#define STRIPE_SHIFT (PAGE_SHIFT - 9)
#define STRIPE_SECTORS (STRIPE_SIZE>>9)
#define IO_THRESHOLD 1
-#define HASH_PAGES 1
-#define HASH_PAGES_ORDER 0
-#define NR_HASH (HASH_PAGES * PAGE_SIZE / sizeof(struct stripe_head *))
+#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])
+#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
static void print_raid6_conf (raid6_conf_t *conf);
-static inline void __release_stripe(raid6_conf_t *conf, struct stripe_head *sh)
+static void __release_stripe(raid6_conf_t *conf, struct stripe_head *sh)
{
if (atomic_dec_and_test(&sh->count)) {
- if (!list_empty(&sh->lru))
- BUG();
- if (atomic_read(&conf->active_stripes)==0)
- BUG();
+ 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);
list_add_tail(&sh->lru, &conf->inactive_list);
atomic_dec(&conf->active_stripes);
if (!conf->inactive_blocked ||
- atomic_read(&conf->active_stripes) < (NR_STRIPES*3/4))
+ atomic_read(&conf->active_stripes) < (conf->max_nr_stripes*3/4))
wake_up(&conf->wait_for_stripe);
}
}
spin_unlock_irqrestore(&conf->device_lock, flags);
}
-static void remove_hash(struct stripe_head *sh)
+static inline void remove_hash(struct stripe_head *sh)
{
PRINTK("remove_hash(), stripe %llu\n", (unsigned long long)sh->sector);
- if (sh->hash_pprev) {
- if (sh->hash_next)
- sh->hash_next->hash_pprev = sh->hash_pprev;
- *sh->hash_pprev = sh->hash_next;
- sh->hash_pprev = NULL;
- }
+ hlist_del_init(&sh->hash);
}
-static __inline__ void insert_hash(raid6_conf_t *conf, struct stripe_head *sh)
+static inline void insert_hash(raid6_conf_t *conf, struct stripe_head *sh)
{
- struct stripe_head **shp = &stripe_hash(conf, sh->sector);
+ struct hlist_head *hp = stripe_hash(conf, sh->sector);
PRINTK("insert_hash(), stripe %llu\n", (unsigned long long)sh->sector);
CHECK_DEVLOCK();
- if ((sh->hash_next = *shp) != NULL)
- (*shp)->hash_pprev = &sh->hash_next;
- *shp = sh;
- sh->hash_pprev = shp;
+ hlist_add_head(&sh->hash, hp);
}
if (!p)
continue;
sh->dev[i].page = NULL;
- page_cache_release(p);
+ put_page(p);
}
}
static void raid6_build_block (struct stripe_head *sh, int i);
-static inline void init_stripe(struct stripe_head *sh, sector_t sector, int pd_idx)
+static void init_stripe(struct stripe_head *sh, sector_t sector, int pd_idx)
{
raid6_conf_t *conf = sh->raid_conf;
int disks = conf->raid_disks, i;
- if (atomic_read(&sh->count) != 0)
- BUG();
- if (test_bit(STRIPE_HANDLE, &sh->state))
- BUG();
+ BUG_ON(atomic_read(&sh->count) != 0);
+ BUG_ON(test_bit(STRIPE_HANDLE, &sh->state));
CHECK_DEVLOCK();
PRINTK("init_stripe called, stripe %llu\n",
static struct stripe_head *__find_stripe(raid6_conf_t *conf, sector_t sector)
{
struct stripe_head *sh;
+ struct hlist_node *hn;
CHECK_DEVLOCK();
PRINTK("__find_stripe, sector %llu\n", (unsigned long long)sector);
- for (sh = stripe_hash(conf, sector); sh; sh = sh->hash_next)
+ hlist_for_each_entry (sh, hn, stripe_hash(conf, sector), hash)
if (sh->sector == sector)
return sh;
PRINTK("__stripe %llu not in cache\n", (unsigned long long)sector);
conf->inactive_blocked = 1;
wait_event_lock_irq(conf->wait_for_stripe,
!list_empty(&conf->inactive_list) &&
- (atomic_read(&conf->active_stripes) < (NR_STRIPES *3/4)
+ (atomic_read(&conf->active_stripes)
+ < (conf->max_nr_stripes *3/4)
|| !conf->inactive_blocked),
conf->device_lock,
unplug_slaves(conf->mddev);
init_stripe(sh, sector, pd_idx);
} else {
if (atomic_read(&sh->count)) {
- if (!list_empty(&sh->lru))
- BUG();
+ BUG_ON(!list_empty(&sh->lru));
