio = r10_bio->devs[m].bio;
rdev_dec_pending(rdev, conf->mddev);
} else if (bio != NULL &&
!test_bit(BIO_UPTODATE, &bio->bi_flags)) {
+ fail = true;
if (!narrow_write_error(r10_bio, m)) {
md_error(conf->mddev, rdev);
set_bit(R10BIO_Degraded,
rdev_dec_pending(rdev, conf->mddev);
}
}
- if (test_bit(R10BIO_WriteError,
- &r10_bio->state))
- close_write(r10_bio);
- raid_end_bio_io(r10_bio);
+ if (fail) {
+ spin_lock_irq(&conf->device_lock);
+ list_add(&r10_bio->retry_list, &conf->bio_end_io_list);
+ conf->nr_queued++;
+ spin_unlock_irq(&conf->device_lock);
+ md_wakeup_thread(conf->mddev->thread);
+ } else {
+ if (test_bit(R10BIO_WriteError,
+ &r10_bio->state))
+ close_write(r10_bio);
+ raid_end_bio_io(r10_bio);
+ }
}
}
-static void raid10d(mddev_t *mddev)
+static void raid10d(struct mddev *mddev)
{
- r10bio_t *r10_bio;
+ struct r10bio *r10_bio;
unsigned long flags;
- conf_t *conf = mddev->private;
+ struct r10conf *conf = mddev->private;
struct list_head *head = &conf->retry_list;
struct blk_plug plug;
md_check_recovery(mddev);
+ if (!list_empty_careful(&conf->bio_end_io_list) &&
+ !test_bit(MD_CHANGE_PENDING, &mddev->flags)) {
+ LIST_HEAD(tmp);
+ spin_lock_irqsave(&conf->device_lock, flags);
+ if (!test_bit(MD_CHANGE_PENDING, &mddev->flags)) {
+ while (!list_empty(&conf->bio_end_io_list)) {
+ list_move(conf->bio_end_io_list.prev, &tmp);
+ conf->nr_queued--;
+ }
+ }
+ spin_unlock_irqrestore(&conf->device_lock, flags);
+ while (!list_empty(&tmp)) {
+ r10_bio = list_first_entry(&conf->bio_end_io_list,
+ struct r10bio, retry_list);
+ list_del(&r10_bio->retry_list);
+ if (mddev->degraded)
+ set_bit(R10BIO_Degraded, &r10_bio->state);
+
+ if (test_bit(R10BIO_WriteError,
+ &r10_bio->state))
+ close_write(r10_bio);
+ raid_end_bio_io(r10_bio);
+ }
+ }
+
blk_start_plug(&plug);
for (;;) {
spin_unlock_irqrestore(&conf->device_lock, flags);
break;
}
- r10_bio = list_entry(head->prev, r10bio_t, retry_list);
+ r10_bio = list_entry(head->prev, struct r10bio, retry_list);
list_del(head->prev);
conf->nr_queued--;
spin_unlock_irqrestore(&conf->device_lock, flags);
}
-static int init_resync(conf_t *conf)
+static int init_resync(struct r10conf *conf)
{
int buffs;
*
*/
-static sector_t sync_request(mddev_t *mddev, sector_t sector_nr,
+static sector_t sync_request(struct mddev *mddev, sector_t sector_nr,
int *skipped, int go_faster)
{
- conf_t *conf = mddev->private;
- r10bio_t *r10_bio;
+ struct r10conf *conf = mddev->private;
+ struct r10bio *r10_bio;
struct bio *biolist = NULL, *bio;
sector_t max_sector, nr_sectors;
int i;
for (i=0 ; i<conf->raid_disks; i++) {
int still_degraded;
- r10bio_t *rb2;
+ struct r10bio *rb2;
sector_t sect;
int must_sync;
int any_working;
/* want to reconstruct this device */
rb2 = r10_bio;
sect = raid10_find_virt(conf, sector_nr, i);
+ if (sect >= mddev->resync_max_sectors) {
+ /* last stripe is not complete - don't
+ * try to recover this sector.
