logalloc.cc

/*
 * Copyright (C) 2015-present ScyllaDB
 */

/*
 * SPDX-License-Identifier: LicenseRef-ScyllaDB-Source-Available-1.0
 */

#include <boost/intrusive/list.hpp>
#include <boost/intrusive/set.hpp>
#include <boost/intrusive/slist.hpp>
#include <stack>
#include <ranges>

#include <seastar/core/memory.hh>
#include <seastar/core/align.hh>
#include <seastar/core/format.hh>
#include <seastar/core/metrics.hh>
#include <seastar/core/reactor.hh>
#include <seastar/core/coroutine.hh>
#include <seastar/core/smp.hh>
#include <seastar/core/with_scheduling_group.hh>
#include <seastar/coroutine/maybe_yield.hh>
#include <seastar/util/alloc_failure_injector.hh>
#include <seastar/util/backtrace.hh>
#include <seastar/util/later.hh>

#include "utils/assert.hh"
#include "utils/logalloc.hh"
#include "utils/log.hh"
#include "utils/dynamic_bitset.hh"
#include "utils/log_heap.hh"
#include "utils/preempt.hh"
#include "utils/vle.hh"
#include "utils/coarse_steady_clock.hh"

#include <random>
#include <chrono>

using namespace std::chrono_literals;

#ifdef SEASTAR_ASAN_ENABLED
#include <sanitizer/asan_interface.h>
// For each aligned 8 byte segment, the algorithm used by address
// sanitizer can represent any addressable prefix followed by a
// poisoned suffix. The details are at:
// https://github.com/google/sanitizers/wiki/AddressSanitizerAlgorithm
// For us this means that:
// * The descriptor must be 8 byte aligned. If it was not, making the
//   descriptor addressable would also make the end of the previous
//   value addressable.
// * Each value must be at least 8 byte aligned. If it was not, making
//   the value addressable would also make the end of the descriptor
//   addressable.
template<typename T>
[[nodiscard]] static T align_up_for_asan(T val) {
    return align_up(val, size_t(8));
}
template<typename T>
void poison(const T* addr, size_t size) {
    // Both values and descriptors must be aligned.
    SCYLLA_ASSERT(uintptr_t(addr) % 8 == 0);
    // This can be followed by
    // * 8 byte aligned descriptor (this is a value)
    // * 8 byte aligned value
    // * dead value
    // * end of segment
    // In all cases, we can align up the size to guarantee that asan
    // is able to poison this.
    ASAN_POISON_MEMORY_REGION(addr, align_up_for_asan(size));
}
void unpoison(const char *addr, size_t size) {
    ASAN_UNPOISON_MEMORY_REGION(addr, size);
}
#else
template<typename T>
[[nodiscard]] static T align_up_for_asan(T val) { return val; }
template<typename T>
void poison(const T* addr, size_t size) { }
void unpoison(const char *addr, size_t size) { }
#endif

namespace bi = boost::intrusive;

standard_allocation_strategy standard_allocation_strategy_instance;

namespace {

class migrators_base {
protected:
    std::vector<const migrate_fn_type*> _migrators;
};

#ifdef DEBUG_LSA_SANITIZER

class migrators : public migrators_base, public enable_lw_shared_from_this<migrators> {
private:
    struct backtrace_entry {
        saved_backtrace _registration;
        saved_backtrace _deregistration;
    };
    std::vector<std::unique_ptr<backtrace_entry>> _backtraces;

    static logging::logger _logger;
private:
    void on_error() { abort(); }
public:
    uint32_t add(const migrate_fn_type* m) {
        _migrators.push_back(m);
        _backtraces.push_back(std::make_unique<backtrace_entry>(backtrace_entry{current_backtrace(), {}}));
        return _migrators.size() - 1;
    }
    void remove(uint32_t idx) {
        if (idx >= _migrators.size()) {
            _logger.error("Attempting to deregister migrator id {} which was never registered:\n{}",
                          idx, current_backtrace());
            on_error();
        }
        if (!_migrators[idx]) {
            _logger.error("Attempting to double deregister migrator id {}:\n{}\n"
                          "Previously deregistered at:\n{}\nRegistered at:\n{}",
                          idx, current_backtrace(), _backtraces[idx]->_deregistration,
                          _backtraces[idx]->_registration);
            on_error();
        }
        _migrators[idx] = nullptr;
        _backtraces[idx]->_deregistration = current_backtrace();
    }
    const migrate_fn_type*& operator[](uint32_t idx) {
        if (idx >= _migrators.size()) {
            _logger.error("Attempting to use migrator id {} that was never registered:\n{}",
                          idx, current_backtrace());
            on_error();
        }
        if (!_migrators[idx]) {
            _logger.error("Attempting to use deregistered migrator id {}:\n{}\n"
                          "Deregistered at:\n{}\nRegistered at:\n{}",
                          idx, current_backtrace(), _backtraces[idx]->_deregistration,
                          _backtraces[idx]->_registration);
            on_error();
        }
        return _migrators[idx];
    }
};

logging::logger migrators::_logger("lsa-migrator-sanitizer");

#else

class migrators : public migrators_base, public enable_lw_shared_from_this<migrators> {
    std::vector<uint32_t> _unused_ids;

public:
    uint32_t add(const migrate_fn_type* m) {
        if (!_unused_ids.empty()) {
            uint32_t idx = _unused_ids.back();
            _unused_ids.pop_back();
            _migrators[idx] = m;
            return idx;
        }
        _migrators.push_back(m);
        return _migrators.size() - 1;
    }
    void remove(uint32_t idx) {
        _unused_ids.push_back(idx);
    }
    const migrate_fn_type*& operator[](uint32_t idx) {
        return _migrators[idx];
    }
};

#endif

static
migrators&
static_migrators() noexcept {
    memory::scoped_critical_alloc_section dfg;
    static thread_local lw_shared_ptr<migrators> obj = make_lw_shared<migrators>();
    return *obj;
}

}

namespace debug {

thread_local migrators* static_migrators = &::static_migrators();

}


uint32_t
migrate_fn_type::register_migrator(migrate_fn_type* m) {
    auto& migrators = *debug::static_migrators;
    auto idx = migrators.add(m);
    // object_descriptor encodes 2 * index() + 1
    SCYLLA_ASSERT(idx * 2 + 1 < utils::uleb64_express_supreme);
    m->_migrators = migrators.shared_from_this();
    return idx;
}

void
migrate_fn_type::unregister_migrator(uint32_t index) {
    static_migrators().remove(index);
}


namespace {

// for printing extra message in reclaim_timer::report() when stall is detected.
//
// this helper struct is deliberately introduced to ensure no dynamic
// allocations in reclaim_timer::report(), which is involved in handling OOMs.
struct extra_msg_when_stall_detected {
    bool stall_detected;
    saved_backtrace backtrace;
    extra_msg_when_stall_detected(bool detected, saved_backtrace&& backtrace)
        : stall_detected{detected}
        , backtrace{std::move(backtrace)}
    {}
};

}

template <>
struct fmt::formatter<extra_msg_when_stall_detected> {
    constexpr auto parse(format_parse_context& ctx) { return ctx.begin(); }
    auto format(const extra_msg_when_stall_detected& msg, fmt::format_context& ctx) const {
        if (msg.stall_detected) {
            return fmt::format_to(ctx.out(), ", at {}", msg.backtrace);
        } else {
            return ctx.out();
        }
    }
};


namespace logalloc {

// LSA-specific bad_alloc variant which allows adding additional information on
// why the allocation failed.
class bad_alloc : public std::bad_alloc {
    sstring _what;
public:
    bad_alloc(sstring what) : _what(std::move(what)) { }
    virtual const char* what() const noexcept override {
        return _what.c_str();
    }
};

#ifdef DEBUG_LSA_SANITIZER

class region_sanitizer {
    struct allocation {
        size_t size;
        saved_backtrace backtrace;
    };
private:
    static logging::logger logger;

    const bool* _report_backtrace = nullptr;
    bool _broken = false;
    std::unordered_map<const void*, allocation> _allocations;
private:
    template<typename Function>
    void run_and_handle_errors(Function&& fn) noexcept {
        memory::scoped_critical_alloc_section dfg;
        if (_broken) {
            return;
        }
        try {
            fn();
        } catch (...) {
            logger.error("Internal error, disabling the sanitizer: {}", std::current_exception());
            _broken = true;
            _allocations.clear();
        }
    }
private:
    void on_error() { abort(); }
public:
    region_sanitizer(const bool& report_backtrace) : _report_backtrace(&report_backtrace) { }
    void on_region_destruction() noexcept {
        run_and_handle_errors([&] {
            if (_allocations.empty()) {
                return;
            }
            for (auto [ptr, alloc] : _allocations) {
                logger.error("Leaked {} byte object at {} allocated from:\n{}",
                             alloc.size, ptr, alloc.backtrace);
            }
            on_error();
        });
    }
    void on_allocation(const void* ptr, size_t size) noexcept {
        run_and_handle_errors([&] {
            auto backtrace = *_report_backtrace ? current_backtrace() : saved_backtrace();
            auto [ it, success ] = _allocations.emplace(ptr, allocation { size, std::move(backtrace) });
            if (!success) {
                logger.error("Attempting to allocate an {} byte object at an already occupied address {}:\n{}\n"
                             "Previous allocation of {} bytes:\n{}",
                             ptr, size, current_backtrace(), it->second.size, it->second.backtrace);
                on_error();
            }
        });
    }
    void on_free(const void* ptr, size_t size) noexcept {
        run_and_handle_errors([&] {
            auto it = _allocations.find(ptr);
            if (it == _allocations.end()) {
                logger.error("Attempting to free an object at {} (size: {}) that does not exist\n{}",
                             ptr, size, current_backtrace());
                on_error();
            }
            if (it->second.size != size) {
                logger.error("Mismatch between allocation and deallocation size of object at {}: {} vs. {}:\n{}\n"
                             "Allocated at:\n{}",
                             ptr, it->second.size, size, current_backtrace(), it->second.backtrace);
                on_error();
            }
            _allocations.erase(it);
        });
    }
    void on_migrate(const void* src, size_t size, const void* dst) noexcept {
        run_and_handle_errors([&] {
            auto it_src = _allocations.find(src);
            if (it_src == _allocations.end()) {
                logger.error("Attempting to migrate an object at {} (size: {}) that does not exist",
                             src, size);
                on_error();
            }
            if (it_src->second.size != size) {
                logger.error("Mismatch between allocation and migration size of object at {}: {} vs. {}\n"
                             "Allocated at:\n{}",
                             src, it_src->second.size, size, it_src->second.backtrace);
                on_error();
            }
            auto [ it_dst, success ] = _allocations.emplace(dst, std::move(it_src->second));
            if (!success) {
                logger.error("Attempting to migrate an {} byte object to an already occupied address {}:\n"
                             "Migrated object allocated from:\n{}\n"
                             "Previous allocation of {} bytes at the destination:\n{}",
                             size, dst, it_src->second.backtrace, it_dst->second.size, it_dst->second.backtrace);
                on_error();
            }
            _allocations.erase(it_src);
        });
    }
    void merge(region_sanitizer& other) noexcept {
        run_and_handle_errors([&] {
            _broken = other._broken;
            if (_broken) {
                _allocations.clear();
            } else {
                _allocations.merge(other._allocations);
                if (!other._allocations.empty()) {
                    for (auto [ptr, o_alloc] : other._allocations) {
                        auto& alloc = _allocations.at(ptr);
                        logger.error("Conflicting allocations at address {} in merged regions\n"
                                     "{} bytes allocated from:\n{}\n"
                                     "{} bytes allocated from:\n{}",
                                     ptr, alloc.size, alloc.backtrace, o_alloc.size, o_alloc.backtrace);
                    }
                    on_error();
                }
            }
        });
    }
};

logging::logger region_sanitizer::logger("lsa-sanitizer");

#else

struct region_sanitizer {
    region_sanitizer(const bool&) { }
    void on_region_destruction() noexcept { }
    void on_allocation(const void*, size_t) noexcept { }
    void on_free(const void* ptr, size_t size) noexcept { }
    void on_migrate(const void*, size_t, const void*) noexcept { }
    void merge(region_sanitizer&) noexcept { }
};

