[PATCH v2] net: stmmac: add explicit check and error on invalid PTP clock rate

[Alexis Lothoré]

Hey Maxime,

On Tue May 27, 2025 at 6:31 PM CEST, Maxime Chevallier wrote:
> Hi Alexis,
>
> On Tue, 27 May 2025 08:33:44 +0200
> Alexis Lothoré <alexis.lothore at bootlin.com> wrote:

[...]

>> +	if (!priv->plat->clk_ptp_rate) {
>> +		netdev_err(priv->dev, "Invalid PTP clock rate");
>> +		return -EINVAL;
>> +	}
>> +
>>  	stmmac_config_hw_tstamping(priv, priv->ptpaddr, systime_flags);
>>  	priv->systime_flags = systime_flags;
>
> This may be some nitpick that can be addressed at a later point, but we
> now have a guarantee that when stmmac_ptp_register() gets called,
> priv->ptp_clk_rate is non-zero, right ? If so, we can drop the test in
> said function :
>
> 	if (priv->plat->has_gmac4 && priv->plat->clk_ptp_rate)
> 		priv->plat->cdc_error_adj = (2 * NSEC_PER_SEC) / priv->plat->clk_ptp_rate;

You are right, my series makes this check a duplicate, I'll remove it.

> There is another spot in the code, like in the EST handling, where we
> divide by priv->plat->ptp_clk_rate :
>
> stmmac_adjust_time(...)
> 	stmmac_est_configure(priv, priv, priv->est,
> 			     priv->plat->clk_ptp_rate)
> 		.est_configure()
> 			ctrl |= ((NSEC_PER_SEC / ptp_rate) [...]
>
> Maybe we should fail EST configuration as well if ptp_clk_rate is 0
> (probably in stmmac_tc.c's tc_taprio_configure or in the
> .est_configure). That can be a step for later as well, as I don't know
> if the setup you found this bug on even supports taprio/EST, and setups
> that do didn't seem to encounter the bug yet.

I guess you are right as well, it may not make sense to try to configure
EST if we have a bogus clk_ptp_rate value. est_configure seems to be called
in both tc_taprio_configure and stmmac_adjust_time, so I'll add the check
in est_configure.

Thanks,

Alexis

-- 
Alexis Lothoré, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com