[PATCH net-next] net: phy: simplify a check in phy_check_link_status

[PATCH net-next] net: phy: simplify a check in phy_check_link_status

[Heiner Kallweit]

Handling case err == 0 in the other branch allows to simplify the
code. In addition I assume in "err & phydev->eee_cfg.tx_lpi_enabled"
it should have been a logical and operator. It works as expected also
with the bitwise and, but using a bitwise and with a bool value looks
ugly to me.

Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
---
 drivers/net/phy/phy.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/phy/phy.c b/drivers/net/phy/phy.c
index c3a0a5ee5..c4236564c 100644
--- a/drivers/net/phy/phy.c
+++ b/drivers/net/phy/phy.c
@@ -985,10 +985,10 @@ static int phy_check_link_status(struct phy_device *phydev)
 		phydev->state = PHY_RUNNING;
 		err = genphy_c45_eee_is_active(phydev,
 					       NULL, NULL, NULL);
-		if (err < 0)
+		if (err <= 0)
 			phydev->enable_tx_lpi = false;
 		else
-			phydev->enable_tx_lpi = (err & phydev->eee_cfg.tx_lpi_enabled);
+			phydev->enable_tx_lpi = phydev->eee_cfg.tx_lpi_enabled;
 
 		phy_link_up(phydev);
 	} else if (!phydev->link && phydev->state != PHY_NOLINK) {
-- 
2.44.0

Re: [PATCH] net: phy: simplify a check in phy_check_link_status

[Ratheesh Kannoth]

On 2024-03-08 at 02:46:12, Heiner Kallweit (hkallweit1@gmail.com) wrote:
> Handling case err == 0 in the other branch allows to simplify the
> code. In addition I assume in "err & phydev->eee_cfg.tx_lpi_enabled"
> it should have been a logical and operator. It works as expected also
> with the bitwise and, but using a bitwise and with a bool value looks
> ugly to me.
>
> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
> ---
>  drivers/net/phy/phy.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/net/phy/phy.c b/drivers/net/phy/phy.c
> index c3a0a5ee5..c4236564c 100644
> --- a/drivers/net/phy/phy.c
> +++ b/drivers/net/phy/phy.c
> @@ -985,10 +985,10 @@ static int phy_check_link_status(struct phy_device *phydev)
>  		phydev->state = PHY_RUNNING;
>  		err = genphy_c45_eee_is_active(phydev,
>  					       NULL, NULL, NULL);
IMO, better to rename "err" to "ret", and do if (ret == true). OR,
we should fix syntax of genphy_c45_eee_is_active() to return bool (true/false)
than doing this, because function name suggest so , xxx_is_active(). err == 0, norm is
for success case.

> -		if (err < 0)
> +		if (err <= 0)
>  			phydev->enable_tx_lpi = false;
>  		else
> -			phydev->enable_tx_lpi = (err & phydev->eee_cfg.tx_lpi_enabled);
> +			phydev->enable_tx_lpi = phydev->eee_cfg.tx_lpi_enabled;
>
>  		phy_link_up(phydev);
>  	} else if (!phydev->link && phydev->state != PHY_NOLINK) {
> --
> 2.44.0

Re: [PATCH] net: phy: simplify a check in phy_check_link_status

[Heiner Kallweit]

On 08.03.2024 05:50, Ratheesh Kannoth wrote:
> On 2024-03-08 at 02:46:12, Heiner Kallweit (hkallweit1@gmail.com) wrote:
>> Handling case err == 0 in the other branch allows to simplify the
>> code. In addition I assume in "err & phydev->eee_cfg.tx_lpi_enabled"
>> it should have been a logical and operator. It works as expected also
>> with the bitwise and, but using a bitwise and with a bool value looks
>> ugly to me.
>>
>> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
>> ---
>>  drivers/net/phy/phy.c | 4 ++--
>>  1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/net/phy/phy.c b/drivers/net/phy/phy.c
>> index c3a0a5ee5..c4236564c 100644
>> --- a/drivers/net/phy/phy.c
>> +++ b/drivers/net/phy/phy.c
>> @@ -985,10 +985,10 @@ static int phy_check_link_status(struct phy_device *phydev)
>>  		phydev->state = PHY_RUNNING;
>>  		err = genphy_c45_eee_is_active(phydev,
>>  					       NULL, NULL, NULL);
> IMO, better to rename "err" to "ret", and do if (ret == true). OR,
> we should fix syntax of genphy_c45_eee_is_active() to return bool (true/false)
> than doing this, because function name suggest so , xxx_is_active(). err == 0, norm is
> for success case.
> 
Return value of genphy_c45_eee_is_active() is tristate int:
<0: error
 0: eee not active
 1: eee active (implicit cast from bool to int)

This tristate behavior isn't unusual, you can find it with other phylib
functions too, another example are several of the rpm functions.
So both, 0 and 1, are success cases (from a technical perspective).

