> On 30-Apr-2015, at 11:37, Vipin K Parashar <vi...@linux.vnet.ibm.com> wrote:
>
> This patch adds support for FSP EPOW (Early Power Off Warning) and
> DPO (Delayed Power Off) events support for PowerNV platform. EPOW events
> are generated by SPCN/FSP due to various critical system conditions that
> need system shutdown. Few examples of these conditions are high ambient
> temperature or system running on UPS power with low UPS battery. DPO event
> is generated in response to admin initiated system shutdown request.
> This patch enables host kernel on PowerNV platform to handle OPAL
> notifications for these events and initiate system poweroff. Since EPOW
> notifications are sent in advance of impending shutdown event and thus this
> patch also adds functionality to wait for EPOW condition to return to
> normal. If EPOW condition doesn't return to normal in estimated time it
> proceeds with graceful system shutdown. System admin can also add host
> userspace scripts to perform any specific actions like graceful guest
> shutdown upon system poweroff.
>
> Signed-off-by: Vipin K Parashar <vi...@linux.vnet.ibm.com>
> ---
> arch/powerpc/include/asm/opal-api.h | 30 ++
> arch/powerpc/include/asm/opal.h | 3 +-
> arch/powerpc/platforms/powernv/Makefile | 1 +
> .../platforms/powernv/opal-poweroff-events.c | 358 +++++++++++++++++++++
> arch/powerpc/platforms/powernv/opal-wrappers.S | 1 +
> 5 files changed, 392 insertions(+), 1 deletion(-)
> create mode 100644 arch/powerpc/platforms/powernv/opal-poweroff-events.c
>
> diff --git a/arch/powerpc/include/asm/opal-api.h
> b/arch/powerpc/include/asm/opal-api.h
> index 0321a90..03b3cef 100644
> --- a/arch/powerpc/include/asm/opal-api.h
> +++ b/arch/powerpc/include/asm/opal-api.h
> @@ -730,6 +730,36 @@ struct opal_i2c_request {
> __be64 buffer_ra; /* Buffer real address */
> };
>
> +/*
> + * EPOW status sharing (OPAL and the host)
> + *
> + * The host will pass on OPAL, a buffer of length OPAL_EPOW_MAX_CLASSES
> + * to fetch system wide EPOW status. Each element in the returned buffer
> + * will contain bitwise EPOW status for each EPOW sub class.
> + */
> +
> +/* EPOW types */
> +enum OpalEpow {
> + OPAL_EPOW_POWER = 0, /* Power EPOW */
> + OPAL_EPOW_TEMP = 1, /* Temperature EPOW */
> + OPAL_EPOW_COOLING = 2, /* Cooling EPOW */
> + OPAL_MAX_EPOW_CLASSES = 3, /* Max EPOW categories */
> +};
> +
> +/* Power EPOW events */
> +enum OpalEpowPower {
> + OPAL_EPOW_POWER_UPS = 0x1, /* System on UPS power */
> + OPAL_EPOW_POWER_UPS_LOW = 0x2, /* System on UPS power with low battery*/
> +};
> +
> +/* Temperature EPOW events */
> +enum OpalEpowTemp {
> + OPAL_EPOW_TEMP_HIGH_AMB = 0x1, /* High ambient temperature */
> + OPAL_EPOW_TEMP_CRIT_AMB = 0x2, /* Critical ambient temperature */
> + OPAL_EPOW_TEMP_HIGH_INT = 0x4, /* High internal temperature */
> + OPAL_EPOW_TEMP_CRIT_INT = 0x8, /* Critical internal temperature */
> +};
> +
> #endif /* __ASSEMBLY__ */
>
> #endif /* __OPAL_API_H */
> diff --git a/arch/powerpc/include/asm/opal.h b/arch/powerpc/include/asm/opal.h
> index 042af1a..0777864 100644
> --- a/arch/powerpc/include/asm/opal.h
> +++ b/arch/powerpc/include/asm/opal.h
> @@ -141,7 +141,6 @@ int64_t opal_pci_fence_phb(uint64_t phb_id);
> int64_t opal_pci_reinit(uint64_t phb_id, uint64_t reinit_scope, uint64_t
> data);
> int64_t opal_pci_mask_pe_error(uint64_t phb_id, uint16_t pe_number, uint8_t
> error_type, uint8_t mask_action);
> int64_t opal_set_slot_led_status(uint64_t phb_id, uint64_t slot_id, uint8_t