#endif

struct segment;

static logging::logger llogger("lsa");
static logging::logger timing_logger("lsa-timing");

static tracker& get_tracker_instance() noexcept {
    memory::scoped_critical_alloc_section dfg;
    static thread_local tracker obj;
    return obj;
}

static thread_local tracker& tracker_instance = get_tracker_instance();

using clock = std::chrono::steady_clock;

class background_reclaimer {
    scheduling_group _sg;
    noncopyable_function<void (size_t target)> _reclaim;
    timer<lowres_clock> _adjust_shares_timer;
    // If engaged, main loop is not running, set_value() to wake it.
    promise<>* _main_loop_wait = nullptr;
    future<> _done;
    bool _stopping = false;
    static constexpr size_t free_memory_threshold = 60'000'000;
private:
    bool have_work() const {
#ifndef SEASTAR_DEFAULT_ALLOCATOR
        return memory::free_memory() < free_memory_threshold;
#else
        return false;
#endif
    }
    void main_loop_wake() {
        llogger.debug("background_reclaimer::main_loop_wake: waking {}", bool(_main_loop_wait));
        if (_main_loop_wait) {
            _main_loop_wait->set_value();
            _main_loop_wait = nullptr;
        }
    }
    future<> main_loop() {
        llogger.debug("background_reclaimer::main_loop: entry");
        while (true) {
            while (!_stopping && !have_work()) {
                promise<> wait;
                _main_loop_wait = &wait;
                llogger.trace("background_reclaimer::main_loop: sleep");
                co_await wait.get_future();
                llogger.trace("background_reclaimer::main_loop: awakened");
                _main_loop_wait = nullptr;
            }
            if (_stopping) {
                break;
            }
            _reclaim(free_memory_threshold - memory::free_memory());
            co_await coroutine::maybe_yield();
        }
        llogger.debug("background_reclaimer::main_loop: exit");
    }
    void adjust_shares() {
        if (have_work()) {
            auto shares = 1 + (1000 * (free_memory_threshold - memory::free_memory())) / free_memory_threshold;
            _sg.set_shares(shares);
            llogger.trace("background_reclaimer::adjust_shares: {}", shares);
            if (_main_loop_wait) {
                main_loop_wake();
            }
        }
    }
public:
    explicit background_reclaimer(scheduling_group sg, noncopyable_function<void (size_t target)> reclaim)
            : _sg(sg)
            , _reclaim(std::move(reclaim))
            , _adjust_shares_timer(default_scheduling_group(), [this] { adjust_shares(); })
            , _done(with_scheduling_group(_sg, [this] { return main_loop(); })) {
        if (sg != default_scheduling_group()) {
            _adjust_shares_timer.arm_periodic(50ms);
        }
    }
    future<> stop() {
        _stopping = true;
        main_loop_wake();
        return std::move(_done);
    }
};

class segment_pool;
struct reclaim_timer;

class tracker::impl {
    std::unique_ptr<logalloc::segment_pool> _segment_pool;
    std::optional<background_reclaimer> _background_reclaimer;
    std::vector<region::impl*> _regions;
    seastar::metrics::metric_groups _metrics;
    unsigned _reclaiming_disabled_depth = 0;
    size_t _reclamation_step = 1;
    bool _abort_on_bad_alloc = false;
    bool _sanitizer_report_backtrace = false;
    reclaim_timer* _active_timer = nullptr;
private:
    // Prevents tracker's reclaimer from running while live. Reclaimer may be
    // invoked synchronously with allocator. This guard ensures that this
    // object is not re-entered while inside one of the tracker's methods.
    struct reclaiming_lock {
        impl& _ref;
        reclaiming_lock(impl& ref) noexcept
            : _ref(ref)
        {
            _ref.disable_reclaim();
        }
        ~reclaiming_lock() {
            _ref.enable_reclaim();
        }
    };
    friend class tracker_reclaimer_lock;
public:
    impl();
    ~impl();
    future<> stop() {
        if (_background_reclaimer) {
            return _background_reclaimer->stop();
        } else {
            return make_ready_future<>();
        }
    }
    void disable_reclaim() noexcept {
        ++_reclaiming_disabled_depth;
    }
    void enable_reclaim() noexcept {
        --_reclaiming_disabled_depth;
    }
    logalloc::segment_pool& segment_pool() {
        return *_segment_pool;
    }
    void register_region(region::impl*);
    void unregister_region(region::impl*) noexcept;
    size_t reclaim(size_t bytes, is_preemptible p);
    // Compacts one segment at a time from sparsest segment to least sparse until work_waiting_on_reactor returns true
    // or there are no more segments to compact.
    idle_cpu_handler_result compact_on_idle(work_waiting_on_reactor check_for_work);
    // Releases whole segments back to the segment pool.
    // After the call, if there is enough evictable memory, the amount of free segments in the pool
    // will be at least reserve_segments + div_ceil(bytes, segment::size).
    // Returns the amount by which segment_pool.total_memory_in_use() has decreased.
    size_t compact_and_evict(size_t reserve_segments, size_t bytes, is_preemptible p);
    void full_compaction();
    void reclaim_all_free_segments();
    occupancy_stats global_occupancy() const noexcept;
    occupancy_stats region_occupancy() const noexcept;
    occupancy_stats occupancy() const noexcept;
    size_t non_lsa_used_space() const noexcept;
    // Set the minimum number of segments reclaimed during single reclamation cycle.
    void set_reclamation_step(size_t step_in_segments) noexcept { _reclamation_step = step_in_segments; }
    size_t reclamation_step() const noexcept { return _reclamation_step; }
    // Abort on allocation failure from LSA
    void enable_abort_on_bad_alloc() noexcept { _abort_on_bad_alloc = true; }
    bool should_abort_on_bad_alloc() const noexcept { return _abort_on_bad_alloc; }
    void setup_background_reclaim(scheduling_group sg) {
        SCYLLA_ASSERT(!_background_reclaimer);
        _background_reclaimer.emplace(sg, [this] (size_t target) {
            reclaim(target, is_preemptible::yes);
        });
    }
    // const bool&, so interested parties can save a reference and see updates.
    const bool& sanitizer_report_backtrace() const { return _sanitizer_report_backtrace; }
    void set_sanitizer_report_backtrace(bool rb) { _sanitizer_report_backtrace = rb; }
    bool try_set_active_timer(reclaim_timer& timer) {
        if (_active_timer) {
            return false;
        }
        _active_timer = &timer;
        return true;
    }
    bool try_reset_active_timer(reclaim_timer& timer) {
        if (_active_timer == &timer) {
            _active_timer = nullptr;
            return true;
        }
        return false;
    }
private:
    // Like compact_and_evict() but assumes that reclaim_lock is held around the operation.
    size_t compact_and_evict_locked(size_t reserve_segments, size_t bytes, is_preemptible preempt);
    // Like reclaim() but assumes that reclaim_lock is held around the operation.
    size_t reclaim_locked(size_t bytes, is_preemptible p);
};

tracker_reclaimer_lock::tracker_reclaimer_lock(tracker::impl& impl) noexcept : _tracker_impl(impl) {
    _tracker_impl.disable_reclaim();
}

tracker_reclaimer_lock::~tracker_reclaimer_lock() {
    _tracker_impl.enable_reclaim();
}

tracker::tracker()
    : _impl(std::make_unique<impl>())
    , _reclaimer([this] (seastar::memory::reclaimer::request r) { return reclaim(r); }, memory::reclaimer_scope::sync)
{ }

tracker::~tracker() {
}

future<>
tracker::stop() {
    return _impl->stop();
}

size_t tracker::reclaim(size_t bytes) {
    return _impl->reclaim(bytes, is_preemptible::no);
}

occupancy_stats tracker::global_occupancy() const noexcept {
    return _impl->global_occupancy();
}

occupancy_stats tracker::region_occupancy() const noexcept {
    return _impl->region_occupancy();
}

occupancy_stats tracker::occupancy() const noexcept {
    return _impl->occupancy();
}

size_t tracker::non_lsa_used_space() const noexcept {
    return _impl->non_lsa_used_space();
}

void tracker::full_compaction() {
    return _impl->full_compaction();
}

void tracker::reclaim_all_free_segments() {
    return _impl->reclaim_all_free_segments();
}

tracker& shard_tracker() noexcept {
    return tracker_instance;
}

struct alignas(segment_size) segment {
    static constexpr int size_shift = segment_size_shift;
    static constexpr int size_mask = segment_size | (segment_size - 1);
    using size_type = std::conditional_t<(size_shift < 16), uint16_t, uint32_t>;
    static constexpr size_t size = segment_size;

    uint8_t data[size];

    segment() noexcept { }

    template<typename T = void>
    const T* at(size_t offset) const noexcept {
        return reinterpret_cast<const T*>(data + offset);
    }

    template<typename T = void>
    T* at(size_t offset) noexcept {
        return reinterpret_cast<T*>(data + offset);
    }

    static void* operator new(size_t size) = delete;
    static void* operator new(size_t, void* ptr) noexcept { return ptr; }
    static void operator delete(void* ptr) = delete;
};

static constexpr size_t max_managed_object_size = segment_size * 0.1;
static constexpr auto max_used_space_ratio_for_compaction = 0.85;
static constexpr size_t max_used_space_for_compaction = segment_size * max_used_space_ratio_for_compaction;
static constexpr size_t min_free_space_for_compaction = segment_size - max_used_space_for_compaction;

struct [[gnu::packed]] non_lsa_object_cookie {
    uint64_t value = 0xbadcaffe;
};

static_assert(min_free_space_for_compaction >= max_managed_object_size,
    "Segments which cannot fit max_managed_object_size must not be considered compactible for the sake of forward progress of compaction");

// Since we only compact if there's >= min_free_space_for_compaction of free space,
// we use min_free_space_for_compaction as the histogram's minimum size and put
// everything below that value in the same bucket.
extern constexpr log_heap_options segment_descriptor_hist_options(min_free_space_for_compaction, 3, segment_size);

enum segment_kind : int {
    regular = 0, // Holds objects allocated with region_impl::alloc_small()
    bufs = 1     // Holds objects allocated with region_impl::alloc_buf()
};

struct segment_descriptor : public log_heap_hook<segment_descriptor_hist_options> {
    static constexpr segment::size_type free_space_mask = segment::size_mask;
    static constexpr unsigned bits_for_free_space = segment::size_shift + 1;
    static constexpr segment::size_type segment_kind_mask = 1 << bits_for_free_space;
    static constexpr unsigned bits_for_segment_kind = 1;
    static constexpr unsigned shift_for_segment_kind = bits_for_free_space;
    static_assert(sizeof(segment::size_type) * 8 >= bits_for_free_space + bits_for_segment_kind);

    segment::size_type _free_space;
    region::impl* _region;

    segment::size_type free_space() const noexcept {
        return _free_space & free_space_mask;
    }

    void set_free_space(segment::size_type free_space) noexcept {
        _free_space = (_free_space & ~free_space_mask) | free_space;
    }

    segment_kind kind() const noexcept {
        return static_cast<segment_kind>((_free_space & segment_kind_mask) >> shift_for_segment_kind);
    }

    void set_kind(segment_kind kind) noexcept {
        _free_space = (_free_space & ~segment_kind_mask)
                | static_cast<segment::size_type>(kind) << shift_for_segment_kind;
    }

    // Valid if kind() == segment_kind::bufs.
    //
    // _buf_pointers holds links to lsa_buffer objects (paired with lsa_buffer::_link)
    // of live objects in the segment. The purpose of this is so that segment compaction
    // can update the pointers when it moves the objects.
    // The order of entangled objects in the vector is irrelevant.
    // Also, not all entangled objects may be engaged.
    std::vector<entangled> _buf_pointers;

    segment_descriptor() noexcept
        : _region(nullptr)
    { }

    bool is_empty() const noexcept {
        return free_space() == segment::size;
    }

    occupancy_stats occupancy() const noexcept {
        return { free_space(), segment::size };
    }

    void record_alloc(segment::size_type size) noexcept {
        _free_space -= size;
    }

    void record_free(segment::size_type size) noexcept {
        _free_space += size;
    }
};

using segment_descriptor_hist = log_heap<segment_descriptor, segment_descriptor_hist_options>;

class segment_store_backend {
protected:
    memory::memory_layout _layout;
    // Whether freeing segments actually increases availability of non-lsa memory.
    bool _freed_segment_increases_general_memory_availability;
    // Aligned (to segment::size) address of the first segment.
    uintptr_t _segments_base;

public:
    explicit segment_store_backend(memory::memory_layout layout, bool freed_segment_increases_general_memory_availability) noexcept
        : _layout(layout)
        , _freed_segment_increases_general_memory_availability(freed_segment_increases_general_memory_availability)
        , _segments_base(align_up(_layout.start, static_cast<uintptr_t>(segment::size)))
    { }
    virtual ~segment_store_backend() = default;
    memory::memory_layout memory_layout() const noexcept { return _layout; }
    uintptr_t segments_base() const noexcept { return _segments_base; }
    virtual void* alloc_segment_memory() noexcept = 0;
    virtual void free_segment_memory(void* seg) noexcept = 0;
    virtual size_t free_memory() const noexcept = 0;
    bool can_allocate_more_segments(size_t non_lsa_reserve) const noexcept {
        if (_freed_segment_increases_general_memory_availability) {
            return free_memory() >= non_lsa_reserve + segment::size;
        } else {
            return free_memory() >= segment::size;
        }
    }
};

// Segments are allocated from the seastar allocator.
// The entire memory area of the local shard is used as a segment store, i.e.
// segments are allocated from the same memory area regular objeces are.
class seastar_memory_segment_store_backend : public segment_store_backend {
public:
    seastar_memory_segment_store_backend()
        : segment_store_backend(memory::get_memory_layout(), true)
    { }
    virtual void* alloc_segment_memory() noexcept override {
        return aligned_alloc(segment::size, segment::size);
    }
    virtual void free_segment_memory(void* seg) noexcept override {
        ::free(seg);
    }
    virtual size_t free_memory() const noexcept override {
        return memory::free_memory();
    }
};

