From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754725AbXEBFeo (ORCPT ); Wed, 2 May 2007 01:34:44 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754819AbXEBFeo (ORCPT ); Wed, 2 May 2007 01:34:44 -0400 Received: from haxent.com ([65.99.219.155]:2153 "EHLO haxent.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754725AbXEBFem (ORCPT ); Wed, 2 May 2007 01:34:42 -0400 Message-Id: <20070502052235.914764000@haxent.com.br> User-Agent: quilt/0.45-1 Date: Wed, 02 May 2007 02:22:35 -0300 From: Davi Arnaut To: Andrew Morton Cc: Davide Libenzi , Linus Torvalds , Linux Kernel Mailing List Subject: [patch 00/22] pollfs: filesystem abstraction for pollable objects Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org This patch set introduces a new file system for the delivery of pollable events through file descriptors. To the detriment of debugability, pollable objects are a nice adjunct to nonblocking/epoll/event-based servers. The pollfs filesystem abstraction provides better mechanisms needed for creating and maintaining pollable objects. Also the pollable futex approach is far superior (send and receive events from userspace or kernel) to eventfd and fixes (supercedes) FUTEX_FD at the same time. The (non) blocking and object size (user <-> kernel) semantics and are handled internally, decoupling the core filesystem from the "subsystems" (mere push and pop operations). Currently implemented waitable "objects" are: signals, futexes, ai/o blocks and timers. More details at each patch. http://haxent.com/~davi/pollfs/ Comments are welcome. -- Davi Arnaut