Coverage Report

Created: 2026-04-29 19:21

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
/tmp/bitcoin/src/txgraph.cpp
Line
Count
Source
1
// Copyright (c) The Bitcoin Core developers
2
// Distributed under the MIT software license, see the accompanying
3
// file COPYING or http://www.opensource.org/licenses/mit-license.php.
4
5
#include <txgraph.h>
6
7
#include <cluster_linearize.h>
8
#include <random.h>
9
#include <util/bitset.h>
10
#include <util/check.h>
11
#include <util/feefrac.h>
12
#include <util/vector.h>
13
14
#include <compare>
15
#include <functional>
16
#include <memory>
17
#include <set>
18
#include <span>
19
#include <unordered_set>
20
#include <utility>
21
22
namespace {
23
24
using namespace cluster_linearize;
25
26
/** The maximum number of levels a TxGraph can have (0 = main, 1 = staging). */
27
static constexpr int MAX_LEVELS{2};
28
29
// Forward declare the TxGraph implementation class.
30
class TxGraphImpl;
31
32
/** Position of a DepGraphIndex within a Cluster::m_linearization. */
33
using LinearizationIndex = uint32_t;
34
/** Position of a Cluster within TxGraphImpl::ClusterSet::m_clusters. */
35
using ClusterSetIndex = uint32_t;
36
37
/** Quality levels for cached cluster linearizations. */
38
enum class QualityLevel
39
{
40
    /** This is a singleton cluster consisting of a transaction that individually exceeds the
41
     *  cluster size limit. It cannot be merged with anything. */
42
    OVERSIZED_SINGLETON,
43
    /** This cluster may have multiple disconnected components, which are all NEEDS_FIX. */
44
    NEEDS_SPLIT_FIX,
45
    /** This cluster may have multiple disconnected components, which are all NEEDS_RELINEARIZE. */
46
    NEEDS_SPLIT,
47
    /** This cluster may be non-topological. */
48
    NEEDS_FIX,
49
    /** This cluster has undergone changes that warrant re-linearization. */
50
    NEEDS_RELINEARIZE,
51
    /** The minimal level of linearization has been performed, but it is not known to be optimal. */
52
    ACCEPTABLE,
53
    /** The linearization is known to be optimal. */
54
    OPTIMAL,
55
    /** This cluster is not registered in any ClusterSet::m_clusters.
56
     *  This must be the last entry in QualityLevel as ClusterSet::m_clusters is sized using it. */
57
    NONE,
58
};
59
60
/** Information about a transaction inside TxGraphImpl::Trim. */
61
struct TrimTxData
62
{
63
    // Fields populated by Cluster::AppendTrimData(). These are immutable after TrimTxData
64
    // construction.
65
    /** Chunk feerate for this transaction. */
66
    FeePerWeight m_chunk_feerate;
67
    /** GraphIndex of the transaction. */
68
    TxGraph::GraphIndex m_index;
69
    /** Size of the transaction. */
70
    uint32_t m_tx_size;
71
72
    // Fields only used internally by TxGraphImpl::Trim():
73
    /** Number of unmet dependencies this transaction has. -1 if the transaction is included. */
74
    uint32_t m_deps_left;
75
    /** Number of dependencies that apply to this transaction as child. */
76
    uint32_t m_parent_count;
77
    /** Where in deps_by_child those dependencies begin. */
78
    uint32_t m_parent_offset;
79
    /** Number of dependencies that apply to this transaction as parent. */
80
    uint32_t m_children_count;
81
    /** Where in deps_by_parent those dependencies begin. */
82
    uint32_t m_children_offset;
83
84
    // Fields only used internally by TxGraphImpl::Trim()'s union-find implementation, and only for
85
    // transactions that are definitely included or definitely rejected.
86
    //
87
    // As transactions get processed, they get organized into trees which form partitions
88
    // representing the would-be clusters up to that point. The root of each tree is a
89
    // representative for that partition. See
90
    // https://en.wikipedia.org/wiki/Disjoint-set_data_structure.
91
    //
92
    /** Pointer to another TrimTxData, towards the root of the tree. If this is a root, m_uf_parent
93
     *  is equal to this itself. */
94
    TrimTxData* m_uf_parent;
95
    /** If this is a root, the total number of transactions in the partition. */
96
    uint32_t m_uf_count;
97
    /** If this is a root, the total size of transactions in the partition. */
98
    uint64_t m_uf_size;
99
};
100
101
/** A grouping of connected transactions inside a TxGraphImpl::ClusterSet. */
102
class Cluster
103
{
104
    friend class TxGraphImpl;
105
    friend class BlockBuilderImpl;
106
107
protected:
108
    using GraphIndex = TxGraph::GraphIndex;
109
    using SetType = BitSet<MAX_CLUSTER_COUNT_LIMIT>;
110
    /** The quality level of m_linearization. */
111
    QualityLevel m_quality{QualityLevel::NONE};
112
    /** Which position this Cluster has in TxGraphImpl::ClusterSet::m_clusters[m_quality]. */
113
    ClusterSetIndex m_setindex{ClusterSetIndex(-1)};
114
    /** Sequence number for this Cluster (for tie-breaking comparison between equal-chunk-feerate
115
        transactions in distinct clusters). */
116
    uint64_t m_sequence;
117
118
130k
    explicit Cluster(uint64_t sequence) noexcept : m_sequence(sequence) {}
119
120
public:
121
    // Provide virtual destructor, for safe polymorphic usage inside std::unique_ptr.
