Re: [PATCH 05/24] SQUASHME: ensure pnfs_update_lseg clears lsegp on error

[Boaz Harrosh <bharrosh@panasas.com>]

On 06/08/2010 07:19 AM, Fred Isaman wrote:
> This should be squashed into my (or alexandros's)submission patches for
> version 2. Compensate for Alexandros returning error but assigning lseg.
> 
> Signed-off-by: Fred Isaman <iisaman@netapp.com>
> ---
>  fs/nfs/pnfs.c |    3 +++
>  1 files changed, 3 insertions(+), 0 deletions(-)
> 
> diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
> index 2f8fa3c..b990471 100644
> --- a/fs/nfs/pnfs.c
> +++ b/fs/nfs/pnfs.c
> @@ -1064,6 +1064,8 @@ pnfs_update_layout(struct inode *ino,
>  	DEFINE_WAIT(__wait);
>  	int result = 0;
>  
> +	if (take_ref)
> +		*lsegpp = NULL;
>  	lo = get_lock_alloc_layout(ino);
>  	if (IS_ERR(lo)) {
>  		dprintk("%s ERROR: can't get pnfs_layout_type\n", __func__);
> @@ -1078,6 +1080,7 @@ pnfs_update_layout(struct inode *ino,
>  			put_lseg(lseg);
>  
>  		/* someone is cleaning the layout */
> +		lseg = NULL;
>  		result = -EAGAIN;
>  		goto out_put;
>  	}

Please get these patchset organised.
* Cleanup which go in permanently - first / separate.
* Degrading B to A patches - later / separate.
  Plus the A to B patch if needed or a note that a simple patch -R will do.
  (Or a TODO: note that an A to B patch is missing)

This kind of patch kills any ability to really see what is happening.

And, as far as code progression and understanding goes:
* Fred's patches should go first - We first get a layout.
* Then Alexandros patches ontop. - Then we return / recall a layout.

Boaz