>> -		if (err < 0)
>> +		if (err <= 0)
>>  			phydev->enable_tx_lpi = false;
>>  		else
>> -			phydev->enable_tx_lpi = (err & phydev->eee_cfg.tx_lpi_enabled);
>> +			phydev->enable_tx_lpi = phydev->eee_cfg.tx_lpi_enabled;
>>
>>  		phy_link_up(phydev);
>>  	} else if (!phydev->link && phydev->state != PHY_NOLINK) {
>> --
>> 2.44.0

RE: [EXTERNAL] [PATCH net-next] net: phy: simplify a check in phy_check_link_status

[Suman Ghosh]

Hi Heiner,

To me it looks like both patches,
r8169: switch to new function phy_support_eee and net: phy: simplify a check in phy_check_link_status is related and can be pushed as a series. This will make change more harmonic. Because, you are moving setting of enable_tx_lpi in one patch and removing from the other one.

Regards,
Suman

>-----Original Message-----
>From: Heiner Kallweit <hkallweit1@gmail.com>
>Sent: Friday, March 8, 2024 2:46 AM
>To: Russell King - ARM Linux <linux@armlinux.org.uk>; Andrew Lunn
><andrew@lunn.ch>; Paolo Abeni <pabeni@redhat.com>; Eric Dumazet
><edumazet@google.com>; David Miller <davem@davemloft.net>; Jakub
>Kicinski <kuba@kernel.org>
>Cc: netdev@vger.kernel.org
>Subject: [EXTERNAL] [PATCH net-next] net: phy: simplify a check in
>phy_check_link_status
>
>Handling case err == 0 in the other branch allows to simplify the code.
>In addition I assume in "err & phydev->eee_cfg.tx_lpi_enabled"
>it should have been a logical and operator. It works as expected also
>with the bitwise and, but using a bitwise and with a bool value looks
>ugly to me.
>
>Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
>---
> drivers/net/phy/phy.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
>diff --git a/drivers/net/phy/phy.c b/drivers/net/phy/phy.c index
>c3a0a5ee5..c4236564c 100644
>--- a/drivers/net/phy/phy.c
>+++ b/drivers/net/phy/phy.c
>@@ -985,10 +985,10 @@ static int phy_check_link_status(struct phy_device
>*phydev)
> 		phydev->state = PHY_RUNNING;
> 		err = genphy_c45_eee_is_active(phydev,
> 					       NULL, NULL, NULL);
>-		if (err < 0)
>+		if (err <= 0)
> 			phydev->enable_tx_lpi = false;
> 		else
>-			phydev->enable_tx_lpi = (err & phydev-
>>eee_cfg.tx_lpi_enabled);
>+			phydev->enable_tx_lpi = phydev->eee_cfg.tx_lpi_enabled;
>
> 		phy_link_up(phydev);
> 	} else if (!phydev->link && phydev->state != PHY_NOLINK) {
>--
>2.44.0
>

Re: [PATCH net-next] net: phy: simplify a check in phy_check_link_status

[Heiner Kallweit]

On 07.03.2024 22:16, Heiner Kallweit wrote:
> Handling case err == 0 in the other branch allows to simplify the
> code. In addition I assume in "err & phydev->eee_cfg.tx_lpi_enabled"
> it should have been a logical and operator. It works as expected also
> with the bitwise and, but using a bitwise and with a bool value looks
> ugly to me.
> 
> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
> ---
>  drivers/net/phy/phy.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/net/phy/phy.c b/drivers/net/phy/phy.c
> index c3a0a5ee5..c4236564c 100644
> --- a/drivers/net/phy/phy.c
> +++ b/drivers/net/phy/phy.c
> @@ -985,10 +985,10 @@ static int phy_check_link_status(struct phy_device *phydev)
>  		phydev->state = PHY_RUNNING;
>  		err = genphy_c45_eee_is_active(phydev,
>  					       NULL, NULL, NULL);
> -		if (err < 0)
> +		if (err <= 0)
>  			phydev->enable_tx_lpi = false;
>  		else
> -			phydev->enable_tx_lpi = (err & phydev->eee_cfg.tx_lpi_enabled);
> +			phydev->enable_tx_lpi = phydev->eee_cfg.tx_lpi_enabled;
>  
>  		phy_link_up(phydev);
>  	} else if (!phydev->link && phydev->state != PHY_NOLINK) {

This patch was set to "changes requested" in patchwork. The comment that this refers to
(make two patches a series) isn't applicable and I answered to it here:
https://lore.kernel.org/netdev/d6ee6353-5cb0-4751-9b69-255ab62e6b56@gmail.com/T/
Whatever is better for you: If it can still be applied, fine. Otherwise I'd resubmit
after the merge window.

Re: [PATCH net-next] net: phy: simplify a check in phy_check_link_status

[patchwork-bot+netdevbpf]

Hello:

This patch was applied to netdev/net-next.git (main)
by Jakub Kicinski <kuba@kernel.org>:

On Thu, 7 Mar 2024 22:16:12 +0100 you wrote:
> Handling case err == 0 in the other branch allows to simplify the
> code. In addition I assume in "err & phydev->eee_cfg.tx_lpi_enabled"
> it should have been a logical and operator. It works as expected also
> with the bitwise and, but using a bitwise and with a bool value looks
> ugly to me.
> 
> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
> 
> [...]

Here is the summary with links:
  - [net-next] net: phy: simplify a check in phy_check_link_status
    https://git.kernel.org/netdev/net-next/c/c786459fc827

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html