} else {
if (!test_bit(STRIPE_HANDLE, &sh->state))
atomic_inc(&conf->active_stripes);
- if (list_empty(&sh->lru))
- BUG();
+ BUG_ON(list_empty(&sh->lru));
list_del_init(&sh->lru);
}
}
return sh;
}
-static int grow_stripes(raid6_conf_t *conf, int num)
+static int grow_one_stripe(raid6_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;
+ }
+ /* 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(raid6_conf_t *conf, int num)
+{
kmem_cache_t *sc;
int devs = conf->raid_disks;
- sprintf(conf->cache_name, "raid6/%s", mdname(conf->mddev));
+ sprintf(conf->cache_name[0], "raid6/%s", mdname(conf->mddev));
- sc = kmem_cache_create(conf->cache_name,
+ sc = kmem_cache_create(conf->cache_name[0],
sizeof(struct stripe_head)+(devs-1)*sizeof(struct r5dev),
0, 0, NULL, NULL);
if (!sc)
return 1;
conf->slab_cache = sc;
- while (num--) {
- sh = kmem_cache_alloc(sc, GFP_KERNEL);
- if (!sh)
- return 1;
- memset(sh, 0, sizeof(*sh) + (devs-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(sc, sh);
+ while (num--)
+ if (!grow_one_stripe(conf))
return 1;
- }
- /* 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 0;
}
-static void shrink_stripes(raid6_conf_t *conf)
+static int drop_one_stripe(raid6_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->raid_disks);
+ kmem_cache_free(conf->slab_cache, sh);
+ atomic_dec(&conf->active_stripes);
+ return 1;
+}
- while (1) {
- spin_lock_irq(&conf->device_lock);
- sh = get_free_stripe(conf);
- spin_unlock_irq(&conf->device_lock);
- if (!sh)
- break;
- if (atomic_read(&sh->count))
- BUG();
- shrink_buffers(sh, conf->raid_disks);
- kmem_cache_free(conf->slab_cache, sh);
- atomic_dec(&conf->active_stripes);
- }
- kmem_cache_destroy(conf->slab_cache);
+static void shrink_stripes(raid6_conf_t *conf)
+{
+ while (drop_one_stripe(conf))
+ ;
+
+ if (conf->slab_cache)
+ kmem_cache_destroy(conf->slab_cache);
conf->slab_cache = NULL;
}
-static int raid6_end_read_request (struct bio * bi, unsigned int bytes_done,
- int error)
+static int raid6_end_read_request(struct bio * bi, unsigned int bytes_done,
+ int error)
{
struct stripe_head *sh = bi->bi_private;
raid6_conf_t *conf = sh->raid_conf;
#else
set_bit(R5_UPTODATE, &sh->dev[i].flags);
#endif
+ if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
+ printk(KERN_INFO "raid6: read error corrected!!\n");
+ 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 {
- md_error(conf->mddev, conf->disks[i].rdev);
+ int retry = 0;
clear_bit(R5_UPTODATE, &sh->dev[i].flags);
+ atomic_inc(&conf->disks[i].rdev->read_errors);
+ if (conf->mddev->degraded)
+ printk(KERN_WARNING "raid6: read error not correctable.\n");
+ else if (test_bit(R5_ReWrite, &sh->dev[i].flags))
+ /* Oh, no!!! */
+ printk(KERN_WARNING "raid6: read error NOT corrected!!\n");
+ else if (atomic_read(&conf->disks[i].rdev->read_errors)
+ > conf->max_nr_stripes)
+ printk(KERN_WARNING
+ "raid6: Too many read errors, failing device.\n");
+ 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, conf->disks[i].rdev);
+ }
}
rdev_dec_pending(conf->disks[i].rdev, conf->mddev);
#if 0
raid6_conf_t *conf = (raid6_conf_t *) mddev->private;
PRINTK("raid6: error called\n");
- if (!rdev->faulty) {
+ if (!test_bit(Faulty, &rdev->flags)) {
mddev->sb_dirty = 1;
- if (rdev->in_sync) {
+ if (test_bit(In_sync, &rdev->flags)) {
conf->working_disks--;
mddev->degraded++;
conf->failed_disks++;
- rdev->in_sync = 0;
+ clear_bit(In_sync, &rdev->flags);
/*
* if recovery was running, make sure it aborts.