> led_type, uint8_t led_action);
> -int64_t opal_get_epow_status(__be64 *status);
> int64_t opal_set_system_attention_led(uint8_t led_action);
> int64_t opal_pci_next_error(uint64_t phb_id, __be64 *first_frozen_pe,
> __be16 *pci_error_type, __be16 *severity);
> @@ -200,6 +199,8 @@ int64_t opal_flash_write(uint64_t id, uint64_t offset,
> uint64_t buf,
> uint64_t size, uint64_t token);
> int64_t opal_flash_erase(uint64_t id, uint64_t offset, uint64_t size,
> uint64_t token);
> +int32_t opal_get_epow_status(__be32 *status, __be32 *num_classes);
> +int32_t opal_get_dpo_status(__be32 *timeout);
>
> /* Internal functions */
> extern int early_init_dt_scan_opal(unsigned long node, const char *uname,
> diff --git a/arch/powerpc/platforms/powernv/Makefile
> b/arch/powerpc/platforms/powernv/Makefile
> index 33e44f3..b817bdb 100644
> --- a/arch/powerpc/platforms/powernv/Makefile
> +++ b/arch/powerpc/platforms/powernv/Makefile
> @@ -2,6 +2,7 @@ obj-y += setup.o opal-wrappers.o opal.o
> opal-async.o
> obj-y += opal-rtc.o opal-nvram.o opal-lpc.o opal-flash.o
> obj-y += rng.o opal-elog.o opal-dump.o opal-sysparam.o
> opal-sensor.o
> obj-y += opal-msglog.o opal-hmi.o opal-power.o
> +obj-y += opal-poweroff-events.o
>
> obj-$(CONFIG_SMP) += smp.o subcore.o subcore-asm.o
> obj-$(CONFIG_PCI) += pci.o pci-p5ioc2.o pci-ioda.o
> diff --git a/arch/powerpc/platforms/powernv/opal-poweroff-events.c
> b/arch/powerpc/platforms/powernv/opal-poweroff-events.c
> new file mode 100644
> index 0000000..9b169e2
> --- /dev/null
> +++ b/arch/powerpc/platforms/powernv/opal-poweroff-events.c
> @@ -0,0 +1,358 @@
Instead of creating a new file can you merge this functionality with existing
code in opal-power.c. It seems to be doing work similar to what you are doing
in this code.
> +/*
> + * PowerNV poweroff events support
> + *
> + * Copyright 2015 IBM Corp.
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License
> + * as published by the Free Software Foundation; either version
> + * 2 of the License, or (at your option) any later version.
> + */
> +
> +#define pr_fmt(fmt) "POWEROFF_EVENT: " fmt
> +
> +#include <linux/kernel.h>
> +#include <linux/spinlock.h>
> +#include <linux/timer.h>
> +#include <linux/reboot.h>
> +#include <asm/opal.h>
> +#include <asm/machdep.h>
> +
> +/* System EPOW status */
> +u32 epow_status[OPAL_MAX_EPOW_CLASSES];
> +int num_epow_classes;
> +
> +/* EPOW event timer and corresponding locks */
> +static struct timer_list epow_timer;
> +static DEFINE_SPINLOCK(epow_timer_spinlock);
> +
> +/* EPOW, DPO event status values */
> +#define DPO_DETECTED 1
> +#define EPOW_DETECTED 1
Instead of macros for this can you simply return a bool from corresponding
functions.
> +
> +/* EPOW events supported */
> +#define EPOW_POWER_UPS 0
> +#define EPOW_POWER_UPS_LOW 1
> +#define EPOW_TEMP_HIGH_AMB 2
> +#define EPOW_TEMP_CRIT_AMB 3
> +#define EPOW_TEMP_HIGH_INT 4
> +#define EPOW_TEMP_CRIT_INT 5
> +#define MAX_EPOW_EVENTS 6
> +
> +/* EPOW events description */
> +static const char * const epow_events_map[] = {
> + [EPOW_POWER_UPS] = "UPS",
> + [EPOW_POWER_UPS_LOW] = "UPS-low",
> + [EPOW_TEMP_HIGH_AMB] = "high-ambient-temp",
> + [EPOW_TEMP_CRIT_AMB] = "crit-ambient-temp",
> + [EPOW_TEMP_HIGH_INT] = "high-internal-temp",
> + [EPOW_TEMP_CRIT_INT] = "crit-internal-temp",
> +};
> +
> +/* EPOW events timeout values */
> +static int epow_timeout[MAX_EPOW_EVENTS];
> +
> +/*
> + * TODO: Export various event timeout values via device tree.