// Segments storage is allocated via `mmap()`.
// This area cannot be shrunk or enlarged, so freeing segments doesn't increase
// memory availability.
class standard_memory_segment_store_backend : public segment_store_backend {
    struct free_segment {
        free_segment* next = nullptr;
    };

private:
    uintptr_t _segments_offset = 0;
    free_segment* _freelist = nullptr;
    size_t _available_segments; // for fast free_memory()

private:
    static memory::memory_layout allocate_memory(size_t segments) {
        const auto size = segments * segment_size;
        auto p = mmap(nullptr, size,
                PROT_READ | PROT_WRITE,
                MAP_PRIVATE | MAP_ANONYMOUS,
                -1, 0);
        if (p == MAP_FAILED) {
            std::abort();
        }
        madvise(p, size, MADV_HUGEPAGE);
        auto start = reinterpret_cast<uintptr_t>(p);
        return {start, start + size};
    }
public:
    standard_memory_segment_store_backend(size_t segments)
        : segment_store_backend(allocate_memory(segments), false)
        , _available_segments((_layout.end - _segments_base) / segment_size)
    { }
    ~standard_memory_segment_store_backend() {
        munmap(reinterpret_cast<void*>(_layout.start), _layout.end - _layout.start);
    }
    virtual void* alloc_segment_memory() noexcept override {
        if (_freelist) {
            --_available_segments;
            return std::exchange(_freelist, _freelist->next);
        }
        auto seg = _segments_base + _segments_offset * segment_size;
        if (seg + segment_size > _layout.end) {
            return nullptr;
        }
        ++_segments_offset;
        --_available_segments;
        return reinterpret_cast<void*>(seg);
    }
    virtual void free_segment_memory(void* seg) noexcept override {
        unpoison(reinterpret_cast<char*>(seg), sizeof(free_segment));
        auto fs = new (seg) free_segment;
        fs->next = _freelist;
        _freelist = fs;
        ++_available_segments;
    }
    virtual size_t free_memory() const noexcept override {
        return _available_segments * segment_size;
    }
};

static constexpr size_t segment_npos = size_t(-1);

// Segments are allocated from a large contiguous memory area.
class contiguous_memory_segment_store {
    std::unique_ptr<segment_store_backend> _backend;

public:
    size_t non_lsa_reserve = 0;
    contiguous_memory_segment_store()
        : _backend(std::make_unique<seastar_memory_segment_store_backend>())
    { }
    struct with_standard_memory_backend {};
    contiguous_memory_segment_store(with_standard_memory_backend, size_t available_memory) {
        use_standard_allocator_segment_pool_backend(available_memory);
    }
    void use_standard_allocator_segment_pool_backend(size_t available_memory) {
        _backend = std::make_unique<standard_memory_segment_store_backend>(available_memory / segment::size);
        llogger.debug("using the standard allocator segment pool backend with {} available memory", available_memory);
    }
    const segment* segment_from_idx(size_t idx) const noexcept {
        return reinterpret_cast<segment*>(_backend->segments_base()) + idx;
    }
    segment* segment_from_idx(size_t idx) noexcept {
        return reinterpret_cast<segment*>(_backend->segments_base()) + idx;
    }
    size_t idx_from_segment(const segment* seg) const noexcept {
        const auto seg_uint = reinterpret_cast<uintptr_t>(seg);
        if (seg_uint < _backend->memory_layout().start || seg_uint > _backend->memory_layout().end) [[unlikely]] {
            return segment_npos;
        }
        return seg - reinterpret_cast<segment*>(_backend->segments_base());
    }
    std::pair<segment*, size_t> allocate_segment() noexcept {
        auto p = _backend->alloc_segment_memory();
        if (!p) {
            return {nullptr, 0};
        }
        auto seg = new (p) segment;
        poison(seg, sizeof(segment));
        return {seg, idx_from_segment(seg)};
    }
    void free_segment(segment *seg) noexcept {
        seg->~segment();
        _backend->free_segment_memory(seg);
    }
    size_t max_segments() const noexcept {
        return (_backend->memory_layout().end - _backend->segments_base()) / segment::size;
    }
    bool can_allocate_more_segments() const noexcept {
        return _backend->can_allocate_more_segments(non_lsa_reserve);
    }
};
#ifndef SEASTAR_DEFAULT_ALLOCATOR
using segment_store = contiguous_memory_segment_store;
#else
class segment_store {
    std::unique_ptr<contiguous_memory_segment_store> _delegate_store;
    std::vector<segment*> _segments;
    std::unordered_map<segment*, size_t> _segment_indexes;
    static constexpr size_t _std_memory_available = size_t(1) << 30; // emulate 1GB per shard
    std::vector<segment*>::iterator find_empty() noexcept {
        return std::find(_segments.begin(), _segments.end(), nullptr);
    }
    std::vector<segment*>::const_iterator find_empty() const noexcept {
        return std::find(_segments.cbegin(), _segments.cend(), nullptr);
    }
    void free_segments() noexcept {
        for (segment *seg : _segments) {
            if (seg) {
                seg->~segment();
                free(seg);
            }
        }
    }

public:
    size_t non_lsa_reserve = 0;
    segment_store() : _segments(max_segments()) {
        _segment_indexes.reserve(max_segments());
    }
    void use_standard_allocator_segment_pool_backend(size_t available_memory) {
        _delegate_store = std::make_unique<contiguous_memory_segment_store>(contiguous_memory_segment_store::with_standard_memory_backend{}, available_memory);
        free_segments();
        _segment_indexes = {};
        llogger.debug("using the standard allocator segment pool backend with {} available memory", available_memory);
    }
    const segment* segment_from_idx(size_t idx) const noexcept {
        if (_delegate_store) {
            return _delegate_store->segment_from_idx(idx);
        }
        SCYLLA_ASSERT(idx < _segments.size());
        return _segments[idx];
    }
    segment* segment_from_idx(size_t idx) noexcept {
        if (_delegate_store) {
            return _delegate_store->segment_from_idx(idx);
        }
        SCYLLA_ASSERT(idx < _segments.size());
        return _segments[idx];
    }
    size_t idx_from_segment(const segment* seg) const noexcept {
        if (_delegate_store) {
            return _delegate_store->idx_from_segment(seg);
        }
        auto i = _segment_indexes.find(const_cast<segment*>(seg));
        if (i == _segment_indexes.end()) {
            return segment_npos;
        }
        return i->second;
    }
    std::pair<segment*, size_t> allocate_segment() noexcept {
        if (_delegate_store) {
            return _delegate_store->allocate_segment();
        }
        auto p = aligned_alloc(segment::size, segment::size);
        if (!p) {
            return {nullptr, 0};
        }
        auto seg = new (p) segment;
        poison(seg, sizeof(segment));
        auto i = find_empty();
        SCYLLA_ASSERT(i != _segments.end());
        *i = seg;
        size_t ret = i - _segments.begin();
        _segment_indexes[seg] = ret;
        return {seg, ret};
    }
    void free_segment(segment *seg) noexcept {
        if (_delegate_store) {
            return _delegate_store->free_segment(seg);
        }
        seg->~segment();
        ::free(seg);
        size_t i = idx_from_segment(seg);
        _segment_indexes.erase(seg);
        _segments[i] = nullptr;
    }
    ~segment_store() {
        free_segments();
    }
    size_t max_segments() const noexcept {
        if (_delegate_store) {
            return _delegate_store->max_segments();
        }
        return _std_memory_available / segment::size;
    }
    bool can_allocate_more_segments() const noexcept {
        if (_delegate_store) {
            return _delegate_store->can_allocate_more_segments();
        }
        auto i = find_empty();
        return i != _segments.end();
    }
};
#endif

// Segment pool implementation for the seastar allocator.
// Stores segment descriptors in a vector which is indexed using most significant
// bits of segment address.
//
// We prefer using high-address segments, and returning low-address segments to the seastar
// allocator in order to segregate lsa and non-lsa memory, to reduce fragmentation.
class segment_pool {
    logalloc::tracker::impl& _tracker;
    segment_store _store;
    std::vector<segment_descriptor> _segments;
    size_t _segments_in_use{};
    utils::dynamic_bitset _lsa_owned_segments_bitmap; // owned by this
    utils::dynamic_bitset _lsa_free_segments_bitmap;  // owned by this, but not in use
    size_t _free_segments = 0;

    // Invariant: _free_segments > _current_emergency_reserve_goal.
    // Used to ensure that some critical allocations won't fail.
    // (We grow _current_emergency_reserve_goal in advance and shrink it right
    // before the critical allocations, which allows them to utilize the pre-reserved
    // segments).
    size_t _current_emergency_reserve_goal = 1;
    // Used by allocating_section to request a certain number of free segments
    // to be prepared for usage when the section is entered.
    // This is more of a side-channel argument to refill_emergency_reserve() than a real piece of state.
    // Passing it via a variable makes it easier to debug.
    size_t _emergency_reserve_max = 30;
    bool _allocation_failure_flag = false;
    bool _allocation_enabled = true;

    struct allocation_lock {
        segment_pool& _pool;
        bool _prev;
        allocation_lock(segment_pool& p) noexcept
            : _pool(p)
            , _prev(p._allocation_enabled)
        {
            _pool._allocation_enabled = false;
        }
        ~allocation_lock() {
            _pool._allocation_enabled = _prev;
        }
    };

    size_t _non_lsa_memory_in_use = 0;
    // Invariants - a segment is in one of the following states:
    //   In use by some region
    //     - set in _lsa_owned_segments_bitmap
    //     - clear in _lsa_free_segments_bitmap
    //     - counted in _segments_in_use
    //   Free:
    //     - set in _lsa_owned_segments_bitmap
    //     - set in _lsa_free_segments_bitmap
    //     - counted in _unreserved_free_segments
    //   Non-lsa:
    //     - clear everywhere
private:
    segment* allocate_segment(size_t reserve);
    void deallocate_segment(segment* seg) noexcept;
    friend void* segment::operator new(size_t);
    friend void segment::operator delete(void*);

    segment* allocate_or_fallback_to_reserve();
    const segment* segment_from_idx(size_t idx) const noexcept {
        return _store.segment_from_idx(idx);
    }
    segment* segment_from_idx(size_t idx) noexcept {
        return _store.segment_from_idx(idx);
    }
    size_t idx_from_segment(const segment* seg) const noexcept {
        return _store.idx_from_segment(seg);
    }
    size_t max_segments() const noexcept {
        return _store.max_segments();
    }
    bool can_allocate_more_segments() const noexcept {
        return _allocation_enabled && _store.can_allocate_more_segments();
    }
    bool compact_segment(segment* seg);
public:
    explicit segment_pool(logalloc::tracker::impl& tracker);
    logalloc::tracker::impl& tracker() { return _tracker; }
    void prime(size_t available_memory, size_t min_free_memory);
    void use_standard_allocator_segment_pool_backend(size_t available_memory);
    segment* new_segment(region::impl* r);
    const segment_descriptor& descriptor(const segment* seg) const noexcept {
        uintptr_t index = idx_from_segment(seg);
        return _segments[index];
    }
    segment_descriptor& descriptor(segment* seg) noexcept {
        uintptr_t index = idx_from_segment(seg);
        return _segments[index];
    }
    // Returns segment containing given object or nullptr.
    segment* containing_segment(const void* obj) noexcept;
    segment* segment_from(const segment_descriptor& desc) noexcept;
    void free_segment(segment*) noexcept;
    void free_segment(segment*, segment_descriptor&) noexcept;
    size_t segments_in_use() const noexcept;
    size_t current_emergency_reserve_goal() const noexcept { return _current_emergency_reserve_goal; }
    void set_emergency_reserve_max(size_t new_size) noexcept { _emergency_reserve_max = new_size; }
    size_t emergency_reserve_max() const noexcept { return _emergency_reserve_max; }
    void set_current_emergency_reserve_goal(size_t goal) noexcept { _current_emergency_reserve_goal = goal; }
    void clear_allocation_failure_flag() noexcept { _allocation_failure_flag = false; }
    bool allocation_failure_flag() const noexcept { return _allocation_failure_flag; }
    void refill_emergency_reserve();
    void ensure_free_segments(size_t n_segments);
    void add_non_lsa_memory_in_use(size_t n) noexcept {
        _non_lsa_memory_in_use += n;
    }
    void subtract_non_lsa_memory_in_use(size_t n) noexcept {
        SCYLLA_ASSERT(_non_lsa_memory_in_use >= n);
        _non_lsa_memory_in_use -= n;
    }
    size_t non_lsa_memory_in_use() const noexcept {
        return _non_lsa_memory_in_use;
    }
    size_t total_memory_in_use() const noexcept {
        return _non_lsa_memory_in_use + _segments_in_use * segment::size;
    }
    size_t total_free_memory() const noexcept {
        return _free_segments * segment::size;
    }
    struct reservation_goal;
    void set_region(segment* seg, region::impl* r) noexcept {
        set_region(descriptor(seg), r);
    }
    void set_region(segment_descriptor& desc, region::impl* r) noexcept {
        desc._region = r;
    }
    size_t reclaim_segments(size_t target, is_preemptible preempt);
    void reclaim_all_free_segments() {
        reclaim_segments(std::numeric_limits<size_t>::max(), is_preemptible::no);
    }

private:
    tracker::stats _stats{};
public:
    const tracker::stats& statistics() const noexcept { return _stats; }
    inline void on_segment_compaction(size_t used_size) noexcept;
    inline void on_memory_allocation(size_t size) noexcept;
    inline void on_memory_deallocation(size_t size) noexcept;
    inline void on_memory_eviction(size_t size) noexcept;
    size_t unreserved_free_segments() const noexcept { return _free_segments - std::min(_free_segments, _emergency_reserve_max); }
    size_t free_segments() const noexcept { return _free_segments; }
};

struct reclaim_timer {
    using extra_logger = noncopyable_function<void(log_level)>;
private:
    // CLOCK_MONOTONIC_COARSE is not quite what we want -- to look for stalls,
    // we want thread time, not wall time. Wall time will give false positives
    // if the process is descheduled.
    // For this reason Seastar uses CLOCK_THREAD_CPUTIME_ID in its stall detector.
    // Unfortunately, CLOCK_THREAD_CPUTIME_ID_COARSE does not exist.
    // It's not an important problem, though.
    using clock = utils::coarse_steady_clock;
    struct stats {
        occupancy_stats region_occupancy;
        tracker::stats pool_stats;