*/
set_bit(MD_RECOVERY_ERR, &mddev->recovery);
}
- rdev->faulty = 1;
+ set_bit(Faulty, &rdev->flags);
printk (KERN_ALERT
"raid6: Disk failure on %s, disabling device."
" Operation continuing on %d devices\n",
if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
wake_up(&conf->wait_for_overlap);
- if (sh->dev[i].written) BUG();
+ BUG_ON(sh->dev[i].written);
sh->dev[i].written = chosen;
}
break;
}
/* Compute one missing block */
-static void compute_block_1(struct stripe_head *sh, int dd_idx)
+static void compute_block_1(struct stripe_head *sh, int dd_idx, int nozero)
{
raid6_conf_t *conf = sh->raid_conf;
int i, count, disks = conf->raid_disks;
compute_parity(sh, UPDATE_PARITY);
} else {
ptr[0] = page_address(sh->dev[dd_idx].page);
- memset(ptr[0], 0, STRIPE_SIZE);
+ if (!nozero) memset(ptr[0], 0, STRIPE_SIZE);
count = 1;
for (i = disks ; i--; ) {
if (i == dd_idx || i == qd_idx)
}
if (count != 1)
xor_block(count, STRIPE_SIZE, ptr);
- set_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
+ if (!nozero) set_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
+ else clear_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
}
}
return;
} else {
/* We're missing D+Q; recompute D from P */
- compute_block_1(sh, (dd_idx1 == qd_idx) ? dd_idx2 : dd_idx1);
+ compute_block_1(sh, (dd_idx1 == qd_idx) ? dd_idx2 : dd_idx1, 0);
compute_parity(sh, UPDATE_PARITY); /* Is this necessary? */
return;
}
if (*bip && (*bip)->bi_sector < bi->bi_sector + ((bi->bi_size)>>9))
goto overlap;
- if (*bip && bi->bi_next && (*bip) != bi->bi_next)
- BUG();
+ BUG_ON(*bip && bi->bi_next && (*bip) != bi->bi_next);
if (*bip)
bi->bi_next = *bip;
*bip = bi;
}
+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);
+}
/*
* handle_stripe - do things to a stripe.
