[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <62df2808e5c50_1f5536294f2@dwillia2-xfh.jf.intel.com.notmuch>
Date: Mon, 25 Jul 2022 16:32:24 -0700
From: Dan Williams <dan.j.williams@...el.com>
To: "Darrick J. Wong" <djwong@...nel.org>,
Li Jinlin <lijinlin3@...wei.com>
CC: <viro@...iv.linux.org.uk>, <dan.j.williams@...el.com>,
<willy@...radead.org>, <jack@...e.cz>,
<linux-fsdevel@...r.kernel.org>, <nvdimm@...ts.linux.dev>,
<linux-kernel@...r.kernel.org>, <linfeilong@...wei.com>,
<liuzhiqiang26@...wei.com>
Subject: Re: [PATCH] fsdax: Fix infinite loop in dax_iomap_rw()
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@...wei.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@...nel.org>
Thanks, will get this merged up for v5.19-final.
Powered by blists - more mailing lists