usb: typec: tcpm: fix cc role at port reset

usb: typec: tcpm: fix cc role at port reset

[Frank Wang]

In the current implementation, the tcpm set CC1/CC2 role to open when
it do port reset would cause the VBUS removed by the Type-C partner.

This sets CC1/CC2 according to the default state of port to fix it.

Comments are suggested by Guenter Roeck.

Signed-off-by: Frank Wang <frank.wang@rock-chips.com>
---
 drivers/usb/typec/tcpm/tcpm.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/typec/tcpm/tcpm.c b/drivers/usb/typec/tcpm/tcpm.c
index 3c6b0c8e2d3ae..9f6aaa3e70ca8 100644
--- a/drivers/usb/typec/tcpm/tcpm.c
+++ b/drivers/usb/typec/tcpm/tcpm.c
@@ -4885,7 +4885,8 @@ static void run_state_machine(struct tcpm_port *port)
 		break;
 	case PORT_RESET:
 		tcpm_reset_port(port);
-		tcpm_set_cc(port, TYPEC_CC_OPEN);
+		tcpm_set_cc(port, tcpm_default_state(port) == SNK_UNATTACHED ?
+			    TYPEC_CC_RD : tcpm_rp_cc(port));
 		tcpm_set_state(port, PORT_RESET_WAIT_OFF,
 			       PD_T_ERROR_RECOVERY);
 		break;
-- 
2.17.1

Re: usb: typec: tcpm: fix cc role at port reset

[Guenter Roeck]

On 6/15/23 19:20, Frank Wang wrote:
> In the current implementation, the tcpm set CC1/CC2 role to open when
> it do port reset would cause the VBUS removed by the Type-C partner.
> 
> This sets CC1/CC2 according to the default state of port to fix it.
> 
> Comments are suggested by Guenter Roeck.
> 

I have no idea (recollection) of what I suggested here :-(

Guenter

> Signed-off-by: Frank Wang <frank.wang@rock-chips.com>
> ---
>   drivers/usb/typec/tcpm/tcpm.c | 3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/usb/typec/tcpm/tcpm.c b/drivers/usb/typec/tcpm/tcpm.c
> index 3c6b0c8e2d3ae..9f6aaa3e70ca8 100644
> --- a/drivers/usb/typec/tcpm/tcpm.c
> +++ b/drivers/usb/typec/tcpm/tcpm.c
> @@ -4885,7 +4885,8 @@ static void run_state_machine(struct tcpm_port *port)
>   		break;
>   	case PORT_RESET:
>   		tcpm_reset_port(port);
> -		tcpm_set_cc(port, TYPEC_CC_OPEN);
> +		tcpm_set_cc(port, tcpm_default_state(port) == SNK_UNATTACHED ?
> +			    TYPEC_CC_RD : tcpm_rp_cc(port));
>   		tcpm_set_state(port, PORT_RESET_WAIT_OFF,
>   			       PD_T_ERROR_RECOVERY);
>   		break;

Re: usb: typec: tcpm: fix cc role at port reset

[Frank Wang]

Hi Guenter,

On 2023/6/16 10:27, Guenter Roeck wrote:
> On 6/15/23 19:20, Frank Wang wrote:
>> In the current implementation, the tcpm set CC1/CC2 role to open when
>> it do port reset would cause the VBUS removed by the Type-C partner.
>>
>> This sets CC1/CC2 according to the default state of port to fix it.
>>
>> Comments are suggested by Guenter Roeck.
>>
>
> I have no idea (recollection) of what I suggested here :-(

Aha, this is also an old patch I have sent before, you suggested that 
the cc role should require some tweaking based on the port type or 
default state, do you remember it now?

Anyway, Can you help to review this patch again?


BR.
Frank

>
> Guenter
>
>> Signed-off-by: Frank Wang <frank.wang@rock-chips.com>
>> ---
>>   drivers/usb/typec/tcpm/tcpm.c | 3 ++-
>>   1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/usb/typec/tcpm/tcpm.c 
>> b/drivers/usb/typec/tcpm/tcpm.c
>> index 3c6b0c8e2d3ae..9f6aaa3e70ca8 100644
>> --- a/drivers/usb/typec/tcpm/tcpm.c
>> +++ b/drivers/usb/typec/tcpm/tcpm.c
>> @@ -4885,7 +4885,8 @@ static void run_state_machine(struct tcpm_port 
>> *port)
>>           break;
>>       case PORT_RESET:
>>           tcpm_reset_port(port);
>> -        tcpm_set_cc(port, TYPEC_CC_OPEN);
>> +        tcpm_set_cc(port, tcpm_default_state(port) == SNK_UNATTACHED ?
>> +                TYPEC_CC_RD : tcpm_rp_cc(port));
>>           tcpm_set_state(port, PORT_RESET_WAIT_OFF,
>>                      PD_T_ERROR_RECOVERY);
>>           break;
>

Re: usb: typec: tcpm: fix cc role at port reset

[Greg KH]

On Fri, Jun 16, 2023 at 11:03:14AM +0800, Frank Wang wrote:
> Hi Guenter,
> 
> On 2023/6/16 10:27, Guenter Roeck wrote:
> > On 6/15/23 19:20, Frank Wang wrote:
> > > In the current implementation, the tcpm set CC1/CC2 role to open when
> > > it do port reset would cause the VBUS removed by the Type-C partner.
> > > 
> > > This sets CC1/CC2 according to the default state of port to fix it.
> > > 
> > > Comments are suggested by Guenter Roeck.
> > > 
> > 
> > I have no idea (recollection) of what I suggested here :-(
> 
> Aha, this is also an old patch I have sent before, you suggested that the cc
> role should require some tweaking based on the port type or default state,
> do you remember it now?

Please version the patch, and provide the change information as
documented.

thanks,

greg k-h