*
*
*/
-static void handle_stripe(struct stripe_head *sh)
+static void handle_stripe(struct stripe_head *sh, struct page *tmp_page)
{
raid6_conf_t *conf = sh->raid_conf;
int disks = conf->raid_disks;
syncing = test_bit(STRIPE_SYNCING, &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);
- clear_bit(R5_Syncio, &dev->flags);
PRINTK("check %d: state 0x%lx read %p write %p written %p\n",
i, dev->flags, dev->toread, dev->towrite, dev->written);
non_overwrite++;
}
if (dev->written) written++;
- rdev = conf->disks[i].rdev; /* FIXME, should I be looking rdev */
- if (!rdev || !rdev->in_sync) {
+ rdev = rcu_dereference(conf->disks[i].rdev);
+ 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 ( failed < 2 )
failed_num[failed] = i;
failed++;
} else
set_bit(R5_Insync, &dev->flags);
}
+ rcu_read_unlock();
PRINTK("locked=%d uptodate=%d to_read=%d"
" to_write=%d failed=%d failed_num=%d,%d\n",
locked, uptodate, to_read, to_write, failed,
if (failed > 2 && to_read+to_write+written) {
for (i=disks; i--; ) {
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;
}
/* fail any reads if this device is non-operational */
- if (!test_bit(R5_Insync, &sh->dev[i].flags)) {
+ if (!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))
if (uptodate == disks-1) {
PRINTK("Computing stripe %llu block %d\n",
(unsigned long long)sh->sector, i);
- compute_block_1(sh, i);
+ compute_block_1(sh, i, 0);
uptodate++;
} else if ( uptodate == disks-2 && failed >= 2 ) {
/* Computing 2-failure is *very* expensive; only do it if failed >= 2 */
locked++;
PRINTK("Reading block %d (sync=%d)\n",
i, syncing);
- if (syncing)
- md_sync_acct(conf->disks[i].rdev->bdev,
- STRIPE_SECTORS);
}
}
}
/* We have failed blocks and need to compute them */
switch ( failed ) {
case 0: BUG();
- case 1: compute_block_1(sh, failed_num[0]); break;
+ case 1: compute_block_1(sh, failed_num[0], 0); break;
case 2: compute_block_2(sh, failed_num[0], failed_num[1]); break;
default: BUG(); /* This request should have been failed? */
}
(unsigned long long)sh->sector, i);
locked++;
set_bit(R5_Wantwrite, &sh->dev[i].flags);
-#if 0 /**** FIX: I don't understand the logic here... ****/
- if (!test_bit(R5_Insync, &sh->dev[i].flags)
- || ((i==pd_idx || i==qd_idx) && failed == 0)) /* FIX? */
- set_bit(STRIPE_INSYNC, &sh->state);
-#endif
}
+ /* 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)
* Any reads will already have been scheduled, so we just see if enough data
* is available
*/
- if (syncing && locked == 0 &&
- !test_bit(STRIPE_INSYNC, &sh->state) && failed <= 2) {
+ if (syncing && locked == 0 && !test_bit(STRIPE_INSYNC, &sh->state)) {
+ int update_p = 0, update_q = 0;
+ struct r5dev *dev;
+
set_bit(STRIPE_HANDLE, &sh->state);
-#if 0 /* RAID-6: Don't support CHECK PARITY yet */
- if (failed == 0) {
- char *pagea;
- if (uptodate != disks)
- BUG();
- compute_parity(sh, CHECK_PARITY);
- uptodate--;
- pagea = page_address(sh->dev[pd_idx].page);
- if ((*(u32*)pagea) == 0 &&
- !memcmp(pagea, pagea+4, STRIPE_SIZE-4)) {
- /* parity is correct (on disc, not in buffer any more) */
- set_bit(STRIPE_INSYNC, &sh->state);
- }
- }
-#endif
- if (!test_bit(STRIPE_INSYNC, &sh->state)) {
- int failed_needupdate[2];
