On Tue, 2017-09-12 at 10:53 +0800, Yan, Zheng wrote:
> When session get evicted, all file locks associated with the session
> get released remotely by mds. File locks tracked by kernel become
> stale. In this situation, set an error flag on inode. The flag makes
> further file locks return -EIO.
>
> Another option to handle this situation is cleanup file locks tracked
> kernel. I do not choose it because it is inconvenient to notify user
> program about the error.
>
> Signed-off-by: "Yan, Zheng" <zyan@xxxxxxxxxx>
> ---
> fs/ceph/locks.c | 52 ++++++++++++++++++++++++++++++++++++++++------------
> fs/ceph/mds_client.c | 21 ++++++++++++++++-----
> fs/ceph/super.h | 2 ++
> 3 files changed, 58 insertions(+), 17 deletions(-)
>
> diff --git a/fs/ceph/locks.c b/fs/ceph/locks.c
> index 4addd18ac6f9..e240ac0903e4 100644
> --- a/fs/ceph/locks.c
> +++ b/fs/ceph/locks.c
> @@ -38,7 +38,13 @@ static void ceph_fl_copy_lock(struct file_lock *dst, struct file_lock *src)
> static void ceph_fl_release_lock(struct file_lock *fl)
> {
> struct inode *inode = file_inode(fl->fl_file);
> - atomic_dec(&ceph_inode(inode)->i_filelock_ref);
> + struct ceph_inode_info *ci = ceph_inode(inode);
> + if (atomic_dec_and_test(&ci->i_filelock_ref)) {
> + /* clear error when all locks are released */
> + spin_lock(&ci->i_ceph_lock);
> + ci->i_ceph_flags &= ~CEPH_I_ERROR_FILELOCK;
> + spin_unlock(&ci->i_ceph_lock);
> + }
> }
>
> static const struct file_lock_operations ceph_fl_lock_ops = {
> @@ -207,10 +213,11 @@ static int ceph_lock_wait_for_completion(struct ceph_mds_client *mdsc,
> int ceph_lock(struct file *file, int cmd, struct file_lock *fl)
> {
> struct inode *inode = file_inode(file);
> - int err;
> + struct ceph_inode_info *ci = ceph_inode(inode);
> + int err = 0;
> u16 op = CEPH_MDS_OP_SETFILELOCK;
> - u8 lock_cmd;
> u8 wait = 0;
> + u8 lock_cmd;
>
> if (!(fl->fl_flags & FL_POSIX))
> return -ENOLCK;
> @@ -226,7 +233,10 @@ int ceph_lock(struct file *file, int cmd, struct file_lock *fl)
> else if (IS_SETLKW(cmd))
> wait = 1;
>
> - if (op == CEPH_MDS_OP_SETFILELOCK) {
> + spin_lock(&ci->i_ceph_lock);
> + if (ci->i_ceph_flags & CEPH_I_ERROR_FILELOCK) {
> + err = -EIO;
> + } else if (op == CEPH_MDS_OP_SETFILELOCK) {
> /*
> * increasing i_filelock_ref closes race window between
> * handling request reply and adding file_lock struct to
> @@ -234,7 +244,13 @@ int ceph_lock(struct file *file, int cmd, struct file_lock *fl)
> * window. Caller function will decrease the counter.
> */
> fl->fl_ops = &ceph_fl_lock_ops;
> - atomic_inc(&ceph_inode(inode)->i_filelock_ref);
> + atomic_inc(&ci->i_filelock_ref);
> + }
> + spin_unlock(&ci->i_ceph_lock);
> + if (err < 0) {
> + if (op == CEPH_MDS_OP_SETFILELOCK && F_UNLCK == fl->fl_type)
> + posix_lock_file(file, fl, NULL);
> + return err;
> }
>
> if (F_RDLCK == fl->fl_type)
> @@ -246,10 +262,10 @@ int ceph_lock(struct file *file, int cmd, struct file_lock *fl)
>
> err = ceph_lock_message(CEPH_LOCK_FCNTL, op, inode, lock_cmd, wait, fl);
> if (!err) {
> - if (op != CEPH_MDS_OP_GETFILELOCK) {
> + if (op == CEPH_MDS_OP_SETFILELOCK) {
> dout("mds locked, locking locally");
> err = posix_lock_file(file, fl, NULL);
> - if (err && (CEPH_MDS_OP_SETFILELOCK == op)) {
> + if (err) {
> /* undo! This should only happen if
> * the kernel detects local
> * deadlock. */
> @@ -266,9 +282,10 @@ int ceph_lock(struct file *file, int cmd, struct file_lock *fl)
> int ceph_flock(struct file *file, int cmd, struct file_lock *fl)
> {
> struct inode *inode = file_inode(file);
> - int err;
> - u8 lock_cmd;
> + struct ceph_inode_info *ci = ceph_inode(inode);
> + int err = 0;
> u8 wait = 0;
> + u8 lock_cmd;
>
> if (!(fl->fl_flags & FL_FLOCK))
> return -ENOLCK;
> @@ -278,9 +295,20 @@ int ceph_flock(struct file *file, int cmd, struct file_lock *fl)
>
> dout("ceph_flock, fl_file: %p", fl->fl_file);
>
> - /* see comment in ceph_lock */
> - fl->fl_ops = &ceph_fl_lock_ops;
