diff options
author | Blagovest Kolenichev <bkolenichev@codeaurora.org> | 2017-09-21 13:19:38 -0700 |
---|---|---|
committer | Blagovest Kolenichev <bkolenichev@codeaurora.org> | 2017-09-21 13:19:38 -0700 |
commit | c988eaaeaf5f1194a7366ecfab9209d0fda13b0e (patch) | |
tree | 81b5cf30dc473c449718e30ba5f526f2c409213e /kernel/sched | |
parent | 6f777b2385c98a17d69bbeead6edbc7ad7470f72 (diff) | |
parent | a8935c98cf97201e9ee50f22a27264a3220c82e8 (diff) |
Merge android-4.4@a8935c9 (v4.4.87) into msm-4.4
* refs/heads/tmp-a8935c9:
Linux 4.4.87
crypto: algif_skcipher - only call put_page on referenced and used pages
epoll: fix race between ep_poll_callback(POLLFREE) and ep_free()/ep_remove()
kvm: arm/arm64: Force reading uncached stage2 PGD
kvm: arm/arm64: Fix race in resetting stage2 PGD
drm/ttm: Fix accounting error when fail to get pages for pool
xfrm: policy: check policy direction value
wl1251: add a missing spin_lock_init()
CIFS: remove endian related sparse warning
CIFS: Fix maximum SMB2 header size
alpha: uapi: Add support for __SANE_USERSPACE_TYPES__
cpuset: Fix incorrect memory_pressure control file mapping
cpumask: fix spurious cpumask_of_node() on non-NUMA multi-node configs
ceph: fix readpage from fscache
i2c: ismt: Return EMSGSIZE for block reads with bogus length
i2c: ismt: Don't duplicate the receive length for block reads
irqchip: mips-gic: SYNC after enabling GIC region
ANDROID: cpufreq-dt: Set sane defaults for schedutil rate limits
BACKPORT: cpufreq: schedutil: Use policy-dependent transition delays
FROMLIST: binder: fix an ret value override
FROMLIST: binder: fix memory corruption in binder_transaction binder
Linux 4.4.86
drm/i915: fix compiler warning in drivers/gpu/drm/i915/intel_uncore.c
scsi: sg: reset 'res_in_use' after unlinking reserved array
scsi: sg: protect accesses to 'reserved' page array
arm64: fpsimd: Prevent registers leaking across exec
x86/io: Add "memory" clobber to insb/insw/insl/outsb/outsw/outsl
arm64: mm: abort uaccess retries upon fatal signal
lpfc: Fix Device discovery failures during switch reboot test.
p54: memset(0) whole array
lightnvm: initialize ppa_addr in dev_to_generic_addr()
gcov: support GCC 7.1
gcov: add support for gcc version >= 6
i2c: jz4780: drop superfluous init
btrfs: remove duplicate const specifier
ALSA: au88x0: Fix zero clear of stream->resources
scsi: isci: avoid array subscript warning
sched: WALT: fix window mis-alignment
sched: EAS: kill incorrect nohz idle cpu kick
sched: EAS: fix incorrect energy delta calculation due to rounding error
sched: EAS/WALT: take into account of waking task's load
cpufreq: sched: WALT: don't apply capacity margin twice
sched: WALT: fix potential overflow
sched: EAS: schedfreq: fix CPU util over estimation
sched: EAS/WALT: use cr_avg instead of prev_runnable_sum
sched: WALT: fix broken cumulative runnable average accounting
sched: deadline: WALT: account cumulative runnable avg
FROMLIST: android: binder: Add page usage in binder stats
FROMLIST: android: binder: Add shrinker tracepoints
FROMLIST: android: binder: Add global lru shrinker to binder
FROMLIST: android: binder: Move buffer out of area shared with user space
