From owner-freebsd-hackers Thu Apr 6 09:59:51 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id JAA02090 for hackers-outgoing; Thu, 6 Apr 1995 09:59:51 -0700 Received: from cs.weber.edu (cs.weber.edu [137.190.16.16]) by freefall.cdrom.com (8.6.10/8.6.6) with SMTP id JAA02080 for ; Thu, 6 Apr 1995 09:59:49 -0700 Received: by cs.weber.edu (4.1/SMI-4.1.1) id AA13427; Thu, 6 Apr 95 10:51:55 MDT From: terry@cs.weber.edu (Terry Lambert) Message-Id: <9504061651.AA13427@cs.weber.edu> Subject: Re: devices going offline To: joerg_wunsch@uriah.heep.sax.de Date: Thu, 6 Apr 95 10:51:54 MDT Cc: freebsd-hackers@FreeBSD.org In-Reply-To: <199504052232.AAA07605@uriah.heep.sax.de> from "J Wunsch" at Apr 6, 95 00:32:50 am X-Mailer: ELM [version 2.4dev PL52] Sender: hackers-owner@FreeBSD.org Precedence: bulk > > When an external device has the capability of being manually > > taken offline during an operation, should the driver automatically > > resume when the device returns online? Or, abort the operation > > in progress (signalling an error)? Or, prompt (how?) for user > > intervention?? > > It simply can't do the latter. (A device driver does not have an > interactive terminal, and it does not need to have an attached process > while operating in the ``bottom half'' of the kernel.) > > I'd say: EIO. I'd say Plug-N-Play device management to get a handler that can do the right thing with devices appearing and disappearing in a general way. Like Win95's "Volume Manager", actually a more generic "Resource Manager". The biggie for implementation is callback registration by the drivers, which is a pain for loading/unloading them without doing yet more callback registration. Win95 has some clever ideas, especially in their file system, as silly as that might seem. Terry Lambert terry@cs.weber.edu --- Any opinions in this posting are my own and not those of my present or previous employers.