| Message ID | 20210713211306.50593-1-junxiao.bi@oracle.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [1/2] ocfs2: fix zero out valid data | expand |
Anybody help review this? Thanks, Junxiao. On 7/13/21 2:13 PM, Junxiao Bi wrote: > If append-dio feature is enabled, direct-io write and fallocate could run > in parallel to extend file size, fallocate used "orig_isize" to record > i_size before taking "ip_alloc_sem", when ocfs2_zeroout_partial_cluster() > zeroout EOF blocks, i_size maybe already extended by ocfs2_dio_end_io_write(), > that will cause valid data zeroed out. > > Fixes: 1733438b1d06 ("ocfs2: fix data corruption by fallocate") > Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com> > Cc: <stable@vger.kernel.org> > --- > fs/ocfs2/file.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c > index 775657943057..53bb46ce3cbb 100644 > --- a/fs/ocfs2/file.c > +++ b/fs/ocfs2/file.c > @@ -1935,7 +1935,6 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode, > goto out_inode_unlock; > } > > - orig_isize = i_size_read(inode); > switch (sr->l_whence) { > case 0: /*SEEK_SET*/ > break; > @@ -1943,7 +1942,7 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode, > sr->l_start += f_pos; > break; > case 2: /*SEEK_END*/ > - sr->l_start += orig_isize; > + sr->l_start += i_size_read(inode); > break; > default: > ret = -EINVAL; > @@ -1998,6 +1997,7 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode, > ret = -EINVAL; > } > > + orig_isize = i_size_read(inode); > /* zeroout eof blocks in the cluster. */ > if (!ret && change_size && orig_isize < size) { > ret = ocfs2_zeroout_partial_cluster(inode, orig_isize,
On 7/14/21 5:13 AM, Junxiao Bi wrote: > If append-dio feature is enabled, direct-io write and fallocate could run > in parallel to extend file size, fallocate used "orig_isize" to record > i_size before taking "ip_alloc_sem", when ocfs2_zeroout_partial_cluster() > zeroout EOF blocks, i_size maybe already extended by ocfs2_dio_end_io_write(), > that will cause valid data zeroed out. > > Fixes: 1733438b1d06 ("ocfs2: fix data corruption by fallocate") Fixes: 6bba4471f0cc ("ocfs2: fix data corruption by fallocate") With the above fixed, Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com> > Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com>> Cc: <stable@vger.kernel.org> > --- > fs/ocfs2/file.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c > index 775657943057..53bb46ce3cbb 100644 > --- a/fs/ocfs2/file.c > +++ b/fs/ocfs2/file.c > @@ -1935,7 +1935,6 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode, > goto out_inode_unlock; > } > > - orig_isize = i_size_read(inode); > switch (sr->l_whence) { > case 0: /*SEEK_SET*/ > break; > @@ -1943,7 +1942,7 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode, > sr->l_start += f_pos; > break; > case 2: /*SEEK_END*/ > - sr->l_start += orig_isize; > + sr->l_start += i_size_read(inode); > break; > default: > ret = -EINVAL; > @@ -1998,6 +1997,7 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode, > ret = -EINVAL; > } > > + orig_isize = i_size_read(inode); > /* zeroout eof blocks in the cluster. */ > if (!ret && change_size && orig_isize < size) { > ret = ocfs2_zeroout_partial_cluster(inode, orig_isize, >
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 775657943057..53bb46ce3cbb 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -1935,7 +1935,6 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode, goto out_inode_unlock; } - orig_isize = i_size_read(inode); switch (sr->l_whence) { case 0: /*SEEK_SET*/ break; @@ -1943,7 +1942,7 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode, sr->l_start += f_pos; break; case 2: /*SEEK_END*/ - sr->l_start += orig_isize; + sr->l_start += i_size_read(inode); break; default: ret = -EINVAL; @@ -1998,6 +1997,7 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode, ret = -EINVAL; } + orig_isize = i_size_read(inode); /* zeroout eof blocks in the cluster. */ if (!ret && change_size && orig_isize < size) { ret = ocfs2_zeroout_partial_cluster(inode, orig_isize,
If append-dio feature is enabled, direct-io write and fallocate could run in parallel to extend file size, fallocate used "orig_isize" to record i_size before taking "ip_alloc_sem", when ocfs2_zeroout_partial_cluster() zeroout EOF blocks, i_size maybe already extended by ocfs2_dio_end_io_write(), that will cause valid data zeroed out. Fixes: 1733438b1d06 ("ocfs2: fix data corruption by fallocate") Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com> Cc: <stable@vger.kernel.org> --- fs/ocfs2/file.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)