#include "ordered-data.h"
#include "transaction.h"
#include "backref.h"
+#include "extent_io.h"
/*
* This is only the first step towards a full-features scrub. It reads all
* any can be found.
*
* Future enhancements:
- * - To enhance the performance, better read-ahead strategies for the
- * extent-tree can be employed.
* - In case an unrepairable extent is encountered, track which files are
* affected and report them
* - In case of a read error on files with nodatasum, map the file and read
* the extent to trigger a writeback of the good copy
* - track and record media errors, throw out bad devices
* - add a mode to also read unallocated space
- * - make the prefetch cancellable
*/
struct scrub_bio;
atomic_set(&sdev->in_flight, 0);
atomic_set(&sdev->fixup_cnt, 0);
atomic_set(&sdev->cancel_req, 0);
- sdev->csum_size = btrfs_super_csum_size(&fs_info->super_copy);
+ sdev->csum_size = btrfs_super_csum_size(fs_info->super_copy);
INIT_LIST_HEAD(&sdev->csum_list);
spin_lock_init(&sdev->list_lock);
static int scrub_fixup_readpage(u64 inum, u64 offset, u64 root, void *ctx)
{
- struct page *page;
+ struct page *page = NULL;
unsigned long index;
struct scrub_fixup_nodatasum *fixup = ctx;
int ret;
- int corrected;
+ int corrected = 0;
struct btrfs_key key;
- struct inode *inode;
+ struct inode *inode = NULL;
u64 end = offset + PAGE_SIZE - 1;
struct btrfs_root *local_root;
if (IS_ERR(inode))
return PTR_ERR(inode);
- ret = set_extent_bit(&BTRFS_I(inode)->io_tree, offset, end,
- EXTENT_DAMAGED, 0, NULL, NULL, GFP_NOFS);
-
- /* set_extent_bit should either succeed or give proper error */
- WARN_ON(ret > 0);
- if (ret)
- return ret < 0 ? ret : -EFAULT;
-
index = offset >> PAGE_CACHE_SHIFT;
page = find_or_create_page(inode->i_mapping, index, GFP_NOFS);
- if (!page)
- return -ENOMEM;
+ if (!page) {
+ ret = -ENOMEM;
+ goto out;
+ }
- ret = extent_read_full_page(&BTRFS_I(inode)->io_tree, page,
- btrfs_get_extent, fixup->mirror_num);
- wait_on_page_locked(page);
- corrected = !test_range_bit(&BTRFS_I(inode)->io_tree, offset, end,
- EXTENT_DAMAGED, 0, NULL);
+ if (PageUptodate(page)) {
+ struct btrfs_mapping_tree *map_tree;
+ if (PageDirty(page)) {
+ /*
+ * we need to write the data to the defect sector. the
+ * data that was in that sector is not in memory,
+ * because the page was modified. we must not write the
+ * modified page to that sector.
+ *
+ * TODO: what could be done here: wait for the delalloc
+ * runner to write out that page (might involve
+ * COW) and see whether the sector is still
+ * referenced afterwards.
+ *
+ * For the meantime, we'll treat this error
+ * incorrectable, although there is a chance that a
+ * later scrub will find the bad sector again and that
+ * there's no dirty page in memory, then.
+ */
+ ret = -EIO;
+ goto out;
+ }
+ map_tree = &BTRFS_I(inode)->root->fs_info->mapping_tree;
+ ret = repair_io_failure(map_tree, offset, PAGE_SIZE,
+ fixup->logical, page,
+ fixup->mirror_num);
+ unlock_page(page);
+ corrected = !ret;
+ } else {
+ /*
+ * we need to get good data first. the general readpage path
+ * will call repair_io_failure for us, we just have to make
+ * sure we read the bad mirror.