- struct r5dev *adev, *bdev;
- if ( failed < 1 )
- failed_num[0] = pd_idx;
- if ( failed < 2 )
- failed_num[1] = (failed_num[0] == qd_idx) ? pd_idx : qd_idx;
-
- failed_needupdate[0] = !test_bit(R5_UPTODATE, &sh->dev[failed_num[0]].flags);
- failed_needupdate[1] = !test_bit(R5_UPTODATE, &sh->dev[failed_num[1]].flags);
-
- PRINTK("sync: failed=%d num=%d,%d fnu=%u%u\n",
- failed, failed_num[0], failed_num[1], failed_needupdate[0], failed_needupdate[1]);
-
-#if 0 /* RAID-6: This code seems to require that CHECK_PARITY destroys the uptodateness of the parity */
- /* should be able to compute the missing block(s) and write to spare */
- if ( failed_needupdate[0] ^ failed_needupdate[1] ) {
- if (uptodate+1 != disks)
- BUG();
- compute_block_1(sh, failed_needupdate[0] ? failed_num[0] : failed_num[1]);
- uptodate++;
- } else if ( failed_needupdate[0] & failed_needupdate[1] ) {
- if (uptodate+2 != disks)
- BUG();
- compute_block_2(sh, failed_num[0], failed_num[1]);
- uptodate += 2;
+ BUG_ON(failed>2);
+ BUG_ON(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 (failed == 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 (!q_failed && 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_parity(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;
}
-#else
- compute_block_2(sh, failed_num[0], failed_num[1]);
- uptodate += failed_needupdate[0] + failed_needupdate[1];
-#endif
- if (uptodate != disks)
- BUG();
+ /* now write out any block on a failed drive,
+ * or P or Q if they need it
+ */
- PRINTK("Marking for sync stripe %llu blocks %d,%d\n",
- (unsigned long long)sh->sector, failed_num[0], failed_num[1]);
+ if (failed == 2) {
+ dev = &sh->dev[failed_num[1]];
+ locked++;
+ set_bit(R5_LOCKED, &dev->flags);
+ set_bit(R5_Wantwrite, &dev->flags);
+ }
+ if (failed >= 1) {
+ dev = &sh->dev[failed_num[0]];
+ locked++;
+ set_bit(R5_LOCKED, &dev->flags);
+ set_bit(R5_Wantwrite, &dev->flags);
+ }
- /**** FIX: Should we really do both of these unconditionally? ****/
- adev = &sh->dev[failed_num[0]];
- locked += !test_bit(R5_LOCKED, &adev->flags);
- set_bit(R5_LOCKED, &adev->flags);
- set_bit(R5_Wantwrite, &adev->flags);
- bdev = &sh->dev[failed_num[1]];
- locked += !test_bit(R5_LOCKED, &bdev->flags);
- set_bit(R5_LOCKED, &bdev->flags);
+ if (update_p) {
+ dev = &sh->dev[pd_idx];
+ locked ++;
+ set_bit(R5_LOCKED, &dev->flags);
+ set_bit(R5_Wantwrite, &dev->flags);
+ }
+ if (update_q) {
+ dev = &sh->dev[qd_idx];
+ locked++;
+ set_bit(R5_LOCKED, &dev->flags);
+ set_bit(R5_Wantwrite, &dev->flags);
+ }
clear_bit(STRIPE_DEGRADED, &sh->state);
- set_bit(R5_Wantwrite, &bdev->flags);
set_bit(STRIPE_INSYNC, &sh->state);
- set_bit(R5_Syncio, &adev->flags);
- set_bit(R5_Syncio, &bdev->flags);
}
}
+
if (syncing && 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 (failed <= 2 && ! conf->mddev->ro)
+ for (i=0; i<failed;i++) {
+ dev = &sh->dev[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);
+ }
+ }
+ }
spin_unlock(&sh->lock);
while ((bi=return_bi)) {
rcu_read_lock();
rdev = rcu_dereference(conf->disks[i].rdev);
- if (rdev && rdev->faulty)
+ if (rdev && test_bit(Faulty, &rdev->flags))
rdev = NULL;
if (rdev)
atomic_inc(&rdev->nr_pending);
