Re: [PATCH 3/3] net: phy: mediatek-ge: do not disable EEE advertisement

Re: [PATCH 3/3] net: phy: mediatek-ge: do not disable EEE advertisement

[Florian Fainelli]

On 3/18/2024 12:46 AM, Arınç ÜNAL via B4 Relay wrote:
> From: Arınç ÜNAL <arinc.unal@arinc9.com>
> 
> There's no need to disable Energy-Efficient Ethernet (EEE) advertisement on
> the MT7530 and MT7531 switch PHYs. EEE works fine on MT7530 and MT7531
> switch PHYs. Remove the code where EEE advertisement is disabled.
> 
> This is a bugfix because there's a possible race condition where the
> mediatek-ge driver would kick in after the MT7530 DSA subdriver which would
> have EEE disabled until manually enabled.
> 
> Fixes: e40d2cca0189 ("net: phy: add MediaTek Gigabit Ethernet PHY driver")
> Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>

Could we get Sean to chime in, maybe there was a valid reason to disable 
EEE advertisement in the first place?
-- 
Florian

Re: [PATCH 1/3] net: dsa: mt7530: fix enabling EEE on MT7531 switch on all boards

[Florian Fainelli]

On 3/18/2024 12:46 AM, Arınç ÜNAL via B4 Relay wrote:
> From: Arınç ÜNAL <arinc.unal@arinc9.com>
> 
> The commit 40b5d2f15c09 ("net: dsa: mt7530: Add support for EEE features")
> brought EEE support but did not enable EEE on MT7531 switch MACs. EEE is
> enabled on MT7531 switch MACs either by pulling the LAN2LED0 pin low on the
> board (bootstrapping), or unsetting the EEE_DIS bit on the trap register.
> 
> There are existing boards that were not designed to pull the pin low.
> Therefore, unset the EEE_DIS bit on the trap register.
> 
> Unlike MT7530, the modifiable trap register won't be populated identical to
> the trap status register after reset. Therefore, read from the trap status
> register, modify the bits, then write to the modifiable trap register.
> 
> The disable EEE bit on the trap pertains to the LAN2LED0 pin which is
> usually used to control an LED. Once the bit is unset, the pin will be low.
> That will make the active low LED turn on.
> 
> The pin is controlled by the switch PHY. It seems that the PHY controls the
> pin in the way that it inverts the pin state. That means depending on the
> wiring of the LED connected to LAN2LED0 on the board, the LED may be on
> without an active link.
> 
> Fixes: 40b5d2f15c09 ("net: dsa: mt7530: Add support for EEE features")
> Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>

Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
-- 
Florian

Re: [PATCH 0/3] Fix EEE support for MT7531 and MT7988 SoC switch

[Arınç ÜNAL]

On 18.03.2024 10:46, Arınç ÜNAL via B4 Relay wrote:
> Hi.
> 
> This patch series fixes EEE support for MT7531 and the switch on the MT7988
> SoC. EEE did not work on MT7531 on most boards before this, it is unclear
> what's the status on MT7988 SoC switch as I don't have the hardware.
> 
> Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>

I see the state of this patch series is deferred on patchwork. I see that I
forgot to delegate this to the net tree. As I don't see any objections in
this series, I'll send v2 with it tomorrow.

Arınç

Re: [PATCH 0/3] Fix EEE support for MT7531 and MT7988 SoC switch

[Daniel Golle]

On Wed, Mar 20, 2024 at 11:10:19AM +0300, Arınç ÜNAL wrote:
> On 18.03.2024 10:46, Arınç ÜNAL via B4 Relay wrote:
> > Hi.
> > 
> > This patch series fixes EEE support for MT7531 and the switch on the MT7988
> > SoC. EEE did not work on MT7531 on most boards before this, it is unclear
> > what's the status on MT7988 SoC switch as I don't have the hardware.
> > 
> > Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
> 
> I see the state of this patch series is deferred on patchwork. I see that I
> forgot to delegate this to the net tree. As I don't see any objections in
> this series, I'll send v2 with it tomorrow.

