Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:7 as permitted sender) client-ip=2620:137:e000::3:7;
From:   Rohan G Thomas <rohan.g.thomas@intel.com>
To:     fancer.lancer@gmail.com
Cc:     alexandre.torgue@foss.st.com, andriy.shevchenko@linux.intel.com,
        davem@davemloft.net, edumazet@google.com, joabreu@synopsys.com,
        kuba@kernel.org, linux-arm-kernel@lists.infradead.org,
        linux-kernel@vger.kernel.org,
        linux-stm32@st-md-mailman.stormreply.com,
        mcoquelin.stm32@gmail.com, netdev@vger.kernel.org,
        pabeni@redhat.com, rohan.g.thomas@intel.com
Subject: Re: [PATCH net-next 1/1] net: stmmac: xgmac: EST interrupts handling
Date:   Fri,  1 Dec 2023 14:49:09 +0800
Message-Id: <20231201064909.28399-1-rohan.g.thomas@intel.com>
In-Reply-To: <cjgx6e3agc6gpvs75nhkf6wlztk73epmct6tcuooyqvk2nx2o2@vr5buyk637t3>
References: <cjgx6e3agc6gpvs75nhkf6wlztk73epmct6tcuooyqvk2nx2o2@vr5buyk637t3>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

On Fri, 6 Oct 2023 13:08:33 +0300 Serge Semin wrote:
> Hi Rohan, Jakub
> ...
> Interesting thing. My DW QoS Eth _v5.10a_ HW manual explicitly states that
> it's multiplied by _6_ in nanoseconds (just rechecked). So either there is a
> difference between the minor DW QoS Eth IP-core releases or the older HW-
> manuals have had a typo in the MTL_EST_CONTROL.PTOV field description.
> Synopsys normally describes such changes (whether it was a mistake or a
> functional change) in the IP-core release notes. The release notes document
> is supplied as a separate pdf file. Alas I don't have one.( Even if I had it it
> would have been useless since the change was introduced in the newer QoS
> IP-cores. Rohan, do you happen to have the release notes for DW QoS Eth IP-
> core v5.30 at hands?
> Something like DWC_ether_qos_rc_relnotes.pdf?

Hi Serge,

Sorry for the delay. Sends out another version with the suggested changes.

Managed to get DWC_ether_qos_relnotes.pdf for v5.20a and v5.30a. But I couldn't
find anything related to this. So for refactoring, I'm keeping the logic as in
the upstream code to avoid any regression.

> 
> Also please double check that your DW QoS Eth v5.30a for sure states that
> MTL_EST_CONTROL.PTOV contains value multiplied by _6_. So we wouldn't
> be wasting time trying to workaround a more complex problem than we
> already have.

Yes, I checked this again. For DW QoS Eth v5.30a the multiplier for 
MTL_EST_CONTROL.PTOV is _9_ as per the databook.

Also noticed a similar difference for MTL_EST_Status.BTRL field length. As per
the upstream code and DW QoS Eth v5.10a databook this field covers bit 8 to bit
11. But for the xgmac IP and DW QoS Eth v5.30a databook this field covers bit 8
to bit 15. Again nothing mentioned in the release notes. Here also I'm keeping
the logic as in the upstream code to avoid any regression.

> 
> -Serge(y)
> 

Best Regards,
Rohan