path: root/arch/powerpc/platforms/pseries/ras.c

// SPDX-License-Identifier: GPL-2.0-or-later
/*
 * Copyright (C) 2001 Dave Engebretsen IBM Corporation
 */

#include <linux/sched.h>
#include <linux/interrupt.h>
#include <linux/irq.h>
#include <linux/of.h>
#include <linux/fs.h>
#include <linux/reboot.h>
#include <linux/irq_work.h>

#include <asm/machdep.h>
#include <asm/rtas.h>
#include <asm/firmware.h>
#include <asm/mce.h>

#include "pseries.h"

static unsigned char ras_log_buf[RTAS_ERROR_LOG_MAX];
static DEFINE_SPINLOCK(ras_log_buf_lock);

static int ras_check_exception_token;

static void mce_process_errlog_event(struct irq_work *work);
static struct irq_work mce_errlog_process_work = {
	.func = mce_process_errlog_event,
};

#define EPOW_SENSOR_TOKEN	9
#define EPOW_SENSOR_INDEX	0

/* EPOW events counter variable */
static int num_epow_events;

static irqreturn_t ras_hotplug_interrupt(int irq, void *dev_id);
static irqreturn_t ras_epow_interrupt(int irq, void *dev_id);
static irqreturn_t ras_error_interrupt(int irq, void *dev_id);

/* RTAS pseries MCE errorlog section. */
struct pseries_mc_errorlog {
	__be32	fru_id;
	__be32	proc_id;
	u8	error_type;
	/*
	 * sub_err_type (1 byte). Bit fields depends on error_type
	 *
	 *   MSB0
	 *   |
	 *   V
	 *   01234567
	 *   XXXXXXXX
	 *
	 * For error_type == MC_ERROR_TYPE_UE
	 *   XXXXXXXX
	 *   X		1: Permanent or Transient UE.
	 *    X		1: Effective address provided.
	 *     X	1: Logical address provided.
	 *      XX	2: Reserved.
	 *        XXX	3: Type of UE error.
	 *
	 * For error_type != MC_ERROR_TYPE_UE
	 *   XXXXXXXX
	 *   X		1: Effective address provided.
	 *    XXXXX	5: Reserved.
	 *         XX	2: Type of SLB/ERAT/TLB error.
	 */
	u8	sub_err_type;
	u8	reserved_1[6];
	__be64	effective_address;
	__be64	logical_address;
} __packed;

/* RTAS pseries MCE error types */
#define MC_ERROR_TYPE_UE		0x00
#define MC_ERROR_TYPE_SLB		0x01
#define MC_ERROR_TYPE_ERAT		0x02
#define MC_ERROR_TYPE_UNKNOWN		0x03
#define MC_ERROR_TYPE_TLB		0x04
#define MC_ERROR_TYPE_D_CACHE		0x05
#define MC_ERROR_TYPE_I_CACHE		0x07

/* RTAS pseries MCE error sub types */
#define MC_ERROR_UE_INDETERMINATE		0
#define MC_ERROR_UE_IFETCH			1
#define MC_ERROR_UE_PAGE_TABLE_WALK_IFETCH	2
#define MC_ERROR_UE_LOAD_STORE			3
#define MC_ERROR_UE_PAGE_TABLE_WALK_LOAD_STORE	4

#define UE_EFFECTIVE_ADDR_PROVIDED		0x40
#define UE_LOGICAL_ADDR_PROVIDED		0x20

#define MC_ERROR_SLB_PARITY		0
#define MC_ERROR_SLB_MULTIHIT		1
#define MC_ERROR_SLB_INDETERMINATE	2

#define MC_ERROR_ERAT_PARITY		1
#define MC_ERROR_ERAT_MULTIHIT		2
#define MC_ERROR_ERAT_INDETERMINATE	3

#define MC_ERROR_TLB_PARITY		1
#define MC_ERROR_TLB_MULTIHIT		2
#define MC_ERROR_TLB_INDETERMINATE	3

static inline u8 rtas_mc_error_sub_type(const struct pseries_mc_errorlog *mlog)
{
	switch (mlog->error_type) {
	case	MC_ERROR_TYPE_UE:
		return (mlog->sub_err_type & 0x07);
	case	MC_ERROR_TYPE_SLB:
	case	MC_ERROR_TYPE_ERAT:
	case	MC_ERROR_TYPE_TLB:
		return (mlog->sub_err_type & 0x03);
	default:
		return 0;
	}
}

