From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jason Gunthorpe Subject: Re: if/how to dictate IB device name per PCI BDF Date: Fri, 12 Oct 2012 16:43:32 -0600 Message-ID: <20121012224332.GB25541@obsidianresearch.com> References: <5076A755.30605@mellanox.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <5076A755.30605-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org> Sender: linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Or Gerlitz Cc: Roland Dreier , "linux-rdma (linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org)" , Shlomo Pongratz , Eli Cohen List-Id: linux-rdma@vger.kernel.org On Thu, Oct 11, 2012 at 01:02:45PM +0200, Or Gerlitz wrote: > Thinking about this a bit, will it be possible to provide a udev > rulethat can dictate to the IB core what name / suffix digit to > assign for a device with certain BDF? FWIW, I reflected on this once (I belive it was for the netlink discussion).. Allowing for a device to be renamed creates a serious problem for RDMA since there isn't a stable 'if-index' like way for software to refer to it. The udev rules all rely on a means to rename the device once the kernel has created it. For this reason, all the software I've built uses the port GID to ID the resource for bind/etc and strongly discourages use of the RDMA device name. The ipoib devices should already be renamable via udev rules, though has anyone checked that udev doesn't have a problem with the long 'MAC'? Jason -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html