Re: [PATCH] Bluetooth: Fix skb allocation in mgmt_remote_name()

[Luiz Augusto von Dentz]

Hi Radosław,

On Thu, Jan 13, 2022 at 2:07 PM Radosław Biernacki <rad@xxxxxxxxxxxx> wrote:
>
> Hi Luiz,
>
> czw., 13 sty 2022 o 17:17 Luiz Augusto von Dentz
> <luiz.dentz@xxxxxxxxx> napisał(a):
> >
> > Hi Radoslaw,
> >
> > On Thu, Jan 13, 2022 at 7:09 AM Radoslaw Biernacki <rad@xxxxxxxxxxxx> wrote:
> > >
> > > From: Radoslaw Biernacki <rad@xxxxxxxxxxxx>
> > >
> > > This patch fixes skb allocation, as lack of space for ev might push skb
> > > tail beyond its end.
> > > Also introduce eir_precalc_len() that can be used instead of magic
> > > numbers for similar eir operations on skb.
> > >
> > > Fixes: cf1bce1de7eeb ("Bluetooth: mgmt: Make use of mgmt_send_event_skb in MGMT_EV_DEVICE_FOUND")
> > > Signed-off-by: Angela Czubak <acz@xxxxxxxxxxxx>
> > > Signed-off-by: Marek Maslanka <mm@xxxxxxxxxxxx>
> > > Signed-off-by: Radoslaw Biernacki <rad@xxxxxxxxxxxx>
> > > ---
> > >  net/bluetooth/eir.h  |  5 +++++
> > >  net/bluetooth/mgmt.c | 12 ++++--------
> > >  2 files changed, 9 insertions(+), 8 deletions(-)
> > >
> > > diff --git a/net/bluetooth/eir.h b/net/bluetooth/eir.h
> > > index 05e2e917fc25..e5876751f07e 100644
> > > --- a/net/bluetooth/eir.h
> > > +++ b/net/bluetooth/eir.h
> > > @@ -15,6 +15,11 @@ u8 eir_create_scan_rsp(struct hci_dev *hdev, u8 instance, u8 *ptr);
> > >  u8 eir_append_local_name(struct hci_dev *hdev, u8 *eir, u8 ad_len);
> > >  u8 eir_append_appearance(struct hci_dev *hdev, u8 *ptr, u8 ad_len);
> > >
> > > +static inline u16 eir_precalc_len(u8 data_len)
> > > +{
> > > +       return sizeof(u8) * 2 + data_len;
> > > +}
> > > +
> > >  static inline u16 eir_append_data(u8 *eir, u16 eir_len, u8 type,
> > >                                   u8 *data, u8 data_len)
> > >  {
> > > diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c
> > > index 37087cf7dc5a..d517fd847730 100644
> > > --- a/net/bluetooth/mgmt.c
> > > +++ b/net/bluetooth/mgmt.c
> > > @@ -9680,13 +9680,11 @@ void mgmt_remote_name(struct hci_dev *hdev, bdaddr_t *bdaddr, u8 link_type,
> > >  {
> > >         struct sk_buff *skb;
> > >         struct mgmt_ev_device_found *ev;
> > > -       u16 eir_len;
> > > -       u32 flags;
> > > +       u16 eir_len = 0;
> > > +       u32 flags = 0;
> > >
> > > -       if (name_len)
> > > -               skb = mgmt_alloc_skb(hdev, MGMT_EV_DEVICE_FOUND, 2 + name_len);
> > > -       else
> > > -               skb = mgmt_alloc_skb(hdev, MGMT_EV_DEVICE_FOUND, 0);
> > > +       skb = mgmt_alloc_skb(hdev, MGMT_EV_DEVICE_FOUND,
> > > +                            sizeof(*ev) + (name ? eir_precalc_len(name_len) : 0));
> >
> > Looks like mgmt_device_connected also has a similar problem.
>
> Yes, I was planning to send a patch to this one though it will not be as slick.
> It would be nice to have a helper which will call skb_put() and add
> eir data at once.
> Basically skb operation in pair to, what eir_append_data() does with
> help of eir_len but without awkwardness when passing return value to
> skb_put() (as it returns offset not size).

Hmm, that might be a good idea indeed something like eir_append_skb,
if only we could grow the skb with skb_put directly that would
eliminate the problem with having to reserve enough space for the
worse case.

> I will send V2 with two patches. I hope they will align with your
> original goal of eliminating the necessity of intermediary buffers at
> some point in future.
>
> >
> > >         ev = skb_put(skb, sizeof(*ev));
> > >         bacpy(&ev->addr.bdaddr, bdaddr);
> > > @@ -9696,10 +9694,8 @@ void mgmt_remote_name(struct hci_dev *hdev, bdaddr_t *bdaddr, u8 link_type,
> > >         if (name) {
> > >                 eir_len = eir_append_data(ev->eir, 0, EIR_NAME_COMPLETE, name,
> > >                                           name_len);
> > > -               flags = 0;
> > >                 skb_put(skb, eir_len);
> > >         } else {
> > > -               eir_len = 0;
> > >                 flags = MGMT_DEV_FOUND_NAME_REQUEST_FAILED;
> > >         }
> >
> > These changes would leave flags and eir_len uninitialized.
>
> Both are initialized to 0 by this patch.

Sorry, I must be blind that I didn't see you had changed that to be
initialized in their declaration.

> >
> > > --
> > > 2.34.1.703.g22d0c6ccf7-goog
> > >
> >
> >
> > --
> > Luiz Augusto von Dentz



-- 
Luiz Augusto von Dentz