Handle thermal events to mask CPUs
The Hardware Feedback Interface event is delivered as the Netlink's THERMAL_GENL_ATTR_CPU_CAPABILITY attribute. Add code to receive and parse these attributes: logical CPU number, performance, and efficiency enumeration values (0-1023). When an event indicates CPU performance and efficiency are zeros, then the CPU needs to be banned from IRQs. Rebuild object tree to make this banned list effective. Creating the banned CPU list here is a bit tricky here because the amount of data that each Netlink notification can carry out is not enough in some systems with many cores. This means the handler has to wait for multiple notifications to determine banned CPUs per thermal event. Establish a logic to maintain the list based on the kernel behaviors. Also, always push the current list as of the banned CPUs, because there is no clear line whether each notification is at the end of a thermal event or not. Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
This commit is contained in:
parent
c65cda1836
commit
560291389e
1 changed files with 165 additions and 7 deletions
172
thermal.c
172
thermal.c
|
@ -6,6 +6,7 @@
|
|||
|
||||
#include <stdio.h>
|
||||
|
||||
#include <sys/sysinfo.h>
|
||||
#include <netlink/genl/genl.h>
|
||||
#include <netlink/genl/family.h>
|
||||
#include <netlink/genl/ctrl.h>
|
||||
|
@ -14,8 +15,62 @@
|
|||
|
||||
cpumask_t thermal_banned_cpus;
|
||||
|
||||
#define INVALID_NL_FD -1
|
||||
#define MAX_RECV_ERRS 2
|
||||
/* Events of thermal_genl_family */
|
||||
enum thermal_genl_event {
|
||||
THERMAL_GENL_EVENT_UNSPEC,
|
||||
THERMAL_GENL_EVENT_TZ_CREATE, /* Thermal zone creation */
|
||||
THERMAL_GENL_EVENT_TZ_DELETE, /* Thermal zone deletion */
|
||||
THERMAL_GENL_EVENT_TZ_DISABLE, /* Thermal zone disabled */
|
||||
THERMAL_GENL_EVENT_TZ_ENABLE, /* Thermal zone enabled */
|
||||
THERMAL_GENL_EVENT_TZ_TRIP_UP, /* Trip point crossed the way up */
|
||||
THERMAL_GENL_EVENT_TZ_TRIP_DOWN, /* Trip point crossed the way down */
|
||||
THERMAL_GENL_EVENT_TZ_TRIP_CHANGE, /* Trip point changed */
|
||||
THERMAL_GENL_EVENT_TZ_TRIP_ADD, /* Trip point added */
|
||||
THERMAL_GENL_EVENT_TZ_TRIP_DELETE, /* Trip point deleted */
|
||||
THERMAL_GENL_EVENT_CDEV_ADD, /* Cdev bound to the thermal zone */
|
||||
THERMAL_GENL_EVENT_CDEV_DELETE, /* Cdev unbound */
|
||||
THERMAL_GENL_EVENT_CDEV_STATE_UPDATE, /* Cdev state updated */
|
||||
THERMAL_GENL_EVENT_TZ_GOV_CHANGE, /* Governor policy changed */
|
||||
THERMAL_GENL_EVENT_CAPACITY_CHANGE, /* CPU capacity changed */
|
||||
__THERMAL_GENL_EVENT_MAX,
|
||||
};
|
||||
#define THERMAL_GENL_EVENT_MAX (__THERMAL_GENL_EVENT_MAX - 1)
|
||||
|
||||
/* Attributes of thermal_genl_family */
|
||||
enum thermal_genl_attr {
|
||||
THERMAL_GENL_ATTR_UNSPEC,
|
||||
THERMAL_GENL_ATTR_TZ,
|
||||
