Re: [PATCH] fsdax: Fix infinite loop in dax_iomap_rw()

From: Dan Williams <dan.j.williams@intel.com>
To: "Darrick J. Wong" <djwong@kernel.org>, Li Jinlin <lijinlin3@huawei.com>
Cc: <viro@zeniv.linux.org.uk>, <dan.j.williams@intel.com>,
	<willy@infradead.org>, <jack@suse.cz>,
	<linux-fsdevel@vger.kernel.org>, <nvdimm@lists.linux.dev>,
	<linux-kernel@vger.kernel.org>, <linfeilong@huawei.com>,
	<liuzhiqiang26@huawei.com>
Subject: Re: [PATCH] fsdax: Fix infinite loop in dax_iomap_rw()
Date: Mon, 25 Jul 2022 16:32:24 -0700	[thread overview]
Message-ID: <62df2808e5c50_1f5536294f2@dwillia2-xfh.jf.intel.com.notmuch> (raw)
In-Reply-To: <Yt8Hw2cXPz1ScQ1y@magnolia>

Darrick J. Wong wrote:
> On Mon, Jul 25, 2022 at 11:20:50AM +0800, Li Jinlin wrote:
> > I got an infinite loop and a WARNING report when executing a tail command
> > in virtiofs.
> > 
> >   WARNING: CPU: 10 PID: 964 at fs/iomap/iter.c:34 iomap_iter+0x3a2/0x3d0
> >   Modules linked in:
> >   CPU: 10 PID: 964 Comm: tail Not tainted 5.19.0-rc7
> >   Call Trace:
> >   <TASK>
> >   dax_iomap_rw+0xea/0x620
> >   ? __this_cpu_preempt_check+0x13/0x20
> >   fuse_dax_read_iter+0x47/0x80
> >   fuse_file_read_iter+0xae/0xd0
> >   new_sync_read+0xfe/0x180
> >   ? 0xffffffff81000000
> >   vfs_read+0x14d/0x1a0
> >   ksys_read+0x6d/0xf0
> >   __x64_sys_read+0x1a/0x20
> >   do_syscall_64+0x3b/0x90
> >   entry_SYSCALL_64_after_hwframe+0x63/0xcd
> > 
> > The tail command will call read() with a count of 0. In this case,
> > iomap_iter() will report this WARNING, and always return 1 which casuing
> > the infinite loop in dax_iomap_rw().
> > 
> > Fixing by checking count whether is 0 in dax_iomap_rw().
> > 
> > Fixes: ca289e0b95af ("fsdax: switch dax_iomap_rw to use iomap_iter")
> > Signed-off-by: Li Jinlin <lijinlin3@huawei.com>
> 
> Huh, I didn't know FUSE supports DAX and iomap now...

Yeah, it came in via DAX support for virtio-fs.

> > ---
> >  fs/dax.c | 3 +++
> >  1 file changed, 3 insertions(+)
> > 
> > diff --git a/fs/dax.c b/fs/dax.c
> > index 4155a6107fa1..7ab248ed21aa 100644
> > --- a/fs/dax.c
> > +++ b/fs/dax.c
> > @@ -1241,6 +1241,9 @@ dax_iomap_rw(struct kiocb *iocb, struct iov_iter *iter,
> >  	loff_t done = 0;
> >  	int ret;
> >  
> > +	if (!iomi.len)
> > +		return 0;
> 
> Hmm, most of the callers of dax_iomap_rw skip the whole call if
> iov_iter_count(to)==0, so I wonder if fuse_dax_read_iter should do the
> same?
> 
> That said, iomap_dio_rw bails early if you pass it iomi.len, so I don't
> have any real objections to this.

That was the same conclusion I came to...

> Reviewed-by: Darrick J. Wong <djwong@kernel.org>

Thanks, will get this merged up for v5.19-final.