Commit 0eb3ff12 authored by NeilBrown's avatar NeilBrown Committed by Linus Torvalds

[PATCH] md: raid10 read-error handling - resync and read-only

Add in correct read-error handling for resync and read-only situations.

When read-only, we don't over-write, so we need to mark the failed drive in
the r10_bio so we don't re-try it.  During resync, we always read all blocks,
so if there is a read error, we simply over-write it with the good block that
we found (assuming we found one).

Note that the recovery case still isn't handled in an interesting way.  There
is nothing useful to do for the 2-copies case.  If there are 3 or more copies,
then we could try reading from one of the non-missing copies, but this is a
bit complicated and very rarely would be used, so I'm leaving it for now.
Signed-off-by: default avatarNeil Brown <neilb@suse.de>
Signed-off-by: default avatarAndrew Morton <akpm@osdl.org>
Signed-off-by: default avatarLinus Torvalds <torvalds@osdl.org>
parent 4443ae10
...@@ -172,7 +172,7 @@ static void put_all_bios(conf_t *conf, r10bio_t *r10_bio) ...@@ -172,7 +172,7 @@ static void put_all_bios(conf_t *conf, r10bio_t *r10_bio)
for (i = 0; i < conf->copies; i++) { for (i = 0; i < conf->copies; i++) {
struct bio **bio = & r10_bio->devs[i].bio; struct bio **bio = & r10_bio->devs[i].bio;
if (*bio) if (*bio && *bio != IO_BLOCKED)
bio_put(*bio); bio_put(*bio);
*bio = NULL; *bio = NULL;
} }
...@@ -500,6 +500,7 @@ static int read_balance(conf_t *conf, r10bio_t *r10_bio) ...@@ -500,6 +500,7 @@ static int read_balance(conf_t *conf, r10bio_t *r10_bio)
disk = r10_bio->devs[slot].devnum; disk = r10_bio->devs[slot].devnum;
while ((rdev = rcu_dereference(conf->mirrors[disk].rdev)) == NULL || while ((rdev = rcu_dereference(conf->mirrors[disk].rdev)) == NULL ||
r10_bio->devs[slot].bio == IO_BLOCKED ||
!test_bit(In_sync, &rdev->flags)) { !test_bit(In_sync, &rdev->flags)) {
slot++; slot++;
if (slot == conf->copies) { if (slot == conf->copies) {
...@@ -517,6 +518,7 @@ static int read_balance(conf_t *conf, r10bio_t *r10_bio) ...@@ -517,6 +518,7 @@ static int read_balance(conf_t *conf, r10bio_t *r10_bio)
slot = 0; slot = 0;
disk = r10_bio->devs[slot].devnum; disk = r10_bio->devs[slot].devnum;
while ((rdev=rcu_dereference(conf->mirrors[disk].rdev)) == NULL || while ((rdev=rcu_dereference(conf->mirrors[disk].rdev)) == NULL ||
r10_bio->devs[slot].bio == IO_BLOCKED ||
!test_bit(In_sync, &rdev->flags)) { !test_bit(In_sync, &rdev->flags)) {
slot ++; slot ++;
if (slot == conf->copies) { if (slot == conf->copies) {
...@@ -537,6 +539,7 @@ static int read_balance(conf_t *conf, r10bio_t *r10_bio) ...@@ -537,6 +539,7 @@ static int read_balance(conf_t *conf, r10bio_t *r10_bio)
if ((rdev=rcu_dereference(conf->mirrors[ndisk].rdev)) == NULL || if ((rdev=rcu_dereference(conf->mirrors[ndisk].rdev)) == NULL ||
r10_bio->devs[nslot].bio == IO_BLOCKED ||
!test_bit(In_sync, &rdev->flags)) !test_bit(In_sync, &rdev->flags))
continue; continue;
...@@ -1104,7 +1107,6 @@ static int raid10_remove_disk(mddev_t *mddev, int number) ...@@ -1104,7 +1107,6 @@ static int raid10_remove_disk(mddev_t *mddev, int number)
static int end_sync_read(struct bio *bio, unsigned int bytes_done, int error) static int end_sync_read(struct bio *bio, unsigned int bytes_done, int error)
{ {
int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
r10bio_t * r10_bio = (r10bio_t *)(bio->bi_private); r10bio_t * r10_bio = (r10bio_t *)(bio->bi_private);
conf_t *conf = mddev_to_conf(r10_bio->mddev); conf_t *conf = mddev_to_conf(r10_bio->mddev);
int i,d; int i,d;
...@@ -1119,7 +1121,10 @@ static int end_sync_read(struct bio *bio, unsigned int bytes_done, int error) ...@@ -1119,7 +1121,10 @@ static int end_sync_read(struct bio *bio, unsigned int bytes_done, int error)
