[Open-FCoE] [PATCH v2] libfc, fcoe, fnic: Separate rport and lport max retry counts

Abhijeet Arvind Joglekar (abjoglek) abjoglek at cisco.com
Tue Apr 21 16:52:58 UTC 2009


> -----Original Message-----
> From: Love, Robert W [mailto:robert.w.love at intel.com] 
> Sent: Thursday, April 16, 2009 2:03 PM
> To: Abhijeet Arvind Joglekar (abjoglek)
> Cc: devel at open-fcoe.org
> Subject: RE: [Open-FCoE] [PATCH v2] libfc, fcoe, fnic: 
> Separate rport and lport max retry counts
> 
> Abhijeet Arvind Joglekar (abjoglek) wrote:
> >> -----Original Message-----
> >> From: Robert Love [mailto:robert.w.love at intel.com]
> >> Sent: Thursday, April 16, 2009 12:44 PM
> >> To: Abhijeet Arvind Joglekar (abjoglek)
> >> Cc: devel at open-fcoe.org
> >> Subject: Re: [Open-FCoE] [PATCH v2] libfc, fcoe, fnic:
> >> Separate rport and lport max retry counts
> >> 
> >> On Thu, 2009-04-16 at 12:40 -0700, Robert Love wrote:
> >>> From: Abhijeet Joglekar <abjoglek at cisco.com>
> >>> 
> >>> This allows number of retries to be configured separately 
> for lport 
> >>> and rport. This is preparation for fnic which will allow 
> independent 
> >>> configuration of lport and rport retries.
> >>> 
> >> This is just an update to the patch. I'm confused what you 
> want me to 
> >> do with this patch- push it for 30-rc or 31 merge?
> > 
> >> I didn't OK the patch as a fix before since it's not a fix. I 
> >> suggested that it should go with fnic since this is to 
> support a fnic 
> >> feature.
> > 
> > I will submit fnic for 30-merge tomorrow. However, fnic won't have 
> > this separate rport retry counter, since the libFC patch to 
> enable it 
> > (which is this patch) is not in scsi-misc yet.
> > 
> > Once fnic goes in scsi-misc, and gets pulled into fcoe-features and 
> > fcoe-fixes through scsi-misc, we could push this patch out for 31 
> > merge or 30-rc (if James is ok with it).
> > 
> Would it work for you to just add this patch to your fnic set 
> and I'll pull your fnic set into -features when you've put it 
> out to linux-scsi to make sure our other features don't 
> conflict? Otherwise I'd want to push this with updated 
> logging patches for the 31 merge window.
> 
> I like it as a part of the fnic patch set for two reasons. 1) 
> it's supporting a fnic feature and 2) it doesn't make that 
> feature dependent on this dangling patch. This prep work 
> patch and the corresponding fnic feature can then go in at 
> the same time.
> 
> I'm fine with this change, so you can add my Signed-off line 
> to the patch if it's in the fnic set.

I submitted fnic without this patch, since the libFC part of it was not
in scsi-misc.

Once fnic makes it into fcoe-features via scsi-misc, I can re-submit
this patch with changes to all 3 components: libfc,fcoe,fnic. If James
B. is ok with it, it can go in 30-rc or in 31-merge.

-- abhijeet 



More information about the devel mailing list