From: David Howells Subject: Re: [git pull] more vfs bits Date: Sun, 22 Feb 2015 12:44:12 +0000 Message-ID: <27542.1424609052@warthog.procyon.org.uk> References: <20150221033452.GU29656@ZenIV.linux.org.uk> <20150222005125.GX29656@ZenIV.linux.org.uk> <20150222020207.GA29656@ZenIV.linux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Cc: dhowells@redhat.com, Al Viro , Linux Kernel Mailing List , linux-fsdevel To: Linus Torvalds Return-path: In-Reply-To: Content-ID: <27541.1424609052.1@warthog.procyon.org.uk> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org Linus Torvalds wrote: > So the ACCESS_ONCE() thing is more special than just "done under RCU". > It's more like "really special case done without any of the normal > locking _or_ any of the normal RCU checks". > > That said, the overhead of using ACCESS_ONCE() is basically nil, so > it's not like we couldn't just start doing more of them, and make it > be more of a "any time we're under RCU" kind of thing. Some functions access ->d_inode more than once. Wouldn't that potentially increase the number of load instructions? Admittedly, calls to dentry->d_inode could be replaced with inode = dentry->d_inode, then use inode. > Yeah, I think "d_backing_store_inode()" would probably be more along > the lines, but that's a mouthful. Maybe shortened to > "d_backing_inode()"? Sounds more reasonable than d_opened_inode(). d_actual_inode() might also work. d_lower_inode() might work too. David