[pnfs] [PATCH 16/20] 2.6-latest update nfs4_pnfs_dserver_get

Dean Hildebrand seattleplus at gmail.com
Mon Dec 3 16:16:23 EST 2007



William A. (Andy) Adamson wrote:
>
>
> On 12/2/07, *Dean Hildebrand* <seattleplus at gmail.com 
> <mailto:seattleplus at gmail.com>> wrote:
>
>
>     >
>     > -     /* For debugging */
>     >       tmp = offset + count - 1;
>     >       do_div(tmp, layout->stripe_unit);
>     > -     dbg_stripe_idx = do_div(tmp, layout->num_fh) +
>     layout->first_stripe_index;
>     > +     end_idx = do_div(tmp, di->stripe_count) +
>     layout->first_stripe_index;
>     >
>     > -     dprintk("%s: offset=%Lu, count=%u, si=%u, dsi=%u, "
>     > -                "num_devs=%u, stripe_unit=%Lu\n",
>     > -                   __FUNCTION__,
>     > -                offset, count, stripe_idx, dbg_stripe_idx,
>     layout->num_fh,
>     > -                layout->stripe_unit);
>     > +     dprintk("%s: offset=%Lu, count=%Zu, si=%u, dsi=%u, "
>     > +             "stripe_count=%u, stripe_unit=%Lu
>     first_stripe_index %d\n",
>     > +             __FUNCTION__,
>     > +             offset, count, stripe_idx, end_idx, di->stripe_count,
>     > +             layout->stripe_unit, layout->first_stripe_index);
>     >
>     > -     BUG_ON(dbg_stripe_idx != stripe_idx);
>     > +     BUG_ON(end_idx != stripe_idx);
>     >
>
>     Since we are going through the code, isn't a bug_on too harsh for
>     this
>     error?  Do we still want this debugging code?  If so, maybe we should
>     make it just return an error.
>
>
> something to consider - i'm following benny's suggestion of not 
> changing too much,
> just port forward.
>
> this code will dramatically change as we move to draft-17 (18, 19??).
Dramatically change?  How dramatic could it be after the last move to 
draft-13?  Has the file layout really changed so much?
Dean
>
> -->Andy
>
>     Dean
>
>     >
>     > -     /* TODO: rewrite this function!
>     > -     dserver->dev = nfs4_pnfs_device_get(inode, layout->dev_id,
>     stripe_idx);
>     > +     dserver->dev = &di->stripe_devs[stripe_idx];
>     >       if (dserver->dev == NULL)
>     >               return 1;
>     > -     */
>     > -     dserver->fh = &layout->fh_array[stripe_idx];
>     > +     if (layout->num_fh == 1)
>     > +             dserver->fh = &layout->fh_array[0];
>     > +     else
>     > +             dserver->fh = &layout->fh_array[stripe_idx];
>     >
>     > -     dprintk("%s: dev_id=%u, idx=%u, offset=%Lu, count=%u\n",
>     > -                  __FUNCTION__, layout->dev_id, stripe_idx,
>     offset, count);
>     > +     dprintk("%s: dev_id=%u, idx=%u, offset=%Lu, count=%Zu\n",
>     > +             __FUNCTION__, layout->dev_id, stripe_idx, offset,
>     count);
>     >
>     >       return 0;
>     > +
>     >  }
>     >
>     >  /* Currently not used.
>     >
>     _______________________________________________
>     pNFS mailing list
>     pNFS at linux-nfs.org <mailto:pNFS at linux-nfs.org>
>     http://linux-nfs.org/cgi-bin/mailman/listinfo/pnfs
>
>


More information about the pNFS mailing list