rcu_read_unlock();
if (rdev) {
- if (test_bit(R5_Syncio, &sh->dev[i].flags))
+ if (syncing)
md_sync_acct(rdev->bdev, STRIPE_SECTORS);
bi->bi_bdev = rdev->bdev;
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 == 1)
}
}
-static inline void raid6_activate_delayed(raid6_conf_t *conf)
+static void raid6_activate_delayed(raid6_conf_t *conf)
{
if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD) {
while (!list_empty(&conf->delayed_list)) {
}
}
-static inline void activate_bit_delay(raid6_conf_t *conf)
+static void activate_bit_delay(raid6_conf_t *conf)
{
/* device_lock is held */
struct list_head head;
rcu_read_lock();
for (i=0; i<mddev->raid_disks; i++) {
mdk_rdev_t *rdev = rcu_dereference(conf->disks[i].rdev);
- if (rdev && !rdev->faulty && atomic_read(&rdev->nr_pending)) {
+ if (rdev && !test_bit(Faulty, &rdev->flags) && atomic_read(&rdev->nr_pending)) {
request_queue_t *r_queue = bdev_get_queue(rdev->bdev);
atomic_inc(&rdev->nr_pending);
rcu_read_lock();
for (i=0; i<mddev->raid_disks && ret == 0; i++) {
mdk_rdev_t *rdev = rcu_dereference(conf->disks[i].rdev);
- if (rdev && !rdev->faulty) {
+ if (rdev && !test_bit(Faulty, &rdev->flags)) {
struct block_device *bdev = rdev->bdev;
request_queue_t *r_queue = bdev_get_queue(bdev);
}
finish_wait(&conf->wait_for_overlap, &w);
raid6_plug_device(conf);
- handle_stripe(sh);
+ handle_stripe(sh, NULL);
release_stripe(sh);
} else {
/* cannot get stripe for read-ahead, just give-up */
int data_disks = raid_disks - 2;
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 */
if (mddev->curr_resync < max_sector) /* aborted */
bitmap_end_sync(mddev->bitmap, mddev->curr_resync,
&sync_blocks, 1);
- else /* compelted sync */
+ else /* completed sync */
conf->fullsync = 0;
bitmap_close_sync(mddev->bitmap);
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;
*/
schedule_timeout_uninterruptible(1);
}
- bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, 0);
+ /* 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);
- handle_stripe(sh);
+ handle_stripe(sh, NULL);
release_stripe(sh);
return STRIPE_SECTORS;
if (conf->seq_flush - conf->seq_write > 0) {
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);
}
list_del_init(first);
atomic_inc(&sh->count);
- if (atomic_read(&sh->count)!= 1)
- BUG();
+ BUG_ON(atomic_read(&sh->count)!= 1);
spin_unlock_irq(&conf->device_lock);
handled++;
- handle_stripe(sh);
+ handle_stripe(sh, conf->spare_page);
release_stripe(sh);
spin_lock_irq(&conf->device_lock);
PRINTK("--- raid6d inactive\n");
}
+static ssize_t
+raid6_show_stripe_cache_size(mddev_t *mddev, char *page)
+{
+ raid6_conf_t *conf = mddev_to_conf(mddev);
+ if (conf)
+ return sprintf(page, "%d\n", conf->max_nr_stripes);
+ else
+ return 0;
+}
+
+static ssize_t
+raid6_store_stripe_cache_size(mddev_t *mddev, const char *page, size_t len)
+{
+ raid6_conf_t *conf = mddev_to_conf(mddev);
+ char *end;
+ int new;
+ if (len >= PAGE_SIZE)
+ return -EINVAL;
+ if (!conf)
+ return -ENODEV;
+
+ new = simple_strtoul(page, &end, 10);
+ if (!*page || (*end && *end != '\n') )
+ 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;
+ }
+ while (new > conf->max_nr_stripes) {
+ if (grow_one_stripe(conf))
