From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: From: Miklos Szeredi To: linux-fsdevel@vger.kernel.org Cc: linux-kernel@vger.kernel.org, Al Viro Subject: [PATCH] proc/locks: getattr dev and ino Date: Wed, 20 Sep 2017 11:25:25 +0200 Message-Id: <1505899525-21443-1-git-send-email-mszeredi@redhat.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: /proc/locks reports incorrect device/ino for overlayfs. Same is true for several filesystems that use a different value for i_ino as for st_ino. Also btrfs uses different st_dev values for subvolumes. Do a getattr for getting the correct dev/ino information. Make sure filesystem doesn't start network traffic and does not return fields which are expensive to calculate. Signed-off-by: Miklos Szeredi --- fs/locks.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/fs/locks.c b/fs/locks.c index 1bd71c4d663a..d4caea5268dc 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -2683,11 +2683,20 @@ static void lock_get_status(struct seq_file *f, struct file_lock *fl, : (fl->fl_type == F_WRLCK) ? "WRITE" : "READ "); } if (inode) { + struct kstat st; + int err; + + err = vfs_getattr(&fl->fl_file->f_path, &st, + STATX_INO, AT_STATX_DONT_SYNC); + if (err) + goto nostat; + /* userspace relies on this representation of dev_t */ - seq_printf(f, "%d %02x:%02x:%ld ", fl_pid, - MAJOR(inode->i_sb->s_dev), - MINOR(inode->i_sb->s_dev), inode->i_ino); + seq_printf(f, "%d %02x:%02x:%lld ", fl_pid, + MAJOR(st.dev), MINOR(st.dev), + (unsigned long long) st.ino); } else { +nostat: seq_printf(f, "%d :0 ", fl_pid); } if (IS_POSIX(fl)) { -- 2.5.5