Skip to contentSkip to conversation

Btrfs: do not inc uncorrectable_errors counter on ro scrubs

PATCH New Ilya Dryomov 1 participant 3 messages Nov 3, 2013 Btrfs file system

lore
Share & tools
lore
#0 lore raw

PATCH 1/3 Btrfs: do not inc uncorrectable_errors counter on ro scrubs

Currently if we discover an error when scrubbing in ro mode we a) blindly increment the uncorrectable_errors counter, and b) spam the dmesg with the 'unable to fixup (regular) error at ...' message, even though a) we haven't tried to determine if the error is correctable or not, and b) we haven't tried to fixup anything. Fix this.

Cc Stefan Behrens <[email protected]>
---
 fs/btrfs/scrub.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
fs/btrfs/scrub.c +4 -2
index f21e2df..f94b98d 100644
@@ -938,8 +938,10 @@ static int scrub_handle_errored_block(struct scrub_block *sblock_to_check)
 				BTRFS_DEV_STAT_CORRUPTION_ERRS);
 	}
 
-	if (sctx->readonly && !sctx->is_dev_replace)
-		goto did_not_correct_error;
+	if (sctx->readonly) {
+		ASSERT(!sctx->is_dev_replace);
+		goto out;
+	}
 
 	if (!is_metadata && !have_csum) {
 		struct scrub_fixup_nodatasum *fixup_nodatasum;
Signature · 1 line
1.7.10.4
#1 lore raw

PATCH 2/3 Btrfs: disable online raid-repair on ro mounts

This disables the "if needed, write the good copy back before the read is completed" part of the read sequence for read-only mounts.

---
 fs/btrfs/extent_io.c |   11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)
fs/btrfs/extent_io.c +8 -3
index c09a40d..a0b7403 100644
@@ -1979,6 +1979,7 @@ int repair_io_failure(struct btrfs_fs_info *fs_info, u64 start,
 	struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree;
 	int ret;
 
+	ASSERT(!(fs_info->sb->s_flags & MS_RDONLY));
 	BUG_ON(!mirror_num);
 
 	/* we can't repair anything in raid56 yet */
@@ -2035,6 +2036,9 @@ int repair_eb_io_failure(struct btrfs_root *root, struct extent_buffer *eb,
 	unsigned long i, num_pages = num_extent_pages(eb->start, eb->len);
 	int ret = 0;
 
+	if (root->fs_info->sb->s_flags & MS_RDONLY)
+		return -EROFS;
+
 	for (i = 0; i < num_pages; i++) {
 		struct page *p = extent_buffer_page(eb, i);
 		ret = repair_io_failure(root->fs_info, start, PAGE_CACHE_SIZE,
@@ -2056,12 +2060,12 @@ static int clean_io_failure(u64 start, struct page *page)
 	u64 private;
 	u64 private_failure;
 	struct io_failure_record *failrec;
-	struct btrfs_fs_info *fs_info;
+	struct inode *inode = page->mapping->host;
+	struct btrfs_fs_info *fs_info = BTRFS_I(inode)->root->fs_info;
 	struct extent_state *state;
 	int num_copies;
 	int did_repair = 0;
 	int ret;
-	struct inode *inode = page->mapping->host;
 
 	private = 0;
 	ret = count_range_bits(&BTRFS_I(inode)->io_failure_tree, &private,
@@ -2084,6 +2088,8 @@ static int clean_io_failure(u64 start, struct page *page)
 		did_repair = 1;
 		goto out;
 	}
+	if (fs_info->sb->s_flags & MS_RDONLY)
+		goto out;
 
 	spin_lock(&BTRFS_I(inode)->io_tree.lock);
 	state = find_first_extent_bit_state(&BTRFS_I(inode)->io_tree,
@@ -2093,7 +2099,6 @@ static int clean_io_failure(u64 start, struct page *page)
 
 	if (state && state->start <= failrec->start &&
 	    state->end >= failrec->start + failrec->len - 1) {
-		fs_info = BTRFS_I(inode)->root->fs_info;
 		num_copies = btrfs_num_copies(fs_info, failrec->logical,
 					      failrec->len);
 		if (num_copies > 1)  {
Signature · 1 line
1.7.10.4
#2 lore raw

PATCH 3/3 Btrfs: fix __btrfs_start_workers retval

In reply to Ilya Dryomov

__btrfs_start_workers returns 0 in case it raced with btrfs_stop_workers and lost the race. This is wrong because worker in this case is not allowed to start and is in fact destroyed. Return -EINVAL instead.

---
 fs/btrfs/async-thread.c |    1 +
 1 file changed, 1 insertion(+)
fs/btrfs/async-thread.c +1 -0
index 08cc08f..1d5f3d7 100644
@@ -495,6 +495,7 @@ static int __btrfs_start_workers(struct btrfs_workers *workers)
 	spin_lock_irq(&workers->lock);
 	if (workers->stopping) {
 		spin_unlock_irq(&workers->lock);
+		ret = -EINVAL;
 		goto fail_kthread;
 	}
 	list_add_tail(&worker->worker_list, &workers->idle_list);
Signature · 1 line
1.7.10.4