xfs
[Top] [All Lists]

Re: [RFC PATCH 1/2] xfs: add basic per-mount sysfs attribute support

To: Brian Foster <bfoster@xxxxxxxxxx>
Subject: Re: [RFC PATCH 1/2] xfs: add basic per-mount sysfs attribute support
From: Dave Chinner <david@xxxxxxxxxxxxx>
Date: Tue, 20 May 2014 22:16:47 +1000
Cc: xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <1400263966-35541-2-git-send-email-bfoster@xxxxxxxxxx>
References: <1400263966-35541-1-git-send-email-bfoster@xxxxxxxxxx> <1400263966-35541-2-git-send-email-bfoster@xxxxxxxxxx>
User-agent: Mutt/1.5.21 (2010-09-15)
On Fri, May 16, 2014 at 02:12:45PM -0400, Brian Foster wrote:
> Initialize/destroy a kset on module init/fini for XFS. The XFS attribute
> directory is represented as /sys/fs/xfs. Create a subdirectory per-mount
> based on the system device name.
> 
> Also add some basic macros to aid in the addition of sysfs attributes.
> This code is modeled after the equivalent mechanism for ext4.
> 
> Signed-off-by: Brian Foster <bfoster@xxxxxxxxxx>
> ---
>  fs/xfs/xfs_mount.c | 79 
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++
>  fs/xfs/xfs_mount.h |  2 ++
>  fs/xfs/xfs_super.c | 12 ++++++++-
>  3 files changed, 92 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c
> index 944f3d9..9ed9dd0 100644
> --- a/fs/xfs/xfs_mount.c
> +++ b/fs/xfs/xfs_mount.c
> @@ -60,6 +60,9 @@ static DEFINE_MUTEX(xfs_uuid_table_mutex);
>  static int xfs_uuid_table_size;
>  static uuid_t *xfs_uuid_table;
>  
> +extern struct kset *xfs_kset;
> +static struct kobj_type xfs_ktype;
> +
>  /*
>   * See if the UUID is unique among mounted XFS filesystems.
>   * Mount fails if UUID is nil or a FS with the same UUID is already mounted.
> @@ -955,6 +958,13 @@ xfs_mountfs(
>       "Unable to allocate reserve blocks. Continuing without reserve pool.");
>       }
>  
> +     mp->m_kobject.kset = xfs_kset;
> +     init_completion(&mp->m_kobject_complete);
> +     error = kobject_init_and_add(&mp->m_kobject, &xfs_ktype, NULL,
> +                                  "%s", mp->m_fsname);
> +     if (error)
> +             goto out_rtunmount;
> +

I think this should be wrapped in a helper, same for the destruction
code. It shoul dalso be called earlier rather than later, because we
are going to want to add children to this object (e.g. per-ag sysfs
attribute directories), and we probably want to isolate information
from different subsystems to different subdirectories under this.
e.g;

        /sysfs/fs/xfs/vda/....
                     /vdb/log/...
                         /trans/...
                         /ag/0/sb/....
                              /agf/....
                              /agi/....
                              /agfl/....
                            /1/sb/...
                            ...
                            /n/agfl/....
                         /quota/....

Dumping everything in the root directory is probably not a good
idea, and hence we shoul dbe creating the heirarchy as we initialise
subsystems during mount, and freeing them as we tear them down from
unmount...

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx

<Prev in Thread] Current Thread [Next in Thread>