Sorry, but imho it should go to net-next, and you might have overlook
it but there have been some concerns.

For sure it should not go to net tree as you are enabling a new feature
and not fixing anything which is broken. EEE on MT7988 works fine as
of now (see my reply), EEE on MT7531 was supposedly intentionally
disabled for reasons we should ask MTK people about.

Re: [PATCH 0/3] Fix EEE support for MT7531 and MT7988 SoC switch

[Arınç ÜNAL]

On 20.03.2024 14:08, Daniel Golle wrote:
> On Wed, Mar 20, 2024 at 11:10:19AM +0300, Arınç ÜNAL wrote:
>> On 18.03.2024 10:46, Arınç ÜNAL via B4 Relay wrote:
>>> Hi.
>>>
>>> This patch series fixes EEE support for MT7531 and the switch on the MT7988
>>> SoC. EEE did not work on MT7531 on most boards before this, it is unclear
>>> what's the status on MT7988 SoC switch as I don't have the hardware.
>>>
>>> Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
>>
>> I see the state of this patch series is deferred on patchwork. I see that I
>> forgot to delegate this to the net tree. As I don't see any objections in
>> this series, I'll send v2 with it tomorrow.
> 
> Sorry, but imho it should go to net-next, and you might have overlook
> it but there have been some concerns.

I don't believe I overlooked anything. I've come to this conclusion after
reading every response in this thread.

> 
> For sure it should not go to net tree as you are enabling a new feature
> and not fixing anything which is broken. EEE on MT7988 works fine as

EEE support exists since the commit which I've mentioned on my patches
here. I am fixing it. I thought I had explained this clearly on the
patches.

> of now (see my reply), EEE on MT7531 was supposedly intentionally
> disabled for reasons we should ask MTK people about.

Are you talking about the EEE_DIS bit on the trap? There's no default
setting there, it's up to the board vendor to enable/disable EEE by pulling
the affine pin low or high. So there's no intentional disabling by MediaTek
there. I see no need to ask the corporate regarding this.

This patch series does not in any way enable EEE on the switch PHYs and
MACs when it's disabled by default.

Arınç

Re: [PATCH 3/3] net: phy: mediatek-ge: do not disable EEE advertisement

[Arınç ÜNAL]

On 18.03.2024 10:46, Arınç ÜNAL via B4 Relay wrote:
> From: Arınç ÜNAL <arinc.unal@arinc9.com>
> 
> There's no need to disable Energy-Efficient Ethernet (EEE) advertisement on
> the MT7530 and MT7531 switch PHYs. EEE works fine on MT7530 and MT7531
> switch PHYs. Remove the code where EEE advertisement is disabled.
> 
> This is a bugfix because there's a possible race condition where the
> mediatek-ge driver would kick in after the MT7530 DSA subdriver which would
> have EEE disabled until manually enabled.

Can I get an opinion on this? Is it actually possible that the PHY driver
would start probing after the DSA subdriver? On the console logs for the
DSA subdriver, I can see that the name of the PHY driver will appear, which
makes me believe the PHY driver would actually never probe after the DSA
subdriver.

[    4.402641] mt7530-mdio mdio-bus:1f wan (uninitialized): PHY [mt7530-0:00] driver [MediaTek MT7530 PHY] (irq=POLL)
[    4.420392] mt7530-mdio mdio-bus:1f lan0 (uninitialized): PHY [mt7530-0:01] driver [MediaTek MT7530 PHY] (irq=POLL)
[    4.437791] mt7530-mdio mdio-bus:1f lan1 (uninitialized): PHY [mt7530-0:02] driver [MediaTek MT7530 PHY] (irq=POLL)
[    4.455096] mt7530-mdio mdio-bus:1f lan2 (uninitialized): PHY [mt7530-0:03] driver [MediaTek MT7530 PHY] (irq=POLL)
[    4.472422] mt7530-mdio mdio-bus:1f lan3 (uninitialized): PHY [mt7530-0:04] driver [MediaTek MT7530 PHY] (irq=POLL)

I don't want to submit a bugfix to the net tree if the bug won't ever
happen in real life.

