[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20240227110802.552bff55@device-28.home>
Date: Tue, 27 Feb 2024 11:08:02 +0100
From: Maxime Chevallier <maxime.chevallier@...tlin.com>
To: Bastien Curutchet <bastien.curutchet@...tlin.com>
Cc: "David S. Miller" <davem@...emloft.net>, Eric Dumazet
<edumazet@...gle.com>, Jakub Kicinski <kuba@...nel.org>, Paolo Abeni
<pabeni@...hat.com>, Rob Herring <robh+dt@...nel.org>, Krzysztof Kozlowski
<krzysztof.kozlowski+dt@...aro.org>, Conor Dooley <conor+dt@...nel.org>,
Pavel Machek <pavel@....cz>, Lee Jones <lee@...nel.org>, Richard Cochran
<richardcochran@...il.com>, Andrew Lunn <andrew@...n.ch>, Heiner Kallweit
<hkallweit1@...il.com>, Russell King <linux@...linux.org.uk>,
linux-kernel@...r.kernel.org, netdev@...r.kernel.org,
devicetree@...r.kernel.org, linux-leds@...r.kernel.org, Thomas Petazzoni
<thomas.petazzoni@...tlin.com>, herve.codina@...tlin.com,
christophercordahi@...ometrics.ca
Subject: Re: [PATCH v2 5/6] net: phy: DP83640: Explicitly disabling PHY
Control Frames
Hi Bastien,
On Tue, 27 Feb 2024 10:39:44 +0100
Bastien Curutchet <bastien.curutchet@...tlin.com> wrote:
> The PHY offers a PHY control frame feature that allows to access PHY
> registers through the MAC transmit data interface. This functionality
> is not handled by the driver but can be enabled via hardware strap or
> register access.
>
> Disable the feature in config_init() to save some latency on MII packets.
>
> Signed-off-by: Bastien Curutchet <bastien.curutchet@...tlin.com>
> ---
> drivers/net/phy/dp83640.c | 6 ++++++
> drivers/net/phy/dp83640_reg.h | 4 ++++
> 2 files changed, 10 insertions(+)
>
> diff --git a/drivers/net/phy/dp83640.c b/drivers/net/phy/dp83640.c
> index 16c9fda50b19..b371dea23937 100644
> --- a/drivers/net/phy/dp83640.c
> +++ b/drivers/net/phy/dp83640.c
> @@ -1120,6 +1120,7 @@ static int dp83640_config_init(struct phy_device *phydev)
> {
> struct dp83640_private *dp83640 = phydev->priv;
> struct dp83640_clock *clock = dp83640->clock;
> + int val;
>
> if (clock->chosen && !list_empty(&clock->phylist))
> recalibrate(clock);
> @@ -1135,6 +1136,11 @@ static int dp83640_config_init(struct phy_device *phydev)
> ext_write(0, phydev, PAGE4, PTP_CTL, PTP_ENABLE);
> mutex_unlock(&clock->extreg_lock);
>
> + /* Disable unused PHY control frames */
> + phy_write(phydev, PAGESEL, 0);
> + val = phy_read(phydev, PCFCR) & ~PCF_EN;
> + phy_write(phydev, PCFCR, val);
Use phy_modify instead, and you might also want to look at the paging.
The ext_write before apparently does some page-management itself through
the clock struct (?).
> +
> return 0;
> }
>
> diff --git a/drivers/net/phy/dp83640_reg.h b/drivers/net/phy/dp83640_reg.h
> index bf34d422d91e..b5adb8958c08 100644
> --- a/drivers/net/phy/dp83640_reg.h
> +++ b/drivers/net/phy/dp83640_reg.h
> @@ -10,6 +10,7 @@
> #define PHYCR 0x0019 /* PHY Control Register */
> #define PHYCR2 0x001c /* PHY Control Register 2 */
> #define EDCR 0x001D /* Energy Detect Control Register */
> +#define PCFCR 0x001F /* PHY Control Frames Control Register */
>
> #define PAGE4 0x0004
> #define PTP_CTL 0x0014 /* PTP Control Register */
> @@ -68,6 +69,9 @@
> /* Bit definitions for the EDCR register */
> #define ED_EN BIT(15) /* Enable Energy Detect Mode */
>
> +/* Bit definitions for the PCFCR register */
> +#define PCF_EN BIT(0) /* Enable PHY Control Frames */
> +
> /* Bit definitions for the PTP_CTL register */
> #define TRIG_SEL_SHIFT (10) /* PTP Trigger Select */
> #define TRIG_SEL_MASK (0x7)
Thanks,
Maxime
Powered by blists - more mailing lists