[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1828884A29C6694DAF28B7E6B8A82373A8FC24E3@ORSMSX109.amr.corp.intel.com>
Date: Thu, 16 Apr 2015 17:21:18 +0000
From: "Hefty, Sean" <sean.hefty@...el.com>
To: Roland Dreier <roland@...nel.org>,
Jason Gunthorpe <jgunthorpe@...idianresearch.com>
CC: Michael Wang <yun.wang@...fitbricks.com>,
"Weiny, Ira" <ira.weiny@...el.com>,
Hal Rosenstock <hal.rosenstock@...il.com>,
"linux-rdma@...r.kernel.org" <linux-rdma@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
Tom Tucker <tom@...ngridcomputing.com>,
Steve Wise <swise@...ngridcomputing.com>,
Hoang-Nam Nguyen <hnguyen@...ibm.com>,
Christoph Raisch <raisch@...ibm.com>,
infinipath <infinipath@...el.com>, Eli Cohen <eli@...lanox.com>,
"Latif, Faisal" <faisal.latif@...el.com>,
Jack Morgenstein <jackm@....mellanox.co.il>,
Or Gerlitz <ogerlitz@...lanox.com>,
Haggai Eran <haggaie@...lanox.com>,
Tom Talpey <tom@...pey.com>, Doug Ledford <dledford@...hat.com>
Subject: RE: [PATCH v3 07/28] IB/Verbs: Reform IB-ulp ipoib
> > No idea, but having ib_register_device fail and unwind if a client
> > fails to attach makes sense to me.
>
> It seems a bit unfriendly to fail an entire device if one ULP has a
> problem. Let's say you have a system whose main network connection is
> IPoIB. Would you want that connection to come up even if, say, the
> NFS/RDMA server fails to find the memory registration type it likes?
What's missing is some way for the device to indicate which modules are actually necessary for it to run. Without having something like that, I agree with Roland.
- Sean
Powered by blists - more mailing lists