From: Theodore Ts'o Subject: Re: [PATCH] e4defrag: fix build when posix_fadvise is missing Date: Wed, 1 Jan 2014 12:37:44 -0500 Message-ID: <20140101173744.GD17519@thunk.org> References: <33245e3808058c72b66931ac14aea8d5dc6d1ba5.1388572103.git.baruch@tkos.co.il> <20140101172205.GC17519@thunk.org> <20140101173146.GM4470@tarshish> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-ext4@vger.kernel.org To: Baruch Siach Return-path: Received: from imap.thunk.org ([74.207.234.97]:44957 "EHLO imap.thunk.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754072AbaAARhs (ORCPT ); Wed, 1 Jan 2014 12:37:48 -0500 Content-Disposition: inline In-Reply-To: <20140101173146.GM4470@tarshish> Sender: linux-ext4-owner@vger.kernel.org List-ID: On Wed, Jan 01, 2014 at 07:31:46PM +0200, Baruch Siach wrote: > > Or do you mean that for some reason, uClibc is not providing > > posix_fadvise on all architectures, even though the kernel supports it? > > > > That seems wierd. > > The xtensa architecture has __NR_fadvise64_64 but not __NR_fadvise64. Should I > clarify this in the commit log? Is uClibc providing a posix_fadvise64()? Maybe the better fix would be to try to use posix_fadvise64 if it is exists, with posix_fadvise() and the locally defined posix_fadvise() as fallbacks. If not, why is uClibc not providing any userspace access to posix_fadvise, if the kernel suppots it, and it's insisting on providing a function delcaration for it? - Ted