NFS: Fix a race between mmap() and O_DIRECT
commite231c6879c
upstream. When locking the file in order to do O_DIRECT on it, we must unmap any mmapped ranges on the pagecache so that we can flush out the dirty data. Fixes:a5864c999d
("NFS: Do not serialise O_DIRECT reads and writes") Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This commit is contained in:
parent
1c38ad2f26
commit
f1e31607e7
|
@ -99,7 +99,7 @@ static void nfs_block_buffered(struct nfs_inode *nfsi, struct inode *inode)
|
|||
{
|
||||
if (!test_bit(NFS_INO_ODIRECT, &nfsi->flags)) {
|
||||
set_bit(NFS_INO_ODIRECT, &nfsi->flags);
|
||||
nfs_wb_all(inode);
|
||||
nfs_sync_mapping(inode->i_mapping);
|
||||
}
|
||||
}
|
||||
|
||||
|
|
Loading…
Reference in a new issue