+ conf->max_nr_stripes++;
+ else break;
+ }
+ return len;
+}
+
+static struct md_sysfs_entry
+raid6_stripecache_size = __ATTR(stripe_cache_size, S_IRUGO | S_IWUSR,
+ raid6_show_stripe_cache_size,
+ raid6_store_stripe_cache_size);
+
+static ssize_t
+stripe_cache_active_show(mddev_t *mddev, char *page)
+{
+ raid6_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
+raid6_stripecache_active = __ATTR_RO(stripe_cache_active);
+
+static struct attribute *raid6_attrs[] = {
+ &raid6_stripecache_size.attr,
+ &raid6_stripecache_active.attr,
+ NULL,
+};
+static struct attribute_group raid6_attrs_group = {
+ .name = NULL,
+ .attrs = raid6_attrs,
+};
+
static int run(mddev_t *mddev)
{
raid6_conf_t *conf;
return -EIO;
}
- mddev->private = kmalloc (sizeof (raid6_conf_t)
- + mddev->raid_disks * sizeof(struct disk_info),
- GFP_KERNEL);
+ mddev->private = kzalloc(sizeof (raid6_conf_t), GFP_KERNEL);
if ((conf = mddev->private) == NULL)
goto abort;
- memset (conf, 0, sizeof (*conf) + mddev->raid_disks * sizeof(struct disk_info) );
+ conf->disks = kzalloc(mddev->raid_disks * sizeof(struct disk_info),
+ GFP_KERNEL);
+ if (!conf->disks)
+ goto abort;
+
conf->mddev = mddev;
- if ((conf->stripe_hashtbl = (struct stripe_head **) __get_free_pages(GFP_ATOMIC, HASH_PAGES_ORDER)) == NULL)
+ if ((conf->stripe_hashtbl = kzalloc(PAGE_SIZE, GFP_KERNEL)) == NULL)
+ goto abort;
+
+ conf->spare_page = alloc_page(GFP_KERNEL);
+ if (!conf->spare_page)
goto abort;
- memset(conf->stripe_hashtbl, 0, HASH_PAGES * PAGE_SIZE);
spin_lock_init(&conf->device_lock);
init_waitqueue_head(&conf->wait_for_stripe);
disk->rdev = rdev;
- if (rdev->in_sync) {
+ if (test_bit(In_sync, &rdev->flags)) {
char b[BDEVNAME_SIZE];
printk(KERN_INFO "raid6: device %s operational as raid"
" disk %d\n", bdevname(rdev->bdev,b),
goto abort;
}
-#if 0 /* FIX: For now */
if (mddev->degraded > 0 &&
mddev->recovery_cp != MaxSector) {
- printk(KERN_ERR "raid6: cannot start dirty degraded array for %s\n", mdname(mddev));
- goto abort;
+ if (mddev->ok_start_degraded)
+ printk(KERN_WARNING "raid6: starting dirty degraded array:%s"
+ "- data corruption possible.\n",
+ mdname(mddev));
+ else {
+ printk(KERN_ERR "raid6: cannot start dirty degraded array"
+ " for %s\n", mdname(mddev));
+ goto abort;
+ }
}
-#endif
{
mddev->thread = md_register_thread(raid6d, mddev, "%s_raid6");
* 2 * (n-2) * chunksize where 'n' is the number of raid devices
*/
{
- int stripe = (mddev->raid_disks-2) * mddev->chunk_size
- / PAGE_CACHE_SIZE;
+ int stripe = (mddev->raid_disks-2) *
+ (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 */
- mddev->array_size = mddev->size * (mddev->raid_disks - 2);
+ sysfs_create_group(&mddev->kobj, &raid6_attrs_group);
- if (mddev->bitmap)
- mddev->thread->timeout = mddev->bitmap->daemon_sleep * HZ;
+ mddev->array_size = mddev->size * (mddev->raid_disks - 2);
mddev->queue->unplug_fn = raid6_unplug_device;
mddev->queue->issue_flush_fn = raid6_issue_flush;
abort:
if (conf) {
print_raid6_conf(conf);
- if (conf->stripe_hashtbl)
- free_pages((unsigned long) conf->stripe_hashtbl,
- HASH_PAGES_ORDER);
+ safe_put_page(conf->spare_page);
+ kfree(conf->stripe_hashtbl);
+ kfree(conf->disks);
kfree(conf);
