Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Wed, 26 Jul 2023 13:57:28 -0700
From:   Richard Cochran <richardcochran@gmail.com>
To:     Johannes Zink <j.zink@pengutronix.de>
Cc:     Giuseppe Cavallaro <peppe.cavallaro@st.com>,
        Alexandre Torgue <alexandre.torgue@foss.st.com>,
        Jose Abreu <joabreu@synopsys.com>,
        "David S. Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>,
        Maxime Coquelin <mcoquelin.stm32@gmail.com>,
        Russell King <linux@armlinux.org.uk>,
        patchwork-jzi@pengutronix.de, netdev@vger.kernel.org,
        linux-stm32@st-md-mailman.stormreply.com,
        linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org,
        kernel@pengutronix.de, kernel test robot <lkp@intel.com>
Subject: Re: [PATCH v2] net: stmmac: correct MAC propagation delay
Message-ID: <ZMGIuKVP7BEotbrn@hoboy.vegasvil.org>
References: <20230719-stmmac_correct_mac_delay-v2-1-3366f38ee9a6@pengutronix.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20230719-stmmac_correct_mac_delay-v2-1-3366f38ee9a6@pengutronix.de>
Precedence: bulk

On Mon, Jul 24, 2023 at 12:01:31PM +0200, Johannes Zink wrote:

Earlier versions of the IP core return zero from these...

> +#define	PTP_TS_INGR_LAT	0x68	/* MAC internal Ingress Latency */
> +#define	PTP_TS_EGR_LAT	0x6c	/* MAC internal Egress Latency */

and so...

> +static void correct_latency(struct stmmac_priv *priv)
> +{
> +	void __iomem *ioaddr = priv->ptpaddr;
> +	u32 reg_tsic, reg_tsicsns;
> +	u32 reg_tsec, reg_tsecsns;
> +	u64 scaled_ns;
> +	u32 val;
> +
> +	/* MAC-internal ingress latency */
> +	scaled_ns = readl(ioaddr + PTP_TS_INGR_LAT);
> +
> +	/* See section 11.7.2.5.3.1 "Ingress Correction" on page 4001 of
> +	 * i.MX8MP Applications Processor Reference Manual Rev. 1, 06/2021
> +	 */
> +	val = readl(ioaddr + PTP_TCR);
> +	if (val & PTP_TCR_TSCTRLSSR)
> +		/* nanoseconds field is in decimal format with granularity of 1ns/bit */
> +		scaled_ns = ((u64)NSEC_PER_SEC << 16) - scaled_ns;
> +	else
> +		/* nanoseconds field is in binary format with granularity of ~0.466ns/bit */
> +		scaled_ns = ((1ULL << 31) << 16) -
> +			DIV_U64_ROUND_CLOSEST(scaled_ns * PSEC_PER_NSEC, 466U);
> +
> +	reg_tsic = scaled_ns >> 16;
> +	reg_tsicsns = scaled_ns & 0xff00;
> +
> +	/* set bit 31 for 2's compliment */
> +	reg_tsic |= BIT(31);
> +
> +	writel(reg_tsic, ioaddr + PTP_TS_INGR_CORR_NS);

here reg_tsic = 0x80000000 for a correction of -2.15 seconds!

@Jakub  Can you please revert this patch?

Thanks,
Richard