        friend stats operator+(const stats& s1, const stats& s2) {
            stats result(s1);
            result += s2;
            return result;
        }
        friend stats operator-(const stats& s1, const stats& s2) {
            stats result(s1);
            result -= s2;
            return result;
        }
        stats& operator+=(const stats& other) {
            region_occupancy += other.region_occupancy;
            pool_stats += other.pool_stats;
            return *this;
        }
        stats& operator-=(const stats& other) {
            region_occupancy -= other.region_occupancy;
            pool_stats -= other.pool_stats;
            return *this;
        }
    };

    clock::duration _duration_threshold;

    const char* _name;

    const is_preemptible _preemptible;
    const size_t _memory_to_release;
    const size_t _segments_to_release;
    const size_t _reserve_goal, _reserve_max;
    tracker::impl& _tracker;
    segment_pool& _segment_pool;
    extra_logger _extra_logs;

    const bool _debug_enabled;
    bool _stall_detected = false;

    size_t _memory_released = 0;

    clock::time_point _start;
    stats _start_stats, _end_stats, _stat_diff;

    clock::duration _duration;

    inline reclaim_timer(const char* name, is_preemptible preemptible, size_t memory_to_release, size_t segments_to_release, tracker::impl& tracker, segment_pool& segment_pool, extra_logger extra_logs);

public:
    inline reclaim_timer(const char* name, is_preemptible preemptible, size_t memory_to_release, size_t segments_to_release, tracker::impl& tracker, extra_logger extra_logs = [](log_level){})
        : reclaim_timer(name, preemptible, memory_to_release, segments_to_release, tracker, tracker.segment_pool(), std::move(extra_logs))
    {}

    inline reclaim_timer(const char* name, is_preemptible preemptible, size_t memory_to_release, size_t segments_to_release, segment_pool& segment_pool, extra_logger extra_logs = [](log_level){})
        : reclaim_timer(name, preemptible, memory_to_release, segments_to_release, segment_pool.tracker(), segment_pool, std::move(extra_logs))
    {}

    inline ~reclaim_timer();

    size_t set_memory_released(size_t memory_released) noexcept {
        return this->_memory_released = memory_released;
    }

private:
    void sample_stats(stats& data);
    template <typename T>
    void log_if_changed(log_level level, const char* name, T before, T now) const noexcept {
        if (now != before) {
            timing_logger.log(level, "- {}: {:.3f} -> {:.3f}", name, before, now);
        }
    }
    template <typename T>
    void log_if_any(log_level level, const char* name, T value) const noexcept {
        if (value != 0) {
            timing_logger.log(level, "- {}: {}", name, value);
        }
    }
    template <typename T>
    void log_if_any_mem(log_level level, const char* name, T value) const noexcept {
        if (value != 0) {
            timing_logger.log(level, "- {}: {:.3f} MiB", name, (float)value / (1024*1024));
        }
    }

    void report() const noexcept;
};

tracker::stats tracker::statistics() const {
    return _impl->segment_pool().statistics();
}

size_t segment_pool::reclaim_segments(size_t target, is_preemptible preempt) {
    // Reclaimer tries to release segments occupying lower parts of the address
    // space.
    llogger.debug("Trying to reclaim {} segments", target);

    // Reclamation. Migrate segments to higher addresses and shrink segment pool.
    size_t reclaimed_segments = 0;

    reclaim_timer timing_guard("reclaim_segments", preempt, target * segment::size, target, *this, [&] (log_level level) {
        timing_logger.log(level, "- reclaimed {} out of requested {} segments", reclaimed_segments, target);
    });

    // We may fail to reclaim because a region has reclaim disabled (usually because
    // it is in an allocating_section. Failed reclaims can cause high CPU usage
    // if all of the lower addresses happen to be in a reclaim-disabled region (this
    // is somewhat mitigated by the fact that checking for reclaim disabled is very
    // cheap), but worse, failing a segment reclaim can lead to reclaimed memory
    // being fragmented.  This results in the original allocation continuing to fail.
    //
    // To combat that, we limit the number of failed reclaims. If we reach the limit,
    // we fail the reclaim.  The surrounding allocating_section will release the
    // reclaim_lock, and increase reserves, which will result in reclaim being
    // retried with all regions being reclaimable, and succeed in allocating
    // contiguous memory.
    size_t failed_reclaims_allowance = 10;

    for (size_t src_idx = _lsa_owned_segments_bitmap.find_first_set();
            reclaimed_segments != target && src_idx != utils::dynamic_bitset::npos
                    && _free_segments > _current_emergency_reserve_goal;
            src_idx = _lsa_owned_segments_bitmap.find_next_set(src_idx)) {
        auto src = segment_from_idx(src_idx);
        if (!_lsa_free_segments_bitmap.test(src_idx)) {
            if (!compact_segment(src)) {
                if (--failed_reclaims_allowance == 0) {
                    break;
                }
                continue;
            }
        }
        _lsa_free_segments_bitmap.clear(src_idx);
        _lsa_owned_segments_bitmap.clear(src_idx);
        _store.free_segment(src);
        ++reclaimed_segments;
        --_free_segments;
        if (preempt && need_preempt()) {
            break;
        }
    }

    llogger.debug("Reclaimed {} segments (requested {})", reclaimed_segments, target);
    timing_guard.set_memory_released(reclaimed_segments * segment::size);
    return reclaimed_segments;
}

segment* segment_pool::allocate_segment(size_t reserve)
{
    //
    // When allocating a segment we want to avoid:
    //  - LSA and general-purpose allocator shouldn't constantly fight each
    //    other for every last bit of memory
    //
    // allocate_segment() always works with LSA reclaimer disabled.
    // 1. Firstly, the algorithm tries to allocate an lsa-owned but free segment
    // 2. If no free segmented is available, a new segment is allocated from the
    //    system allocator. However, if the free memory is below set threshold
    //    this step is skipped.
    // 3. Finally, the algorithm ties to compact and evict data stored in LSA
    //    memory in order to reclaim enough segments.
    //
    do {
        tracker_reclaimer_lock rl(_tracker);
        if (_free_segments > reserve) {
            auto free_idx = _lsa_free_segments_bitmap.find_last_set();
            _lsa_free_segments_bitmap.clear(free_idx);
            auto seg = segment_from_idx(free_idx);
            --_free_segments;
            return seg;
        }
        if (can_allocate_more_segments()) {
            memory::disable_abort_on_alloc_failure_temporarily dfg;
            auto [seg, idx] = _store.allocate_segment();
            if (!seg) {
                continue;
            }
            _lsa_owned_segments_bitmap.set(idx);
            return seg;
        }
    } while (_tracker.compact_and_evict(reserve, _tracker.reclamation_step() * segment::size, is_preemptible::no));
    return nullptr;
}

void segment_pool::deallocate_segment(segment* seg) noexcept
{
    SCYLLA_ASSERT(_lsa_owned_segments_bitmap.test(idx_from_segment(seg)));
    _lsa_free_segments_bitmap.set(idx_from_segment(seg));
    _free_segments++;
}

void segment_pool::refill_emergency_reserve() {
    try {
        ensure_free_segments(_emergency_reserve_max);
    } catch (const std::bad_alloc&) {
        throw bad_alloc(format("failed to refill emergency reserve of {} (have {} free segments)", _emergency_reserve_max, _free_segments));
    }
}

void segment_pool::ensure_free_segments(size_t n_segments) {
    while (_free_segments < n_segments) {
        auto seg = allocate_segment(n_segments);
        if (!seg) {
            throw std::bad_alloc();
        }
        ++_segments_in_use;
        free_segment(seg);
    }
}

segment*
segment_pool::containing_segment(const void* obj) noexcept {
    auto addr = reinterpret_cast<uintptr_t>(obj);
    auto offset = addr & (segment::size - 1);
    auto seg = reinterpret_cast<segment*>(addr - offset);
    auto index = idx_from_segment(seg);
    if (index == segment_npos) {
        return nullptr;
    }
    auto& desc = _segments[index];
    if (desc._region) {
        return seg;
    } else {
        return nullptr;
    }
}

segment*
segment_pool::segment_from(const segment_descriptor& desc) noexcept {
    SCYLLA_ASSERT(desc._region);
    auto index = &desc - &_segments[0];
    return segment_from_idx(index);
}

segment*
segment_pool::allocate_or_fallback_to_reserve() {
    auto seg = allocate_segment(_current_emergency_reserve_goal);
    if (!seg) {
        _allocation_failure_flag = true;
        throw bad_alloc(format("failed to allocate segment (_current_emergency_reserve_goal={})", _current_emergency_reserve_goal));
    }
    return seg;
}

segment*
segment_pool::new_segment(region::impl* r) {
    auto seg = allocate_or_fallback_to_reserve();
    ++_segments_in_use;
    segment_descriptor& desc = descriptor(seg);
    desc.set_free_space(segment::size);
    desc.set_kind(segment_kind::regular);
    desc._region = r;
    return seg;
}

void segment_pool::free_segment(segment* seg) noexcept {
    free_segment(seg, descriptor(seg));
}

void segment_pool::free_segment(segment* seg, segment_descriptor& desc) noexcept {
    llogger.trace("Releasing segment {}", fmt::ptr(seg));
    desc._region = nullptr;
    deallocate_segment(seg);
    --_segments_in_use;
}

segment_pool::segment_pool(tracker::impl& tracker)
    : _tracker(tracker)
    , _segments(max_segments())
    , _lsa_owned_segments_bitmap(max_segments())
    , _lsa_free_segments_bitmap(max_segments())
{
}

void segment_pool::prime(size_t available_memory, size_t min_free_memory) {
    auto old_emergency_reserve = std::exchange(_emergency_reserve_max, std::numeric_limits<size_t>::max());
    try {
        // Allocate all of memory so that we occupy the top part. Afterwards, we'll start
        // freeing from the bottom.
        _store.non_lsa_reserve = 0;
        refill_emergency_reserve();
    } catch (std::bad_alloc&) {
        _emergency_reserve_max = old_emergency_reserve;
    }
    // We want to leave more free memory than just min_free_memory() in order to reduce
    // the frequency of expensive segment-migrating reclaim() called by the seastar allocator.
    size_t min_gap = 1 * 1024 * 1024;
    size_t max_gap = 32 * 1024 * 1024;
    size_t gap = std::min(max_gap, std::max(available_memory / 16, min_gap));
    _store.non_lsa_reserve = min_free_memory + gap;
    // Since the reclaimer is not yet in place, free some low memory for general use
    reclaim_segments(_store.non_lsa_reserve / segment::size, is_preemptible::no);
}

void segment_pool::use_standard_allocator_segment_pool_backend(size_t available_memory) {
    if (_segments_in_use) {
        throw std::runtime_error("cannot change segment store backend after segments are in use");
    }
    _store.use_standard_allocator_segment_pool_backend(available_memory);
    _segments = std::vector<segment_descriptor>(max_segments());
    _lsa_owned_segments_bitmap = utils::dynamic_bitset(max_segments());
    _lsa_free_segments_bitmap = utils::dynamic_bitset(max_segments());
}

inline void segment_pool::on_segment_compaction(size_t used_size) noexcept {
    _stats.segments_compacted++;
    _stats.memory_compacted += used_size;
}

inline void segment_pool::on_memory_allocation(size_t size) noexcept {
    _stats.memory_allocated += size;
    ++_stats.num_allocations;
}

inline void segment_pool::on_memory_deallocation(size_t size) noexcept {
    _stats.memory_freed += size;
}

inline void segment_pool::on_memory_eviction(size_t size) noexcept {
    _stats.memory_evicted += size;
}

// RAII wrapper to maintain segment_pool::current_emergency_reserve_goal()
class segment_pool::reservation_goal {
    segment_pool& _sp;
    size_t _old_goal;
public:
    reservation_goal(segment_pool& sp, size_t goal) noexcept
            : _sp(sp), _old_goal(_sp.current_emergency_reserve_goal()) {
        _sp.set_current_emergency_reserve_goal(goal);
    }
    ~reservation_goal() {
        _sp.set_current_emergency_reserve_goal(_old_goal);
    }
};

size_t segment_pool::segments_in_use() const noexcept {
    return _segments_in_use;
}

reclaim_timer::reclaim_timer(const char* name, is_preemptible preemptible, size_t memory_to_release, size_t segments_to_release, tracker::impl& tracker, segment_pool& segment_pool, extra_logger extra_logs)
    : _duration_threshold(
            // We only report reclaim stalls when their measured duration is
            // bigger than the threshold by at least one measurement error
            // (clock resolution). This prevents false positives.
            //
            // Explanation for the 10us: The clock value is not always an
            // integral multiply of its resolution. In the case of coarse
            // clocks, resolution only describes the frequency of syncs with
            // the hardware clock -- no effort is made to round the values to
            // resolution. Therefore, tick durations vary slightly in both
            // directions. We subtract something slightly bigger than these
            // variations, to accommodate blocked-reactor-notify-ms values which
            // are multiplies of resolution.
            // E.g. with kernel CONFIG_HZ=250, coarse clock resolution is 4ms.
            // If also we also have blocked-reactor-notify-ms=4, then we would
            // like to report two-tick stalls, since they have durations of
            // 4ms-8ms. But two-tick durations can be just slightly smaller
            // than 8ms (e.g. 7999us) due to the inaccuracy. So we set the
            // threshold not to (blocked_reactor_notify_ms + resolution) = 8000us,
            // but to (blocked_reactor_notify_ms + resolution - 10us) = 7990us,
            // to account for this.
            engine().get_blocked_reactor_notify_ms() + std::max(0ns, clock::get_resolution() - 10us))
    , _name(name)
    , _preemptible(preemptible)
    , _memory_to_release(memory_to_release)
    , _segments_to_release(segments_to_release)
    , _reserve_goal(segment_pool.current_emergency_reserve_goal())
    , _reserve_max(segment_pool.emergency_reserve_max())
    , _tracker(tracker)
    , _segment_pool(segment_pool)
    , _extra_logs(std::move(extra_logs))
    , _debug_enabled(timing_logger.is_enabled(logging::log_level::debug))
{
    if (!_tracker.try_set_active_timer(*this)) {
        return;
    }

