Received-SPF: pass (google.com: domain of linux-kernel+bounces-88283-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Message-ID: <ef948e8e-b1ae-476f-b07f-4bda6d33c51b@ti.com>
Date: Fri, 1 Mar 2024 16:18:32 +0530
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH net-next 2/2] net: ethernet: ti: am65-cpsw: Enable RX HW
 timestamp only for PTP packets
Content-Language: en-US
To: Jacob Keller <jacob.e.keller@intel.com>,
        Dan Carpenter
	<dan.carpenter@linaro.org>,
        Roger Quadros <rogerq@kernel.org>,
        Siddharth
 Vadapalli <s-vadapalli@ti.com>,
        Richard Cochran <richardcochran@gmail.com>,
        Paolo Abeni <pabeni@redhat.com>, Jakub Kicinski <kuba@kernel.org>,
        Eric
 Dumazet <edumazet@google.com>,
        "David S. Miller" <davem@davemloft.net>
CC: <linux-kernel@vger.kernel.org>, <netdev@vger.kernel.org>
References: <20240215110953.3225099-1-c-vankar@ti.com>
 <20240215110953.3225099-2-c-vankar@ti.com>
 <3e81297c-2821-4af4-b13d-dc33ae8f85cc@intel.com>
From: Chintan Vankar <c-vankar@ti.com>
In-Reply-To: <3e81297c-2821-4af4-b13d-dc33ae8f85cc@intel.com>
Content-Type: text/plain; charset="UTF-8"; format=flowed
Content-Transfer-Encoding: 7bit


On 17/02/24 03:51, Jacob Keller wrote:
> 
> 
> On 2/15/2024 3:09 AM, Chintan Vankar wrote:
>> The CPSW peripherals on J7AHP, J7VCL, J7AEP, J7ES, AM64 SoCs have
>> an errata i2401 "CPSW: Host Timestamps Cause CPSW Port to Lock up".
>>
> 
> What's different about timestamping only PTP packets that prevents this
> port lock up?

The difference is the way we are timestamping the packets. Instead of
getting the timestamp from CPTS module, we are getting the timestamp
from CPTS Event FIFO.

In the current mechanism of timestamping, am65-cpsw-nuss driver
timestamps all received packets by setting the TSTAMP_EN bit in
CPTS_CONTROL register, which directs the CPTS module to timestamp all
received packets, followed by passing timestamp via DMA descriptors.
This mechanism was responsible for the CPSW port to lock up in certain
condition. We are preventing port lock up by disabling TSTAMP_EN bit in
CPTS_CONTROL register.

The mechanism we are following in this patch, utilizes the CPTS Event
FIFO that records timestamps corresponding to certain events, with one
such event being the reception of an Ethernet packet with EtherType
field set to PTP.