Re: [PATCH v3 4/5] drm/xe/xe_drm_ras: Wire get-counter support into xe_drm_ras for CRI

["Tauro, Riana" <riana.tauro@intel.com>]

On 4/23/2026 4:39 PM, Raag Jadav wrote:
> On Tue, Apr 21, 2026 at 08:21:00PM +0530, Riana Tauro wrote:
>> Hook CRI get-counter support to xe_drm_ras to allow userspace to query
>> counters if supported. When userspace queries a drm_ras error counter,
>> fetch the latest counter value from system controller.
>>
>> Integrate this with xe_drm_ras.
>>
>> Usage :
>>
>> Query all error counter value using ynl
>>
>> $ sudo ynl --family drm_ras --dump get-error-counter --json \
>> '{"node-id":0}'
>> [{'error-id': 1, 'error-name': 'core-compute', 'error-value': 0},
>>   {'error-id': 2, 'error-name': 'soc-internal', 'error-value': 0},
>>   {'error-id': 3, 'error-name': 'device-memory', 'error-value': 0},
>>   {'error-id': 4, 'error-name': 'pcie', 'error-value': 0},
>>   {'error-id': 5, 'error-name': 'fabric', 'error-value': 0}]
>>
>> Query single error counter value using ynl
>>
>> $ sudo ynl --family drm_ras  --do get-error-counter --json \
>>    '{"node-id":1, "error-id":1}'
> Nit: Align spacing with above command.
>
>> {'error-id': 1, 'error-name': 'core-compute', 'error-value': 2}
>>
>> Signed-off-by: Riana Tauro <riana.tauro@intel.com>
>> ---
>> v2: split patches (Raag)
>> ---
>>   drivers/gpu/drm/xe/xe_drm_ras.c | 19 ++++++++++++-------
>>   1 file changed, 12 insertions(+), 7 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/xe/xe_drm_ras.c b/drivers/gpu/drm/xe/xe_drm_ras.c
>> index e07dc23a155e..7040ff0f9539 100644
>> --- a/drivers/gpu/drm/xe/xe_drm_ras.c
>> +++ b/drivers/gpu/drm/xe/xe_drm_ras.c
>> @@ -11,19 +11,28 @@
>>   
>>   #include "xe_device_types.h"
>>   #include "xe_drm_ras.h"
>> +#include "xe_ras.h"
>>   
>>   static const char * const error_components[] = DRM_XE_RAS_ERROR_COMPONENT_NAMES;
>>   static const char * const error_severity[] = DRM_XE_RAS_ERROR_SEVERITY_NAMES;
>>   
>> -static int hw_query_error_counter(struct xe_drm_ras_counter *info,
>> +static int hw_query_error_counter(struct xe_device *xe,
>> +				  const enum drm_xe_ras_error_severity severity,
>>   				  u32 error_id, const char **name, u32 *val)
>>   {
>> +	struct xe_drm_ras *ras = &xe->ras;
>> +	struct xe_drm_ras_counter *info = ras->info[severity];
> Use int ret = 0 ...
>
>>   	if (!info || !info[error_id].name)
>>   		return -ENOENT;
>>   
>>   	*name = info[error_id].name;
>>   	*val = atomic_read(&info[error_id].counter);
> Move this as an else case below and ...

My bad!! made a mistake here. Will fix it

Thanks
Riana

>
>> +	/* Fetch counter from system controller if supported */
>> +	if (xe->info.has_sysctrl)
>> +		return xe_ras_get_error_counter(xe, severity, error_id, val);
> ret = get_counter();
>
>>   	return 0;
> return ret;
>
> Raag
>
>>   }