    _start = clock::now();
    sample_stats(_start_stats);
}

reclaim_timer::~reclaim_timer() {
    if (!_tracker.try_reset_active_timer(*this)) {
        return;
    }

    _duration = clock::now() - _start;
    _stall_detected = _duration >= _duration_threshold;
    if (_debug_enabled || _stall_detected) {
        sample_stats(_end_stats);
        _stat_diff = _end_stats - _start_stats;
        report();
    }
}

void reclaim_timer::sample_stats(stats& data) {
    if (_debug_enabled) {
        data.region_occupancy = _tracker.region_occupancy();
    }
    data.pool_stats = _segment_pool.statistics();
}

void reclaim_timer::report() const noexcept {
    // The logger can allocate (and will recover from allocation failure), and
    // we're in a memory-sensitive situation here and allocation can easily fail.
    // Prevent --abort-on-seastar-bad-alloc from crashing us in a situation that
    // we're likely to recover from, by reclaiming more.
    auto guard = memory::disable_abort_on_alloc_failure_temporarily();

    auto time_level = _stall_detected ? log_level::warn : log_level::debug;
    auto info_level = _stall_detected ? log_level::info : log_level::debug;
    auto MiB = 1024*1024;
    auto msg_extra = extra_msg_when_stall_detected(_stall_detected,
                                                   _stall_detected ? current_backtrace() : saved_backtrace{});

    timing_logger.log(time_level, "{} took {} us, trying to release {:.3f} MiB {}preemptibly, reserve: {{goal: {}, max: {}}}{}",
                        _name, (_duration + 500ns) / 1us, (float)_memory_to_release / MiB, _preemptible ? "" : "non-",
                        _reserve_goal, _reserve_max,
                        msg_extra);
    log_if_any(info_level, "segments to release", _segments_to_release);
    _extra_logs(info_level);
    if (_memory_released > 0) {
        auto bytes_per_second =
            static_cast<float>(_memory_released) / std::chrono::duration_cast<std::chrono::duration<float>>(_duration).count();
        timing_logger.log(info_level, "- reclamation rate = {} MiB/s", format("{:.3f}", bytes_per_second / MiB));
    }

    if (_debug_enabled) {
        log_if_changed(info_level, "occupancy of regions",
                        _start_stats.region_occupancy.used_fraction(), _end_stats.region_occupancy.used_fraction());
    }

    auto pool_diff = _stat_diff.pool_stats;
    log_if_any_mem(info_level, "evicted memory", pool_diff.memory_evicted);
    log_if_any(info_level, "compacted segments", pool_diff.segments_compacted);
    log_if_any_mem(info_level, "compacted memory", pool_diff.memory_compacted);
    log_if_any_mem(info_level, "allocated memory", pool_diff.memory_allocated);
}

region_listener::~region_listener() = default;

//
// For interface documentation see logalloc::region and allocation_strategy.
//
// Allocation dynamics.
//
// Objects are allocated inside fixed-size segments. Objects don't cross
// segment boundary. Active allocations are served from a single segment using
// bump-the-pointer method. That segment is called the active segment. When
// active segment fills up, it is closed. Closed segments are kept in a heap
// which orders them by occupancy. As objects are freed, the segment become
// sparser and are eventually released. Objects which are too large are
// allocated using standard allocator.
//
// Segment layout.
//
// Objects in a segment are laid out sequentially. Each object is preceded by
// a descriptor (see object_descriptor). Object alignment is respected, so if
// there is a gap between the end of current object and the next object's
// descriptor, a trunk of the object descriptor is left right after the
// current object with the flags byte indicating the amount of padding.
//
// Per-segment metadata is kept in a separate array, managed by segment_pool
// object.
//
class region_impl final : public basic_region_impl {
    // Serialized object descriptor format:
    //  byte0 byte1 ... byte[n-1]
    //  bit0-bit5: ULEB64 significand
    //  bit6: 1 iff first byte
    //  bit7: 1 iff last byte
    // This format allows decoding both forwards and backwards (by scanning for bit7/bit6 respectively);
    // backward decoding is needed to recover the descriptor from the object pointer when freeing.
    //
    // Significand interpretation (value = n):
    //     even:  dead object, size n/2 (including descriptor)
    //     odd:   migrate_fn_type at index n/2, from static_migrators()
    class object_descriptor {
    private:
        uint32_t _n;
    private:
        explicit object_descriptor(uint32_t n) noexcept : _n(n) {}
    public:
        object_descriptor(allocation_strategy::migrate_fn migrator) noexcept
                : _n(migrator->index() * 2 + 1)
        { }

        static object_descriptor make_dead(size_t size) noexcept {
            return object_descriptor(size * 2);
        }

        allocation_strategy::migrate_fn migrator() const noexcept {
            return static_migrators()[_n / 2];
        }

        uint8_t alignment() const noexcept {
            return migrator()->align();
        }

        // excluding descriptor
        segment::size_type live_size(const void* obj) const noexcept {
            return migrator()->size(obj);
        }

        // including descriptor
        segment::size_type dead_size() const noexcept {
            return _n / 2;
        }

        bool is_live() const noexcept {
            return (_n & 1) == 1;
        }

        segment::size_type encoded_size() const noexcept {
            return utils::uleb64_encoded_size(_n); // 0 is illegal
        }

        void encode(char*& pos) const noexcept {
            utils::uleb64_encode(pos, _n, poison<char>, unpoison);
        }

        // non-canonical encoding to allow padding (for alignment); encoded_size must be
        // sufficient (greater than this->encoded_size()), _n must be the migrator's
        // index() (i.e. -- suitable for express encoding)
        void encode(char*& pos, size_t encoded_size, size_t size) const noexcept {
            utils::uleb64_express_encode(pos, _n, encoded_size, size, poison<char>, unpoison);
        }

        static object_descriptor decode_forwards(const char*& pos) noexcept {
            return object_descriptor(utils::uleb64_decode_forwards(pos, poison<char>, unpoison));
        }

        static object_descriptor decode_backwards(const char*& pos) noexcept {
            return object_descriptor(utils::uleb64_decode_bacwards(pos, poison<char>, unpoison));
        }

        friend std::ostream& operator<<(std::ostream& out, const object_descriptor& desc) {
            if (!desc.is_live()) {
                return out << format("{{free {:d}}}", desc.dead_size());
            } else {
                auto m = desc.migrator();
                auto x = reinterpret_cast<uintptr_t>(&desc) + sizeof(desc);
                x = align_up(x, m->align());
                auto obj = reinterpret_cast<const void*>(x);
                return out << format("{{migrator={:p}, alignment={:d}, size={:d}}}",
                                      (void*)m, m->align(), m->size(obj));
            }
        }
    };
private: // lsa_buffer allocator
    segment* _buf_active = nullptr;
    size_t _buf_active_offset;
    static constexpr size_t buf_align = 4096; // All lsa_buffer:s will have addresses aligned to this value.
    // Emergency storage to ensure forward progress during segment compaction,
    // by ensuring that _buf_pointers allocation inside new_buf_active() does not fail.
    std::vector<entangled> _buf_ptrs_for_compact_segment;
private:
    region* _region = nullptr;
    region_listener* _listener = nullptr;
    segment* _active = nullptr;
    size_t _active_offset;
    segment_descriptor_hist _segment_descs; // Contains only closed segments
    occupancy_stats _closed_occupancy;
    occupancy_stats _non_lsa_occupancy;
    // This helps us updating out region_listener*. That's because we call update before
    // we have a chance to update the occupancy stats - mainly because at this point we don't know
    // what will we do with the new segment. Also, because we are not ever interested in the
    // fraction used, we'll keep it as a scalar and convert when we need to present it as an
    // occupancy. We could actually just present this as a scalar as well and never use occupancies,
    // but consistency is good.
    size_t _evictable_space = 0;
    // This is a mask applied to _evictable_space with bitwise-and before it's returned from evictable_space().
    // Used for forcing the result to zero without using conditionals.
    size_t _evictable_space_mask = std::numeric_limits<size_t>::max();
    bool _evictable = false;
    region_sanitizer _sanitizer;
    uint64_t _id;
    eviction_fn _eviction_fn;
private:
    struct compaction_lock {
        region_impl& _region;
        bool _prev;
        compaction_lock(region_impl& r) noexcept
            : _region(r)
            , _prev(r._reclaiming_enabled)
        {
            _region._reclaiming_enabled = false;
        }
        ~compaction_lock() {
            _region._reclaiming_enabled = _prev;
        }
    };

    void* alloc_small(const object_descriptor& desc, segment::size_type size, size_t alignment) {
        if (!_active) {
            _active = new_segment();
            _active_offset = 0;
        }

        auto desc_encoded_size = desc.encoded_size();

        size_t obj_offset = align_up_for_asan(align_up(_active_offset + desc_encoded_size, alignment));
        if (obj_offset + size > segment::size) {
            close_and_open();
            return alloc_small(desc, size, alignment);
        }

        auto old_active_offset = _active_offset;
        auto pos = _active->at<char>(_active_offset);
        // Use non-canonical encoding to allow for alignment pad
        desc.encode(pos, obj_offset - _active_offset, size);
        unpoison(pos, size);
        _active_offset = obj_offset + size;

        // Align the end of the value so that the next descriptor is aligned
        _active_offset = align_up_for_asan(_active_offset);
        segment_pool().descriptor(_active).record_alloc(_active_offset - old_active_offset);
        return pos;
    }

    template<typename Func>
    requires std::is_invocable_r_v<void, Func, const object_descriptor*, void*, size_t>
    void for_each_live(segment* seg, Func&& func) {
        // scylla-gdb.py:scylla_lsa_segment is coupled with this implementation.
        auto pos = align_up_for_asan(seg->at<const char>(0));
        while (pos < seg->at<const char>(segment::size)) {
            auto old_pos = pos;
            const auto desc = object_descriptor::decode_forwards(pos);
            if (desc.is_live()) {
                auto size = desc.live_size(pos);
                func(&desc, const_cast<char*>(pos), size);
                pos += size;
            } else {
                pos = old_pos + desc.dead_size();
            }
            pos = align_up_for_asan(pos);
        }
    }

    void close_active() {
        if (!_active) {
            return;
        }
        if (_active_offset < segment::size) {
            auto desc = object_descriptor::make_dead(segment::size - _active_offset);
            auto pos =_active->at<char>(_active_offset);
            desc.encode(pos);
        }
        auto& desc = segment_pool().descriptor(_active);
        llogger.trace("Closing segment {}, used={}, waste={} [B]", fmt::ptr(_active), desc.occupancy(), segment::size - _active_offset);
        _closed_occupancy += desc.occupancy();

        _segment_descs.push(desc);
        _active = nullptr;
    }

    void close_buf_active() {
        if (!_buf_active) {
            return;
        }
        auto& desc = segment_pool().descriptor(_buf_active);
        llogger.trace("Closing buf segment {}, used={}, waste={} [B]", fmt::ptr(_buf_active), desc.occupancy(), segment::size - _buf_active_offset);
        _closed_occupancy += desc.occupancy();

        _segment_descs.push(desc);
        _buf_active = nullptr;
    }

    void free_segment(segment_descriptor& desc) noexcept {
        free_segment(segment_pool().segment_from(desc), desc);
    }

    void free_segment(segment* seg) noexcept {
        free_segment(seg, segment_pool().descriptor(seg));
    }

    void free_segment(segment* seg, segment_descriptor& desc) noexcept {
        segment_pool().free_segment(seg, desc);
        if (_listener) {
            _evictable_space -= segment_size;
            _listener->decrease_usage(_region, -segment::size);
        }
    }

    segment* new_segment() {
        segment* seg = segment_pool().new_segment(this);
        if (_listener) {
            _evictable_space += segment_size;
            _listener->increase_usage(_region, segment::size);
        }
        return seg;
    }

    lsa_buffer alloc_buf(size_t buf_size) {
        // Note: Can be re-entered from allocation sites below due to memory reclamation which
        // invokes segment compaction.
        static_assert(segment::size % buf_align == 0);
        if (buf_size > segment::size) {
            throw_with_backtrace<std::runtime_error>(format("Buffer size {} too large", buf_size));
        }

        if (_buf_active_offset + buf_size > segment::size) {
            close_buf_active();
        }

        if (!_buf_active) {
            new_buf_active();
        }

        lsa_buffer ptr;
        ptr._buf = _buf_active->at<char>(_buf_active_offset);
        ptr._size = buf_size;
        unpoison(ptr._buf, buf_size);

        segment_descriptor& desc = segment_pool().descriptor(_buf_active);
        ptr._desc = &desc;
        desc._buf_pointers.emplace_back(entangled::make_paired_with(ptr._link));
        auto alloc_size = align_up(buf_size, buf_align);
        desc.record_alloc(alloc_size);
        _buf_active_offset += alloc_size;

        return ptr;
    }

    void free_buf(lsa_buffer& buf) noexcept {
        segment_descriptor &desc = *buf._desc;
        segment *seg = segment_pool().segment_from(desc);

        if (seg != _buf_active) {
            _closed_occupancy -= desc.occupancy();
        }

        auto alloc_size = align_up(buf._size, buf_align);
        desc.record_free(alloc_size);
        poison(buf._buf, buf._size);

