From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Bottomley Subject: Re: [RFC] put symbolic links between drivers and modules in the sysfs tree Date: 26 Sep 2004 09:09:03 -0400 Sender: linux-scsi-owner@vger.kernel.org Message-ID: <1096204149.10924.2.camel@mulgrave> References: Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: Received: from stat16.steeleye.com ([209.192.50.48]:18371 "EHLO hancock.sc.steeleye.com") by vger.kernel.org with ESMTP id S269530AbUIZNKH (ORCPT ); Sun, 26 Sep 2004 09:10:07 -0400 In-Reply-To: List-Id: linux-scsi@vger.kernel.org To: Herbert Xu Cc: viro@parcelfarce.linux.theplanet.co.uk, greg@kroah.com, Rusty Russell , Linux Kernel , SCSI Mailing List On Sun, 2004-09-26 at 06:37, Herbert Xu wrote: > James Bottomley wrote: > > > >> So what will your userland code do when you run it on a system with > >> non-modular kernel currently running? > > > > Not put a module in the initial ramdisk, since it would be unnecessary. > > The only information the patch seeks to add is the linkage between > > driver and module. So you can work back from sysfs to know which > > devices have which modules > > You're assuming that the kernel before/after the reboot have the same > configuration. This is false in general. No I'm not. For an initrd/initramfs the only assumption would be that the boot device's driver is compiled in or modular. If this isn't true, the system won't boot anyway. James