[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

[dm-devel] Re: [PATCH 3/3] Add timeout feature

Hi, Andrew and Oleg.

Andrew Morton wrote:
On Mon, 18 Aug 2008 21:28:56 +0900
Takashi Sato <t-sato yk jp nec com> wrote:

The timeout feature is added to freeze ioctl.  And new ioctl
to reset the timeout period is added.
o Freeze the filesystem
  int ioctl(int fd, int FIFREEZE, long *timeout_sec)
    fd: The file descriptor of the mountpoint
    FIFREEZE: request code for the freeze
    timeout_sec: the timeout period in seconds
             If it's 0 or 1, the timeout isn't set.
             This special case of "1" is implemented to keep
             the compatibility with XFS applications.
    Return value: 0 if the operation succeeds. Otherwise, -1

o Reset the timeout period
  int ioctl(int fd, int FIFREEZE_RESET_TIMEOUT, long *timeout_sec)
    fd:file descriptor of mountpoint
    FIFREEZE_RESET_TIMEOUT: request code for reset of timeout period
    timeout_sec: new timeout period in seconds
    Return value: 0 if the operation succeeds. Otherwise, -1
    Error number: If the filesystem has already been unfrozen,
                  errno is set to EINVAL.

I don't think the changelogs actually explained why this feature is
being added?

I will fix the changelogs as following.

The timeout feature is added to "freeze ioctl" to solve a deadlock
when the freezer accesses a frozen filesystem. And new ioctl
to reset the timeout period is added to extend the timeout period.
For example, the freezer resets the timeout period to 10 seconds every 5
seconds.  In this approach, even if the freezer causes a deadlock by
accessing the frozen filesystem, it will be solved by the timeout
in 10 seconds and the freezer will be able to recognize that
at the next reset of timeout period.

Which userspace tools are expected to send these ioctls?  Something in
util-linux?  dm-utils?  Are patches to those packages planned?

I think the management software for the storage device
will use these ioctls to take a consistent backup.


+ * ioctl_freeze_reset_timeout - Reset timeout for freeze.
+ *
+ * @filp:       target file
+ * @argp:       timeout value(sec)
+ *
+ * Reset timeout for freeze.
+ */
+static int
+ioctl_freeze_reset_timeout(struct file *filp, int __user *argp)
+    int timeout_sec;
+    unsigned int timeout_msec;
+    struct super_block *sb = filp->f_path.dentry->d_inode->i_sb;
+    struct block_device *bdev = sb->s_bdev;
+    int error;
+    if (!capable(CAP_SYS_ADMIN))
+            return -EPERM;
+    /* If a regular file or a directory isn't specified, return EINVAL. */
+    if (bdev == NULL)
+            return -EINVAL;
+    /* arg(sec) to tick value */
+    error = get_user(timeout_sec, argp);
+    if (error)
+            return error;
+    if (timeout_sec <= 0 || timeout_sec > UINT_MAX/1000)
+            return -EINVAL;
+    timeout_msec = timeout_sec * 1000;
+    down(&bdev->bd_freeze_sem);
+    if (!bdev->bd_freeze_count) {
+            up(&bdev->bd_freeze_sem);
+            return -EINVAL;
+    }
+    /* setup unfreeze timer */
+    add_freeze_timeout(bdev, timeout_msec);
+    up(&bdev->bd_freeze_sem);
+    return 0;

This duplicates quite a bit of code from ioctl_freeze().  Can this be
cleaned up?

I will clean it up.

  * When you add any new common ioctls to the switches above and below
  * please update compat_sys_ioctl() too.
@@ -235,13 +302,17 @@ int do_vfs_ioctl(struct file *filp, unsi
-            error = ioctl_freeze(filp);
+            error = ioctl_freeze(filp, argp);
case FITHAW:
             error = ioctl_thaw(filp);
+            error = ioctl_freeze_reset_timeout(filp, argp);
+            break;
             if (S_ISREG(filp->f_path.dentry->d_inode->i_mode))
                     error = file_ioctl(filp, cmd, arg);


+ * freeze_timeout - Thaw the filesystem.
+ *
+ * @work:   work queue (delayed_work.work)
+ *
+ * Called by the delayed work when elapsing the timeout period.
+ * Thaw the filesystem.
+ */
+void freeze_timeout(struct work_struct *work)
+    struct block_device *bd = container_of(work,
+                    struct block_device, bd_freeze_timeout.work);
+    struct super_block *sb = get_super(bd);
+    thaw_bdev(bd, sb);
+    if (sb)
+            drop_super(sb);

I can't see why this was exported.

It isn't needed, I will delete it.

+ * add_freeze_timeout - Add timeout for freeze.
+ *
+ * @bdev:           block device struct
+ * @timeout_msec:   timeout period
+ *
+ * Add the delayed work for freeze timeout to the delayed work queue.
+ */
+void add_freeze_timeout(struct block_device *bdev, unsigned int timeout_msec)
+    s64 timeout_jiffies = msecs_to_jiffies(timeout_msec);
+    /* Set delayed work queue */
+    cancel_delayed_work_sync(&bdev->bd_freeze_timeout);
+    schedule_delayed_work(&bdev->bd_freeze_timeout, timeout_jiffies);

I don't particularly like the names of these new global symbols.  The
kernel already has a "freezer" thing, part of power-management. Introducing another one is a bit confusing.

otoh, freezer seems to have consistently used "freezer", so the 'r'
arguable saves us.

Still, I'd have thought that "fsfreeze" would have been a clearer, more
specific identifier for the whole project.

I will rename the names of these global symbols.
For example, "add_fsfreeze_timeout"...

+ * del_freeze_timeout - Delete timeout for freeze.
+ *
+ * @bdev:   block device struct
+ *
+ * Delete the delayed work for freeze timeout from the delayed work queue.
+ */
+void del_freeze_timeout(struct block_device *bdev)
+    /*
+     * It's possible that the delayed work task (freeze_timeout()) calls
+     * del_freeze_timeout().  If the delayed work task calls
+     * cancel_delayed_work_sync((), the deadlock will occur.
+     * So we need this check (delayed_work_pending()).
+     */
+    if (delayed_work_pending(&bdev->bd_freeze_timeout))
+            cancel_delayed_work_sync(&bdev->bd_freeze_timeout);

So if the calling task is keventd via run_workqueue() then
delayed_work_pending() should return false due to run_workqueue()
ordering, so we avoid the deadlock.

Seems a bit racy if some other process starts the delayed-work while
this function is running but I guess the new semaphore prevents that.

Perhaps cancel_delayed_work_sync() shouldn't hang up if called from the
work handler?

I think so.

In my current implementation,
the delayed work task calls thaw_bdev() to unfreeze a filesystem
and it calls del_freeze_timeout().  So, the deadlock occurs.
But I've found that the delayed work task doesn't need to call
del_freeze_timeout() because it is removed
from the delayed work queue automatically after the work finishes.
So I will fix thaw_bdev() so that it doesn't call
del_freeze_timeout() only when it's called by the delayed work task.
And I will delete delayed_work_pending() in del_freeze_timeout().

Cheers, Takashi

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]