Date:   Mon, 27 Mar 2017 12:02:33 +0200
From:   Ladi Prosek <lprosek@...hat.com>
To:     "Michael S. Tsirkin" <mst@...hat.com>
Cc:     Arnd Bergmann <arnd@...db.de>,
        David Hildenbrand <david@...hat.com>,
        Jason Wang <jasowang@...hat.com>,
        Yisheng Xie <xieyisheng1@...wei.com>,
        Konstantin Neumoin <kneumoin@...tuozzo.com>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        virtualization@...ts.linux-foundation.org,
        Minchan Kim <minchan@...nel.org>,
        "Denis V. Lunev" <den@...nvz.org>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Ingo Molnar <mingo@...nel.org>
Subject: Re: [PATCH] virtio_balloon: prevent uninitialized variable use

On Fri, Mar 24, 2017 at 9:59 PM, Michael S. Tsirkin <mst@...hat.com> wrote:
> On Fri, Mar 24, 2017 at 09:40:07PM +0100, Arnd Bergmann wrote:
>> On Fri, Mar 24, 2017 at 9:11 PM, Ladi Prosek <lprosek@...hat.com> wrote:
>> > On Fri, Mar 24, 2017 at 7:38 PM, David Hildenbrand <david@...hat.com> wrote:
>> >> On 23.03.2017 16:17, Arnd Bergmann wrote:
>> >>> The latest gcc-7.0.1 snapshot reports a new warning:
>> >>>
>> >>> virtio/virtio_balloon.c: In function 'update_balloon_stats':
>> >>> virtio/virtio_balloon.c:258:26: error: 'events[2]' is used uninitialized in this function [-Werror=uninitialized]
>> >>> virtio/virtio_balloon.c:260:26: error: 'events[3]' is used uninitialized in this function [-Werror=uninitialized]
>> >>> virtio/virtio_balloon.c:261:56: error: 'events[18]' is used uninitialized in this function [-Werror=uninitialized]
>> >>> virtio/virtio_balloon.c:262:56: error: 'events[17]' is used uninitialized in this function [-Werror=uninitialized]
>> >>>
>> >>> This seems absolutely right, so we should add an extra check to
>> >>> prevent copying uninitialized stack data into the statistics.
>> >>> From all I can tell, this has been broken since the statistics code
>> >>> was originally added in 2.6.34.
>> >>>
>> >>> Fixes: 9564e138b1f6 ("virtio: Add memory statistics reporting to the balloon driver (V4)")
>> >>> Signed-off-by: Arnd Bergmann <arnd@...db.de>
>> >>> ---
>> >>>  drivers/virtio/virtio_balloon.c | 2 ++
>> >>>  1 file changed, 2 insertions(+)
>> >>>
>> >>> diff --git a/drivers/virtio/virtio_balloon.c b/drivers/virtio/virtio_balloon.c
>> >>> index 4e1191508228..cd5c54e2003d 100644
>> >>> --- a/drivers/virtio/virtio_balloon.c
>> >>> +++ b/drivers/virtio/virtio_balloon.c
>> >>> @@ -254,12 +254,14 @@ static void update_balloon_stats(struct virtio_balloon *vb)
>> >>>
>> >>>       available = si_mem_available();
>> >>>
>> >>> +#ifdef CONFIG_VM_EVENT_COUNTERS
>> >>>       update_stat(vb, idx++, VIRTIO_BALLOON_S_SWAP_IN,
>> >>>                               pages_to_bytes(events[PSWPIN]));
>> >>>       update_stat(vb, idx++, VIRTIO_BALLOON_S_SWAP_OUT,
>> >>>                               pages_to_bytes(events[PSWPOUT]));
>> >>>       update_stat(vb, idx++, VIRTIO_BALLOON_S_MAJFLT, events[PGMAJFAULT]);
>> >>>       update_stat(vb, idx++, VIRTIO_BALLOON_S_MINFLT, events[PGFAULT]);
>> >>> +#endif
>> >>>       update_stat(vb, idx++, VIRTIO_BALLOON_S_MEMFREE,
>> >>>                               pages_to_bytes(i.freeram));
>> >>>       update_stat(vb, idx++, VIRTIO_BALLOON_S_MEMTOT,
>> >
>> > This will leave four uninitialized slots in vb->stats if
>> > CONFIG_VM_EVENT_COUNTERS is not defined. update_balloon_stats should
>> > have
>> >
>> >   BUG_ON(idx < VIRTIO_BALLOON_S_NR);
>> >
>> > at the end.
>> >
>> > You need to make sure that vb->stats is smaller, either by using
>> > something else than VIRTIO_BALLOON_S_NR for its size or something else
>> > than sizeof(vb->stats) as the last argument to sg_init_one in this
>> > file.
>>
>> Ah, got it. Would one of you create a fixed patch for this, or should I?

I will do it. If we go the route of returning 'idx' from
update_balloon_stats, my patch

   virtio_balloon: don't push uninitialized buffers to stats virtqueue

needs to be redone anyway. Thanks!

>> An easy way to solve it would be to preinitialize the events array
>> and return zeroes to the host, but I don't know if that's the right
>> solution.
>
>
> I think that's out of spec.
>
>> Another option would be to return 'idx' from update_balloon_stats,
>> and use that for the size:
>>
>> diff --git a/drivers/virtio/virtio_balloon.c b/drivers/virtio/virtio_balloon.c
>> index cd5c54e2003d..bea3cfb88e53 100644
>> --- a/drivers/virtio/virtio_balloon.c
>> +++ b/drivers/virtio/virtio_balloon.c
>> @@ -242,7 +242,7 @@ static inline void update_stat(struct
>> virtio_balloon *vb, int idx,
>>
>>  #define pages_to_bytes(x) ((u64)(x) << PAGE_SHIFT)
>>
>> -static void update_balloon_stats(struct virtio_balloon *vb)
>> +static unsigned int update_balloon_stats(struct virtio_balloon *vb)
>>  {
>>   unsigned long events[NR_VM_EVENT_ITEMS];
>>   struct sysinfo i;
>> @@ -268,6 +268,8 @@ static void update_balloon_stats(struct virtio_balloon *vb)
>>   pages_to_bytes(i.totalram));
>>   update_stat(vb, idx++, VIRTIO_BALLOON_S_AVAIL,
>>   pages_to_bytes(available));
>> +
>> + return idx;
>>  }
>>
>>  /*
>> @@ -294,13 +296,14 @@ static void stats_handle_request(struct
>> virtio_balloon *vb)
>>   struct virtqueue *vq;
>>   struct scatterlist sg;
>>   unsigned int len;
>> + unsigned int num_stats;
>>
>> - update_balloon_stats(vb);
>> + num_stats = update_balloon_stats(vb);
>>
>>   vq = vb->stats_vq;
>>   if (!virtqueue_get_buf(vq, &len))
>>   return;
>> - sg_init_one(&sg, vb->stats, sizeof(vb->stats));
>> + sg_init_one(&sg, vb->stats, sizeof(vb->stats[0] * num_stats));
>>   virtqueue_add_outbuf(vq, &sg, 1, vb, GFP_KERNEL);
>>   virtqueue_kick(vq);
>>  }
>>
>>      Arnd
>
> Sounds reasonable.
>
> --
> MST

Powered by blists - more mailing lists