        // Pack links so that segment compaction only has to walk live objects.
        // This procedure also ensures that the link for buf is destroyed, either
        // by replacing it with the last entangled, or by popping it from the back
        // if it is the last element.
        // Moving entangled links around is fine so we can move last_link.
        entangled& last_link = desc._buf_pointers.back();
        entangled& buf_link = *buf._link.get();
        std::swap(last_link, buf_link);
        desc._buf_pointers.pop_back();

        if (seg != _buf_active) {
            if (desc.is_empty()) {
                SCYLLA_ASSERT(desc._buf_pointers.empty());
                _segment_descs.erase(desc);
                desc._buf_pointers = std::vector<entangled>();
                free_segment(seg, desc);
            } else {
                _segment_descs.adjust_up(desc);
                _closed_occupancy += desc.occupancy();
            }
        }
    }

    void compact_segment_locked(segment* seg, segment_descriptor& desc) noexcept {
        auto seg_occupancy = desc.occupancy();
        llogger.debug("Compacting segment {} from region {}, {}", fmt::ptr(seg), id(), seg_occupancy);

        ++_invalidate_counter;

        if (desc.kind() == segment_kind::bufs) {
            // This will free the storage of _buf_ptrs_for_compact_segment
            // making sure that alloc_buf() makes progress.
            // Also, empties desc._buf_pointers, making it back a generic segment, which
            // we need to do before freeing it.
            _buf_ptrs_for_compact_segment = std::move(desc._buf_pointers);
            for (entangled& e : _buf_ptrs_for_compact_segment) {
                if (e) {
                    lsa_buffer* old_ptr = e.get(&lsa_buffer::_link);
                    SCYLLA_ASSERT(&desc == old_ptr->_desc);
                    lsa_buffer dst = alloc_buf(old_ptr->_size);
                    memcpy(dst._buf, old_ptr->_buf, dst._size);
                    old_ptr->_link = std::move(dst._link);
                    old_ptr->_buf = dst._buf;
                    old_ptr->_desc = dst._desc;
                }
            }
        } else {
            for_each_live(seg, [this](const object_descriptor *desc, void *obj, size_t size) {
                auto dst = alloc_small(*desc, size, desc->alignment());
                _sanitizer.on_migrate(obj, size, dst);
                desc->migrator()->migrate(obj, dst, size);
            });
        }

        free_segment(seg, desc);
        segment_pool().on_segment_compaction(seg_occupancy.used_space());
    }

    void close_and_open() {
        segment* new_active = new_segment();
        close_active();
        _active = new_active;
        _active_offset = 0;
    }

    void new_buf_active() {
        std::vector<entangled> ptrs;
        ptrs.reserve(segment::size / buf_align);
        segment* new_active = new_segment();
        if (_buf_active) [[unlikely]] {
            // Memory allocation above could allocate active buffer during segment compaction.
            close_buf_active();
        }
        SCYLLA_ASSERT((uintptr_t)new_active->at(0) % buf_align == 0);
        segment_descriptor& desc = segment_pool().descriptor(new_active);
        desc._buf_pointers = std::move(ptrs);
        desc.set_kind(segment_kind::bufs);
        _buf_active = new_active;
        _buf_active_offset = 0;
    }

    static uint64_t next_id() noexcept {
        static std::atomic<uint64_t> id{0};
        return id.fetch_add(1);
    }
    struct unlisten_temporarily {
        region_impl* impl;
        region_listener* listener;
        explicit unlisten_temporarily(region_impl* impl)
                : impl(impl), listener(impl->_listener) {
            if (listener) {
                listener->del(impl->_region);
            }
        }
        ~unlisten_temporarily() {
            if (listener) {
                listener->add(impl->_region);
            }
        }
    };

public:
    explicit region_impl(tracker& tracker, region* region)
        : basic_region_impl(tracker), _region(region), _sanitizer(tracker.get_impl().sanitizer_report_backtrace()), _id(next_id())
    {
        _buf_ptrs_for_compact_segment.reserve(segment::size / buf_align);
        _preferred_max_contiguous_allocation = max_managed_object_size;
        tracker_instance._impl->register_region(this);
    }

    virtual ~region_impl() {
        _sanitizer.on_region_destruction();

        _tracker.get_impl().unregister_region(this);

        while (!_segment_descs.empty()) {
            auto& desc = _segment_descs.one_of_largest();
            _segment_descs.pop_one_of_largest();
            SCYLLA_ASSERT(desc.is_empty());
            free_segment(desc);
        }
        _closed_occupancy = {};
        if (_active) {
            SCYLLA_ASSERT(segment_pool().descriptor(_active).is_empty());
            free_segment(_active);
            _active = nullptr;
        }
        if (_buf_active) {
            SCYLLA_ASSERT(segment_pool().descriptor(_buf_active).is_empty());
            free_segment(_buf_active);
            _buf_active = nullptr;
        }
    }

    region_impl(region_impl&&) = delete;
    region_impl(const region_impl&) = delete;

    logalloc::segment_pool& segment_pool() const {
        return _tracker.get_impl().segment_pool();
    }

    bool empty() const noexcept {
        return occupancy().used_space() == 0;
    }

    void listen(region_listener* listener) {
        _listener = listener;
        _listener->add(_region);
    }

    void unlisten() {
        // _listener may have been removed be merge(), so check for that.
        // Yes, it's awkward, we should have the caller unlisten before merge
        // to remove implicit behavior.
        if (_listener) {
            _listener->del(_region);
            _listener = nullptr;
        }
    }

    void moved(region* new_region) {
        if (_listener) {
            _listener->moved(_region, new_region);
        }
        _region = new_region;
    }

    // Note: allocation is disallowed in this path
    // since we don't instantiate reclaiming_lock
    // while traversing _regions
    occupancy_stats occupancy() const noexcept {
        occupancy_stats total = _non_lsa_occupancy;
        total += _closed_occupancy;
        if (_active) {
            total += segment_pool().descriptor(_active).occupancy();
        }
        if (_buf_active) {
            total += segment_pool().descriptor(_buf_active).occupancy();
        }
        return total;
    }

    occupancy_stats compactible_occupancy() const noexcept {
        return _closed_occupancy;
    }

    occupancy_stats evictable_occupancy() const noexcept {
        return occupancy_stats(0, _evictable_space & _evictable_space_mask);
    }

    void ground_evictable_occupancy() {
        _evictable_space_mask = 0;
        if (_listener) {
            _listener->decrease_evictable_usage(_region);
        }
    }

    //
    // Returns true if this region can be compacted and compact() will make forward progress,
    // so that this will eventually stop:
    //
    //    while (is_compactible()) { compact(); }
    //
    bool is_compactible() const noexcept {
        return _reclaiming_enabled
            // We require 2 segments per allocation segregation group to ensure forward progress during compaction.
            // There are currently two fixed groups, one for the allocation_strategy implementation and one for lsa_buffer:s.
            && (_closed_occupancy.free_space() >= 4 * segment::size)
            && _segment_descs.contains_above_min();
    }

    bool is_idle_compactible() const noexcept {
        return is_compactible();
    }

    virtual void* alloc(allocation_strategy::migrate_fn migrator, size_t size, size_t alignment) override {
        compaction_lock _(*this);
        memory::on_alloc_point();
        auto& pool = segment_pool();
        pool.on_memory_allocation(size);
        if (size > max_managed_object_size) {
            auto ptr = standard_allocator().alloc(migrator, size + sizeof(non_lsa_object_cookie), alignment);
            // This isn't very acurrate, the correct free_space value would be
            // malloc_usable_size(ptr) - size, but there is no way to get
            // the exact object size at free.
            auto allocated_size = malloc_usable_size(ptr);
            new ((char*)ptr + allocated_size - sizeof(non_lsa_object_cookie)) non_lsa_object_cookie();
            _non_lsa_occupancy += occupancy_stats(0, allocated_size);
            if (_listener) {
                 _evictable_space += allocated_size;
                _listener->increase_usage(_region, allocated_size);
            }
            pool.add_non_lsa_memory_in_use(allocated_size);
            return ptr;
        } else {
            auto ptr = alloc_small(object_descriptor(migrator), (segment::size_type) size, alignment);
            _sanitizer.on_allocation(ptr, size);
            return ptr;
        }
    }

private:
    void on_non_lsa_free(void* obj) noexcept {
        auto allocated_size = malloc_usable_size(obj);
        auto cookie = (non_lsa_object_cookie*)((char*)obj + allocated_size) - 1;
        SCYLLA_ASSERT(cookie->value == non_lsa_object_cookie().value);
        _non_lsa_occupancy -= occupancy_stats(0, allocated_size);
        if (_listener) {
            _evictable_space -= allocated_size;
            _listener->decrease_usage(_region, allocated_size);
        }
        segment_pool().subtract_non_lsa_memory_in_use(allocated_size);
    }
public:
    virtual void free(void* obj) noexcept override {
        compaction_lock _(*this);
        segment* seg = segment_pool().containing_segment(obj);
        if (!seg) {
            on_non_lsa_free(obj);
            standard_allocator().free(obj);
            return;
        }

        auto pos = reinterpret_cast<const char*>(obj);
        auto desc = object_descriptor::decode_backwards(pos);
        free(obj, desc.live_size(obj));
    }

    virtual void free(void* obj, size_t size) noexcept override {
        compaction_lock _(*this);
        auto& pool = segment_pool();
        segment* seg = pool.containing_segment(obj);

        if (!seg) {
            on_non_lsa_free(obj);
            standard_allocator().free(obj, size);
            return;
        }

        _sanitizer.on_free(obj, size);

        segment_descriptor& seg_desc = pool.descriptor(seg);

        auto pos = reinterpret_cast<const char*>(obj);
        auto old_pos = pos;
        auto desc = object_descriptor::decode_backwards(pos);
        auto dead_size = align_up_for_asan(size + (old_pos - pos));
        desc = object_descriptor::make_dead(dead_size);
        auto npos = const_cast<char*>(pos);
        desc.encode(npos);
        poison(pos, dead_size);

        if (seg != _active) {
            _closed_occupancy -= seg_desc.occupancy();
        }

        seg_desc.record_free(dead_size);
        pool.on_memory_deallocation(dead_size);

        if (seg != _active) {
            if (seg_desc.is_empty()) {
                _segment_descs.erase(seg_desc);
                free_segment(seg, seg_desc);
            } else {
                _segment_descs.adjust_up(seg_desc);
                _closed_occupancy += seg_desc.occupancy();
            }
        }
    }

    virtual size_t object_memory_size_in_allocator(const void* obj) const noexcept override {
        segment* seg = segment_pool().containing_segment(obj);

        if (!seg) {
            return standard_allocator().object_memory_size_in_allocator(obj);
        } else {
            auto pos = reinterpret_cast<const char*>(obj);
            auto desc = object_descriptor::decode_backwards(pos);
            return desc.encoded_size() + desc.live_size(obj);
        }
    }

    // Merges another region into this region. The other region is made
    // to refer to this region.
    // Doesn't invalidate references to allocated objects.
    void merge(region_impl& other) noexcept {
        // unlisten_temporarily may allocate via region_listener callbacks, which should not
        // fail, because we have a matching deallocation before that and we don't
        // allocate between them.
        memory::scoped_critical_alloc_section dfg;

        compaction_lock dct1(*this);
        compaction_lock dct2(other);
        unlisten_temporarily ult1(this);
        unlisten_temporarily ult2(&other);

        auto& pool = segment_pool();

        if (_active && pool.descriptor(_active).is_empty()) {
            pool.free_segment(_active);
            _active = nullptr;
        }
        if (!_active) {
            _active = other._active;
            other._active = nullptr;
            _active_offset = other._active_offset;
            if (_active) {
                pool.set_region(_active, this);
            }
        } else {
            other.close_active();
        }
        other.close_buf_active();

        for (auto& desc : other._segment_descs) {
            pool.set_region(desc, this);
        }
        _segment_descs.merge(other._segment_descs);

        _closed_occupancy += other._closed_occupancy;
        _non_lsa_occupancy += other._non_lsa_occupancy;
        other._closed_occupancy = {};
        other._non_lsa_occupancy = {};

        // Make sure both regions will notice a future increment
        // to the reclaim counter
        _invalidate_counter = std::max(_invalidate_counter, other._invalidate_counter);

        _sanitizer.merge(other._sanitizer);
        other._sanitizer = region_sanitizer(_tracker.get_impl().sanitizer_report_backtrace());
    }

    std::unordered_map<std::string, uint64_t> collect_stats() const {
        std::unordered_map<std::string, uint64_t> sizes;
        for (auto& desc : _segment_descs) {
            const_cast<region_impl&>(*this).for_each_live(segment_pool().segment_from(desc), [&sizes] (const object_descriptor* desc, void* obj, size_t size) {
                auto n = desc->migrator()->name();
                if (sizes.contains(n)) {
                    sizes[n] += size;
                } else {
                    sizes.emplace(n, size);
                }
            });
        }
        return sizes;
    }

    // Returns occupancy of the sparsest compactible segment.
    occupancy_stats min_occupancy() const noexcept {
        if (_segment_descs.empty()) {
            return {};
        }
        return _segment_descs.one_of_largest().occupancy();
    }

    // Compacts a single segment, most appropriate for it
    void compact() noexcept {
        compaction_lock _(*this);
        auto& desc = _segment_descs.one_of_largest();
        _segment_descs.pop_one_of_largest();
        _closed_occupancy -= desc.occupancy();
        segment* seg = segment_pool().segment_from(desc);
        compact_segment_locked(seg, desc);
    }