Arınç

Re: [PATCH 3/3] net: phy: mediatek-ge: do not disable EEE advertisement

[Russell King (Oracle)]

On Wed, Mar 20, 2024 at 11:40:56PM +0300, Arınç ÜNAL wrote:
> On 18.03.2024 10:46, Arınç ÜNAL via B4 Relay wrote:
> > From: Arınç ÜNAL <arinc.unal@arinc9.com>
> > 
> > There's no need to disable Energy-Efficient Ethernet (EEE) advertisement on
> > the MT7530 and MT7531 switch PHYs. EEE works fine on MT7530 and MT7531
> > switch PHYs. Remove the code where EEE advertisement is disabled.
> > 
> > This is a bugfix because there's a possible race condition where the
> > mediatek-ge driver would kick in after the MT7530 DSA subdriver which would
> > have EEE disabled until manually enabled.
> 
> Can I get an opinion on this? Is it actually possible that the PHY driver
> would start probing after the DSA subdriver? On the console logs for the
> DSA subdriver, I can see that the name of the PHY driver will appear, which
> makes me believe the PHY driver would actually never probe after the DSA
> subdriver.
> 
> [    4.402641] mt7530-mdio mdio-bus:1f wan (uninitialized): PHY [mt7530-0:00] driver [MediaTek MT7530 PHY] (irq=POLL)
> [    4.420392] mt7530-mdio mdio-bus:1f lan0 (uninitialized): PHY [mt7530-0:01] driver [MediaTek MT7530 PHY] (irq=POLL)
> [    4.437791] mt7530-mdio mdio-bus:1f lan1 (uninitialized): PHY [mt7530-0:02] driver [MediaTek MT7530 PHY] (irq=POLL)
> [    4.455096] mt7530-mdio mdio-bus:1f lan2 (uninitialized): PHY [mt7530-0:03] driver [MediaTek MT7530 PHY] (irq=POLL)
> [    4.472422] mt7530-mdio mdio-bus:1f lan3 (uninitialized): PHY [mt7530-0:04] driver [MediaTek MT7530 PHY] (irq=POLL)
> 
> I don't want to submit a bugfix to the net tree if the bug won't ever
> happen in real life.

It would be really great if you could tell us which bug fixes you're
submitting are for a real problem that you or a user have encountered,
and which are down to essentially code inspection and things that
"aren't correct". Basically, don't do this.

It isn't true that the PHY specific driver will be probed before DSA
initialises - consider the case where the DSA driver is built-in but
the PHY specific driver is modular and on the not-yet-mounted rootfs.
That would result in the generic PHY driver being used even when the
PHY specific driver were to be loaded later - and thus only basic
standard 802.3 PHY behaviour will be supported.

That's not specific to mt7530, it applies to everything that uses
phylib. It isn't something that really warrants "bug fixing" in each
and every driver.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 80Mbps down 10Mbps up. Decent connectivity at last!

Re: [PATCH 3/3] net: phy: mediatek-ge: do not disable EEE advertisement

[Arınç ÜNAL]