+ */
+ ret = set_extent_bits(&BTRFS_I(inode)->io_tree, offset, end,
+ EXTENT_DAMAGED, GFP_NOFS);
+ if (ret) {
+ /* set_extent_bits should give proper error */
+ WARN_ON(ret > 0);
+ if (ret > 0)
+ ret = -EFAULT;
+ goto out;
+ }
- if (corrected)
- WARN_ON(!PageUptodate(page));
- else
- clear_extent_bit(&BTRFS_I(inode)->io_tree, offset, end,
- EXTENT_DAMAGED, 0, 0, NULL, GFP_NOFS);
+ ret = extent_read_full_page(&BTRFS_I(inode)->io_tree, page,
+ btrfs_get_extent,
+ fixup->mirror_num);
+ wait_on_page_locked(page);
+
+ corrected = !test_range_bit(&BTRFS_I(inode)->io_tree, offset,
+ end, EXTENT_DAMAGED, 0, NULL);
+ if (!corrected)
+ clear_extent_bits(&BTRFS_I(inode)->io_tree, offset, end,
+ EXTENT_DAMAGED, GFP_NOFS);
+ }
- put_page(page);
- iput(inode);
+out:
+ if (page)
+ put_page(page);
+ if (inode)
+ iput(inode);
if (ret < 0)
return ret;
struct scrub_dev *sdev = sbio->sdev;
struct btrfs_fs_info *fs_info = sdev->dev->dev_root->fs_info;
struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree;
- struct btrfs_multi_bio *multi = NULL;
+ struct btrfs_bio *bbio = NULL;
struct scrub_fixup_nodatasum *fixup;
u64 logical = sbio->logical + ix * PAGE_SIZE;
u64 length;
length = PAGE_SIZE;
ret = btrfs_map_block(map_tree, REQ_WRITE, logical, &length,
- &multi, 0);
- if (ret || !multi || length < PAGE_SIZE) {
+ &bbio, 0);
+ if (ret || !bbio || length < PAGE_SIZE) {
printk(KERN_ERR
"scrub_fixup: btrfs_map_block failed us for %llu\n",
(unsigned long long)logical);
return;
}
- if (multi->num_stripes == 1)
+ if (bbio->num_stripes == 1)
/* there aren't any replicas */
goto uncorrectable;
/*
* first find a good copy
*/
- for (i = 0; i < multi->num_stripes; ++i) {
+ for (i = 0; i < bbio->num_stripes; ++i) {
if (i + 1 == sbio->spag[ix].mirror_num)
continue;
- if (scrub_fixup_io(READ, multi->stripes[i].dev->bdev,
- multi->stripes[i].physical >> 9,
+ if (scrub_fixup_io(READ, bbio->stripes[i].dev->bdev,
+ bbio->stripes[i].physical >> 9,
sbio->bio->bi_io_vec[ix].bv_page)) {
/* I/O-error, this is not a good copy */
continue;
if (scrub_fixup_check(sbio, ix) == 0)
break;
}
- if (i == multi->num_stripes)
+ if (i == bbio->num_stripes)
goto uncorrectable;
if (!sdev->readonly) {
}
}
- kfree(multi);
+ kfree(bbio);
spin_lock(&sdev->stat_lock);
++sdev->stat.corrected_errors;
spin_unlock(&sdev->stat_lock);
return;
uncorrectable:
- kfree(multi);
+ kfree(bbio);
spin_lock(&sdev->stat_lock);
++sdev->stat.uncorrectable_errors;
spin_unlock(&sdev->stat_lock);
int slot;
int i;
u64 nstripes;
- int start_stripe;
struct extent_buffer *l;
struct btrfs_key key;
u64 physical;
u64 logical;
u64 generation;
int mirror_num;
+ struct reada_control *reada1;
+ struct reada_control *reada2;
+ struct btrfs_key key_start;
+ struct btrfs_key key_end;
u64 increment = map->stripe_len;
u64 offset;
if (!path)
return -ENOMEM;
- path->reada = 2;
path->search_commit_root = 1;
path->skip_locking = 1;
/*
- * find all extents for each stripe and just read them to get
- * them into the page cache
- * FIXME: we can do better. build a more intelligent prefetching
+ * trigger the readahead for extent tree csum tree and wait for
+ * completion. During readahead, the scrub is officially paused
+ * to not hold off transaction commits
*/
logical = base + offset;
- physical = map->stripes[num].physical;
- ret = 0;
- for (i = 0; i < nstripes; ++i) {
- key.objectid = logical;