> + * Zero timeout value for any event suggests that it needs
> + * immediate shutdown.
> + */
> +#define TIMEOUT_EPOW_POWER_UPS 450
> +#define TIMEOUT_EPOW_TEMP_HIGH_AMB 450
> +
> +/*
> + * Get various EPOW event timeouts.
> + * TODO: For now hardcoding timeout values but they need to be
> + * obtained via firmware device-tree.
> + */
> +void get_epow_timeouts(void)
> +{
> + epow_timeout[EPOW_POWER_UPS] = TIMEOUT_EPOW_POWER_UPS;
> + epow_timeout[EPOW_TEMP_HIGH_AMB] = TIMEOUT_EPOW_TEMP_HIGH_AMB;
> +}
> +
Instead of having a separate function function for initialising the array you
can use array initialisers.
> +/* EPOW poweroff function. */
> +static void epow_poweroff(unsigned long event)
> +{
> + pr_info("Powering off system due to %s EPOW event\n",
> + epow_events_map[event]);
> + orderly_poweroff(true);
> +}
> +
> +/* Start EPOW poweroff timer */
> +static void start_epow_timer(unsigned long event, int32_t timeout)
> +{
> + unsigned long flags;
> +
> + spin_lock_irqsave(&epow_timer_spinlock, flags);
> + /* Check for already running epow poweroff timer */
> + if (timer_pending(&epow_timer)) {
> + /* Timer for same event */
> + if (epow_timer.data == event) {
> + spin_unlock_irqrestore(&epow_timer_spinlock, flags);
> + return;
> + }
> +
> + /* Timer with early poweroff timeout */
> + if (epow_timer.expires < (jiffies + timeout * HZ)) {
> + event = epow_timer.data;
> + spin_unlock_irqrestore(&epow_timer_spinlock, flags);
> + pr_info("Poweroff already scheduled for %s EPOW event "
> + "with earlier timeout.\n",
> + epow_events_map[event]);
> + return;
> + }
> + }
> +
> + /* Start a new timer/modify existing timer with new timeout value */
> + epow_timer.data = event;
> + mod_timer(&epow_timer, jiffies + timeout * HZ);
> + spin_unlock_irqrestore(&epow_timer_spinlock, flags);
> + pr_info("Scheduled system poweroff due to %s EPOW event "
> + "after %d seconds\n", epow_events_map[event], timeout);
> +}
> +
> +/* Stop poweroff timer */
> +static void stop_epow_timer(void)
> +{
> + int rc;
> + unsigned long flags;
> +
> + spin_lock_irqsave(&epow_timer_spinlock, flags);
> + rc = del_timer(&epow_timer);
> + spin_unlock_irqrestore(&epow_timer_spinlock, flags);
> +
> + if (rc)
> + pr_info("Poweroff timer deactivated\n");
> +}
> +
> +/* Get DPO status */
> +static int get_dpo_status(int32_t *dpo_timeout)
> +{
> + int rc;
> + __be32 opal_dpo_timeout;
> +
> + rc = opal_get_dpo_status(&opal_dpo_timeout);
> + if (rc == OPAL_WRONG_STATE) {
> + *dpo_timeout = 0;
> + return 0;
> + }
> +
> + *dpo_timeout = be32_to_cpu(opal_dpo_timeout);
> + return DPO_DETECTED;
> +}
> +
> +/* Process DPO event */
> +void process_dpo(void)
> +{
> + pr_info("Powering off system due to poweroff request.\n");
> + orderly_poweroff(true);
> +}
Function too small and can be combined with function opal_dpo_event.
> +
> +/* Get EPOW status */
> +static int get_epow_status(void)
> +{
> + int i;
> + bool epow_detected = false;
> +
> + __be32 opal_epow_status[OPAL_MAX_EPOW_CLASSES];
> + __be32 opal_epow_classes;
> +
> + opal_epow_classes = cpu_to_be32(OPAL_MAX_EPOW_CLASSES);
> + for (i = 0; i < OPAL_MAX_EPOW_CLASSES; i++)
> + opal_epow_status[i] = cpu_to_be32(0);
0 is represented same in be or le. So instead use static initialiser for the
array.