+ */
+ continue;
+ }
/* Unless we are doing a full sync, we only need
* to recover the block if it is set in the bitmap
*/
int k;
int d = r10_bio->devs[j].devnum;
sector_t from_addr, to_addr;
- mdk_rdev_t *rdev;
+ struct md_rdev *rdev;
sector_t sector, first_bad;
int bad_sectors;
if (!conf->mirrors[d].rdev ||
if (j == conf->copies) {
/* Cannot recover, so abort the recovery or
* record a bad block */
- put_buf(r10_bio);
- if (rb2)
- atomic_dec(&rb2->remaining);
- r10_bio = rb2;
if (any_working) {
/* problem is that there are bad blocks
* on other device(s)
conf->mirrors[i].recovery_disabled
= mddev->recovery_disabled;
}
+ put_buf(r10_bio);
+ if (rb2)
+ atomic_dec(&rb2->remaining);
+ r10_bio = rb2;
break;
}
}
if (biolist == NULL) {
while (r10_bio) {
- r10bio_t *rb2 = r10_bio;
- r10_bio = (r10bio_t*) rb2->master_bio;
+ struct r10bio *rb2 = r10_bio;
+ r10_bio = (struct r10bio*) rb2->master_bio;
rb2->master_bio = NULL;
put_buf(rb2);
}
else {
bad_sectors -= (sector - first_bad);
if (max_sync > bad_sectors)
- max_sync = max_sync;
+ max_sync = bad_sectors;
continue;
}
}
}
static sector_t
-raid10_size(mddev_t *mddev, sector_t sectors, int raid_disks)
+raid10_size(struct mddev *mddev, sector_t sectors, int raid_disks)
{
sector_t size;
- conf_t *conf = mddev->private;
+ struct r10conf *conf = mddev->private;
if (!raid_disks)
raid_disks = conf->raid_disks;
}
-static conf_t *setup_conf(mddev_t *mddev)
+static struct r10conf *setup_conf(struct mddev *mddev)
{
- conf_t *conf = NULL;
+ struct r10conf *conf = NULL;
int nc, fc, fo;
sector_t stride, size;
int err = -EINVAL;
}
err = -ENOMEM;
- conf = kzalloc(sizeof(conf_t), GFP_KERNEL);
+ conf = kzalloc(sizeof(struct r10conf), GFP_KERNEL);
if (!conf)
goto out;
spin_lock_init(&conf->device_lock);
INIT_LIST_HEAD(&conf->retry_list);
+ INIT_LIST_HEAD(&conf->bio_end_io_list);
spin_lock_init(&conf->resync_lock);
init_waitqueue_head(&conf->wait_barrier);
return ERR_PTR(err);
}
-static int run(mddev_t *mddev)
+static int run(struct mddev *mddev)
{
- conf_t *conf;
+ struct r10conf *conf;
int i, disk_idx, chunk_size;
- mirror_info_t *disk;
- mdk_rdev_t *rdev;
+ struct mirror_info *disk;
+ struct md_rdev *rdev;
sector_t size;
/*
if (disk->rdev)
conf->fullsync = 1;
}
+ disk->recovery_disabled = mddev->recovery_disabled - 1;
}
if (mddev->recovery_cp != MaxSector)
return -EIO;
}
-static int stop(mddev_t *mddev)
+static int stop(struct mddev *mddev)
{
- conf_t *conf = mddev->private;
+ struct r10conf *conf = mddev->private;
raise_barrier(conf, 0);
lower_barrier(conf);
return 0;
}
-static void raid10_quiesce(mddev_t *mddev, int state)
+static void raid10_quiesce(struct mddev *mddev, int state)
{
- conf_t *conf = mddev->private;
+ struct r10conf *conf = mddev->private;
switch(state) {
case 1:
}
}
-static void *raid10_takeover_raid0(mddev_t *mddev)
+static void *raid10_takeover_raid0(struct mddev *mddev)
{
- mdk_rdev_t *rdev;
- conf_t *conf;
+ struct md_rdev *rdev;
+ struct r10conf *conf;
if (mddev->degraded > 0) {
printk(KERN_ERR "md/raid10:%s: Error: degraded raid0!\n",
return conf;
}
-static void *raid10_takeover(mddev_t *mddev)
+static void *raid10_takeover(struct mddev *mddev)
{
- struct raid0_private_data *raid0_priv;
+ struct r0conf *raid0_conf;
/* raid10 can take over:
* raid0 - providing it has only two drives
*/
if (mddev->level == 0) {
/* for raid0 takeover only one zone is supported */
- raid0_priv = mddev->private;
- if (raid0_priv->nr_strip_zones > 1) {
+ raid0_conf = mddev->private;
+ if (raid0_conf->nr_strip_zones > 1) {
printk(KERN_ERR "md/raid10:%s: cannot takeover raid 0"
" with more than one zone.\n",
mdname(mddev));
return ERR_PTR(-EINVAL);
}
-static struct mdk_personality raid10_personality =
+static struct md_personality raid10_personality =
{
.name = "raid10",
.level = 10,
MODULE_ALIAS("md-personality-9"); /* RAID10 */
MODULE_ALIAS("md-raid10");
MODULE_ALIAS("md-level-10");
+
+module_param(max_queued_requests, int, S_IRUGO|S_IWUSR);
git.openpandora.org / pandora-kernel.git / blobdiff