> - atomic_inc(&ceph_inode(inode)->i_filelock_ref);
> + spin_lock(&ci->i_ceph_lock);
> + if (ci->i_ceph_flags & CEPH_I_ERROR_FILELOCK) {
> + err = -EIO;
> + } else {
> + /* see comment in ceph_lock */
> + fl->fl_ops = &ceph_fl_lock_ops;
> + atomic_inc(&ci->i_filelock_ref);
> + }
> + spin_unlock(&ci->i_ceph_lock);
> + if (err < 0) {
> + if (F_UNLCK == fl->fl_type)
> + locks_lock_file_wait(file, fl);
> + return err;
> + }
>
> if (IS_SETLKW(cmd))
> wait = 1;
> diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
> index 817ea438aa19..26893cc1fbee 100644
> --- a/fs/ceph/mds_client.c
> +++ b/fs/ceph/mds_client.c
> @@ -1214,6 +1214,13 @@ static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap,
> }
> spin_unlock(&mdsc->cap_dirty_lock);
>
> + if (atomic_read(&ci->i_filelock_ref) > 0) {
> + /* make further file lock syscall return -EIO */
> + ci->i_ceph_flags |= CEPH_I_ERROR_FILELOCK;
> + pr_warn_ratelimited(" dropping file locks for %p %lld\n",
> + inode, ceph_ino(inode));
> + }
> +
> if (!ci->i_dirty_caps && ci->i_prealloc_cap_flush) {
> list_add(&ci->i_prealloc_cap_flush->i_list, &to_remove);
> ci->i_prealloc_cap_flush = NULL;
> @@ -2828,7 +2835,7 @@ static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap,
> struct ceph_mds_cap_reconnect v2;
> struct ceph_mds_cap_reconnect_v1 v1;
> } rec;
> - struct ceph_inode_info *ci;
> + struct ceph_inode_info *ci = cap->ci;
> struct ceph_reconnect_state *recon_state = arg;
> struct ceph_pagelist *pagelist = recon_state->pagelist;
> char *path;
> @@ -2837,8 +2844,6 @@ static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap,
> u64 snap_follows;
> struct dentry *dentry;
>
> - ci = cap->ci;
> -
> dout(" adding %p ino %llx.%llx cap %p %lld %s\n",
> inode, ceph_vinop(inode), cap, cap->cap_id,
> ceph_cap_string(cap->issued));
> @@ -2871,7 +2876,8 @@ static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap,
> rec.v2.issued = cpu_to_le32(cap->issued);
> rec.v2.snaprealm = cpu_to_le64(ci->i_snap_realm->ino);
> rec.v2.pathbase = cpu_to_le64(pathbase);
> - rec.v2.flock_len = 0;
> + rec.v2.flock_len =
> + (ci->i_ceph_flags & CEPH_I_ERROR_FILELOCK) ? 0 : 1;
> } else {
> rec.v1.cap_id = cpu_to_le64(cap->cap_id);
> rec.v1.wanted = cpu_to_le32(__ceph_caps_wanted(ci));
> @@ -2900,7 +2906,12 @@ static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap,
> u8 struct_v = 0;
>
> encode_again:
> - ceph_count_locks(inode, &num_fcntl_locks, &num_flock_locks);
> + if (rec.v2.flock_len) {
> + ceph_count_locks(inode, &num_fcntl_locks, &num_flock_locks);
> + } else {
> + num_fcntl_locks = 0;
> + num_flock_locks = 0;
> + }
> if (num_fcntl_locks + num_flock_locks > 0) {
> flocks = kmalloc((num_fcntl_locks + num_flock_locks) *
> sizeof(struct ceph_filelock), GFP_NOFS);
> diff --git a/fs/ceph/super.h b/fs/ceph/super.h
> index 88cbc0981c23..6fe4a2c5fd24 100644
> --- a/fs/ceph/super.h
> +++ b/fs/ceph/super.h
> @@ -487,6 +487,8 @@ static inline struct inode *ceph_find_inode(struct super_block *sb,
> #define CEPH_I_KICK_FLUSH (1 << 9) /* kick flushing caps */
> #define CEPH_I_FLUSH_SNAPS (1 << 10) /* need flush snapss */
> #define CEPH_I_ERROR_WRITE (1 << 11) /* have seen write errors */
> +#define CEPH_I_ERROR_FILELOCK (1 << 12) /* have seen file lock errors */
> +
>
> /*
> * We set the ERROR_WRITE bit when we start seeing write errors on an inode
This is an ugly situation. We don't really have a good notification
mechanism for when this occurs. Older UNIXes would sometimes issue a
SIGLOST to the application when something like this happens.
We don't have anything like that in Linux at the moment, though I know
Anna Schumaker proposed a mechanism like that for NFS a few years ago.
Maybe we should consider resurrecting that effort? You can certainly hit
the same problem with CIFS as well...
In any case, this is as good a way to handle this as any for now. You
can add:
Acked-by: Jeff Layton <jlayton@xxxxxxxxxx>
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
