sascha/esphome
1
0
Fork 0
mirror of https://github.com/esphome/esphome.git synced 2025-09-26 15:12:21 +01:00
Code Issues Releases Wiki Activity

[scheduler] Reduce SchedulerItem memory usage by 7.4% on 32-bit platforms

Browse Source
This commit is contained in:
J. Nick Koston
2025-09-03 16:13:07 -05:00
parent 23c6650902
commit d505f5ecaa
2 changed files with 37 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

27
esphome/core/scheduler.cpp
View File

@@ -14,7 +14,7 @@ namespace esphome {
static const char *const TAG = "scheduler";
static const uint32_t MAX_LOGICALLY_DELETED_ITEMS = 10;
static constexpr uint32_t MAX_LOGICALLY_DELETED_ITEMS = 10;
// Half the 32-bit range - used to detect rollovers vs normal time progression
static constexpr uint32_t HALF_MAX_UINT32 = std::numeric_limits<uint32_t>::max() / 2;
// max delay to start an interval sequence
@@ -117,12 +117,12 @@ void HOT Scheduler::set_timer_common_(Component *component, SchedulerItem::Type
// first execution happens immediately after a random smallish offset
// Calculate random offset (0 to min(interval/2, 5s))
uint32_t offset = (uint32_t) (std::min(delay / 2, MAX_INTERVAL_DELAY) * random_float());
item->next_execution_ = now + offset;
item->set_next_execution(now + offset);
ESP_LOGV(TAG, "Scheduler interval for %s is %" PRIu32 "ms, offset %" PRIu32 "ms", name_cstr ? name_cstr : "", delay,
offset);
} else {
item->interval = 0;
item->next_execution_ = now + delay;
item->set_next_execution(now + delay);
}
#ifdef ESPHOME_DEBUG_SCHEDULER
@@ -138,7 +138,8 @@ void HOT Scheduler::set_timer_common_(Component *component, SchedulerItem::Type
name_cstr ? name_cstr : "(null)", type_str, delay);
} else {
ESP_LOGD(TAG, "set_%s(name='%s/%s', %s=%" PRIu32 ", offset=%" PRIu32 ")", type_str, item->get_source(),
name_cstr ? name_cstr : "(null)", type_str, delay, static_cast<uint32_t>(item->next_execution_ - now));
name_cstr ? name_cstr : "(null)", type_str, delay,
static_cast<uint32_t>(item->get_next_execution() - now));
}
#endif /* ESPHOME_DEBUG_SCHEDULER */
@@ -285,9 +286,10 @@ optional<uint32_t> HOT Scheduler::next_schedule_in(uint32_t now) {
auto &item = this->items_[0];
// Convert the fresh timestamp from caller (usually Application::loop()) to 64-bit
const auto now_64 = this->millis_64_(now); // 'now' from parameter - fresh from caller
if (item->next_execution_ < now_64)
const uint64_t next_exec = item->get_next_execution();
if (next_exec < now_64)
return 0;
return item->next_execution_ - now_64;
return next_exec - now_64;
}
void HOT Scheduler::call(uint32_t now) {
#ifndef ESPHOME_THREAD_SINGLE
@@ -354,7 +356,7 @@ void HOT Scheduler::call(uint32_t now) {
const char *name = item->get_name();
ESP_LOGD(TAG, " %s '%s/%s' interval=%" PRIu32 " next_execution in %" PRIu64 "ms at %" PRIu64,
item->get_type_str(), item->get_source(), name ? name : "(null)", item->interval,
item->next_execution_ - now_64, item->next_execution_);
item->get_next_execution() - now_64, item->get_next_execution());
old_items.push_back(std::move(item));
}
@@ -401,7 +403,7 @@ void HOT Scheduler::call(uint32_t now) {
{
// Don't copy-by value yet
auto &item = this->items_[0];
if (item->next_execution_ > now_64) {
if (item->get_next_execution() > now_64) {
// Not reached timeout yet, done for this call
break;
}
@@ -440,7 +442,7 @@ void HOT Scheduler::call(uint32_t now) {
const char *item_name = item->get_name();
ESP_LOGV(TAG, "Running %s '%s/%s' with interval=%" PRIu32 " next_execution=%" PRIu64 " (now=%" PRIu64 ")",
item->get_type_str(), item->get_source(), item_name ? item_name : "(null)", item->interval,
item->next_execution_, now_64);
item->get_next_execution(), now_64);
#endif /* ESPHOME_DEBUG_SCHEDULER */
// Warning: During callback(), a lot of stuff can happen, including:
@@ -465,7 +467,7 @@ void HOT Scheduler::call(uint32_t now) {
}
if (item->type == SchedulerItem::INTERVAL) {
item->next_execution_ = now_64 + item->interval;
item->set_next_execution(now_64 + item->interval);
// Add new item directly to to_add_
// since we have the lock held
this->to_add_.push_back(std::move(item));
@@ -744,7 +746,10 @@ uint64_t Scheduler::millis_64_(uint32_t now) {
bool HOT Scheduler::SchedulerItem::cmp(const std::unique_ptr<SchedulerItem> &a,
const std::unique_ptr<SchedulerItem> &b) {
return a->next_execution_ > b->next_execution_;
// High bits are almost always equal (change only on 32-bit rollover ~49 days)
// Optimize for common case: check low bits first when high bits are equal
return (a->next_execution_high_ == b->next_execution_high_) ? (a->next_execution_low_ > b->next_execution_low_)
: (a->next_execution_high_ > b->next_execution_high_);
}
} // namespace esphome

31
esphome/core/scheduler.h
View File

@@ -88,19 +88,19 @@ class Scheduler {
struct SchedulerItem {
// Ordered by size to minimize padding
Component *component;
uint32_t interval;
// 64-bit time to handle millis() rollover. The scheduler combines the 32-bit millis()
// with a 16-bit rollover counter to create a 64-bit time that won't roll over for
// billions of years. This ensures correct scheduling even when devices run for months.
uint64_t next_execution_;
// Optimized name storage using tagged union
union {
const char *static_name; // For string literals (no allocation)
char *dynamic_name; // For allocated strings
} name_;
uint32_t interval;
// Split 64-bit time to handle millis() rollover. The scheduler combines the 32-bit millis()
// with a 16-bit rollover counter to create a 64-bit time that won't roll over for
// billions of years. This ensures correct scheduling even when devices run for months.
// Split into two fields for better memory alignment on 32-bit systems.
uint32_t next_execution_low_; // Lower 32 bits of next execution time
std::function<void()> callback;
uint16_t next_execution_high_; // Upper 16 bits of next execution time
#ifdef ESPHOME_THREAD_MULTI_ATOMICS
// Multi-threaded with atomics: use atomic for lock-free access
@@ -126,7 +126,8 @@ class Scheduler {
SchedulerItem()
: component(nullptr),
interval(0),
next_execution_(0),
next_execution_low_(0),
next_execution_high_(0),
#ifdef ESPHOME_THREAD_MULTI_ATOMICS
// remove is initialized in the member declaration as std::atomic<bool>{false}
type(TIMEOUT),
@@ -157,7 +158,7 @@ class Scheduler {
SchedulerItem &operator=(SchedulerItem &&) = delete;
// Helper to get the name regardless of storage type
const char *get_name() const { return name_is_dynamic ? name_.dynamic_name : name_.static_name; }
constexpr const char *get_name() const { return name_is_dynamic ? name_.dynamic_name : name_.static_name; }
// Helper to set name with proper ownership
void set_name(const char *name, bool make_copy = false) {
@@ -183,7 +184,17 @@ class Scheduler {
}
static bool cmp(const std::unique_ptr<SchedulerItem> &a, const std::unique_ptr<SchedulerItem> &b);
const char *get_type_str() const { return (type == TIMEOUT) ? "timeout" : "interval"; }
// Helper methods to work with split execution time (constexpr for optimization)
constexpr uint64_t get_next_execution() const {
return (static_cast<uint64_t>(next_execution_high_) << 32) | next_execution_low_;
}
constexpr void set_next_execution(uint64_t value) {
next_execution_low_ = static_cast<uint32_t>(value);
next_execution_high_ = static_cast<uint16_t>(value >> 32);
}
constexpr const char *get_type_str() const { return (type == TIMEOUT) ? "timeout" : "interval"; }
const char *get_source() const { return component ? component->get_component_source() : "unknown"; }
};