From 0e08fa789d39aa01923e3ba144bd808291895c3c Mon Sep 17 00:00:00 2001 From: Steve French Date: Sat, 23 Aug 2025 21:15:59 -0500 Subject: [PATCH 1/2] smb3 client: fix return code mapping of remap_file_range We were returning -EOPNOTSUPP for various remap_file_range cases but for some of these the copy_file_range_syscall() requires -EINVAL to be returned (e.g. where source and target file ranges overlap when source and target are the same file). This fixes xfstest generic/157 which was expecting EINVAL for that (and also e.g. for when the src offset is beyond end of file). Cc: stable@vger.kernel.org Acked-by: Paulo Alcantara (Red Hat) Signed-off-by: Steve French --- fs/smb/client/cifsfs.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/fs/smb/client/cifsfs.c b/fs/smb/client/cifsfs.c index 3bd85ab2deb1..e1848276bab4 100644 --- a/fs/smb/client/cifsfs.c +++ b/fs/smb/client/cifsfs.c @@ -1358,6 +1358,20 @@ static loff_t cifs_remap_file_range(struct file *src_file, loff_t off, truncate_setsize(target_inode, new_size); fscache_resize_cookie(cifs_inode_cookie(target_inode), new_size); + } else if (rc == -EOPNOTSUPP) { + /* + * copy_file_range syscall man page indicates EINVAL + * is returned e.g when "fd_in and fd_out refer to the + * same file and the source and target ranges overlap." + * Test generic/157 was what showed these cases where + * we need to remap EOPNOTSUPP to EINVAL + */ + if (off >= src_inode->i_size) { + rc = -EINVAL; + } else if (src_inode == target_inode) { + if (off + len > destoff) + rc = -EINVAL; + } } if (rc == 0 && new_size > target_cifsi->netfs.zero_point) target_cifsi->netfs.zero_point = new_size; From ab529e6ca1f67bcf31f3ea80c72bffde2e9e053e Mon Sep 17 00:00:00 2001 From: Shuhao Fu Date: Thu, 28 Aug 2025 02:24:19 +0800 Subject: [PATCH 2/2] fs/smb: Fix inconsistent refcnt update A possible inconsistent update of refcount was identified in `smb2_compound_op`. Such inconsistent update could lead to possible resource leaks. Why it is a possible bug: 1. In the comment section of the function, it clearly states that the reference to `cfile` should be dropped after calling this function. 2. Every control flow path would check and drop the reference to `cfile`, except the patched one. 3. Existing callers would not handle refcount update of `cfile` if -ENOMEM is returned. To fix the bug, an extra goto label "out" is added, to make sure that the cleanup logic would always be respected. As the problem is caused by the allocation failure of `vars`, the cleanup logic between label "finished" and "out" can be safely ignored. According to the definition of function `is_replayable_error`, the error code of "-ENOMEM" is not recoverable. Therefore, the replay logic also gets ignored. Signed-off-by: Shuhao Fu Acked-by: Paulo Alcantara (Red Hat) Cc: stable@vger.kernel.org Signed-off-by: Steve French --- fs/smb/client/smb2inode.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/fs/smb/client/smb2inode.c b/fs/smb/client/smb2inode.c index 2a0316c514e4..31c13fb5b85b 100644 --- a/fs/smb/client/smb2inode.c +++ b/fs/smb/client/smb2inode.c @@ -207,8 +207,10 @@ static int smb2_compound_op(const unsigned int xid, struct cifs_tcon *tcon, server = cifs_pick_channel(ses); vars = kzalloc(sizeof(*vars), GFP_ATOMIC); - if (vars == NULL) - return -ENOMEM; + if (vars == NULL) { + rc = -ENOMEM; + goto out; + } rqst = &vars->rqst[0]; rsp_iov = &vars->rsp_iov[0]; @@ -864,6 +866,7 @@ static int smb2_compound_op(const unsigned int xid, struct cifs_tcon *tcon, smb2_should_replay(tcon, &retries, &cur_sleep)) goto replay_again; +out: if (cfile) cifsFileInfo_put(cfile);