> +
> + /* Get EPOW events information from OPAL */
> + opal_get_epow_status(opal_epow_status, &opal_epow_classes);
Since you get the number of classes copied to the array back from opal, you may
not need to initialise it to zero in the loop above.
> +
> + /* Copy EPOW status */
> + memset(epow_status, 0, sizeof(epow_status[0] * OPAL_MAX_EPOW_CLASSES);
Simple memset(epow_status, 0, sizeof(epow_status)) should have the same effect
> + num_epos_classes = be32_to_cpu(opal_epow_classes);
> + for (i = 0; i < num_epow_classes; i++) {
> + epow_status[i] = be32_to_cpu(opal_epow_status[i]);
> + if (epow_status[i])
> + epow_detected = true;
> + }
> +
> + pr_info("EPOW classes supported OPAL = %d, Host = %d "
> + "EPOW Status = 0x%x, 0x%x, 0x%x\n",
> + num_epow_classes, OPAL_MAX_EPOW_CLASSES,
> + epow_status[0], epow_status[1], epow_status[2]);
> +
> + if (epow_detected)
> + return EPOW_DETECTED;
> +
> + return 0;
> +}
> +
> +/* Process EPOW information */
> +void process_epow(void)
> +{
> + int i, timeout = 0, event = -1;
> + bool epow_normal = false;
> +
> + /* Check for EPOW return to normal state */
> + for (i = 0; i < OPAL_MAX_EPOW_CLASSES; i++) {
> + if (epow_status[i])
> + break;
> + }
This is the same check you do in func get_epow_status. Instead refactor the
code to do this check only once.
> +
> + if (i == OPAL_MAX_EPOW_CLASSES)
> + epow_normal = true;
> +
> + /* Cancel any pending shutdown timer due to EPOW normal state.*/
> + if (epow_normal) {
> + stop_epow_timer();
> + return;
> + }
Can merge the two if conditions above to eliminate variable epow_normal.
> +
> + /* Determine EPOW events and poweroff timeouts */
> + if (epow_status[OPAL_EPOW_POWER] & OPAL_EPOW_POWER_UPS) {
> + pr_info("EPOW due to system running on UPS power\n");
> + event = EPOW_POWER_UPS;
> + timeout = epow_timeout[EPOW_POWER_UPS];
> + }
> +
> + if (epow_status[OPAL_EPOW_POWER] & OPAL_EPOW_POWER_UPS_LOW) {
> + pr_info("EPOW due to system running on UPS power "
> + "with low battery\n
> + event = EPOW_POWER_UPS_LOW;
> + timeout = epow_timeout[EPOW_POWER_UPS_LOW];
> + }
> +
> + if (epow_status[OPAL_EPOW_TEMP] & OPAL_EPOW_TEMP_HIGH_AMB) {
> + pr_info("EPOW due to high ambient temperature\n");
> + event = EPOW_TEMP_HIGH_AMB;
> + timeout = epow_timeout[EPOW_TEMP_HIGH_AMB];
> + }
> +
> + if (epow_status[OPAL_EPOW_TEMP] & OPAL_EPOW_TEMP_CRIT_AMB) {
> + pr_info("EPOW due to critical ambient temperature\n");
> + event = EPOW_TEMP_CRIT_AMB;
> + timeout = epow_timeout[EPOW_TEMP_CRIT_AMB];
> + }
> +
> + if (epow_status[OPAL_EPOW_TEMP] & OPAL_EPOW_TEMP_HIGH_INT) {
> + pr_info("EPOW due to high internal temperature\n");
> + event = EPOW_TEMP_HIGH_INT;
> + timeout = epow_timeout[EPOW_TEMP_HIGH_INT];
> + }
> +
> + if (epow_status[OPAL_EPOW_TEMP] & OPAL_EPOW_TEMP_CRIT_INT) {
> + pr_info("EPOW due to critical internal temperature\n");
> + event = EPOW_TEMP_CRIT_INT;
> + timeout = epow_timeout[EPOW_TEMP_CRIT_INT];
> + }
> +
> + if (event == -1) {
> + pr_err("Unknown EPOW event\n");
> + return;
> + }
> +
> + /* Start EPOW poweroff timer */
> + start_epow_timer(event, timeout);
> +}
> +
> +/* Check for any existing EPOW, DPO events and process them, if existing */
> +static void process_existing_poweroff_events(void)
> +{
> + int rc;
> + int32_t dpo_timeout;
> +
> + /* Check for any existing DPO event */
> + rc = get_dpo_status(&dpo_timeout);
> + if (rc == DPO_DETECTED) {
> + pr_info("Existing DPO event detected\n");