THERMAL_GENL_ATTR_TZ_ID,
|
||||
THERMAL_GENL_ATTR_TZ_TEMP,
|
||||
THERMAL_GENL_ATTR_TZ_TRIP,
|
||||
THERMAL_GENL_ATTR_TZ_TRIP_ID,
|
||||
THERMAL_GENL_ATTR_TZ_TRIP_TYPE,
|
||||
THERMAL_GENL_ATTR_TZ_TRIP_TEMP,
|
||||
THERMAL_GENL_ATTR_TZ_TRIP_HYST,
|
||||
THERMAL_GENL_ATTR_TZ_MODE,
|
||||
THERMAL_GENL_ATTR_TZ_NAME,
|
||||
THERMAL_GENL_ATTR_TZ_CDEV_WEIGHT,
|
||||
THERMAL_GENL_ATTR_TZ_GOV,
|
||||
THERMAL_GENL_ATTR_TZ_GOV_NAME,
|
||||
THERMAL_GENL_ATTR_CDEV,
|
||||
THERMAL_GENL_ATTR_CDEV_ID,
|
||||
THERMAL_GENL_ATTR_CDEV_CUR_STATE,
|
||||
THERMAL_GENL_ATTR_CDEV_MAX_STATE,
|
||||
THERMAL_GENL_ATTR_CDEV_NAME,
|
||||
THERMAL_GENL_ATTR_GOV_NAME,
|
||||
THERMAL_GENL_ATTR_CAPACITY,
|
||||
THERMAL_GENL_ATTR_CAPACITY_CPU_COUNT,
|
||||
THERMAL_GENL_ATTR_CAPACITY_CPU_ID,
|
||||
THERMAL_GENL_ATTR_CAPACITY_CPU_PERF,
|
||||
THERMAL_GENL_ATTR_CAPACITY_CPU_EFF,
|
||||
__THERMAL_GENL_ATTR_MAX,
|
||||
};
|
||||
#define THERMAL_GENL_ATTR_MAX (__THERMAL_GENL_ATTR_MAX - 1)
|
||||
|
||||
#define INVALID_NL_FD -1
|
||||
#define MAX_RECV_ERRS 2
|
||||
#define SYSCONF_ERR -1
|
||||
#define INVALID_EVENT_VALUE -1
|
||||
|
||||
#define THERMAL_GENL_FAMILY_NAME "thermal"
|
||||
#define THERMAL_GENL_EVENT_GROUP_NAME "event"
|
||||
|
@ -254,17 +309,115 @@ err_out:
|
|||
return error;
|
||||
}
|
||||
|
||||
static int handle_thermal_event(struct nl_msg *msg __attribute__((unused)),
|
||||
void *arg __attribute__((unused)))
|
||||
enum {
|
||||
INDEX_CPUNUM,
|
||||
INDEX_PERF,
|
||||
INDEX_EFFI,
|
||||
INDEX_MAX
|
||||
};
|
||||
|
||||
/*
|
||||
* Single netlink notification is not guaranteed to fully deliver all of
|
||||
* the CPU updates per thermal event, due to the implementation choice in
|
||||
* the kernel code. So, this function is intended to manage a CPU list in a
|
||||
* stream of relevant notifications.
|
||||
*/
|
||||
static void update_banned_cpus(int cur_cpuidx, gboolean need_to_ban)
|
||||
{
|
||||
log(TO_ALL, LOG_ERR, "thermal: not yet implemented to process thermal event.\n");
|
||||
return NL_SKIP;
|
||||
static cpumask_t banmask = { 0 }, itrmask = { 0 };
|
||||
long max_cpunum = sysconf(_SC_NPROCESSORS_ONLN);
|
||||
|
||||
if (need_to_ban)
|
||||
cpu_set(cur_cpuidx, banmask);
|
||||
|
||||
cpu_set(cur_cpuidx, itrmask);
|
||||
if (cpus_weight(itrmask) < max_cpunum)
|
||||
return;
|
||||
|
||||
if (cpus_equal(thermal_banned_cpus, banmask))
|
||||
goto out;
|
||||
|
||||
cpus_copy(thermal_banned_cpus, banmask);
|
||||
need_rescan = 1;
|
||||
out:
|
||||
cpus_clear(banmask);
|
||||
cpus_clear(itrmask);
|
||||
}
|
||||
|
||||
static int handle_thermal_event(struct nl_msg *msg, void *arg __attribute__((unused)))
|
||||
{
|
||||
int event_data[INDEX_MAX] = { INVALID_EVENT_VALUE };
|
||||
struct nlattr *attrs[THERMAL_GENL_ATTR_MAX + 1];
|
||||
struct genlmsghdr *genlhdr;
|
||||
