From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1760107AbcIXCAQ (ORCPT ); Fri, 23 Sep 2016 22:00:16 -0400 Received: from [195.159.176.226] ([195.159.176.226]:55836 "EHLO blaine.gmane.org" rhost-flags-FAIL-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1752256AbcIXCAN (ORCPT ); Fri, 23 Sep 2016 22:00:13 -0400 X-Injected-Via-Gmane: http://gmane.org/ To: linux-kernel@vger.kernel.org From: "Herbert, Marc" Subject: Re: [RFC] fs: add userspace critical mounts event support Date: Fri, 23 Sep 2016 18:37:28 -0700 Message-ID: <2deae6da-dd43-7bff-e1fd-ffd26946b928@intel.com> References: <1466117661-22075-3-git-send-email-mcgrof@kernel.org> <20160824203901.GT3296@wotan.suse.de> <20160825194133.GC3296@wotan.suse.de> <20160902235916.GO3296@wotan.suse.de> <20160903002014.GP3296@wotan.suse.de> <20160903174939.GB32345@dtor-ws> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit X-Complaints-To: usenet@blaine.gmane.org User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:45.0) Gecko/20100101 Thunderbird/45.1.1 Cc: linux-doc@vger.kernel.org, cocci@systeme.lip6.fr, linuxppc-dev@ozlabs.org, linux-serial@vger.kernel.org Cc: linux-doc@vger.kernel.org, cocci@systeme.lip6.fr, linux-serial@vger.kernel.org, linux-kernel@vger.kernel.org In-Reply-To: Cc: cocci@systeme.lip6.fr, linuxppc-dev@ozlabs.org, linux-serial@vger.kernel.org, linux-kernel@vger.kernel.org Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 03/09/2016 11:10, Dmitry Torokhov wrote: > I was thinking if we kernel could post > "conditions" (maybe simple stings) that it waits for, and userspace > could unlock these "conditions". One of them might be "firmware > available". On idea offered by Josh Triplett that seems to overlap with this one is to have something similar to the (deprecated) userhelper with *per-blob* requests and notifications except for one major difference: userspace would not anymore be in charge of *providing* the blob but would instead only *signal* when a given blob becomes available and is either found or found missing. Then the kernel loads the blob _by itself_; unlike the userhelper. No new “critical filesystem” concept and a *per-blob basis*, allowing any variation of blob locations across any number of initramfs and filesystems. Could this one fly?