FROMLIST: android: binder: Add allocator selftest
FROMLIST: android: binder: Refactor prev and next buffer into a helper function
android: android-base.config: enable IP6_NF_MATCH_RPFILTER
UPSTREAM: cpufreq: schedutil: Use unsigned int for iowait boost
UPSTREAM: cpufreq: schedutil: Make iowait boost more energy efficient
Conflicts:
drivers/cpufreq/cpufreq-dt.c
kernel/sched/deadline.c
kernel/sched/fair.c
kernel/sched/sched.h
Change-Id: Iee31db3fd1a0d1650ebf3d6de307a4e4637120b4
Signed-off-by: Blagovest Kolenichev <bkolenichev@codeaurora.org>
Diffstat (limited to 'kernel/sched')
-rw-r--r-- | kernel/sched/core.c | 8 | ||||
-rw-r--r-- | kernel/sched/cpufreq_sched.c | 14 | ||||
-rw-r--r-- | kernel/sched/cpufreq_schedutil.c | 62 | ||||
-rw-r--r-- | kernel/sched/deadline.c | 2 | ||||
-rw-r--r-- | kernel/sched/fair.c | 44 | ||||
-rw-r--r-- | kernel/sched/sched.h | 19 | ||||
-rw-r--r-- | kernel/sched/walt.c | 8 |
7 files changed, 116 insertions, 41 deletions
diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 4ecca604e64b..75e15e154cfc 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -3177,20 +3177,20 @@ static void sched_freq_tick_pelt(int cpu) #ifdef CONFIG_SCHED_WALT static void sched_freq_tick_walt(int cpu) { - unsigned long cpu_utilization = cpu_util(cpu); + unsigned long cpu_utilization = cpu_util_freq(cpu); unsigned long capacity_curr = capacity_curr_of(cpu); if (walt_disabled || !sysctl_sched_use_walt_cpu_util) return sched_freq_tick_pelt(cpu); /* - * Add a margin to the WALT utilization. + * Add a margin to the WALT utilization to check if we will need to + * increase frequency. * NOTE: WALT tracks a single CPU signal for all the scheduling * classes, thus this margin is going to be added to the DL class as * well, which is something we do not do in sched_freq_tick_pelt case. */ - cpu_utilization = add_capacity_margin(cpu_utilization); - if (cpu_utilization <= capacity_curr) + if (add_capacity_margin(cpu_utilization) <= capacity_curr) return; /* diff --git a/kernel/sched/cpufreq_sched.c b/kernel/sched/cpufreq_sched.c index f10d9f7d6d07..6ffb23adbcef 100644 --- a/kernel/sched/cpufreq_sched.c +++ b/kernel/sched/cpufreq_sched.c @@ -235,6 +235,18 @@ out: cpufreq_cpu_put(policy); } +#ifdef CONFIG_SCHED_WALT +static inline unsigned long +requested_capacity(struct sched_capacity_reqs *scr) +{ + if (!walt_disabled && sysctl_sched_use_walt_cpu_util) + return scr->cfs; + return scr->cfs + scr->rt; +} +#else +#define requested_capacity(scr) (scr->cfs + scr->rt) +#endif + void update_cpu_capacity_request(int cpu, bool request) { unsigned long new_capacity; @@ -245,7 +257,7 @@ void update_cpu_capacity_request(int cpu, bool request) scr = &per_cpu(cpu_sched_capacity_reqs, cpu); - new_capacity = scr->cfs + scr->rt; + new_capacity = requested_capacity(scr); new_capacity = new_capacity * capacity_margin / SCHED_CAPACITY_SCALE; new_capacity += scr->dl; diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c index e12309c1b07b..b90f7434e13b 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -64,8 +64,9 @@ struct sugov_cpu { struct update_util_data update_util; struct sugov_policy *sg_policy; - unsigned long iowait_boost; - unsigned long iowait_boost_max; + bool iowait_boost_pending; + unsigned int iowait_boost; + unsigned int iowait_boost_max; u64 last_update; /* The fields below are only needed when sharing a policy. */ @@ -224,30 +225,54 @@ static void sugov_set_iowait_boost(struct sugov_cpu *sg_cpu, u64 time, unsigned int flags) { if (flags & SCHED_CPUFREQ_IOWAIT) { - sg_cpu->iowait_boost = sg_cpu->iowait_boost_max; + if (sg_cpu->iowait_boost_pending) + return; + + sg_cpu->iowait_boost_pending = true; + + if (sg_cpu->iowait_boost) { + sg_cpu->iowait_boost <<= 1; + if (sg_cpu->iowait_boost > sg_cpu->iowait_boost_max) + sg_cpu->iowait_boost = sg_cpu->iowait_boost_max; + } else { + sg_cpu->iowait_boost = sg_cpu->sg_policy->policy->min; + } } else if (sg_cpu->iowait_boost) { s64 delta_ns = time - sg_cpu->last_update; /* Clear iowait_boost if the CPU apprears to have been idle. */ - if (delta_ns > TICK_NSEC) + if (delta_ns > TICK_NSEC) { sg_cpu->iowait_boost = 0; + sg_cpu->iowait_boost_pending = false; + } } } static void sugov_iowait_boost(struct sugov_cpu *sg_cpu, unsigned long *util, unsigned long *max) { - unsigned long boost_util = sg_cpu->iowait_boost; - unsigned long boost_max = sg_cpu->iowait_boost_max; + unsigned int boost_util, boost_max; - if (!boost_util) + if (!sg_cpu->iowait_boost) return; + if (sg_cpu->iowait_boost_pending) { + sg_cpu->iowait_boost_pending = false; + } else { + sg_cpu->iowait_boost >>= 1; + if (sg_cpu->iowait_boost < sg_cpu->sg_policy->policy->min) { + sg_cpu->iowait_boost = 0; + return; + } + } + + boost_util = sg_cpu->iowait_boost; + boost_max = sg_cpu->iowait_boost_max; + if (*util * boost_max < *max * boost_util) { *util = boost_util; *max = boost_max; } - sg_cpu->iowait_boost >>= 1; } #ifdef CONFIG_NO_HZ_COMMON @@ -320,6 +345,7 @@ static unsigned int sugov_next_freq_shared(struct sugov_cpu *sg_cpu) delta_ns = last_freq_update_time - j_sg_cpu->last_update; if (delta_ns > TICK_NSEC) { j_sg_cpu->iowait_boost = 0; + j_sg_cpu->iowait_boost_pending = false; continue; } if (j_sg_cpu->flags & SCHED_CPUFREQ_DL) @@ -589,7 +615,6 @@ static int sugov_init(struct cpufreq_policy *policy) { struct sugov_policy *sg_policy; struct sugov_tunables *tunables; - unsigned int lat; int ret = 0; /* State should be equivalent to EXIT */ @@ -628,12 +653,19 @@ static int sugov_init(struct cpufreq_policy *policy) goto stop_kthread; } - tunables->up_rate_limit_us = LATENCY_MULTIPLIER; - tunables->down_rate_limit_us = LATENCY_MULTIPLIER; - lat = policy->cpuinfo.transition_latency / NSEC_PER_USEC; - if (lat) { - tunables->up_rate_limit_us *= lat; - tunables->down_rate_limit_us *= lat; + if (policy->up_transition_delay_us && policy->down_transition_delay_us) { + tunables->up_rate_limit_us = policy->up_transition_delay_us; + tunables->down_rate_limit_us = policy->down_transition_delay_us; + } else { + unsigned int lat; + + tunables->up_rate_limit_us = LATENCY_MULTIPLIER; + tunables->down_rate_limit_us = LATENCY_MULTIPLIER; + lat = policy->cpuinfo.transition_latency / NSEC_PER_USEC; + if (lat) { + tunables->up_rate_limit_us *= lat; + tunables->down_rate_limit_us *= lat; + } } policy->governor_data = sg_policy; diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c