BUG(); BUG();
update_head_pos(i, r10_bio); update_head_pos(i, r10_bio);
d = r10_bio->devs[i].devnum; d = r10_bio->devs[i].devnum;
if (!uptodate)
if (test_bit(BIO_UPTODATE, &bio->bi_flags))
set_bit(R10BIO_Uptodate, &r10_bio->state);
else if (!test_bit(MD_RECOVERY_SYNC, &conf->mddev->recovery))
md_error(r10_bio->mddev, md_error(r10_bio->mddev,
conf->mirrors[d].rdev); conf->mirrors[d].rdev);
...@@ -1209,25 +1214,30 @@ static void sync_request_write(mddev_t *mddev, r10bio_t *r10_bio) ...@@ -1209,25 +1214,30 @@ static void sync_request_write(mddev_t *mddev, r10bio_t *r10_bio)
fbio = r10_bio->devs[i].bio; fbio = r10_bio->devs[i].bio;
/* now find blocks with errors */ /* now find blocks with errors */
for (i=first+1 ; i < conf->copies ; i++) { for (i=0 ; i < conf->copies ; i++) {
int vcnt, j, d; int j, d;
int vcnt = r10_bio->sectors >> (PAGE_SHIFT-9);
if (!test_bit(BIO_UPTODATE, &r10_bio->devs[i].bio->bi_flags))
continue;
/* We know that the bi_io_vec layout is the same for
* both 'first' and 'i', so we just compare them.
* All vec entries are PAGE_SIZE;
*/
tbio = r10_bio->devs[i].bio; tbio = r10_bio->devs[i].bio;
vcnt = r10_bio->sectors >> (PAGE_SHIFT-9);
for (j = 0; j < vcnt; j++) if (tbio->bi_end_io != end_sync_read)
if (memcmp(page_address(fbio->bi_io_vec[j].bv_page), continue;
page_address(tbio->bi_io_vec[j].bv_page), if (i == first)
PAGE_SIZE))
break;
if (j == vcnt)
continue; continue;
mddev->resync_mismatches += r10_bio->sectors; if (test_bit(BIO_UPTODATE, &r10_bio->devs[i].bio->bi_flags)) {
/* We know that the bi_io_vec layout is the same for
* both 'first' and 'i', so we just compare them.
* All vec entries are PAGE_SIZE;
*/
for (j = 0; j < vcnt; j++)
if (memcmp(page_address(fbio->bi_io_vec[j].bv_page),
page_address(tbio->bi_io_vec[j].bv_page),
PAGE_SIZE))
break;
if (j == vcnt)
continue;
mddev->resync_mismatches += r10_bio->sectors;
}
if (test_bit(MD_RECOVERY_CHECK, &mddev->recovery)) if (test_bit(MD_RECOVERY_CHECK, &mddev->recovery))
/* Don't fix anything. */ /* Don't fix anything. */
continue; continue;
...@@ -1308,7 +1318,10 @@ static void recovery_request_write(mddev_t *mddev, r10bio_t *r10_bio) ...@@ -1308,7 +1318,10 @@ static void recovery_request_write(mddev_t *mddev, r10bio_t *r10_bio)
atomic_inc(&conf->mirrors[d].rdev->nr_pending); atomic_inc(&conf->mirrors[d].rdev->nr_pending);
md_sync_acct(conf->mirrors[d].rdev->bdev, wbio->bi_size >> 9); md_sync_acct(conf->mirrors[d].rdev->bdev, wbio->bi_size >> 9);
generic_make_request(wbio); if (test_bit(R10BIO_Uptodate, &r10_bio->state))
generic_make_request(wbio);
else
bio_endio(wbio, wbio->bi_size, -EIO);
} }
...@@ -1445,7 +1458,8 @@ static void raid10d(mddev_t *mddev) ...@@ -1445,7 +1458,8 @@ static void raid10d(mddev_t *mddev)
unfreeze_array(conf); unfreeze_array(conf);
bio = r10_bio->devs[r10_bio->read_slot].bio; bio = r10_bio->devs[r10_bio->read_slot].bio;
r10_bio->devs[r10_bio->read_slot].bio = NULL; r10_bio->devs[r10_bio->read_slot].bio =
mddev->ro ? IO_BLOCKED : NULL;
bio_put(bio); bio_put(bio);
mirror = read_balance(conf, r10_bio); mirror = read_balance(conf, r10_bio);
if (mirror == -1) { if (mirror == -1) {
......
...@@ -104,6 +104,13 @@ struct r10bio_s { ...@@ -104,6 +104,13 @@ struct r10bio_s {
} devs[0]; } devs[0];
}; };
/* when we get a read error on a read-only array, we redirect to another
* device without failing the first device, or trying to over-write to
* correct the read error. To keep track of bad blocks on a per-bio
* level, we store IO_BLOCKED in the appropriate 'bios' pointer
*/
#define IO_BLOCKED ((struct bio*)1)
/* bits for r10bio.state */ /* bits for r10bio.state */
#define R10BIO_Uptodate 0 #define R10BIO_Uptodate 0
#define R10BIO_IsSync 1 #define R10BIO_IsSync 1
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment