net: phy: dp83867: perform soft reset and retain established link
Subsystems (3)
PATCH 1/1 net: phy: dp83867: perform soft reset and retain established link
From: Praneeth Bajjuri <[email protected]>
Current logic is performing hard reset and causing the programmed registers to be wiped out.
as per datasheet: https://www.ti.com/lit/ds/symlink/dp83867cr.pdf
8.6.26 Control Register (CTRL)
do SW_RESTART to perform a reset not including the registers and is acceptable to do this if a link is already present.
--- drivers/net/phy/dp83867.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/phy/dp83867.c +1 -1
@@ -826,7 +826,7 @@ static int dp83867_phy_reset(struct phy_device *phydev) {int err; - err =phy_write (phydev,DP83867_CTRL ,DP83867_SW_RESET ); + err =phy_write (phydev,DP83867_CTRL ,DP83867_SW_RESTART );if (err <0 )return err;
Signature · 1 line
2.17.1
Andrew Lunn
maintainerQuoted 9 lines · On Tue, Mar 23, 2021 at 08:00:06PM -0500, [email protected] wrote:
From: Praneeth Bajjuri <[email protected]>
Current logic is performing hard reset and causing the programmed registers to be wiped out.
as per datasheet: https://www.ti.com/lit/ds/symlink/dp83867cr.pdf
8.6.26 Control Register (CTRL)
doSW_RESTARTto perform a reset not including the registers and is acceptable to do this if a link is already present.
I don't see any code here to determine if the like is present. What if the cable is not plugged in?
drivers/net/phy/dp83867.c static int dp83867_phy_reset(struct phy_device *phydev) · line 831
Quoted 8 lines
drivers/net/phy/dp83867.c +1
-1@@ -826,7 +826,7 @@ static int dp83867_phy_reset(struct phy_device *phydev) {int err; - err =phy_write (phydev,DP83867_CTRL ,DP83867_SW_RESET ); + err =phy_write (phydev,DP83867_CTRL ,DP83867_SW_RESTART );if (err <0 )return err;
The code continues
usleep_range(10, 20);
/* After reset FORCE_LINK_GOOD bit is set. Although the
* default value should be unset. Disable FORCE_LINK_GOOD
* for the phy to work properly.
*/
return phy_modify(phydev, MII_DP83867_PHYCTRL,
DP83867_PHYCR_FORCE_LINK_GOOD, 0);
}Do you still need to clear the FORCE_LINK_GOOD bit after a restart?
Andrew
Andrew Lunn
maintainerQuoted 8 lines
Quoted 5 lines
as per datasheet: https://www.ti.com/lit/ds/symlink/dp83867cr.pdf
8.6.26 Control Register (CTRL)
doSW_RESTARTto perform a reset not including the registers and is acceptable to do this if a link is already present.
Quoted 6 lines
I don't see any code here to determine if the like is present. What if the cable is not plugged in?
This API is primarily used for reset. Link Status is checked thru different
register. This shall not impact the cable plug in/out. With this change, it will align with DP83822 driver API.
So why is there the comment:
Quoted 2 lines
Quoted 2 lines
and is
acceptable to do this if a link is already present.
That kind of says, it is not acceptable to do this if the link is not present. Which is why i'm asking.
Andrew
Hi Andrew,
If Link is present, the current reset will reset the registers including the link and Host will need to re-program all the registers. With the change proposed, it will reset the state machine while keeping the register content.
My understanding is that below API is called to perform the soft reset and not complete wipe of PHY. Do you agree ?
.soft_reset = dp83867_phy_reset,
Regards,
Geet
See below difference between the two reset from DP83867 datasheet.
8.5.5.3 Global Software Reset
A global software reset is accomplished by setting bit 15 of register CTRL (address 0x001F) to 1. This bit resets all the internal circuits in the PHY including IEEE-defined registers and all the extended registers. The global software reset resets the device such that all registers are reset to default values and the hardware configuration values are maintained.
8.5.5.4 Global Software Restart
A global software restart is accomplished by setting bit 14 of register CTRL (0x001F) to 1. This action resets all the PHY circuits except the registers in the Register File.
In DP83822 linux, the API is calling the software reset and we are aligning DP83867 implementaiton with DP83822.
