| Message ID | c3e2f34f0d328196dd7cd6bdae67a8219cef840c.1629780501.git.anand.jain@oracle.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | btrf_show_devname related fixes | expand |
On Tue, Aug 24, 2021 at 01:05:22PM +0800, Anand Jain wrote: > There were few lock dep warnings because btrfs_show_devname() was using > device_list_mutex as recorded in the commits > ccd05285e7f (btrfs: fix a possible umount deadlock) > 779bf3fefa8 (btrfs: fix lock dep warning, move scratch dev out of > device_list_mutex and uuid_mutex) > > And finally, commit 88c14590cdd6 (btrfs: use RCU in btrfs_show_devname > for device list traversal) removed the device_list_mutex from > btrfs_show_devname for performance reasons. > > This patch fixes a stale comment about the function btrfs_show_devname. > > Signed-off-by: Anand Jain <anand.jain@oracle.com> > --- > fs/btrfs/volumes.c | 5 +---- > 1 file changed, 1 insertion(+), 4 deletions(-) > > diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c > index 8470c5b5f35e..1d1204547e72 100644 > --- a/fs/btrfs/volumes.c > +++ b/fs/btrfs/volumes.c > @@ -2278,10 +2278,7 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_device *tgtdev) > > /* > * The update_dev_time() with in btrfs_scratch_superblocks() > - * may lead to a call to btrfs_show_devname() which will try > - * to hold device_list_mutex. And here this device > - * is already out of device list, so we don't have to hold > - * the device_list_mutex lock. > + * may lead to a call to btrfs_show_devname(). I think the whole comment can be deleted, it was there namely for the device_list_mutex and potential problems with show_devname, but this is now sorted. What's remaining is that btrfs_scratch_superblocks happens out of all locks but it's IMO clear from the code that the device is being now disconnected from all structures and freed.
On Tue, Aug 24, 2021 at 01:05:22PM +0800, Anand Jain wrote: > There were few lock dep warnings because btrfs_show_devname() was using > device_list_mutex as recorded in the commits > ccd05285e7f (btrfs: fix a possible umount deadlock) Commit ccd05285e7f does not exist in my tree, but the subject matches 0ccd05285e7f. Please format the references like it's done for the Fixes: tag with HASH ("Subject"). Nobody types that by hand, a simple git alias does that for free: .gitconfig: [alias] one = show -s --pretty='format:%h (\"%s\")'
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 8470c5b5f35e..1d1204547e72 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2278,10 +2278,7 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_device *tgtdev) /* * The update_dev_time() with in btrfs_scratch_superblocks() - * may lead to a call to btrfs_show_devname() which will try - * to hold device_list_mutex. And here this device - * is already out of device list, so we don't have to hold - * the device_list_mutex lock. + * may lead to a call to btrfs_show_devname(). */ btrfs_scratch_superblocks(tgtdev->fs_info, tgtdev->bdev, tgtdev->name->str);
There were few lock dep warnings because btrfs_show_devname() was using device_list_mutex as recorded in the commits ccd05285e7f (btrfs: fix a possible umount deadlock) 779bf3fefa8 (btrfs: fix lock dep warning, move scratch dev out of device_list_mutex and uuid_mutex) And finally, commit 88c14590cdd6 (btrfs: use RCU in btrfs_show_devname for device list traversal) removed the device_list_mutex from btrfs_show_devname for performance reasons. This patch fixes a stale comment about the function btrfs_show_devname. Signed-off-by: Anand Jain <anand.jain@oracle.com> --- fs/btrfs/volumes.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-)