Ilya Dryomov
2011-Apr-09 11:29 UTC
[PATCH] btrfs scrub - restore bios properly after media errors
The current code reallocates a bio after a media error. This is a
temporary measure introduced in v3 after a serious problem related to
bio reuse was found in v2 of scrub patchset.
Basically we did not reset bv_offset and bv_len fields of the bio_vec
structure. They are changed in case I/O error happens, for example, at
offset 512 or 1024 into the page. Also bi_flags field wasn''t properly
setup before reusing the bio.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
---
fs/btrfs/scrub.c | 48 +++++++++++++-----------------------------------
1 files changed, 13 insertions(+), 35 deletions(-)
diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
index e862167..1631327 100644
--- a/fs/btrfs/scrub.c
+++ b/fs/btrfs/scrub.c
@@ -418,45 +418,24 @@ static void scrub_checksum(scrub_work_t *work)
int ret;
if (sbio->err) {
- struct bio *bio;
- struct bio *old_bio;
-
- for (i = 0; i < sbio->count; ++i) {
+ for (i = 0; i < sbio->count; ++i)
scrub_recheck_error(sbio, i);
+
+ sbio->bio->bi_flags &= ~(BIO_POOL_MASK - 1);
+ sbio->bio->bi_flags |= 1 << BIO_UPTODATE;
+ sbio->bio->bi_phys_segments = 0;
+ sbio->bio->bi_idx = 0;
+
+ for (i = 0; i < sbio->count; i++) {
+ struct bio_vec *bi;
+ bi = &sbio->bio->bi_io_vec[i];
+ bi->bv_offset = 0;
+ bi->bv_len = PAGE_SIZE;
}
+
spin_lock(&sdev->stat_lock);
++sdev->stat.read_errors;
spin_unlock(&sdev->stat_lock);
-
- /*
- * FIXME: allocate a new bio after a media error. I haven''t
- * figured out how to reuse this one
- */
- old_bio = sbio->bio;
- bio = bio_kmalloc(GFP_NOFS, SCRUB_PAGES_PER_BIO);
- if (!bio) {
- /*
- * alloc failed. cancel the scrub and don''t requeue
- * this sbio
- */
- printk(KERN_ERR "btrfs scrub: allocation failure, "
- "cancelling scrub\n");
- atomic_inc(&sdev->dev->dev_root->fs_info->
- scrub_cancel_req);
- goto out_no_enqueue;
- }
- sbio->bio = bio;
- bio->bi_private = sbio;
- bio->bi_end_io = scrub_bio_end_io;
- bio->bi_sector = 0;
- bio->bi_bdev = sbio->sdev->dev->bdev;
- bio->bi_size = 0;
- for (i = 0; i < SCRUB_PAGES_PER_BIO; ++i) {
- struct page *page;
- page = old_bio->bi_io_vec[i].bv_page;
- bio_add_page(bio, page, PAGE_SIZE, 0);
- }
- bio_put(old_bio);
goto out;
}
for (i = 0; i < sbio->count; ++i) {
@@ -486,7 +465,6 @@ out:
sbio->next_free = sdev->first_free;
sdev->first_free = sbio->index;
spin_unlock(&sdev->list_lock);
-out_no_enqueue:
atomic_dec(&sdev->in_flight);
wake_up(&sdev->list_wait);
}
--
1.7.2.5
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs"
in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html