Regards,
Geet
Quoted 8 lines · Andrew Lunn
Quoted 5 lines
as per datasheet: https://www.ti.com/lit/ds/symlink/dp83867cr.pdf
8.6.26 Control Register (CTRL)
doSW_RESTARTto perform a reset not including the registers and is acceptable to do this if a link is already present.
Quoted 6 lines
I don't see any code here to determine if the like is present. What if the cable is not plugged in?
This API is primarily used for reset. Link Status is checked thru different
register. This shall not impact the cable plug in/out. With this change, it will align with DP83822 driver API.
So why is there the comment:
> > and is
> > acceptable to do this if a link is already present.
That kind of says, it is not acceptable to do this if the link is not present. Which is why i'm asking.
Andrew
Hi Andrew,
Quoted 15 lines · Andrew Lunn
Quoted 8 lines
> as per datasheet: https://www.ti.com/lit/ds/symlink/dp83867cr.pdf
> 8.6.26 Control Register (CTRL)
> doSW_RESTARTto perform a reset not including the registers and is > acceptable to do this if a link is already present.Quoted 6 lines
I don't see any code here to determine if the like is present. What if the cable is not plugged in?
This API is primarily used for reset. Link Status is checked thru different
register. This shall not impact the cable plug in/out. With this change, it will align with DP83822 driver API.
Quoted 7 lines
So why is there the comment:
Quoted 2 lines
> and is
> acceptable to do this if a link is already present.That kind of says, it is not acceptable to do this if the link is not present. Which is why i'm asking.
Does the feedback from Geet help in clarity you requested. Ref:
https://lore.kernel.org/netdev/[email protected]/
Quoted 3 lines
Andrew
Andrew Lunn
maintainerQuoted 22 lines · Praneeth
Hi Andrew,
Quoted 19 lines · Andrew Lunn
Quoted 7 lines
> as per datasheet: https://www.ti.com/lit/ds/symlink/dp83867cr.pdf
> 8.6.26 Control Register (CTRL)
> doSW_RESTARTto perform a reset not including the registers and is > acceptable to do this if a link is already present.Quoted 6 lines
I don't see any code here to determine if the like is present. What if the cable is not plugged in?
This API is primarily used for reset. Link Status is checked thru different
register. This shall not impact the cable plug in/out. With this change, it will align with DP83822 driver API.So why is there the comment:
Quoted 2 lines
> and is
> acceptable to do this if a link is already present.
Quoted 6 lines
Quoted 2 lines
That kind of says, it is not acceptable to do this if the link is not present. Which is why i'm asking.
Does the feedback from Geet help in clarity you requested. Ref:
https://lore.kernel.org/netdev/[email protected]/
Not really.
Quoted 2 lines
and is
Quoted 1 line
Quoted 1 line
> acceptable to do this if a link is already present.
There needs to be something to either:
1) Ensure there is link, so we known we are within acceptable behaviour.
2) Document what happens when there is no link, meaning we do something which is not acceptable. Is the magic smoke going to be released? Does the link die until the next reboot? Or despite it being unacceptable, nothing really happens, and it is not a problem?
Andrew
Andrew Lunn
maintainerQuoted 16 lines · Geet
Hello Andrew,
Please let me know if you have additional questions/clarifications to approve below change request.
Regards,
Geet
Quoted 8 lines
From: Geet Modi <[email protected]>
Date: Friday, May 28, 2021 at 10:10 AM
To: Andrew Lunn <[email protected]>, "Bajjuri, Praneeth" <[email protected]>
Cc: "David S . Miller" <[email protected]>, Jakub Kicinski <[email protected]>, "[email protected]" <[email protected]>, "[email protected]" <[email protected]>
Subject: Re: [EXTERNAL] Re: [EXTERNAL] Re: [PATCH] net: phy: dp83867: perform soft reset and retain established link
So this all seems to boil down to, it does not matter if it is acceptable or not, you are going to do it. So please just remove that part of the comment. It has no value.
Andrew
Praneeth,
Can you please help edit the comment and resubmit for approval ?
Regards,
Geet
On 6/1/21, 12:18 PM, "Andrew Lunn" <[email protected]> wrote:
On Tue, Jun 01, 2021 at 07:01:04PM +0000, Modi, Geet wrote:
> Hello Andrew,
>
>
>
> Please let me know if you have additional questions/clarifications to approve > below change request.
