Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Sat, 04 Aug 2018 17:07:53 -0700 (PDT)
Message-Id: <20180804.170753.739714742014742080.davem@davemloft.net>
To:     dima@arista.com
Cc:     natechancellor@gmail.com, gregkh@linuxfoundation.org,
        linux-kernel@vger.kernel.org, stable@vger.kernel.org,
        herbert@gondor.apana.org.au, steffen.klassert@secunet.com,
        netdev@vger.kernel.org
Subject: Re: [PATCH 4.4 106/124] netlink: Do not subscribe to non-existent
 groups
From:   David Miller <davem@davemloft.net>
In-Reply-To: <1533426268.2679.152.camel@arista.com>
References: <1533425198.2679.148.camel@arista.com>
        <20180804.163348.896647113172060665.davem@davemloft.net>
        <1533426268.2679.152.camel@arista.com>
Mime-Version: 1.0
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

From: Dmitry Safonov <dima@arista.com>
Date: Sun, 05 Aug 2018 00:44:28 +0100

> On Sat, 2018-08-04 at 16:33 -0700, David Miller wrote:
>> From: Dmitry Safonov <dima@arista.com>
>> Date: Sun, 05 Aug 2018 00:26:38 +0100
>> 
>> > Thanks for the testing, sorry again for the breakage.
>> > Will send the patch shortly (with sizeof(unsigned long long)
>> instead of
>> > hard-coded 64). Hopefully, not too late.
>> 
>> 'groups' is "long unsigned int" not "unsigned long long"
>> 
>>         long unsigned int groups = nladdr->nl_groups;
>> 
>> And nladdr->nl_groups is specified as "__u32"
>> 
>> So the size of the relevant bits is actually strictly 32-bit.
>> 
>> This makes also the "ULL" in your constant specification incorrect as
>> well.
>> 
>> So much crazy stuff going on here :-/
> 
> Right, I misslooked that..
> How does this work on Android then?
> I doubt they changed uabi.

We can have many more than 32 multicast group subscriptions.

That's why the memberships are also stored in the nlk->groups[] array.

I'm --><--- this close to simply reverting all of your changes, the
missed masking seems to be quite harmless.