Re: [RFC PATCH v1 10/18] xfs: Enable async write file modification handling.

[Christian Brauner <brauner@kernel.org>]

On Tue, Apr 26, 2022 at 10:43:27AM -0700, Stefan Roesch wrote:
> This modifies xfs write checks to return -EAGAIN if the request either
> requires to remove privileges or needs to update the file modification
> time. This is required for async buffered writes, so the request gets
> handled in the io worker.
> 
> Signed-off-by: Stefan Roesch <shr@fb.com>
> ---
>  fs/xfs/xfs_file.c | 39 ++++++++++++++++++++++++++++++++++++++-
>  1 file changed, 38 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
> index 5bddb1e9e0b3..6f9da1059e8b 100644
> --- a/fs/xfs/xfs_file.c
> +++ b/fs/xfs/xfs_file.c
> @@ -299,6 +299,43 @@ xfs_file_read_iter(
>  	return ret;
>  }
>  
> +static int xfs_file_modified(struct file *file, int flags)

This should probably be in fs/inode.c as:

int file_modified_async(struct file *file, int flags)

and then file_modified() can simply become:

int file_modified(struct file *file)
{
	return file_modified_async(file, 0);
}

or even:

int file_modified_async(struct file *file, bool async)

int file_modified(struct file *file)
{
	return file_modified_async(file, false);
}

to avoid piecing this together specifically in xfs (as Dave mentioned
this is all pretty generic).