[dm-devel] [PATCH 04/12] libmultipath: cleanup features handling code

[Martin Wilck]

On Thu, 2017-12-07 at 12:48 -0600, Benjamin Marzinski wrote:

> retain_attached_hw_handler was never getting updated before, so
> the output when you created a map was incorrect.

While I've already ACKed your patch, I don't understand what you mean
here. Before your patch, "retain_attached_hw_handler" was set from
config options and correctly copied to the features string in
assemble_map, no?

> diff --git a/libmultipath/configure.c b/libmultipath/configure.c
> index 0dfa250..7ca84b8 100644
> --- a/libmultipath/configure.c
> +++ b/libmultipath/configure.c
> @@ -1060,21 +1062,6 @@ int coalesce_paths (struct vectors * vecs,
> vector newmp, char * refwwid,
>  				remove_feature(&mpp->features,
>  					       "queue_if_no_path");
>  		}
> -		else if (mpp->no_path_retry != NO_PATH_RETRY_UNDEF)
> {
> -			if (mpp->no_path_retry ==
> NO_PATH_RETRY_FAIL) {
> -				condlog(3, "%s: unset
> queue_if_no_path feature",
> -					mpp->alias);
> -				if (!dm_queue_if_no_path(mpp->alias, 
> 0))
> -					remove_feature(&mpp-
> >features,
> -						       "queue_if_no_
> path");
> -			} else {
> -				condlog(3, "%s: set queue_if_no_path
> feature",
> -					mpp->alias);
> -				if (!dm_queue_if_no_path(mpp->alias, 
> 1))
> -					add_feature(&mpp->features,
> -						    "queue_if_no_pat
> h");
> -			}
> -		}

AFAICS we could also get rid of the calls to dm_queue_if_no_path() in 
reload_map(), right?

Martin

-- 
Dr. Martin Wilck <mwilck at suse.com>, Tel. +49 (0)911 74053 2107
SUSE Linux GmbH, GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)