On 20.03.2024 23:53, Russell King (Oracle) wrote:
> On Wed, Mar 20, 2024 at 11:40:56PM +0300, Arınç ÜNAL wrote:
>> On 18.03.2024 10:46, Arınç ÜNAL via B4 Relay wrote:
>> Can I get an opinion on this? Is it actually possible that the PHY driver
>> would start probing after the DSA subdriver? On the console logs for the
>> DSA subdriver, I can see that the name of the PHY driver will appear, which
>> makes me believe the PHY driver would actually never probe after the DSA
>> subdriver.
>>
>> [    4.402641] mt7530-mdio mdio-bus:1f wan (uninitialized): PHY [mt7530-0:00] driver [MediaTek MT7530 PHY] (irq=POLL)
>> [    4.420392] mt7530-mdio mdio-bus:1f lan0 (uninitialized): PHY [mt7530-0:01] driver [MediaTek MT7530 PHY] (irq=POLL)
>> [    4.437791] mt7530-mdio mdio-bus:1f lan1 (uninitialized): PHY [mt7530-0:02] driver [MediaTek MT7530 PHY] (irq=POLL)
>> [    4.455096] mt7530-mdio mdio-bus:1f lan2 (uninitialized): PHY [mt7530-0:03] driver [MediaTek MT7530 PHY] (irq=POLL)
>> [    4.472422] mt7530-mdio mdio-bus:1f lan3 (uninitialized): PHY [mt7530-0:04] driver [MediaTek MT7530 PHY] (irq=POLL)
>>
>> I don't want to submit a bugfix to the net tree if the bug won't ever
>> happen in real life.
> 
> It would be really great if you could tell us which bug fixes you're
> submitting are for a real problem that you or a user have encountered,
> and which are down to essentially code inspection and things that
> "aren't correct". Basically, don't do this.

I agree. Patch 1 fixes a real problem, patch 2 "fixes" a problem found with
code inspection. Though, it would be great if you could review patch 2.

> 
> It isn't true that the PHY specific driver will be probed before DSA
> initialises - consider the case where the DSA driver is built-in but
> the PHY specific driver is modular and on the not-yet-mounted rootfs.
> That would result in the generic PHY driver being used even when the
> PHY specific driver were to be loaded later - and thus only basic
> standard 802.3 PHY behaviour will be supported.
> 
> That's not specific to mt7530, it applies to everything that uses
> phylib. It isn't something that really warrants "bug fixing" in each
> and every driver.

That makes sense. But there's a special case with the MT7530 DSA subdriver
and mediatek-ge driver. The PHY driver is needed for the PHYs to function
properly. So the DSA subdriver forces mediatek-ge to be selected [1]. So
the PHY driver could only be compiled as a module when the DSA subdriver is
also compiled so. And that designates mediatek-ge as a dependency for the
DSA subdriver, if I understand correctly.

[1] https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git/commit/?id=fb4bb62aaac715e50c7c007714af19a2698db88b

[PATCH 3/3] net: phy: mediatek-ge: do not disable EEE advertisement

[Arınç ÜNAL via B4 Relay]

From: Arınç ÜNAL <arinc.unal@arinc9.com>

There's no need to disable Energy-Efficient Ethernet (EEE) advertisement on
the MT7530 and MT7531 switch PHYs. EEE works fine on MT7530 and MT7531
switch PHYs. Remove the code where EEE advertisement is disabled.

This is a bugfix because there's a possible race condition where the
mediatek-ge driver would kick in after the MT7530 DSA subdriver which would
have EEE disabled until manually enabled.

Fixes: e40d2cca0189 ("net: phy: add MediaTek Gigabit Ethernet PHY driver")
Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
---
 drivers/net/phy/mediatek-ge.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/net/phy/mediatek-ge.c b/drivers/net/phy/mediatek-ge.c
index a493ae01b267..54ea64a37ab3 100644
--- a/drivers/net/phy/mediatek-ge.c
+++ b/drivers/net/phy/mediatek-ge.c
@@ -23,9 +23,6 @@ static int mtk_gephy_write_page(struct phy_device *phydev, int page)
 
 static void mtk_gephy_config_init(struct phy_device *phydev)
 {
-	/* Disable EEE */
-	phy_write_mmd(phydev, MDIO_MMD_AN, MDIO_AN_EEE_ADV, 0);
-
 	/* Enable HW auto downshift */
 	phy_modify_paged(phydev, MTK_PHY_PAGE_EXTENDED, 0x14, 0, BIT(4));
 

-- 
2.40.1