From mboxrd@z Thu Jan  1 00:00:00 1970
From: =?iso-8859-1?Q?M=E5ns_Rullg=E5rd?= <mans@mansr.com>
Subject: Re: [PATCH v3 3/4] net: nb8800: Move HW init to ndo_open()
Date: Tue, 14 Nov 2017 16:55:38 +0000
Message-ID: <yw1xbmk4241h.fsf@mansr.com>
References: <230165aa-eaf1-6e2b-7ff3-45b3ee4ffc62@sigmadesigns.com>
        <103bc878-9311-a727-1c21-2070e0a480e5@sigmadesigns.com>
        <yw1xwp2t11ap.fsf@mansr.com>
        <4e22ef54-d823-fc29-fedc-d447fa3d9c88@sigmadesigns.com>
        <yw1xfu9h0xv7.fsf@mansr.com>
        <30042004-4e13-b7e2-b49d-754db54cc5ce@sigmadesigns.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8BIT
Cc: David Miller <davem@davemloft.net>,
        netdev <netdev@vger.kernel.org>,
        "Linux ARM" <linux-arm-kernel@lists.infradead.org>,
        Florian Fainelli <f.fainelli@gmail.com>,
        Thibaud Cornic <thibaud_cornic@sigmadesigns.com>,
        Mason <slash.tmp@free.fr>
To: Marc Gonzalez <marc_gonzalez@sigmadesigns.com>
Return-path: <netdev-owner@vger.kernel.org>
Received: from unicorn.mansr.com ([81.2.72.234]:54390 "EHLO unicorn.mansr.com"
        rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
        id S1753921AbdKNQzl (ORCPT <rfc822;netdev@vger.kernel.org>);
        Tue, 14 Nov 2017 11:55:41 -0500
In-Reply-To: <30042004-4e13-b7e2-b49d-754db54cc5ce@sigmadesigns.com> (Marc
        Gonzalez's message of "Tue, 14 Nov 2017 17:41:09 +0100")
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

Marc Gonzalez <marc_gonzalez@sigmadesigns.com> writes:

> On 14/11/2017 14:54, Måns Rullgård wrote:
>
>> Marc Gonzalez writes:
>> 
>>> On 14/11/2017 13:40, Måns Rullgård wrote:
>>>
>>>> Marc Gonzalez wrote:
>>>>
>>>>> Power entire ethernet block down in ndo_stop().
>>>>> Power it back up in ndo_open() and perform HW init.
>>>>> Delete nb8800_dma_stop.
>>>>
>>>> Leave it alone, please.  Not all chips might have a separate power
>>>> domain for this.  Also, it works just fine on the older chips.
>>>
>>> There is no need for separate power domains. The ethernet block is
>>> clock-gated when it is held in reset.
>> 
>> So you're not powering it down then.  Please be accurate.
>
> Smirk. That looks like trolling.
>
>>> The reset register is implemented on all tango3, tango4, tango5 chips.
>> 
>> It's still not a core feature.
>
> Correct. But it covers 100% of all chips using this driver.
> There is no point in trying to implement support for chips that
> have never existed, do not exist, and never will.

You can't know that.

>>> nb8800_dma_stop() is a hack.
>> 
>> The hack originated from your company.
>
> So why are you so insistent that we keep using it?

Because it's the only way to support some chip variants.  Ones you'd
apparently rather forget, but which nonetheless exist.

>> Also, I have repeated asked you what happens if the tango5 runs out of
>> DMA buffers under normal operation.  Does that also cause it to lock up?
>> If so, you have a much bigger problem on your hands.
>
> I will run iperf3 tests with RX_DESC_COUNT lowered to 2.
> Would that produce conclusive results?
> Do you have other suggestions?

Leave RX_DESC_COUNT alone but add a delay in the nb8800_poll() loop.
That should ensure that queue is drained slowly enough for the buffers
to run out.

-- 
Måns Rullgård