/*
 * Enable the hotplug interrupt late because processing them may touch other
 * devices or systems (e.g. hugepages) that have not been initialized at the
 * subsys stage.
 */
int __init init_ras_hotplug_IRQ(void)
{
	struct device_node *np;

	/* Hotplug Events */
	np = of_find_node_by_path("/event-sources/hot-plug-events");
	if (np != NULL) {
		if (dlpar_workqueue_init() == 0)
			request_event_sources_irqs(np, ras_hotplug_interrupt,
						   "RAS_HOTPLUG");
		of_node_put(np);
	}

	return 0;
}
machine_late_initcall(pseries, init_ras_hotplug_IRQ);

/*
 * Initialize handlers for the set of interrupts caused by hardware errors
 * and power system events.
 */
static int __init init_ras_IRQ(void)
{
	struct device_node *np;

	ras_check_exception_token = rtas_token("check-exception");

	/* Internal Errors */
	np = of_find_node_by_path("/event-sources/internal-errors");
	if (np != NULL) {
		request_event_sources_irqs(np, ras_error_interrupt,
					   "RAS_ERROR");
		of_node_put(np);
	}

	/* EPOW Events */
	np = of_find_node_by_path("/event-sources/epow-events");
	if (np != NULL) {
		request_event_sources_irqs(np, ras_epow_interrupt, "RAS_EPOW");
		of_node_put(np);
	}

	return 0;
}
machine_subsys_initcall(pseries, init_ras_IRQ);

#define EPOW_SHUTDOWN_NORMAL				1
#define EPOW_SHUTDOWN_ON_UPS				2
#define EPOW_SHUTDOWN_LOSS_OF_CRITICAL_FUNCTIONS	3
#define EPOW_SHUTDOWN_AMBIENT_TEMPERATURE_TOO_HIGH	4

static void handle_system_shutdown(char event_modifier)
{
	switch (event_modifier) {
	case EPOW_SHUTDOWN_NORMAL:
		pr_emerg("Power off requested\n");
		orderly_poweroff(true);
		break;

	case EPOW_SHUTDOWN_ON_UPS:
		pr_emerg("Loss of system power detected. System is running on"
			 " UPS/battery. Check RTAS error log for details\n");
		orderly_poweroff(true);
		break;

	case EPOW_SHUTDOWN_LOSS_OF_CRITICAL_FUNCTIONS:
		pr_emerg("Loss of system critical functions detected. Check"
			 " RTAS error log for details\n");
		orderly_poweroff(true);
		break;

	case EPOW_SHUTDOWN_AMBIENT_TEMPERATURE_TOO_HIGH:
		pr_emerg("High ambient temperature detected. Check RTAS"
			 " error log for details\n");
		orderly_poweroff(true);
		break;

	default:
		pr_err("Unknown power/cooling shutdown event (modifier = %d)\n",
			event_modifier);
	}
}

struct epow_errorlog {
	unsigned char sensor_value;
	unsigned char event_modifier;
	unsigned char extended_modifier;
	unsigned char reserved;
	unsigned char platform_reason;
};

#define EPOW_RESET			0
#define EPOW_WARN_COOLING		1
#define EPOW_WARN_POWER			2
#define EPOW_SYSTEM_SHUTDOWN		3
#define EPOW_SYSTEM_HALT		4
#define EPOW_MAIN_ENCLOSURE		5
#define EPOW_POWER_OFF			7

static void rtas_parse_epow_errlog(struct rtas_error_log *log)
{
	struct pseries_errorlog *pseries_log;
	struct epow_errorlog *epow_log;
	char action_code;
	char modifier;

	pseries_log = get_pseries_errorlog(log, PSERIES_ELOG_SECT_ID_EPOW);
	if (pseries_log == NULL)
		return;

	epow_log = (struct epow_errorlog *)pseries_log->data;
	action_code = epow_log->sensor_value & 0xF;	/* bottom 4 bits */
	modifier = epow_log->event_modifier & 0xF;	/* bottom 4 bits */

	switch (action_code) {
	case EPOW_RESET:
		if (num_epow_events) {
			pr_info("Non critical power/cooling issue cleared\n");
			num_epow_events--;
		}
		break;

	case EPOW_WARN_COOLING:
		pr_info("Non-critical cooling issue detected. Check RTAS error"
			" log for details\n");
		break;

	case EPOW_WAR