From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from zeniv.linux.org.uk ([195.92.253.2]:53896 "EHLO ZenIV.linux.org.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753545AbeEWBnY (ORCPT ); Tue, 22 May 2018 21:43:24 -0400 Date: Wed, 23 May 2018 02:43:18 +0100 From: Al Viro To: Linus Torvalds Cc: Avi Kivity , linux-aio@kvack.org, linux-fsdevel@vger.kernel.org, netdev@vger.kernel.org, linux-api@vger.kernel.org, linux-kernel@vger.kernel.org, Kent Overstreet , Christoph Hellwig Subject: YAaioRace (was Re: [PATCH 08/31] aio: implement IOCB_CMD_POLL) Message-ID: <20180523014318.GI30522@ZenIV.linux.org.uk> References: <20180522113108.25713-1-hch@lst.de> <20180522113108.25713-9-hch@lst.de> <20180522220524.GE30522@ZenIV.linux.org.uk> <20180523004530.GG30522@ZenIV.linux.org.uk> <20180523004904.GH30522@ZenIV.linux.org.uk> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20180523004904.GH30522@ZenIV.linux.org.uk> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Wed, May 23, 2018 at 01:49:04AM +0100, Al Viro wrote: > > Looks like we want to call ->ki_cancel() *BEFORE* removing from the list, > > as well as doing fput() after aio_complete(). The same ordering, BTW, goes > > for aio_read() et.al. > > > > Look: > > CPU1: io_cancel() grabs ->ctx_lock, finds iocb and removes it from the list. > > CPU2: aio_rw_complete() on that iocb. Since the sucker is not in the list > > anymore, we do NOT spin on ->ctx_lock and proceed to free iocb > > CPU1: pass freed iocb to ->ki_cancel(). BOOM. > > BTW, it seems that the mainline is vulnerable to this one. I might be > missing something, but... It is, but with a different attack vector - io_cancel(2) won't do it (it does not remove from the list at all), but io_destroy(2) bloody well will. IMO, we need this in mainline; unless somebody has a problem with it, to #fixes it goes: fix io_destroy()/aio_complete() race If io_destroy() gets to cancelling everything that can be cancelled and gets to kiocb_cancel() calling the function driver has left in ->ki_cancel, it becomes vulnerable to a race with IO completion. At that point req is already taken off the list and aio_complete() does *NOT* spin until we (in free_ioctx_users()) releases ->ctx_lock. As the result, it proceeds to kiocb_free(), freing req just it gets passed to ->ki_cancel(). Fix is simple - remove from the list after the call of kiocb_cancel(). All instances of ->ki_cancel() already have to cope with the being called with iocb still on list - that's what happens in io_cancel(2). Cc: stable@kernel.org Fixes: 0460fef2a921 "aio: use cancellation list lazily" Signed-off-by: Al Viro --- diff --git a/fs/aio.c b/fs/aio.c index 8061d9787e54..49f53516eef0 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -634,9 +634,8 @@ static void free_ioctx_users(struct percpu_ref *ref) while (!list_empty(&ctx->active_reqs)) { req = list_first_entry(&ctx->active_reqs, struct aio_kiocb, ki_list); - - list_del_init(&req->ki_list); kiocb_cancel(req); + list_del_init(&req->ki_list); } spin_unlock_irq(&ctx->ctx_lock);