- key.type = BTRFS_EXTENT_ITEM_KEY;
- key.offset = (u64)0;
-
- ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
- if (ret < 0)
- goto out_noplug;
-
- /*
- * we might miss half an extent here, but that doesn't matter,
- * as it's only the prefetch
- */
- while (1) {
- l = path->nodes[0];
- slot = path->slots[0];
- if (slot >= btrfs_header_nritems(l)) {
- ret = btrfs_next_leaf(root, path);
- if (ret == 0)
- continue;
- if (ret < 0)
- goto out_noplug;
- break;
- }
- btrfs_item_key_to_cpu(l, &key, slot);
+ wait_event(sdev->list_wait,
+ atomic_read(&sdev->in_flight) == 0);
+ atomic_inc(&fs_info->scrubs_paused);
+ wake_up(&fs_info->scrub_pause_wait);
- if (key.objectid >= logical + map->stripe_len)
- break;
+ /* FIXME it might be better to start readahead at commit root */
+ key_start.objectid = logical;
+ key_start.type = BTRFS_EXTENT_ITEM_KEY;
+ key_start.offset = (u64)0;
+ key_end.objectid = base + offset + nstripes * increment;
+ key_end.type = BTRFS_EXTENT_ITEM_KEY;
+ key_end.offset = (u64)0;
+ reada1 = btrfs_reada_add(root, &key_start, &key_end);
+
+ key_start.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
+ key_start.type = BTRFS_EXTENT_CSUM_KEY;
+ key_start.offset = logical;
+ key_end.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
+ key_end.type = BTRFS_EXTENT_CSUM_KEY;
+ key_end.offset = base + offset + nstripes * increment;
+ reada2 = btrfs_reada_add(csum_root, &key_start, &key_end);
+
+ if (!IS_ERR(reada1))
+ btrfs_reada_wait(reada1);
+ if (!IS_ERR(reada2))
+ btrfs_reada_wait(reada2);
- path->slots[0]++;
- }
- btrfs_release_path(path);
- logical += increment;
- physical += map->stripe_len;
- cond_resched();
+ mutex_lock(&fs_info->scrub_lock);
+ while (atomic_read(&fs_info->scrub_pause_req)) {
+ mutex_unlock(&fs_info->scrub_lock);
+ wait_event(fs_info->scrub_pause_wait,
+ atomic_read(&fs_info->scrub_pause_req) == 0);
+ mutex_lock(&fs_info->scrub_lock);
}
+ atomic_dec(&fs_info->scrubs_paused);
+ mutex_unlock(&fs_info->scrub_lock);
+ wake_up(&fs_info->scrub_pause_wait);
/*
* collect all data csums for the stripe to avoid seeking during
* the scrub. This might currently (crc32) end up to be about 1MB
*/
- start_stripe = 0;
blk_start_plug(&plug);
-again:
- logical = base + offset + start_stripe * increment;
- for (i = start_stripe; i < nstripes; ++i) {
- ret = btrfs_lookup_csums_range(csum_root, logical,
- logical + map->stripe_len - 1,
- &sdev->csum_list, 1);
- if (ret)
- goto out;
- logical += increment;
- cond_resched();
- }
/*
* now find all extents for each stripe and scrub them
*/
- logical = base + offset + start_stripe * increment;
- physical = map->stripes[num].physical + start_stripe * map->stripe_len;
+ logical = base + offset;
+ physical = map->stripes[num].physical;
ret = 0;
- for (i = start_stripe; i < nstripes; ++i) {
+ for (i = 0; i < nstripes; ++i) {
/*
* canceled?
*/
atomic_dec(&fs_info->scrubs_paused);
mutex_unlock(&fs_info->scrub_lock);
wake_up(&fs_info->scrub_pause_wait);
- scrub_free_csums(sdev);
- start_stripe = i;
- goto again;
}
+ ret = btrfs_lookup_csums_range(csum_root, logical,
+ logical + map->stripe_len - 1,
+ &sdev->csum_list, 1);
+ if (ret)
+ goto out;
+
key.objectid = logical;
key.type = BTRFS_EXTENT_ITEM_KEY;
key.offset = (u64)0;
out:
blk_finish_plug(&plug);
-out_noplug:
btrfs_free_path(path);
return ret < 0 ? ret : 0;
}
git.openpandora.org / pandora-kernel.git / blobdiff