>
>
>
> Regards,
> Geet
>
>
>
>
>
> From: Geet Modi <[email protected]>
> Date: Friday, May 28, 2021 at 10:10 AM
> To: Andrew Lunn <[email protected]>, "Bajjuri, Praneeth" <[email protected]> > Cc: "David S . Miller" <[email protected]>, Jakub Kicinski <[email protected]>, > "[email protected]" <[email protected]>,
> "[email protected]" <[email protected]>
> Subject: Re: [EXTERNAL] Re: [EXTERNAL] Re: [PATCH] net: phy: dp83867: perform > soft reset and retain established link
So this all seems to boil down to, it does not matter if it is acceptable or not, you are going to do it. So please just remove that part of the comment. It has no value.
Andrew
PATCH v2 1/1 net: phy: dp83867: perform soft reset and retain established link
In reply to Bajjuri, PraneethFrom: Praneeth Bajjuri <[email protected]>
Current logic is performing hard reset and causing the programmed registers to be wiped out.
as per datasheet: https://www.ti.com/lit/ds/symlink/dp83867cr.pdf
8.6.26 Control Register (CTRL)
do SW_RESTART to perform a reset not including the registers, If performed when link is already present,
it will drop the link and trigger re-auto negotiation.
--- drivers/net/phy/dp83867.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-)
drivers/net/phy/dp83867.c +1 -5
@@ -826,16 +826,12 @@ static int dp83867_phy_reset(struct phy_device *phydev) {int err; - err =phy_write (phydev,DP83867_CTRL ,DP83867_SW_RESET ); + err =phy_write (phydev,DP83867_CTRL ,DP83867_SW_RESTART );if (err <0 )return err;usleep_range (10 ,20 ); -/* After reset FORCE_LINK_GOOD bit is set. Although the -* default value should be unset. Disable FORCE_LINK_GOOD -* for the phy to work properly. -*/ return phy_modify (phydev,MII_DP83867_PHYCTRL ,DP83867_PHYCR_FORCE_LINK_GOOD ,0 ); }
Signature · 1 line
2.17.1
Andrew,
Quoted 26 lines · Andrew Lunn
Quoted 16 lines · Geet
Hello Andrew,
Please let me know if you have additional questions/clarifications to approve below change request.
Regards,
GeetQuoted 8 lines
From: Geet Modi <[email protected]>
Date: Friday, May 28, 2021 at 10:10 AM
To: Andrew Lunn <[email protected]>, "Bajjuri, Praneeth" <[email protected]>
Cc: "David S . Miller" <[email protected]>, Jakub Kicinski <[email protected]>, "[email protected]" <[email protected]>, "[email protected]" <[email protected]>
Subject: Re: [EXTERNAL] Re: [EXTERNAL] Re: [PATCH] net: phy: dp83867: perform soft reset and retain established link
Quoted 3 lines
So this all seems to boil down to, it does not matter if it is acceptable or not, you are going to do it. So please just remove that part of the comment. It has no value.
Sent v2 addressing comment as per your suggestion.
Ref: https://lore.kernel.org/patchwork/patch/1444281/
Thanks
Praneeth
Quoted 3 lines
Andrew
Quoted 8 lines · On Wed, Jun 09, 2021 at 07:43:42PM -0500, [email protected] wrote:
From: Praneeth Bajjuri <[email protected]>
Current logic is performing hard reset and causing the programmed registers to be wiped out.
as per datasheet: https://www.ti.com/lit/ds/symlink/dp83867cr.pdf
8.6.26 Control Register (CTRL)
Quoted 6 lines
do
SW_RESTARTto perform a reset not including the registers, If performed when link is already present,
it will drop the link and trigger re-auto negotiation.Signed-off-by Praneeth Bajjuri <[email protected]>Signed-off-by Geet Modi <[email protected]>
Andrew
Hello,
Quoted 14 lines · Andrew Lunn
Quoted 12 lines · On Wed, Jun 09, 2021 at 07:43:42PM -0500, [email protected] wrote:
From: Praneeth Bajjuri <[email protected]>
Current logic is performing hard reset and causing the programmed registers to be wiped out.
as per datasheet: https://www.ti.com/lit/ds/symlink/dp83867cr.pdf
8.6.26 Control Register (CTRL)do
SW_RESTARTto perform a reset not including the registers, If performed when link is already present,
it will drop the link and trigger re-auto negotiation.
