The code to overwrite/reread for addressing read errors
in raid1/raid10 currently assumes that the read will
not alter the buffer which could be used to write to
the next device. This is not a safe assumption to make.
So we split the loops into a overwrite loop and a separate re-read
loop, so that the writing is complete before reading is attempted.
Cc: Paul Clements <paul.clements@xxxxxxxxxxxx>
Signed-off-by: Neil Brown <neilb@xxxxxxx>
### Diffstat output
./drivers/md/raid1.c | 38 ++++++++++++++++++++++++++++++--------
./drivers/md/raid10.c | 22 ++++++++++++++++++----
2 files changed, 48 insertions(+), 12 deletions(-)
diff ./drivers/md/raid1.c~current~ ./drivers/md/raid1.c
--- ./drivers/md/raid1.c~current~ 2005-12-01 14:03:25.000000000 +1100
+++ ./drivers/md/raid1.c 2005-12-01 14:03:40.000000000 +1100
@@ -1252,6 +1252,7 @@ static void sync_request_write(mddev_t *
} while (!success && d != r1_bio->read_disk);
if (success) {
+ int start = d;
/* write it back and re-read */
set_bit(R1BIO_Uptodate, &r1_bio->state);
while (d != r1_bio->read_disk) {
@@ -1265,14 +1266,23 @@ static void sync_request_write(mddev_t *
sect + rdev->data_offset,
s<<9,
bio->bi_io_vec[idx].bv_page,
- WRITE) == 0 ||
- sync_page_io(rdev->bdev,
+ WRITE) == 0)
+ md_error(mddev, rdev);
+ }
+ d = start;
+ while (d != r1_bio->read_disk) {
+ if (d == 0)
+ d = conf->raid_disks;
+ d--;
+ if (r1_bio->bios[d]->bi_end_io != end_sync_read)
+ continue;
+ rdev = conf->mirrors[d].rdev;
+ if (sync_page_io(rdev->bdev,
sect + rdev->data_offset,
s<<9,
bio->bi_io_vec[idx].bv_page,
- READ) == 0) {
+ READ) == 0)
md_error(mddev, rdev);
- }
}
} else {
char b[BDEVNAME_SIZE];
@@ -1444,6 +1454,7 @@ static void raid1d(mddev_t *mddev)
if (success) {
/* write it back and re-read */
+ int start = d;
while (d != r1_bio->read_disk) {
if (d==0)
d = conf->raid_disks;
@@ -1453,13 +1464,24 @@ static void raid1d(mddev_t *mddev)
test_bit(In_sync, &rdev->flags)) {
if (sync_page_io(rdev->bdev,
sect + rdev->data_offset,
- s<<9, conf->tmppage, WRITE) == 0 ||
- sync_page_io(rdev->bdev,
+ s<<9, conf->tmppage, WRITE) == 0)
+ /* Well, this device is dead */
+ md_error(mddev, rdev);
+ }
+ }
+ d = start;
+ while (d != r1_bio->read_disk) {
+ if (d==0)
+ d = conf->raid_disks;
+ d--;
+ rdev = conf->mirrors[d].rdev;
+ if (rdev &&
+ test_bit(In_sync, &rdev->flags)) {
+ if (sync_page_io(rdev->bdev,
sect + rdev->data_offset,
- s<<9, conf->tmppage, READ) == 0) {
+ s<<9, conf->tmppage, READ) == 0)
/* Well, this device is dead */
md_error(mddev, rdev);
- }
}
}
} else {
diff ./drivers/md/raid10.c~current~ ./drivers/md/raid10.c
--- ./drivers/md/raid10.c~current~ 2005-12-01 14:03:25.000000000 +1100
+++ ./drivers/md/raid10.c 2005-12-01 14:03:41.000000000 +1100
@@ -1421,6 +1421,7 @@ static void raid10d(mddev_t *mddev)
} while (!success && sl != r10_bio->read_slot);
if (success) {
+ int start = sl;
/* write it back and re-read */
while (sl != r10_bio->read_slot) {
int d;
@@ -1434,14 +1435,27 @@ static void raid10d(mddev_t *mddev)
if (sync_page_io(rdev->bdev,
r10_bio->devs[sl].addr +
sect + rdev->data_offset,
- s<<9, conf->tmppage, WRITE) == 0 ||
- sync_page_io(rdev->bdev,
+ s<<9, conf->tmppage, WRITE) == 0)
+ /* Well, this device is dead */
+ md_error(mddev, rdev);
+ }
+ }
+ sl = start;
+ while (sl != r10_bio->read_slot) {
+ int d;
+ if (sl==0)
+ sl = conf->copies;
+ sl--;
+ d = r10_bio->devs[sl].devnum;
+ rdev = conf->mirrors[d].rdev;
+ if (rdev &&
+ test_bit(In_sync, &rdev->flags)) {
+ if (sync_page_io(rdev->bdev,
r10_bio->devs[sl].addr +
sect + rdev->data_offset,
- s<<9, conf->tmppage, READ) == 0) {
+ s<<9, conf->tmppage, READ) == 0)
/* Well, this device is dead */
md_error(mddev, rdev);
- }
}
}
} else {
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