> + process_dpo();
> + return;
> + } else
> + pr_info("No existing DPO event detected\n");
> +
> + /* Check for any existing EPOW event */
> + rc = get_epow_status();
> + if (rc == EPOW_DETECTED) {
> + pr_info("Existing EPOW event detected.\n");
> + process_epow();
> + } else
> + pr_info("No existing EPOW event detected\n");
> +
> +}
> +
> +/* Platform EPOW message received */
> +static int opal_epow_event(struct notifier_block *nb,
> + unsigned long msg_type, void *msg)
> +{
> + pr_info("EPOW event received\n");
> +
> + /* Get EPOW event details */
> + get_epow_status();
> +
> + /* Process EPOW event information */
> + process_epow();
> +
> + return 0;
> +}
> +
> +
> +/* Platform DPO message received */
> +static int opal_dpo_event(struct notifier_block *nb,
> + unsigned long msg_type, void *msg)
> +{
> + pr_info("DPO event received.\n");
> + process_dpo();
> +
> + return 0;
> +}
> +
> +
> +/* OPAL EPOW event notifier block */
> +static struct notifier_block opal_epow_nb = {
> + .notifier_call = opal_epow_event,
> + .next = NULL,
> + .priority = 0,
> +};
> +
> +/* OPAL DPO event notifier block */
> +static struct notifier_block opal_dpo_nb = {
> + .notifier_call = opal_dpo_event,
> + .next = NULL,
> + .priority = 0,
> +};
> +
> +/* Poweroff events init */
> +static int opal_poweroff_events_init(void)
> +{
> + int ret;
> +
> + /* Initialize poweroff timer */
> + init_timer(&epow_timer);
> + epow_timer.function = epow_poweroff;
> +
> + /* Get EPOW event timeout values */
> + get_epow_timeouts();
> +
> + /* Check for any existing EPOW or DPO events. */
> + process_existing_poweroff_events();
> +
> + /* Register EPOW event notifier */
> + ret = opal_message_notifier_register(OPAL_MSG_EPOW, &opal_epow_nb);
> + if (ret) {
> + pr_err("EPOW event notifier registration failed\n");
> + return ret;
> + }
Timing issue. You may miss some events that occur between process existing
power off event and registering message notifies. Instead first register your
notifier and then process any existing events.
> +
> + /* Register DPO event notifier */
> + ret = opal_message_notifier_register(OPAL_MSG_DPO, &opal_dpo_nb);
> + if (ret) {
> + pr_err("DPO event notifier registration failed\n");
> + opal_notifier_unregister(&opal_epow_nb);
> + return ret;
> + }
> +
> +
> + pr_info("OPAL poweroff events support initialized\n");
> +
> + return 0;
> +}
> +
> +machine_subsys_initcall(powernv, opal_poweroff_events_init);
> diff --git a/arch/powerpc/platforms/powernv/opal-wrappers.S
> b/arch/powerpc/platforms/powernv/opal-wrappers.S
> index a7ade94..5d3c8e3 100644
> --- a/arch/powerpc/platforms/powernv/opal-wrappers.S
> +++ b/arch/powerpc/platforms/powernv/opal-wrappers.S
> @@ -249,6 +249,7 @@ OPAL_CALL(opal_pci_reinit, OPAL_PCI_REINIT);
> OPAL_CALL(opal_pci_mask_pe_error, OPAL_PCI_MASK_PE_ERROR);
> OPAL_CALL(opal_set_slot_led_status, OPAL_SET_SLOT_LED_STATUS);
> OPAL_CALL(opal_get_epow_status, OPAL_GET_EPOW_STATUS);
> +OPAL_CALL(opal_get_dpo_status, OPAL_GET_DPO_STATUS);
> OPAL_CALL(opal_set_system_attention_led, OPAL_SET_SYSTEM_ATTENTION_LED);
> OPAL_CALL(opal_pci_next_error, OPAL_PCI_NEXT_ERROR);
> OPAL_CALL(opal_pci_poll, OPAL_PCI_POLL);
> --
> 1.9.3
>
> _______________________________________________
> Linuxppc-dev mailing list
> Linuxppc-dev@lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/linuxppc-dev
_______________________________________________
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev
