From: Johan Hedberg <johan.hedberg@gmail.com>
To: Marcel Holtmann <marcel@holtmann.org>
Cc: Danny Schweizer <danny.schweizer@proofnet.de>,
linux-bluetooth@vger.kernel.org
Subject: Re: [PATCH] Do not filter multicast addresses by default
Date: Mon, 7 Dec 2015 20:02:34 +0200 [thread overview]
Message-ID: <20151207180234.GA17653@t440s> (raw)
In-Reply-To: <E1BA2ECA-099C-43FF-830F-DD65CCA6963B@holtmann.org>
Hi,
On Mon, Dec 07, 2015, Marcel Holtmann wrote:
> > A Linux PC is connected with another device over Bluetooth PAN using a BNEP interface.
> >
> > Whenever a packet is tried to be sent over the BNEP interface, the function "bnep_net_xmit()" in "net/bluetooth/bnep/netdev.c" is called. This function calls "bnep_net_mc_filter()", which checks (if the destination address is multicast) if the address is set in a certain multicast filter (&s->mc_filter). If it is not, then it is not sent out.
> >
> > This filter is only changed in two other functions, found in net/bluetooth/bnep/core.c": in "bnep_ctrl_set_mc_filter()", which is only called if a message of type "BNEP_FILTER_MULTI_ADDR_SET" is received. Otherwise, it is set in "bnep_add_connection()", where it is set to a default value which only adds the broadcast address to the filter:
> >
> > set_bit(bnep_mc_hash(dev->broadcast), (ulong *) &s->mc_filter);
> >
> > To sum up, if the BNEP interface does not receive any message of type "BNEP_FILTER_MULTI_ADDR_SET", it will not send out any messages with multicast destination addresses except for broadcast.
> >
> > However, in the BNEP specification (page 27 in http://grouper.ieee.org/groups/802/15/Bluetooth/BNEP.pdf), it is said that per default, all multicast addresses should not be filtered, i.e. the BNEP interface should be able to send packets with any multicast destination address.
> >
> > It seems that the default case is wrong: the multicast filter should not block almost all multicast addresses, but should not filter out any.
> >
> > This leads to the problem that e.g. Neighbor Solicitation messages sent with Bluetooth PAN over the BNEP interface to a multicast destination address other than broadcast are blocked and not sent out.
> >
> > Therefore, in the default case, we set the mc_filter to ~0LL to not filter out any multicast addresses.
> > ---
> > net/bluetooth/bnep/core.c | 8 ++++++--
> > 1 file changed, 6 insertions(+), 2 deletions(-)
> >
> > diff --git a/net/bluetooth/bnep/core.c b/net/bluetooth/bnep/core.c
> > index 1641367..8e02289 100644
> > --- a/net/bluetooth/bnep/core.c
> > +++ b/net/bluetooth/bnep/core.c
> > @@ -608,8 +608,12 @@ int bnep_add_connection(struct bnep_connadd_req *req, struct socket *sock)
> > s->msg.msg_flags = MSG_NOSIGNAL;
> >
> > #ifdef CONFIG_BT_BNEP_MC_FILTER
> > - /* Set default mc filter */
> > - set_bit(bnep_mc_hash(dev->broadcast), (ulong *) &s->mc_filter);
> > + /*
> > + * Set default mc filter to not filter out any mc addresses
> > + * as defined in the BNEP specification (revision 0.95a)
> > + * http://grouper.ieee.org/groups/802/15/Bluetooth/BNEP.pdf
> > + */
> > + s->mc_filter = ~0LL;
> > #endif
> >
>
> Applying: Do not filter multicast addresses by default
> /data/kernel/maintainer-bluetooth-next/.git/rebase-apply/patch:15: trailing whitespace.
> /*
> fatal: 1 line adds whitespace errors.
> Patch failed at 0001 Do not filter multicast addresses by default
>
> And please prefix the subject line with Bluetooth:
There's also a missing Signed-off-by and the lines should be wrapped at
some 74 characters (right now it's basically a single line per
paragraph).
Johan
next prev parent reply other threads:[~2015-12-07 18:02 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-12-07 10:31 [PATCH] Do not filter multicast addresses by default Danny Schweizer
2015-12-07 17:38 ` Marcel Holtmann
2015-12-07 18:02 ` Johan Hedberg [this message]
2015-12-11 9:08 ` Danny Schweizer
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20151207180234.GA17653@t440s \
--to=johan.hedberg@gmail.com \
--cc=danny.schweizer@proofnet.de \
--cc=linux-bluetooth@vger.kernel.org \
--cc=marcel@holtmann.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).