Re: [PATCH 2/2] watchdog: sp5100_tco: Enable watchdog on Family 17h devices if disabled

[Guenter Roeck]

Hi Jan,

On 9/10/20 9:34 AM, Jan Kiszka wrote:
> On 10.09.20 18:31, Guenter Roeck wrote:
>> On Family 17h (Ryzen) devices, the WatchdogTmrEn bit of PmDecodeEn not only
>> enables watchdog memory decoding at 0xfeb00000, it also enables the
>> watchdog hardware itself. Use this information to enable the watchdog if
>> it is not already enabled.
>>
>> Cc: Jan Kiszka <jan.kiszka@xxxxxxxxxxx>
>> Signed-off-by: Guenter Roeck <linux@xxxxxxxxxxxx>
>> ---
>>  drivers/watchdog/sp5100_tco.c | 18 ++++++++++++++++++
>>  1 file changed, 18 insertions(+)
>>
>> diff --git a/drivers/watchdog/sp5100_tco.c b/drivers/watchdog/sp5100_tco.c
>> index 85e9664318c9..a730ecbf78cd 100644
>> --- a/drivers/watchdog/sp5100_tco.c
>> +++ b/drivers/watchdog/sp5100_tco.c
>> @@ -17,6 +17,12 @@
>>   *	    AMD Publication 51192 "AMD Bolton FCH Register Reference Guide"
>>   *	    AMD Publication 52740 "BIOS and Kernel Developer’s Guide (BKDG)
>>   *				for AMD Family 16h Models 30h-3Fh Processors"
>> + *	    AMD Publication 55570-B1-PUB "Processor Programming Reference (PPR)
>> + *				for AMD Family 17h Model 18h, Revision B1
>> + *				Processors (PUB)
>> + *	    AMD Publication 55772-A1-PUB "Processor Programming Reference (PPR)
>> + *				for AMD Family 17h Model 20h, Revision A1
>> + *				Processors (PUB)
>>   */
>>  
>>  /*
>> @@ -241,6 +247,18 @@ static int sp5100_tco_setupdevice(struct device *dev,
>>  		break;
>>  	case efch:
>>  		dev_name = SB800_DEVNAME;
>> +		/*
>> +		 * On Family 17h devices, the EFCH_PM_DECODEEN_WDT_TMREN bit of
>> +		 * EFCH_PM_DECODEEN not only enables the EFCH_PM_WDT_ADDR memory
>> +		 * region, it also enables the watchdog itself.
>> +		 */
>> +		if (boot_cpu_data.x86 == 0x17) {
>> +			val = sp5100_tco_read_pm_reg8(EFCH_PM_DECODEEN);
>> +			if (!(val & EFCH_PM_DECODEEN_WDT_TMREN)) {
>> +				sp5100_tco_update_pm_reg8(EFCH_PM_DECODEEN, 0xff,
>> +							  EFCH_PM_DECODEEN_WDT_TMREN);
>> +			}
>> +		}
>>  		val = sp5100_tco_read_pm_reg8(EFCH_PM_DECODEEN);
>>  		if (val & EFCH_PM_DECODEEN_WDT_TMREN)
>>  			mmio_addr = EFCH_PM_WDT_ADDR;
>>
> 
> Won't that bring us EFCH_PM_WDT_ADDR as address, rather than
> EFCH_PM_ACPI_MMIO_ADDR which worked in my case? Or is one an alias of
> the other.
> 

Yes, it does use EFCH_PM_WDT_ADDR. EFCH_PM_ACPI_MMIO_ADDR works as well,
but is meant to be a fallback. Both point to the watchdog memory space.

Thanks,
Guenter