From mboxrd@z Thu Jan  1 00:00:00 1970
From: Christoph Hellwig <hch@infradead.org>
Subject: Re: [PATCH RFC 2/2] IB/hfi1: Fix port ordering issue in a multiport
 device
Date: Mon, 6 Feb 2017 13:04:25 -0800
Message-ID: <20170206210425.GA2062@infradead.org>
References: <148640852878.13713.11121939459524566022.stgit@tstruk-mobl1.ra.intel.com>
 <148640853387.13713.5705865367460168894.stgit@tstruk-mobl1.ra.intel.com>
 <20170206192642.GA22482@obsidianresearch.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Return-path: <linux-kernel-owner@vger.kernel.org>
Content-Disposition: inline
In-Reply-To: <20170206192642.GA22482@obsidianresearch.com>
Sender: linux-kernel-owner@vger.kernel.org
To: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Cc: Tadeusz Struk <tadeusz.struk@intel.com>, gregkh@linuxfoundation.org, linux-rdma@vger.kernel.org, dennis.dalessandro@intel.com, linux-kernel@vger.kernel.org, grant.likely@secretlab.ca, dledford@redhat.com, ira.weiny@intel.com
List-Id: linux-rdma@vger.kernel.org

On Mon, Feb 06, 2017 at 12:26:42PM -0700, Jason Gunthorpe wrote:
> Still no on this from me, this is a horrible abuse of the device core.

Seconded.  Netdevice (and thus RDMA) devices names generated by the
kernel are not stable, period.  Use udev and the MAC / port GUID or
whatever the OFI equivalent is to generate stable device names.