    // Compacts everything. Mainly for testing.
    // Invalidates references to allocated objects.
    void full_compaction() {
        compaction_lock _(*this);
        llogger.debug("Full compaction, {}", occupancy());
        close_and_open();
        close_buf_active();
        segment_descriptor_hist all;
        std::swap(all, _segment_descs);
        _closed_occupancy = {};
        while (!all.empty()) {
            auto& desc = all.one_of_largest();
            all.pop_one_of_largest();
            compact_segment_locked(segment_pool().segment_from(desc), desc);
        }
        llogger.debug("Done, {}", occupancy());
    }

    void compact_segment(segment* seg, segment_descriptor& desc) {
        compaction_lock _(*this);
        if (_active == seg) {
            close_active();
        } else if (_buf_active == seg) {
            close_buf_active();
        }
        _segment_descs.erase(desc);
        _closed_occupancy -= desc.occupancy();
        compact_segment_locked(seg, desc);
    }

    allocation_strategy& allocator() noexcept {
        return *this;
    }

    uint64_t id() const noexcept {
        return _id;
    }

    // Returns true if this pool is evictable, so that evict_some() can be called.
    bool is_evictable() const noexcept {
        return _evictable && _reclaiming_enabled;
    }

    memory::reclaiming_result evict_some() {
        ++_invalidate_counter;
        auto& pool = segment_pool();
        auto freed = pool.statistics().memory_freed;
        auto ret = _eviction_fn();
        pool.on_memory_eviction(pool.statistics().memory_freed - freed);
        return ret;
    }

    void make_not_evictable() noexcept {
        _evictable = false;
        _eviction_fn = {};
    }

    void make_evictable(eviction_fn fn) noexcept {
        _evictable = true;
        _eviction_fn = std::move(fn);
    }

    const eviction_fn& evictor() const noexcept {
        return _eviction_fn;
    }

    // LSA holds an internal "emergency reserve" of free segments that
    // is only "opened" for usage before some critical allocations
    // (in particular: the ones performed during memory compaction)
    // to ensure that they won't fail.
    //
    // Here we hijack this mechanism to let the rest of the application implement
    // some critical sections with infallible LSA allocations.
    //
    // reserve() increments the size of the internal emergency reserve,
    // unreserve() decrements it.
    //
    // When you want to have some critical section that has to do some LSA 
    // allocations infallibly (e.g. to restore some invariants
    // of a LSA-managed data structure in a destructor), you can call reserve()
    // beforehand to ensure that some extra memory will be held unused,
    // and then call unreserve() (with reserve()'s return value as the argument)
    // to make the reserved free segments available to the critical section.
    // 
    uintptr_t reserve(size_t memory) override {
        // We round up the requested reserve to full segments.
        size_t n_segments = (memory + segment::size - 1) >> segment::size_shift;

        auto& pool = segment_pool();
        size_t new_goal = pool.current_emergency_reserve_goal() + n_segments;
        pool.ensure_free_segments(new_goal);
        pool.set_current_emergency_reserve_goal(new_goal);

        static_assert(sizeof(uintptr_t) >= sizeof(size_t));
        return n_segments;
    }

    void unreserve(uintptr_t n_segments) noexcept override {
        auto& pool = segment_pool();
        SCYLLA_ASSERT(pool.current_emergency_reserve_goal() >= n_segments);
        size_t new_goal = pool.current_emergency_reserve_goal() - n_segments;
        pool.set_current_emergency_reserve_goal(new_goal);
    }

    friend class region;
    friend class lsa_buffer;
    friend class region_evictable_occupancy_ascending_less_comparator;
};

lsa_buffer::~lsa_buffer() {
    if (_link) {
        _desc->_region->free_buf(*this);
    }
}

size_t tracker::reclamation_step() const noexcept {
    return _impl->reclamation_step();
}

bool tracker::should_abort_on_bad_alloc() const noexcept {
    return _impl->should_abort_on_bad_alloc();
}

void tracker::configure(const config& cfg) {
    if (cfg.defragment_on_idle) {
        engine().set_idle_cpu_handler([this] (reactor::work_waiting_on_reactor check_for_work) {
            return _impl->compact_on_idle(check_for_work);
        });
    }

    _impl->set_reclamation_step(cfg.lsa_reclamation_step);
    if (cfg.abort_on_lsa_bad_alloc) {
        _impl->enable_abort_on_bad_alloc();
    }
    _impl->setup_background_reclaim(cfg.background_reclaim_sched_group);
    _impl->set_sanitizer_report_backtrace(cfg.sanitizer_report_backtrace);
}

memory::reclaiming_result tracker::reclaim(seastar::memory::reclaimer::request r) {
    return reclaim(std::max(r.bytes_to_reclaim, _impl->reclamation_step() * segment::size))
           ? memory::reclaiming_result::reclaimed_something
           : memory::reclaiming_result::reclaimed_nothing;
}

region::region()
    : _impl(make_shared<impl>(shard_tracker(), this))
{ }

void
region::listen(region_listener* listener) {
    get_impl().listen(listener);
}

void
region::unlisten() {
    get_impl().unlisten();
}

region_impl& region::get_impl() noexcept {
    return *static_cast<region_impl*>(_impl.get());
}
const region_impl& region::get_impl() const noexcept {
    return *static_cast<const region_impl*>(_impl.get());
}

region::region(region&& other) noexcept
    : _impl(std::move(other._impl))
{
    if (_impl) {
        auto r_impl = static_cast<region_impl*>(_impl.get());
        r_impl->moved(this);
    }
}

region& region::operator=(region&& other) noexcept {
    if (this == &other || _impl == other._impl) {
        return *this;
    }
    if (_impl) {
        unlisten();
    }
    this->_impl = std::move(other._impl);
    if (_impl) {
        auto r_impl = static_cast<region_impl*>(_impl.get());
        r_impl->moved(this);
    }
    return *this;
}

region::~region() {
    if (_impl) {
        unlisten();
    }
}

occupancy_stats region::occupancy() const noexcept {
    return get_impl().occupancy();
}

lsa_buffer region::alloc_buf(size_t buffer_size) {
    return get_impl().alloc_buf(buffer_size);
}

void region::merge(region& other) noexcept {
    if (_impl != other._impl) {
        auto& other_impl = other.get_impl();
        // Not very generic, but we know that post-merge the caller
        // (row_cache) isn't interested in listening, and one region
        // can't have many listeners.
        other_impl.unlisten();
        get_impl().merge(other_impl);
        other._impl = _impl;
    }
}

void region::full_compaction() {
    get_impl().full_compaction();
}

memory::reclaiming_result region::evict_some() {
    if (get_impl().is_evictable()) {
        return get_impl().evict_some();
    }
    return memory::reclaiming_result::reclaimed_nothing;
}

void region::make_evictable(eviction_fn fn) noexcept {
    get_impl().make_evictable(std::move(fn));
}

void region::ground_evictable_occupancy() {
    get_impl().ground_evictable_occupancy();
}

occupancy_stats region::evictable_occupancy() const noexcept {
    return get_impl().evictable_occupancy();
}

const eviction_fn& region::evictor() const noexcept {
    return get_impl().evictor();
}

uint64_t region::id() const noexcept {
    return get_impl().id();
}

std::unordered_map<std::string, uint64_t> region::collect_stats() const {
    return get_impl().collect_stats();
}

occupancy_stats tracker::impl::global_occupancy() const noexcept {
    return occupancy_stats(_segment_pool->total_free_memory(), _segment_pool->total_memory_in_use());
}

// Note: allocation is disallowed in this path
// since we don't instantiate reclaiming_lock
// while traversing _regions
occupancy_stats tracker::impl::region_occupancy() const noexcept {
    occupancy_stats total{};
    for (auto&& r: _regions) {
        total += r->occupancy();
    }
    return total;
}

occupancy_stats tracker::impl::occupancy() const noexcept {
    auto occ = region_occupancy();
    {
        auto s = _segment_pool->free_segments() * segment::size;
        occ += occupancy_stats(s, s);
    }
    return occ;
}

size_t tracker::impl::non_lsa_used_space() const noexcept {
#ifdef SEASTAR_DEFAULT_ALLOCATOR
    return 0;
#else
    auto free_space_in_lsa = _segment_pool->free_segments() * segment_size;
    return memory::stats().allocated_memory() - region_occupancy().total_space() - free_space_in_lsa;
#endif
}

void tracker::impl::reclaim_all_free_segments()
{
    llogger.debug("Reclaiming all free segments");
    _segment_pool->reclaim_all_free_segments();
    llogger.debug("Reclamation done");
}

void tracker::impl::full_compaction() {
    reclaiming_lock _(*this);

    llogger.debug("Full compaction on all regions, {}", region_occupancy());

    for (region_impl* r : _regions) {
        if (r->reclaiming_enabled()) {
            r->full_compaction();
        }
    }

    llogger.debug("Compaction done, {}", region_occupancy());
}

static void reclaim_from_evictable(region::impl& r, size_t target_mem_in_use, is_preemptible preempt) {
    llogger.debug("reclaim_from_evictable: total_memory_in_use={} target={}", r.segment_pool().total_memory_in_use(), target_mem_in_use);

    // Before attempting segment compaction, try to evict at least deficit and one segment more so that
    // for workloads in which eviction order matches allocation order we will reclaim full segments
    // without needing to perform expensive compaction.
    auto deficit = r.segment_pool().total_memory_in_use() - target_mem_in_use;
    auto used = r.occupancy().used_space();
    auto used_target = used - std::min(used, deficit + segment::size);

    while (r.segment_pool().total_memory_in_use() > target_mem_in_use) {
        used = r.occupancy().used_space();
        if (used > used_target) {
            llogger.debug("Evicting {} bytes from region {}, occupancy={} in advance",
                    used - used_target, r.id(), r.occupancy());
        } else {
            llogger.debug("Evicting from region {}, occupancy={} until it's compactible", r.id(), r.occupancy());
        }
        while (r.occupancy().used_space() > used_target || !r.is_compactible()) {
            if (r.evict_some() == memory::reclaiming_result::reclaimed_nothing) {
                if (r.is_compactible()) { // Need to make forward progress in case there is nothing to evict.
                    break;
                }
                llogger.debug("Unable to evict more, evicted {} bytes", used - r.occupancy().used_space());
                return;
            }
            if (r.segment_pool().total_memory_in_use() <= target_mem_in_use) {
                llogger.debug("Target met after evicting {} bytes", used - r.occupancy().used_space());
                return;
            }
            if (preempt && need_preempt()) {
                llogger.debug("reclaim_from_evictable preempted");
                return;
            }
        }
        // If there are many compactible segments, we will keep compacting without
        // entering the eviction loop above. So the preemption check there is not
        // sufficient and we also need to check here.
        //
        // Note that a preemptible reclaim_from_evictable may not do any real progress,
        // but it doesn't need to. Preemptible (background) reclaim is an optimization.
        // If the system is overwhelmed, and reclaim_from_evictable keeps getting
        // preempted without doing any useful work, then eventually memory will be
        // exhausted and reclaim will be called synchronously, without preemption.
        if (preempt && need_preempt()) {
            llogger.debug("reclaim_from_evictable preempted");
            return;
        }
        llogger.debug("Compacting after evicting {} bytes", used - r.occupancy().used_space());
        r.compact();
    }
}

idle_cpu_handler_result tracker::impl::compact_on_idle(work_waiting_on_reactor check_for_work) {
    if (_reclaiming_disabled_depth) {
        return idle_cpu_handler_result::no_more_work;
    }
    reclaiming_lock rl(*this);
    if (_regions.empty()) {
        return idle_cpu_handler_result::no_more_work;
    }
    segment_pool::reservation_goal open_emergency_pool(*_segment_pool, 0);

    auto cmp = [] (region::impl* c1, region::impl* c2) {
        if (c1->is_idle_compactible() != c2->is_idle_compactible()) {
            return !c1->is_idle_compactible();
        }
        return c2->min_occupancy() < c1->min_occupancy();
    };

    std::ranges::make_heap(_regions, cmp);

    while (!check_for_work()) {
        std::ranges::pop_heap(_regions, cmp);
        region::impl* r = _regions.back();

        if (!r->is_idle_compactible()) {
            return idle_cpu_handler_result::no_more_work;
        }

        r->compact();

        std::ranges::push_heap(_regions, cmp);
    }
    return idle_cpu_handler_result::interrupted_by_higher_priority_task;
}

size_t tracker::impl::reclaim(size_t memory_to_release, is_preemptible preempt) {
    if (_reclaiming_disabled_depth) {
        return 0;
    }
    reclaiming_lock rl(*this);
    reclaim_timer timing_guard("reclaim", preempt, memory_to_release, 0, *this);
    return timing_guard.set_memory_released(reclaim_locked(memory_to_release, preempt));
}

size_t tracker::impl::reclaim_locked(size_t memory_to_release, is_preemptible preempt) {
    llogger.debug("reclaim_locked({}, preempt={})", memory_to_release, int(bool(preempt)));
    // Reclamation steps:
    // 1. Try to release free segments from segment pool and emergency reserve.
    // 2. Compact used segments and/or evict data.
    constexpr auto max_bytes = std::numeric_limits<size_t>::max() - segment::size;
    auto segments_to_release = align_up(std::min(max_bytes, memory_to_release), segment::size) >> segment::size_shift;
    auto nr_released = _segment_pool->reclaim_segments(segments_to_release, preempt);
    size_t mem_released = nr_released * segment::size;
    if (mem_released >= memory_to_release) {
        llogger.debug("reclaim_locked() = {}", memory_to_release);
        return memory_to_release;
    }
    if (preempt && need_preempt()) {
        llogger.debug("reclaim_locked() = {}", mem_released);
        return mem_released;
    }

    auto compacted = compact_and_evict_locked(_segment_pool->current_emergency_reserve_goal(), memory_to_release - mem_released, preempt);

    if (compacted == 0) {
        llogger.debug("reclaim_locked() = {}", mem_released);
        return mem_released;
    }