Quoted 6 lines
Quoted 2 lines
Signed-off-by Praneeth Bajjuri <[email protected]>Signed-off-by Geet Modi <[email protected]>Reviewed-by Andrew Lunn <[email protected]>Andrew
I reported a few days ago an issue with the DP83822 which I think is caused by a similar change.
https://lore.kernel.org/netdev/CAHvQdo2yzJC89K74c_CZFjPydDQ5i22w36XPR5tKVv_W8a2vcg@mail.gmail.com/
In my case I can't get an link after this change, reverting it fixes the problem for me.
Hannes
Hannes,
Quoted 22 lines · Johannes Pointner
Hello,
Quoted 19 lines · Andrew Lunn
Quoted 8 lines · On Wed, Jun 09, 2021 at 07:43:42PM -0500, [email protected] wrote:
From: Praneeth Bajjuri <[email protected]>
Current logic is performing hard reset and causing the programmed registers to be wiped out.
as per datasheet: https://www.ti.com/lit/ds/symlink/dp83867cr.pdf
8.6.26 Control Register (CTRL)Quoted 6 lines
do
SW_RESTARTto perform a reset not including the registers, If performed when link is already present,
it will drop the link and trigger re-auto negotiation.Signed-off-by Praneeth Bajjuri <[email protected]>Signed-off-by Geet Modi <[email protected]>Reviewed-by Andrew Lunn <[email protected]>
Quoted 7 lines
Quoted 1 line
Andrew
I reported a few days ago an issue with the DP83822 which I think is caused by a similar change.
https://lore.kernel.org/netdev/CAHvQdo2yzJC89K74c_CZFjPydDQ5i22w36XPR5tKVv_W8a2vcg@mail.gmail.com/
In my case I can't get an link after this change, reverting it fixes the problem for me.
Are you saying that instead of reset if sw_restart is done as per this patch, there is no issue?
Quoted 3 lines
Hannes
This patch was applied to netdev/net.git (refs/heads/master):
Hello:
This patch was applied to netdev/net.git (refs/heads/master):
Quoted 9 lines · On Wed, 9 Jun 2021 19:43:42 -0500 you wrote:
From: Praneeth Bajjuri <[email protected]>
Current logic is performing hard reset and causing the programmed registers to be wiped out.
as per datasheet: https://www.ti.com/lit/ds/symlink/dp83867cr.pdf
8.6.26 Control Register (CTRL)[...]
Here is the summary with links:
- [v2] net: phy: dp83867: perform soft reset and retain established link
https://git.kernel.org/netdev/net/c/da9ef50f545f
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
Hello Praneeth,
Quoted 28 lines · Praneeth
Hannes,
Quoted 24 lines · Johannes Pointner
Hello,
Quoted 20 lines · Andrew Lunn
Quoted 8 lines · On Wed, Jun 09, 2021 at 07:43:42PM -0500, [email protected] wrote:
From: Praneeth Bajjuri <[email protected]>
Current logic is performing hard reset and causing the programmed registers to be wiped out.
as per datasheet: https://www.ti.com/lit/ds/symlink/dp83867cr.pdf
8.6.26 Control Register (CTRL)Quoted 6 lines
do
SW_RESTARTto perform a reset not including the registers, If performed when link is already present,
it will drop the link and trigger re-auto negotiation.Signed-off-by Praneeth Bajjuri <[email protected]>Signed-off-by Geet Modi <[email protected]>Reviewed-by Andrew Lunn <[email protected]>Andrew
Quoted 8 lines
Quoted 5 lines
I reported a few days ago an issue with the DP83822 which I think is caused by a similar change.
https://lore.kernel.org/netdev/CAHvQdo2yzJC89K74c_CZFjPydDQ5i22w36XPR5tKVv_W8a2vcg@mail.gmail.com/
In my case I can't get an link after this change, reverting it fixes the problem for me.Are you saying that instead of reset if
sw_restartis done as per this patch, there is no issue?
In my case(DP83822 connected to an i.MX6) if the digital(SW) restart is used (Bit 14) I have the issue that I can' get a link.
ip addr shows:
1: lo: <LOOPBACK> mtu 65536 qdisc noop qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: eth0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000
link/ether 00:60:65:54:32:10 brd ff:ff:ff:ff:ff:ff
If I revert this back to using the SW reset (Bit 15) it works again.
Regards,
Hannes