On Jun 11, 2012, at 10:51 AM, Boaz Harrosh wrote: > On 06/11/2012 05:21 PM, Adamson, Andy wrote: > >> >> On Jun 11, 2012, at 10:04 AM, Benny Halevy wrote: >> >>> On 2012-06-11 13:44, Boaz Harrosh wrote: >>>> Basically my plan is to postpone the LAYOUTRETURN send to >>>> when the segment is last referenced, if a flag was set to >>>> do so, and make sure reference counting is done proper, so >>>> after flag set the last IO_done will send the LAYOUTRETURN. >>>> >>>> If that is OK with all I hope? >>> >>> Sounds OK to me. Thanks! >> >> Sounds fine to me as well, for LAYOUTRETURN during normal behavior. >> >> -->Andy > > > I'm not sure I understood what you meant. > > In the "normal behavior" we don't send a LAYOUTRETURN at all > in the forgetful model. Which proved beneficiary in the light > of concurrent GETs vs RETURNs. > > I thought we are only taking about the error case. Do you > mean there is a 3rd case? please explain. The normal case of calling LAYOUTRETURN on evict inode. -->Andy > > Thanks > Boaz -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
