Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Sun, 30 Aug 2020 20:46:46 -0400
From:   Qian Cai <cai@lca.pw>
To:     "Darrick J. Wong" <darrick.wong@oracle.com>
Cc:     hch@infradead.org, linux-xfs@vger.kernel.org,
        linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] iomap: fix WARN_ON_ONCE() from unprivileged users
Message-ID: <20200831004645.GB4039@lca.pw>
References: <20200830013057.14664-1-cai@lca.pw>
 <20200831003033.GZ6096@magnolia>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20200831003033.GZ6096@magnolia>
User-Agent: Mutt/1.10.1 (2018-07-13)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Sun, Aug 30, 2020 at 05:30:33PM -0700, Darrick J. Wong wrote:
> On Sat, Aug 29, 2020 at 09:30:57PM -0400, Qian Cai wrote:
> > It is trivial to trigger a WARN_ON_ONCE(1) in iomap_dio_actor() by
> > unprivileged users which would taint the kernel, or worse - panic if
> > panic_on_warn or panic_on_taint is set. Hence, just convert it to
> > pr_warn_ratelimited() to let users know their workloads are racing.
> > Thanks Dave Chinner for the initial analysis of the racing reproducers.
> > 
> > Signed-off-by: Qian Cai <cai@lca.pw>
> > ---
> >  fs/iomap/direct-io.c | 9 ++++++++-
> >  1 file changed, 8 insertions(+), 1 deletion(-)
> > 
> > diff --git a/fs/iomap/direct-io.c b/fs/iomap/direct-io.c
> > index c1aafb2ab990..6a6b4bc13269 100644
> > --- a/fs/iomap/direct-io.c
> > +++ b/fs/iomap/direct-io.c
> > @@ -389,7 +389,14 @@ iomap_dio_actor(struct inode *inode, loff_t pos, loff_t length,
> >  	case IOMAP_INLINE:
> >  		return iomap_dio_inline_actor(inode, pos, length, dio, iomap);
> >  	default:
> > -		WARN_ON_ONCE(1);
> > +		/*
> > +		 * DIO is not serialised against mmap() access at all, and so
> > +		 * if the page_mkwrite occurs between the writeback and the
> > +		 * iomap_apply() call in the DIO path, then it will see the
> > +		 * DELALLOC block that the page-mkwrite allocated.
> > +		 */
> > +		pr_warn_ratelimited("page_mkwrite() is racing with DIO read (iomap->type = %u).\n",
> > +				    iomap->type);
> 
> Shouldn't we log the name of triggering process and the file path?  Sort
> of like what dio_warn_stale_pagecache does?

Good idea. I'll add them.