From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dave Chinner Subject: Re: [RFC] killing boilerplate checks in ->link/->mkdir/->rename Date: Tue, 7 Feb 2012 09:49:41 +1100 Message-ID: <20120206224941.GA20305@dastard> References: <20120202012258.GQ23916@ZenIV.linux.org.uk> <20120202212400.GR23916@ZenIV.linux.org.uk> <20120203011612.GS23916@ZenIV.linux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Linus Torvalds , linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org, Joel Becker , Chris Mason , David Miller To: Al Viro Return-path: Content-Disposition: inline In-Reply-To: <20120203011612.GS23916@ZenIV.linux.org.uk> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org On Fri, Feb 03, 2012 at 01:16:12AM +0000, Al Viro wrote: > On Thu, Feb 02, 2012 at 03:46:06PM -0800, Linus Torvalds wrote: > > On Thu, Feb 2, 2012 at 1:24 PM, Al Viro wrote: > * WTF is XFS doing with these checks? It is validating nlink against the maximum supported by the XFS on-disk format. It was originally limited by what could be reported to pathconf() on Irix - a signed int. We have that same problem on Linux, too, because on 32 bit systems the maximum number of links that can be reported via pathconf is 2^31.... > Note that we have them > done _twice_ on all paths - explictly from xfs_create(), xfs_link(), > xfs_rename() and then from xfs_bumplink() called by exactly the same > set of functions. Well, that's a bit stupid, isn't it? Trivial to fix, though... Cheers, Dave. -- Dave Chinner david@fromorbit.com