index a105e97ab6bf..905d8cb5a6eb 100644 --- a/kernel/sched/deadline.c +++ b/kernel/sched/deadline.c @@ -18,6 +18,8 @@ #include <linux/slab.h> +#include "walt.h" + struct dl_bandwidth def_dl_bandwidth; static inline struct task_struct *dl_task_of(struct sched_dl_entity *dl_se) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 853064319b0d..aa016919eab8 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -5809,10 +5809,11 @@ static inline void hrtick_update(struct rq *rq) #endif #ifdef CONFIG_SMP +static bool __cpu_overutilized(int cpu, int delta); static bool cpu_overutilized(int cpu); unsigned long boosted_cpu_util(int cpu); #else -#define boosted_cpu_util(cpu) cpu_util(cpu) +#define boosted_cpu_util(cpu) cpu_util_freq(cpu) #endif #if defined(CONFIG_SMP) && defined(CONFIG_CPU_FREQ_GOV_SCHED) @@ -6626,10 +6627,8 @@ end: */ static int sched_group_energy(struct energy_env *eenv) { - struct sched_domain *sd; - int cpu, total_energy = 0; struct cpumask visit_cpus; - struct sched_group *sg; + u64 total_energy = 0; WARN_ON(!eenv->sg_top->sge); @@ -6637,8 +6636,8 @@ static int sched_group_energy(struct energy_env *eenv) while (!cpumask_empty(&visit_cpus)) { struct sched_group *sg_shared_cap = NULL; - - cpu = cpumask_first(&visit_cpus); + int cpu = cpumask_first(&visit_cpus); + struct sched_domain *sd; /* * Is the group utilization affected by cpus outside this @@ -6650,7 +6649,7 @@ static int sched_group_energy(struct energy_env *eenv) sg_shared_cap = sd->parent->groups; for_each_domain(cpu, sd) { - sg = sd->groups; + struct sched_group *sg = sd->groups; /* Has this sched_domain already been visited? */ if (sd->child && group_first_cpu(sg) != cpu) @@ -6686,11 +6685,9 @@ static int sched_group_energy(struct energy_env *eenv) idle_idx = group_idle_state(eenv, sg); group_util = group_norm_util(eenv, sg); - sg_busy_energy = (group_util * sg->sge->cap_states[cap_idx].power) - >> SCHED_CAPACITY_SHIFT; + sg_busy_energy = (group_util * sg->sge->cap_states[cap_idx].power); sg_idle_energy = ((SCHED_LOAD_SCALE-group_util) - * sg->sge->idle_states[idle_idx].power) - >> SCHED_CAPACITY_SHIFT; + * sg->sge->idle_states[idle_idx].power); total_energy += sg_busy_energy + sg_idle_energy; @@ -6715,7 +6712,7 @@ next_cpu: continue; } - eenv->energy = total_energy; + eenv->energy = total_energy >> SCHED_CAPACITY_SHIFT; return 0; } @@ -7004,9 +7001,14 @@ static inline bool task_fits_max(struct task_struct *p, int cpu) return __task_fits(p, cpu, 0); } +static bool __cpu_overutilized(int cpu, int delta) +{ + return (capacity_of(cpu) * 1024) < ((cpu_util(cpu) + delta) * capacity_margin); +} + static bool cpu_overutilized(int cpu) { - return (capacity_of(cpu) * 1024) < (cpu_util(cpu) * capacity_margin); + return __cpu_overutilized(cpu, 0); } #ifdef CONFIG_SCHED_TUNE @@ -7085,7 +7087,7 @@ schedtune_task_margin(struct task_struct *task) unsigned long boosted_cpu_util(int cpu) { - unsigned long util = cpu_util(cpu); + unsigned long util = cpu_util_freq(cpu); long margin = schedtune_cpu_margin(util, cpu); trace_sched_boost_cpu(cpu, util, margin); @@ -7729,6 +7731,7 @@ static int select_energy_cpu_brute(struct task_struct *p, int prev_cpu, int sync } if (target_cpu != prev_cpu) { + int delta = 0; struct energy_env eenv = { .util_delta = task_util(p), .src_cpu = prev_cpu, @@ -7736,8 +7739,13 @@ static int select_energy_cpu_brute(struct task_struct *p, int prev_cpu, int sync .task = p, }; + +#ifdef CONFIG_SCHED_WALT + if (!walt_disabled && sysctl_sched_use_walt_cpu_util) + delta = task_util(p); +#endif /* Not enough spare capacity on previous cpu */ - if (cpu_overutilized(prev_cpu)) { + if (__cpu_overutilized(prev_cpu, delta)) { schedstat_inc(p, se.statistics.nr_wakeups_secb_insuff_cap); schedstat_inc(this_rq(), eas_stats.secb_insuff_cap); goto unlock; @@ -11348,8 +11356,8 @@ static inline int _nohz_kick_needed(struct rq *rq, int cpu, int *type) return true; /* Do idle load balance if there have misfit task */ - if (energy_aware() && rq->misfit_task) - return 1; + if (energy_aware()) + return rq->misfit_task; return (rq->nr_running >= 2); } @@ -11391,7 +11399,7 @@ static inline bool nohz_kick_needed(struct rq *rq, int *type) #ifndef CONFIG_SCHED_HMP rcu_read_lock(); sd = rcu_dereference(per_cpu(sd_busy, cpu)); - if (sd && !energy_aware()) { + if (sd) { sgc = sd->groups->sgc; nr_busy = atomic_read(&sgc->nr_busy_cpus); diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 33bf0c07e757..c53970b5a8f0 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -2363,6 +2363,12 @@ static inline unsigned long __cpu_util(int cpu, int delta) unsigned long util = cpu_rq(cpu)->cfs.avg.util_avg; unsigned long capacity = capacity_orig_of(cpu); +#ifdef CONFIG_SCHED_WALT + if (!walt_disabled && sysctl_sched_use_walt_cpu_util) + util = div64_u64(cpu_rq(cpu)->cumulative_runnable_avg, + walt_ravg_window >> SCHED_LOAD_SHIFT); +#endif + delta += util; if (delta < 0) return 0; @@ -2375,6 +2381,19 @@ static inline unsigned long cpu_util(int cpu) return __cpu_util(cpu, 0); } +static inline unsigned long cpu_util_freq(int cpu) +{ + unsigned long util = cpu_rq(cpu)->cfs.avg.util_avg; + unsigned long capacity = capacity_orig_of(cpu); + +#ifdef CONFIG_SCHED_WALT + if (!walt_disabled && sysctl_sched_use_walt_cpu_util) + util = div64_u64(cpu_rq(cpu)->prev_runnable_sum, + walt_ravg_window >> SCHED_LOAD_SHIFT); +#endif + return (util >= capacity) ? capacity : util; +} + #endif #ifdef CONFIG_CPU_FREQ_GOV_SCHED diff --git a/kernel/sched/walt.c b/kernel/sched/walt.c index 92c3aae8e056..28e999554463 100644 --- a/kernel/sched/walt.c +++ b/kernel/sched/walt.c @@ -111,8 +111,10 @@ walt_dec_cumulative_runnable_avg(struct rq *rq, static void fixup_cumulative_runnable_avg(struct rq *rq, - struct task_struct *p, s64 task_load_delta) + struct task_struct *p, u64 new_task_load) { + s64 task_load_delta = (s64)new_task_load - task_load(p); + rq->cumulative_runnable_avg += task_load_delta; if ((s64)rq->cumulative_runnable_avg < 0) panic("cra less than zero: tld: %lld, task_load(p) = %u\n", @@ -802,11 +804,11 @@ void walt_set_window_start(struct rq *rq) int cpu = cpu_of(rq); struct rq *sync_rq = cpu_rq(sync_cpu); - if (rq->window_start) + if (likely(rq->window_start)) return; if (cpu == sync_cpu) { - rq->window_start = walt_ktime_clock(); + rq->window_start = 1; } else { raw_spin_unlock(&rq->lock); double_rq_lock(rq, sync_rq); |