    // compact_and_evict_locked() will not return segments to the standard allocator,
    // so do it here:
    nr_released = _segment_pool->reclaim_segments(compacted / segment::size, preempt);
    mem_released += nr_released * segment::size;

    llogger.debug("reclaim_locked() = {}", mem_released);
    return mem_released;
}

size_t tracker::impl::compact_and_evict(size_t reserve_segments, size_t memory_to_release, is_preemptible preempt) {
    if (_reclaiming_disabled_depth) {
        return 0;
    }
    reclaiming_lock rl(*this);
    return compact_and_evict_locked(reserve_segments, memory_to_release, preempt);
}

size_t tracker::impl::compact_and_evict_locked(size_t reserve_segments, size_t memory_to_release, is_preemptible preempt) {
    llogger.debug("compact_and_evict_locked({}, {}, {})", reserve_segments, memory_to_release, int(bool(preempt)));
    //
    // Algorithm outline.
    //
    // Regions are kept in a max-heap ordered so that regions with
    // sparser segments are picked first. Non-compactible regions will be
    // picked last. In each iteration we try to release one whole segment from
    // the region which has the sparsest segment. We do it until we released
    // enough segments or there are no more regions we can compact.
    //
    // When compaction is not sufficient to reclaim space, we evict data from
    // evictable regions.
    //

    // This may run synchronously with allocation, so we should not allocate
    // memory, otherwise we may get std::bad_alloc. Currently we only allocate
    // in the logger when debug level is enabled. It's disabled during normal
    // operation. Having it is still valuable during testing and in most cases
    // should work just fine even if allocates.

    size_t mem_released = 0;

    size_t mem_in_use = _segment_pool->total_memory_in_use();
    memory_to_release += (reserve_segments - std::min(reserve_segments, _segment_pool->free_segments())) * segment::size;
    auto target_mem = mem_in_use - std::min(mem_in_use, memory_to_release - mem_released);

    llogger.debug("Compacting, requested {} bytes, {} bytes in use, target is {}",
        memory_to_release, mem_in_use, target_mem);

    // Allow dipping into reserves while compacting
    segment_pool::reservation_goal open_emergency_pool(*_segment_pool, 0);

    auto cmp = [] (region::impl* c1, region::impl* c2) {
        if (c1->is_compactible() != c2->is_compactible()) {
            return !c1->is_compactible();
        }
        return c2->min_occupancy() < c1->min_occupancy();
    };

    std::ranges::make_heap(_regions, cmp);

    if (llogger.is_enabled(logging::log_level::debug)) {
        llogger.debug("Occupancy of regions:");
        for (region::impl* r : _regions) {
            llogger.debug(" - {}: min={}, avg={}", r->id(), r->min_occupancy(), r->compactible_occupancy());
        }
    }

    {
        int regions = 0, evictable_regions = 0;
        reclaim_timer timing_guard("compact", preempt, memory_to_release, reserve_segments, *this, [&] (log_level level) {
            timing_logger.log(level, "- processed {} regions: reclaimed from {}, compacted {}",
                              regions, evictable_regions, regions - evictable_regions);
        });
        while (_segment_pool->total_memory_in_use() > target_mem) {
            std::ranges::pop_heap(_regions, cmp);
            region::impl* r = _regions.back();

            if (!r->is_compactible()) {
                llogger.trace("Unable to release segments, no compactible pools.");
                break;
            }
            ++regions;

            // Prefer evicting if average occupancy ratio is above the compaction threshold to avoid
            // overhead of compaction in workloads where allocation order matches eviction order, where
            // we can reclaim memory by eviction only. In some cases the cost of compaction on allocation
            // would be higher than the cost of repopulating the region with evicted items.
            if (r->is_evictable() && r->occupancy().used_space() >= max_used_space_ratio_for_compaction * r->occupancy().total_space()) {
                reclaim_from_evictable(*r, target_mem, preempt);
                ++evictable_regions;
            } else {
                r->compact();
            }

            std::ranges::push_heap(_regions, cmp);

            if (preempt && need_preempt()) {
                break;
            }
        }
    }

    auto released_during_compaction = mem_in_use - _segment_pool->total_memory_in_use();

    if (_segment_pool->total_memory_in_use() > target_mem) {
        int regions = 0, evictable_regions = 0;
        reclaim_timer timing_guard("evict", preempt, memory_to_release, reserve_segments, *this, [&] (log_level level) {
            timing_logger.log(level, "- processed {} regions, reclaimed from {}", regions, evictable_regions);
        });
        llogger.debug("Considering evictable regions.");
        // FIXME: Fair eviction
        for (region::impl* r : _regions) {
            if (preempt && need_preempt()) {
                break;
            }
            ++regions;
            if (r->is_evictable()) {
                ++evictable_regions;
                reclaim_from_evictable(*r, target_mem, preempt);
                if (_segment_pool->total_memory_in_use() <= target_mem) {
                    break;
                }
            }
        }
    }

    mem_released += mem_in_use - _segment_pool->total_memory_in_use();

    llogger.debug("Released {} bytes (wanted {}), {} during compaction",
        mem_released, memory_to_release, released_during_compaction);

    return mem_released;
}

void tracker::impl::register_region(region::impl* r) {
    // If needed, increase capacity of regions before taking the reclaim lock,
    // to avoid failing an allocation when push_back() tries to increase
    // capacity.
    //
    // The capacity increase is atomic (wrt _regions) so it cannot be
    // observed
    if (_regions.size() == _regions.capacity()) {
        auto copy = _regions;
        copy.reserve(copy.capacity() * 2);
        _regions = std::move(copy);
    }
    reclaiming_lock _(*this);
    _regions.push_back(r);
    llogger.debug("Registered region @{} with id={}", fmt::ptr(r), r->id());
}

void tracker::impl::unregister_region(region::impl* r) noexcept {
    reclaiming_lock _(*this);
    llogger.debug("Unregistering region, id={}", r->id());
    _regions.erase(std::remove(_regions.begin(), _regions.end(), r), _regions.end());
}

tracker::impl::impl() : _segment_pool(std::make_unique<logalloc::segment_pool>(*this)) {
    namespace sm = seastar::metrics;

    _metrics.add_group("lsa", {
        sm::make_gauge("total_space_bytes", [this] { return region_occupancy().total_space(); },
                       sm::description("Holds a current size of allocated memory in bytes.")),

        sm::make_gauge("used_space_bytes", [this] { return region_occupancy().used_space(); },
                       sm::description("Holds a current amount of used memory in bytes.")),

        sm::make_gauge("small_objects_total_space_bytes", [this] { return region_occupancy().total_space() - _segment_pool->non_lsa_memory_in_use(); },
                       sm::description("Holds a current size of \"small objects\" memory region in bytes.")),

        sm::make_gauge("small_objects_used_space_bytes", [this] { return region_occupancy().used_space() - _segment_pool->non_lsa_memory_in_use(); },
                       sm::description("Holds a current amount of used \"small objects\" memory in bytes.")),

        sm::make_gauge("large_objects_total_space_bytes", [this] { return _segment_pool->non_lsa_memory_in_use(); },
                       sm::description("Holds a current size of allocated non-LSA memory.")),

        sm::make_gauge("non_lsa_used_space_bytes", [this] { return non_lsa_used_space(); },
                       sm::description("Holds a current amount of used non-LSA memory.")),

        sm::make_gauge("free_space", [this] { return _segment_pool->unreserved_free_segments() * segment_size; },
                       sm::description("Holds a current amount of free memory that is under lsa control.")),

        sm::make_gauge("occupancy", [this] { return region_occupancy().used_fraction() * 100; },
                       sm::description("Holds a current portion (in percents) of the used memory.")),

        sm::make_counter("segments_compacted", [this] { return _segment_pool->statistics().segments_compacted; },
                        sm::description("Counts a number of compacted segments.")),

        sm::make_counter("memory_compacted", [this] { return _segment_pool->statistics().memory_compacted; },
                        sm::description("Counts number of bytes which were copied as part of segment compaction.")),

        sm::make_counter("memory_allocated", [this] { return _segment_pool->statistics().memory_allocated; },
                        sm::description("Counts number of bytes which were requested from LSA.")),

        sm::make_counter("memory_evicted", [this] { return _segment_pool->statistics().memory_evicted; },
                        sm::description("Counts number of bytes which were evicted.")),

        sm::make_counter("memory_freed", [this] { return _segment_pool->statistics().memory_freed; },
                        sm::description("Counts number of bytes which were requested to be freed in LSA.")),
    });
}

tracker::impl::~impl() {
    if (!_regions.empty()) {
        for (auto&& r : _regions) {
            llogger.error("Region with id={} not unregistered!", r->id());
        }
        abort();
    }
}

bool segment_pool::compact_segment(segment* seg) {
    auto& desc = descriptor(seg);
    if (!desc._region->reclaiming_enabled()) {
        return false;
    }

    // Called with emergency reserve, open one for
    // region::alloc_small not to throw if it needs
    // one more segment
    reservation_goal open_emergency_pool(*this, 0);
    allocation_lock no_alloc(*this);
    tracker_reclaimer_lock no_reclaim(_tracker);

    desc._region->compact_segment(seg, desc);
    return true;
}

allocating_section::guard::guard(tracker::impl& tracker) noexcept
    : _tracker(tracker), _prev(_tracker.segment_pool().emergency_reserve_max())
{ }

allocating_section::guard::~guard() {
    _tracker.segment_pool().set_emergency_reserve_max(_prev);
}

void allocating_section::maybe_decay_reserve() noexcept {
    // The decay rate is inversely proportional to the reserve
    // (every (s_segments_per_decay/_lsa_reserve) allocations).
    //
    // If the reserve is high, it is expensive since we may need to
    // evict a lot of memory to satisfy the reserve. Hence, we are
    // willing to risk a more frequent bad_alloc in order to decay it.
    // The cost of a bad_alloc is also lower compared to maintaining
    // the reserve.
    //
    // If the reserve is low, it is not expensive to maintain, so we
    // decay it at a lower rate.

    _remaining_lsa_segments_until_decay -= _lsa_reserve;
    if (_remaining_lsa_segments_until_decay < 0) {
        _remaining_lsa_segments_until_decay = s_segments_per_decay;
        _lsa_reserve = std::max(s_min_lsa_reserve, _lsa_reserve / 2);
        llogger.debug("Decaying LSA reserve in section {} to {} segments", static_cast<void*>(this), _lsa_reserve);
    }

    _remaining_std_bytes_until_decay -= _std_reserve;
    if (_remaining_std_bytes_until_decay < 0) {
        _remaining_std_bytes_until_decay = s_bytes_per_decay;
        _std_reserve = std::max(s_min_std_reserve, _std_reserve / 2);
        llogger.debug("Decaying standard allocator head-room in section {} to {} [B]", static_cast<void*>(this), _std_reserve);
    }
}

void allocating_section::reserve(tracker::impl& tracker) {
    auto& pool = tracker.segment_pool();
  try {
    pool.set_emergency_reserve_max(std::max(_lsa_reserve, _minimum_lsa_emergency_reserve));
    pool.refill_emergency_reserve();

    while (true) {
        size_t free = memory::free_memory();
        if (free >= _std_reserve) {
            break;
        }
        if (!tracker.reclaim(_std_reserve - free, is_preemptible::no)) {
            throw bad_alloc(format("failed to reclaim {} bytes of memory, while attempting to ensure an std reserve of {}", _std_reserve - free, _std_reserve));
        }
    }

    pool.clear_allocation_failure_flag();
  } catch (const std::bad_alloc& ex) {
        if (tracker.should_abort_on_bad_alloc()) {
            llogger.error("Aborting due to allocation failure: {}", ex.what());
            abort();
        }
        throw;
  }
}

void allocating_section::on_alloc_failure(logalloc::region& r) {
    r.allocator().invalidate_references();
    if (r.get_tracker().get_impl().segment_pool().allocation_failure_flag()) {
        _lsa_reserve *= 2;
        llogger.info("LSA allocation failure, increasing reserve in section {} to {} segments; trace: {}", fmt::ptr(this), _lsa_reserve, current_backtrace());
    } else {
        _std_reserve *= 2;
        llogger.info("Standard allocator failure, increasing head-room in section {} to {} [B]; trace: {}", fmt::ptr(this), _std_reserve, current_backtrace());
    }
    reserve(r.get_tracker().get_impl());
}

void allocating_section::set_lsa_reserve(size_t reserve) noexcept {
    _lsa_reserve = reserve;
}

void allocating_section::set_std_reserve(size_t reserve) noexcept {
    _std_reserve = reserve;
}

future<> prime_segment_pool(size_t available_memory, size_t min_free_memory) {
    return smp::invoke_on_all([=] {
        shard_tracker().get_impl().segment_pool().prime(available_memory, min_free_memory);
    });
}

future<> use_standard_allocator_segment_pool_backend(size_t available_memory) {
    return smp::invoke_on_all([=] {
        shard_tracker().get_impl().segment_pool().use_standard_allocator_segment_pool_backend(available_memory);
    });
}

}

// Orders segments by free space, assuming all segments have the same size.
// This avoids using the occupancy, which entails extra division operations.
template<>
size_t hist_key<logalloc::segment_descriptor>(const logalloc::segment_descriptor& desc) {
    return desc.free_space();
}