Received-SPF: pass (google.com: best guess record for domain of linux-wireless-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [RFCv3 3/3] nl80211: Send large new_wiphy events
To:     Johannes Berg <johannes@sipsolutions.net>,
        linux-wireless@vger.kernel.org
References: <20190906154303.9303-1-denkenz@gmail.com>
 <20190906154303.9303-3-denkenz@gmail.com>
 <127c19b1855302a467a1bdbf2f25f625a0254d9e.camel@sipsolutions.net>
From:   Denis Kenzior <denkenz@gmail.com>
Message-ID: <f48bdad3-f9f8-20b7-b6ff-6248368a7ae1@gmail.com>
Date:   Wed, 11 Sep 2019 07:20:01 -0500
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101
 Thunderbird/60.8.0
MIME-Version: 1.0
In-Reply-To: <127c19b1855302a467a1bdbf2f25f625a0254d9e.camel@sipsolutions.net>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Sender: linux-wireless-owner@vger.kernel.org
Precedence: bulk

hi Johannes,

On 9/11/19 4:47 AM, Johannes Berg wrote:
> On Fri, 2019-09-06 at 10:43 -0500, Denis Kenzior wrote:
>>
>> + *	There are no limits (outside of netlink protocol limits) on
>> + *	message sizes that can be sent over the "config2" multicast group. It
>> + *	is assumed that applications utilizing "config2" multicast group
>> + *	utilize buffers that are inherently large enough or can utilize
>> + *	MSG_PEEK/MSG_TRUNC in the netlink transport in order to allocate big
>> + *	enough buffers.
> 
> I'm not sure I see how the applications could do buffers that are
> "inherently" large enough, there's no practical message size limit, is
> there (32-bits for the size).

The kernel caps this to 32k right now if I read the code correctly.  But 
fair point.

> 
> I'd argue this should just say that applications should use large
> buffers and still use MSG_PEEK/handle MSG_TRUNC, but I can also edit it
> later.
> 
>> +	msg = nlmsg_new(alloc_size, GFP_KERNEL);
>> +	if (!msg)
>> +		goto legacy;
>> +
>> +	if (WARN_ON(nl80211_send_wiphy(rdev, cmd, msg, 0, 0, 0, &state) < 0)) {
>> +		nlmsg_free(msg);
>> +		goto legacy;
>> +	}
>> +
>> +	genlmsg_multicast_netns(&nl80211_fam, wiphy_net(&rdev->wiphy), msg, 0,
>> +				NL80211_MCGRP_CONFIG2, GFP_KERNEL);
>> +
>> +legacy:
> 
> nit: just use "else" instead of the goto?

I'm not sure I understand?  We want to send both messages here...

Regards,
-Denis