Re: [PATCH 8/9] pnfs/blocklayout: return layouts on setattr

[Peng Tao <bergwolf@primarydata.com>]

On Wed, Sep 10, 2014 at 11:23 PM, Christoph Hellwig <hch@lst.de> wrote:
> This speads up truncate-heavy workloads like fsx by multiple orders of
> magnitude.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
>  fs/nfs/blocklayout/blocklayout.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/fs/nfs/blocklayout/blocklayout.c b/fs/nfs/blocklayout/blocklayout.c
> index a7524c4..d5a2b87 100644
> --- a/fs/nfs/blocklayout/blocklayout.c
> +++ b/fs/nfs/blocklayout/blocklayout.c
> @@ -799,7 +799,8 @@ static struct pnfs_layoutdriver_type blocklayout_type = {
>         .id                             = LAYOUT_BLOCK_VOLUME,
>         .name                           = "LAYOUT_BLOCK_VOLUME",
>         .owner                          = THIS_MODULE,
> -       .flags                          = PNFS_READ_WHOLE_PAGE,
> +       .flags                          = PNFS_LAYOUTRET_ON_SETATTR |
> +                                         PNFS_READ_WHOLE_PAGE,
The reason I didn't add it was because PNFS_LAYOUTRET_ON_SETATTR is
too much for blocks layout. What we really want is to return layouts
on truncate and chown, instead of _all_ setattr requests.

Boaz, does object layout require return on setattr for other reasons?
If not, I'd suggest we change PNFS_LAYOUTRET_ON_SETATTR to return only
on chown/truncate events.

Thanks,
Tao