struct nlmsghdr *msnlh;
|
||||
struct nlattr *cap;
|
||||
int i, remain, rc;
|
||||
void *pos;
|
||||
|
||||
/* get actual netlink message header */
|
||||
msnlh = nlmsg_hdr(msg);
|
||||
|
||||
/* get a pointer to generic netlink header */
|
||||
genlhdr = genlmsg_hdr(msnlh);
|
||||
if (genlhdr->cmd != THERMAL_GENL_EVENT_CAPACITY_CHANGE) {
|
||||
log(TO_ALL, LOG_DEBUG, "thermal: no CPU capacity change.\n");
|
||||
return NL_SKIP;
|
||||
}
|
||||
|
||||
/* parse generic netlink message including attributes */
|
||||
rc = genlmsg_parse(
|
||||
msnlh, /* a pointer to netlink message header */
|
||||
0, /* length of user header */
|
||||
attrs, /* array to store parsed attributes */
|
||||
THERMAL_GENL_ATTR_MAX, /* maximum attribute id as expected */
|
||||
NULL); /* validation policy */
|
||||
if (rc) {
|
||||
log(TO_ALL, LOG_ERR, "thermal: failed to parse message for thermal event.\n");
|
||||
return NL_SKIP;
|
||||
}
|
||||
|
||||
/* get start and length of payload section */
|
||||
cap = attrs[THERMAL_GENL_ATTR_CAPACITY];
|
||||
pos = nla_data(cap);
|
||||
remain = nla_len(cap);
|
||||
|
||||
for (i = 0; nla_ok(pos, remain); pos = nla_next(pos, &remain), i++) {
|
||||
gboolean valid_event = TRUE, need_to_ban;
|
||||
unsigned int value = nla_get_u32(pos);
|
||||
int idx = i % INDEX_MAX;
|
||||
int cur_cpuidx;
|
||||
|
||||
event_data[idx] = value;
|
||||
|
||||
if (idx != INDEX_EFFI)
|
||||
continue;
|
||||
|
||||
cur_cpuidx = event_data[INDEX_CPUNUM];
|
||||
valid_event = !!(cur_cpuidx <= NR_CPUS);
|
||||
if (!valid_event) {
|
||||
log(TO_ALL, LOG_WARNING, "thermal: invalid event - CPU %d\n",
|
||||
cur_cpuidx);
|
||||
continue;
|
||||
}
|
||||
|
||||
/*
|
||||
* A CPU with no performance and no efficiency cannot
|
||||
* handle IRQs:
|
||||
*/
|
||||
need_to_ban = !!(!event_data[INDEX_PERF] && !event_data[INDEX_EFFI]);
|
||||
update_banned_cpus(cur_cpuidx, need_to_ban);
|
||||
|
||||
log(TO_ALL, LOG_DEBUG, "thermal: event - CPU %d, efficiency %d, perf %d.\n",
|
||||
cur_cpuidx, event_data[INDEX_PERF], event_data[INDEX_EFFI]);
|
||||
}
|
||||
|
||||
return NL_OK;
|
||||
}
|
||||
|
||||
static int handler_for_debug(struct nl_msg *msg __attribute__((unused)),
|
||||
void *arg __attribute__((unused)))
|
||||
{
|
||||
return NL_SKIP;
|
||||
return NL_OK;
|
||||
}
|
||||
|
||||
/*
|
||||
|
@ -274,6 +427,11 @@ static gboolean register_netlink_handler(void)
|
|||
{
|
||||
int rc;
|
||||
|
||||
if (sysconf(_SC_NPROCESSORS_ONLN) == SYSCONF_ERR) {
|
||||
log(TO_ALL, LOG_ERR, "thermal: _SC_NPROCESSORS_ONLN not available.\n");
|
||||
return TRUE;
|
||||
}
|
||||
|
||||
rc = nl_cb_set(callback, NL_CB_SEQ_CHECK, NL_CB_CUSTOM, handler_for_debug, NULL);
|
||||
if (rc) {
|
||||
log(TO_ALL, LOG_ERR, "thermal: debug handler registration failed.\n");
|
||||
|
|
Loading…
Reference in a new issue