diff options
| author |   Jaegeuk Kim <jaegeuk@kernel.org> | 2014-08-11 18:37:46 -0700 |
|---|---|---|
| committer | Jaegeuk Kim <jaegeuk@kernel.org> | 2014-08-21 13:55:05 -0700 |
| commit | cf779cab14d50a84b61399f758da269654b863db (patch) | |
| tree | c11a1da5ea17f3e3f86e5cdad7cc57b97427e7fa /fs/f2fs/data.c | |
| parent | f2fs: check s_dirty under cp_mutex (diff) | |
| download | linux-dev-cf779cab14d50a84b61399f758da269654b863db.tar.xz linux-dev-cf779cab14d50a84b61399f758da269654b863db.zip | |
f2fs: handle EIO not to break fs consistency
There are two rules when EIO is occurred.
1. don't write any checkpoint data to preserve the previous checkpoint
2. don't lose the cached dentry/node/meta pages
So, at first, this patch adds set_page_dirty in f2fs_write_end_io's failure.
Then, writing checkpoint/dentry/node blocks is not allowed.
Note that, for the data pages, we can't just throw away by redirtying them.
Otherwise, kworker can fall into infinite loop to flush them.
(Ref. xfstests/019)
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Diffstat (limited to 'fs/f2fs/data.c')
| -rw-r--r-- | fs/f2fs/data.c | 11 |
1 files changed, 10 insertions, 1 deletions
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index ac3ccc25386b..6ba52a393063 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -53,7 +53,7 @@ static void f2fs_write_end_io(struct bio *bio, int err) struct page *page = bvec->bv_page; if (unlikely(err)) { - SetPageError(page); + set_page_dirty(page); set_bit(AS_EIO, &page->mapping->flags); f2fs_stop_checkpoint(sbi); } @@ -836,10 +836,19 @@ write: /* Dentry blocks are controlled by checkpoint */ if (S_ISDIR(inode->i_mode)) { + if (unlikely(f2fs_cp_error(sbi))) + goto redirty_out; err = do_write_data_page(page, &fio); goto done; } + /* we should bypass data pages to proceed the kworkder jobs */ + if (unlikely(f2fs_cp_error(sbi))) { + SetPageError(page); + unlock_page(page); + return 0; + } + if (!wbc->for_reclaim) need_balance_fs = true; else if (has_not_enough_free_secs(sbi, 0)) |