From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jared Hulbert Subject: Re: [PATCH 14/14] Pramfs: XIP operations Date: Tue, 16 Jun 2009 19:15:03 -0700 Message-ID: <6934efce0906161915t280bf69ek450b3c6220a7d379@mail.gmail.com> References: <4A33A841.6080902@gmail.com> <20090613140647.GD30053@uranus.ravnborg.org> <4A34A3C0.80103@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: Sam Ravnborg , Linux Embedded , Linux Kernel , Linux FS Devel , Daniel Walker To: Marco Return-path: Received: from yw-out-2324.google.com ([74.125.46.29]:28203 "EHLO yw-out-2324.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751373AbZFQCXS (ORCPT ); Tue, 16 Jun 2009 22:23:18 -0400 In-Reply-To: <4A34A3C0.80103@gmail.com> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: > I know. It wasn't my intention to introduce it but as I said in my first > patch I've done a porting of this code from 2.6.10 and to remove it I > need time to analyze well the code to avoid deadlock and so on. If > someone would like to help me I'd really appreciate it. However I see > the use of BKL even in other recent "mainlined" fs as ext4, so I > preferred to move the porting effort on other areas. However it's the > first item on my todo list. Why do you need the lock in pram_find_and_alloc_blocks() to begin with? Why wouldn't a mutex work?