}
mddev->private = NULL;
md_unregister_thread(mddev->thread);
mddev->thread = NULL;
shrink_stripes(conf);
- free_pages((unsigned long) conf->stripe_hashtbl, HASH_PAGES_ORDER);
+ kfree(conf->stripe_hashtbl);
blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
+ sysfs_remove_group(&mddev->kobj, &raid6_attrs_group);
kfree(conf);
mddev->private = NULL;
return 0;
static void printall (struct seq_file *seq, raid6_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++) {
sh = conf->stripe_hashtbl[i];
- for (; sh; sh = sh->hash_next) {
+ hlist_for_each_entry(sh, hn, &conf->stripe_hashtbl[i], hash) {
if (sh->raid_conf != conf)
continue;
print_sh(seq, sh);
for (i = 0; i < conf->raid_disks; i++)
seq_printf (seq, "%s",
conf->disks[i].rdev &&
- conf->disks[i].rdev->in_sync ? "U" : "_");
+ test_bit(In_sync, &conf->disks[i].rdev->flags) ? "U" : "_");
seq_printf (seq, "]");
#if RAID6_DUMPSTATE
seq_printf (seq, "\n");
tmp = conf->disks + i;
if (tmp->rdev)
printk(" disk %d, o:%d, dev:%s\n",
- i, !tmp->rdev->faulty,
+ i, !test_bit(Faulty, &tmp->rdev->flags),
bdevname(tmp->rdev->bdev,b));
}
}
for (i = 0; i < conf->raid_disks; i++) {
tmp = conf->disks + i;
if (tmp->rdev
- && !tmp->rdev->faulty
- && !tmp->rdev->in_sync) {
+ && !test_bit(Faulty, &tmp->rdev->flags)
+ && !test_bit(In_sync, &tmp->rdev->flags)) {
mddev->degraded--;
conf->failed_disks--;
conf->working_disks++;
- tmp->rdev->in_sync = 1;
+ set_bit(In_sync, &tmp->rdev->flags);
}
}
print_raid6_conf(conf);
print_raid6_conf(conf);
rdev = p->rdev;
if (rdev) {
- if (rdev->in_sync ||
+ if (test_bit(In_sync, &rdev->flags) ||
atomic_read(&rdev->nr_pending)) {
err = -EBUSY;
goto abort;
/* no point adding a device */
return 0;
/*
- * find the disk ...
+ * find the disk ... but prefer rdev->saved_raid_disk
+ * if possible.
*/
- for (disk=0; disk < mddev->raid_disks; disk++)
+ if (rdev->saved_raid_disk >= 0 &&
+ conf->disks[rdev->saved_raid_disk].rdev == NULL)
+ disk = rdev->saved_raid_disk;
+ else
+ disk = 0;
+ for ( ; disk < mddev->raid_disks; disk++)
if ((p=conf->disks + disk)->rdev == NULL) {
- rdev->in_sync = 0;
+ clear_bit(In_sync, &rdev->flags);
rdev->raid_disk = disk;
found = 1;
if (rdev->saved_raid_disk != disk)
spin_unlock_irq(&conf->device_lock);
break;
}
- if (mddev->thread) {
- if (mddev->bitmap)
- mddev->thread->timeout = mddev->bitmap->daemon_sleep * HZ;
- else
- mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
- md_wakeup_thread(mddev->thread);
- }
}
-static mdk_personality_t raid6_personality=
+
+static struct mdk_personality raid6_personality =
{
.name = "raid6",
+ .level = 6,
.owner = THIS_MODULE,
.make_request = make_request,
.run = run,
.quiesce = raid6_quiesce,
};
-static int __init raid6_init (void)
+static int __init raid6_init(void)
{
int e;
if ( e )
return e;
- return register_md_personality (RAID6, &raid6_personality);
+ return register_md_personality(&raid6_personality);
}
static void raid6_exit (void)
{
- unregister_md_personality (RAID6);
+ unregister_md_personality(&raid6_personality);
}
module_init(raid6_init);
module_exit(raid6_exit);
MODULE_LICENSE("GPL");
MODULE_ALIAS("md-personality-8"); /* RAID6 */
+MODULE_ALIAS("md-raid6");
+MODULE_ALIAS("md-level-6");