Commit 532d7f6b authored by Shashank Gupta's avatar Shashank Gupta Committed by Herbert Xu
Browse files

crypto: qat - add error counters 



Introduce ras counters interface for counting QAT specific device
errors and expose them through the newly created qat_ras sysfs
group attribute.

This adds the following attributes:

- errors_correctable: number of correctable errors
- errors_nonfatal: number of uncorrectable non fatal errors
- errors_fatal: number of uncorrectable fatal errors
- reset_error_counters: resets all counters

These counters are initialized during device bring up and cleared
during device shutdown and are applicable only to QAT GEN4 devices.

Signed-off-by: default avatarShashank Gupta <shashank.gupta@intel.com>
Reviewed-by: default avatarGiovanni Cabiddu <giovanni.cabiddu@intel.com>
Reviewed-by: default avatarTero Kristo <tero.kristo@linux.intel.com>
Signed-off-by: default avatarHerbert Xu <herbert@gondor.apana.org.au>
parent 22289dc9

Documentation/ABI/testing/sysfs-driver-qat_ras

0 → 100644
+41 −0
Original line number Diff line number Diff line 
What:		/sys/bus/pci/devices/<BDF>/qat_ras/errors_correctable

Date:		January 2024

KernelVersion:	6.7

Contact:	qat-linux@intel.com

Description:	(RO) Reports the number of correctable errors detected by the device.



		This attribute is only available for qat_4xxx devices.



What:		/sys/bus/pci/devices/<BDF>/qat_ras/errors_nonfatal

Date:		January 2024

KernelVersion:	6.7

Contact:	qat-linux@intel.com

Description:	(RO) Reports the number of non fatal errors detected by the device.



		This attribute is only available for qat_4xxx devices.



What:		/sys/bus/pci/devices/<BDF>/qat_ras/errors_fatal

Date:		January 2024

KernelVersion:	6.7

Contact:	qat-linux@intel.com

Description:	(RO) Reports the number of fatal errors detected by the device.



		This attribute is only available for qat_4xxx devices.



What:		/sys/bus/pci/devices/<BDF>/qat_ras/reset_error_counters

Date:		January 2024

KernelVersion:	6.7

Contact:	qat-linux@intel.com

Description:	(WO) Write to resets all error counters of a device.



		The following example reports how to reset the counters::



			# echo 1 > /sys/bus/pci/devices/<BDF>/qat_ras/reset_error_counters

			# cat /sys/bus/pci/devices/<BDF>/qat_ras/errors_correctable

			0

			# cat /sys/bus/pci/devices/<BDF>/qat_ras/errors_nonfatal

			0

			# cat /sys/bus/pci/devices/<BDF>/qat_ras/errors_fatal

			0



		This attribute is only available for qat_4xxx devices.

drivers/crypto/intel/qat/qat_4xxx/adf_drv.c

+1 −0
Original line number Diff line number Diff line
@@ -418,6 +418,7 @@ static int adf_probe(struct pci_dev *pdev, const struct pci_device_id *ent) 
		goto out_err;

	}



	accel_dev->ras_errors.enabled = true;

	adf_dbgfs_init(accel_dev);



	ret = adf_dev_up(accel_dev, true);

drivers/crypto/intel/qat/qat_common/Makefile

+1 −0
Original line number Diff line number Diff line
@@ -12,6 +12,7 @@ intel_qat-objs := adf_cfg.o \ 
	adf_admin.o \

	adf_hw_arbiter.o \

	adf_sysfs.o \

	adf_sysfs_ras_counters.o \

	adf_gen2_hw_data.o \

	adf_gen2_config.o \

	adf_gen4_hw_data.o \

drivers/crypto/intel/qat/qat_common/adf_accel_devices.h

+14 −0
Original line number Diff line number Diff line
@@ -7,6 +7,7 @@ 
#include <linux/list.h>

#include <linux/io.h>

#include <linux/ratelimit.h>

#include <linux/types.h>

#include "adf_cfg_common.h"

#include "adf_pfvf_msg.h"
@@ -81,6 +82,18 @@ enum dev_sku_info { 
	DEV_SKU_UNKNOWN,

};



enum ras_errors {

	ADF_RAS_CORR,

	ADF_RAS_UNCORR,

	ADF_RAS_FATAL,

	ADF_RAS_ERRORS,

};



struct adf_error_counters {

	atomic_t counter[ADF_RAS_ERRORS];

	bool enabled;

};



static inline const char *get_sku_info(enum dev_sku_info info)

{

	switch (info) {
@@ -360,6 +373,7 @@ struct adf_accel_dev { 
			u8 pf_compat_ver;

		} vf;

	};

	struct adf_error_counters ras_errors;

	struct mutex state_lock; /* protect state of the device */

	bool is_vf;

	u32 accel_id;

drivers/crypto/intel/qat/qat_common/adf_init.c

+3 −0
Original line number Diff line number Diff line
@@ -9,6 +9,7 @@ 
#include "adf_common_drv.h"

#include "adf_dbgfs.h"

#include "adf_heartbeat.h"

#include "adf_sysfs_ras_counters.h"



static LIST_HEAD(service_table);

static DEFINE_MUTEX(service_lock);
@@ -242,6 +243,7 @@ static int adf_dev_start(struct adf_accel_dev *accel_dev) 
	set_bit(ADF_STATUS_COMP_ALGS_REGISTERED, &accel_dev->status);



	adf_dbgfs_add(accel_dev);

	adf_sysfs_start_ras(accel_dev);



	return 0;

}
@@ -268,6 +270,7 @@ static void adf_dev_stop(struct adf_accel_dev *accel_dev) 
		return;



	adf_dbgfs_rm(accel_dev);

	adf_sysfs_stop_ras(accel_dev);



	clear_bit(ADF_STATUS_STARTING, &accel_dev->status);

	clear_bit(ADF_STATUS_STARTED, &accel_dev->status);