Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756895Ab3CNLPJ (ORCPT ); Thu, 14 Mar 2013 07:15:09 -0400 Received: from mail-la0-f51.google.com ([209.85.215.51]:55955 "EHLO mail-la0-f51.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756693Ab3CNLPH (ORCPT ); Thu, 14 Mar 2013 07:15:07 -0400 MIME-Version: 1.0 X-Originating-IP: [188.6.195.195] In-Reply-To: <20130313154431.e39aed27efb095f9183bc123@linux-foundation.org> References: <1363184193-1796-1-git-send-email-miklos@szeredi.hu> <1363184193-1796-2-git-send-email-miklos@szeredi.hu> <20130313154431.e39aed27efb095f9183bc123@linux-foundation.org> Date: Thu, 14 Mar 2013 12:15:04 +0100 Message-ID: Subject: Re: [PATCH 1/9] vfs: add i_op->dentry_open() From: Miklos Szeredi To: Andrew Morton Cc: viro@zeniv.linux.org.uk, torvalds@linux-foundation.org, linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org, hch@infradead.org, apw@canonical.com, nbd@openwrt.org, neilb@suse.de, jordipujolp@gmail.com, ezk@fsl.cs.sunysb.edu, dhowells@redhat.com, sedat.dilek@googlemail.com, hooanon05@yahoo.co.jp, mszeredi@suse.cz Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2407 Lines: 72 On Wed, Mar 13, 2013 at 11:44 PM, Andrew Morton wrote: > On Wed, 13 Mar 2013 15:16:25 +0100 Miklos Szeredi wrote: > >> From: Miklos Szeredi >> >> Add a new inode operation i_op->dentry_open(). This is for stacked filesystems >> that want to return a struct file from a different filesystem. >> >> ... >> >> +/** >> + * vfs_open - open the file at the given path >> + * @path: path to open >> + * @filp: newly allocated file with f_flag initialized >> + * @cred: credentials to use >> + */ >> +int vfs_open(const struct path *path, struct file *filp, >> + const struct cred *cred) >> +{ >> + struct inode *inode = path->dentry->d_inode; >> + >> + if (inode->i_op->dentry_open) >> + return inode->i_op->dentry_open(path->dentry, filp, cred); >> + else { >> + filp->f_path = *path; >> + return do_dentry_open(filp, NULL, cred); >> + } >> +} > > This looks like it will add some overhead to dentry_open(). That long > walk path->dentry->d_inode->i_op->dentry_open, particularly. Can we do > anything? Using filp->f_inode might save a tad. filp->f_inode is initialized in do_dentry_open. But we can move that to callers. Adding an IOP_DENTRY_OPEN flag should take care of the rest. > It's unfortunate and a bit ugly that ->dentry_open() and > do_dentry_open() don't have the same arguments. One would expect and > like them to do so, and this difference raises suspicions about design > warts. Hmm, the basic reason for the difference is that filesystems should not have access to the vfsmount part of the path. Otherwise it would be trivial to make them match up. > > If they _did_ have the same signature then we could perhaps populate > ->dentry_open with do_dentry_open by default and avoid the `if'. Except there's no way to add a default i_op, AFAIK. > > The test of inode->i_op->dentry_open could do with an unlikely(), but > that won't save us :( > >> +EXPORT_SYMBOL(vfs_open); > > Did you consider EXPORT_SYMBOL_GPL()? It is not clear to me what to base that decision on. Either one is fine by me. Thanks, Miklos -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/