From mboxrd@z Thu Jan 1 00:00:00 1970 From: ebiederm@xmission.com (Eric W. Biederman) Subject: Re: [PATCH 08/11] vfs: Merge check_submounts_and_drop and d_invalidate Date: Tue, 25 Feb 2014 14:03:36 -0800 Message-ID: <87mwhe26kn.fsf@xmission.com> References: <87a9kkax0j.fsf@xmission.com> <8761v7h2pt.fsf@tw-ebiederman.twitter.com> <87li281wx6.fsf_-_@xmission.com> <87ob28kqks.fsf_-_@xmission.com> <87eh34jbsl.fsf_-_@xmission.com> <20140218174053.GE4026@tucsk.piliscsaba.szeredi.hu> <87y510dpra.fsf@xmission.com> <20140225151349.GA19981@fieldses.org> Mime-Version: 1.0 Content-Type: text/plain Cc: Miklos Szeredi , Al Viro , "Serge E. Hallyn" , Linux-Fsdevel , Kernel Mailing List , Andy Lutomirski , Rob Landley , Linus Torvalds , Christoph Hellwig , Karel Zak To: "J. Bruce Fields" Return-path: In-Reply-To: <20140225151349.GA19981@fieldses.org> (J. Bruce Fields's message of "Tue, 25 Feb 2014 10:13:49 -0500") Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org "J. Bruce Fields" writes: > On Mon, Feb 24, 2014 at 04:01:29PM -0800, Eric W. Biederman wrote: >> Miklos Szeredi writes: >> >> > >> > You can optimize this by including the negative check within the above d_locked >> > region and calling __d_drop() instead. >> >> For this patch just moving the code and not changing it is the corret >> thing to do because it helps with review and understanding the code. >> >> There are two ways I could see going with optimizing the preamble. >> Simply dropping the d_lock from around the d_unhashed test as a pointer >> dereference should be atomic, and the test is racy against >> d_materialise_unique. > > Could you explain? What's the race, and what are the consequences? >> (We don't always hold the parent directories inode mutex when d_invalidate is called). d_unhashed is not a permanent condition because of d_materialise_unique, and d_splice_alias. d_invalidate can be called on an unhashed dentry in one of two ways (either d_revalidate dropped the dentry or another routine that drops the dentry beat the current invocation of d_invalidate to the job). There are 3 places d_revalidate is called. Once on the rcu path with with the appropriate flag set. Once without out the parent i_mutex held, just off of the rcu path, on that path d_invalidate is when d_revalidate fails. Once during lookup with the parent directory i_mutex held. Because the parent direcories i_mutex is not always held accross d_revalidate and the following d_invalidate it happens that d_invalidate is not always an atomic operation. At worst the race results in a dentry that is dropped when it could be hashed, that we will resurrect next time someone attempts to look it up and d_materialise_unique/d_splice_alias is called. None of that really matters for optimizing d_invalidate, but it is part of the background in which d_invalidate lives. All that is significant in d_invalidate is knowing that d_materialise_unique, and possibly d_splice_alias may run concurrently with d_invalidate. It is unlikely and essentially harmless. After my patchset (because I removed all of the d_drop's from .d_revalidate) the only race that should remain is between two parallel calls of d_invalidate. Which probably means we can remove the test for d_unhashed altogether. Right now I just want to make this first big step and make certain the code is solid. After that optimization is easy. Eric