122
130k
    virtual ~Cluster() = default;
123
124
    // Cannot move or copy (would invalidate Cluster* in Locator and ClusterSet). */
125
    Cluster(const Cluster&) = delete;
126
    Cluster& operator=(const Cluster&) = delete;
127
    Cluster(Cluster&&) = delete;
128
    Cluster& operator=(Cluster&&) = delete;
129
130
    // Generic helper functions.
131
132
    /** Whether the linearization of this Cluster can be exposed. */
133
    bool IsAcceptable() const noexcept
134
285M
    {
135
285M
        return m_quality == QualityLevel::ACCEPTABLE || m_quality == QualityLevel::OPTIMAL;
136
285M
    }
137
    /** Whether the linearization of this Cluster is topological. */
138
    bool IsTopological() const noexcept
139
238k
    {
140
238k
        return m_quality != QualityLevel::NEEDS_FIX && m_quality != QualityLevel::NEEDS_SPLIT_FIX;
141
238k
    }
142
    /** Whether the linearization of this Cluster is optimal. */
143
    bool IsOptimal() const noexcept
144
10.6M
    {
145
10.6M
        return m_quality == QualityLevel::OPTIMAL;
146
10.6M
    }
147
    /** Whether this cluster is oversized. Note that no changes that can cause oversizedness are
148
     *  ever applied, so the only way a materialized Cluster object can be oversized is by being
149
     *  an individually oversized transaction singleton. */
150
21.4M
    bool IsOversized() const noexcept { return m_quality == QualityLevel::OVERSIZED_SINGLETON; }
151
    /** Whether this cluster requires splitting. */
152
    bool NeedsSplitting() const noexcept
153
285M
    {
154
285M
        return m_quality == QualityLevel::NEEDS_SPLIT || m_quality == QualityLevel::NEEDS_SPLIT_FIX;
155
285M
    }
156
157
    /** Get the smallest number of transactions this Cluster is intended for. */
158
    virtual DepGraphIndex GetMinIntendedTxCount() const noexcept = 0;
159
    /** Get the maximum number of transactions this Cluster supports. */
160
    virtual DepGraphIndex GetMaxTxCount() const noexcept = 0;
161
    /** Total memory usage currently for this Cluster, including all its dynamic memory, plus Cluster
162
     *  structure itself, and ClusterSet::m_clusters entry. */
163
    virtual size_t TotalMemoryUsage() const noexcept = 0;
164
    /** Determine the range of DepGraphIndexes used by this Cluster. */
165
    virtual DepGraphIndex GetDepGraphIndexRange() const noexcept = 0;
166
    /** Get the number of transactions in this Cluster. */
167
    virtual LinearizationIndex GetTxCount() const noexcept = 0;
168
    /** Get the total size of the transactions in this Cluster. */
169
    virtual uint64_t GetTotalTxSize() const noexcept = 0;
170
    /** Given a DepGraphIndex into this Cluster, find the corresponding GraphIndex. */
171
    virtual GraphIndex GetClusterEntry(DepGraphIndex index) const noexcept = 0;
172
    /** Append a transaction with given GraphIndex at the end of this Cluster and its
173
     *  linearization. Return the DepGraphIndex it was placed at. */
174
    virtual DepGraphIndex AppendTransaction(GraphIndex graph_idx, FeePerWeight feerate) noexcept = 0;
175
    /** Add dependencies to a given child in this cluster. */
176
    virtual void AddDependencies(SetType parents, DepGraphIndex child) noexcept = 0;
177
    /** Invoke visit1_fn for each transaction in the cluster, in linearization order, then
178
     *  visit2_fn in the same order, then wipe this Cluster. */
179
    virtual void ExtractTransactions(const std::function<void (DepGraphIndex, GraphIndex, FeePerWeight)>& visit1_fn, const std::function<void (DepGraphIndex, GraphIndex, SetType)>& visit2_fn) noexcept = 0;
180
    /** Figure out what level this Cluster exists at in the graph. In most cases this is known by
181
     *  the caller already (see all "int level" arguments below), but not always. */
182
    virtual int GetLevel(const TxGraphImpl& graph) const noexcept = 0;
183
    /** Only called by TxGraphImpl::SwapIndexes. */
184
    virtual void UpdateMapping(DepGraphIndex cluster_idx, GraphIndex graph_idx) noexcept = 0;
185
    /** Push changes to Cluster and its linearization to the TxGraphImpl Entry objects. Main chunk
186
     *  information is computed if the cluster is acceptable, or when rename is set. Rename is used
187
     *  when called from Compact, to recompute after GraphIndexes may have changed; in this case,
188
     *  no chunk index objects are removed or created either. */
189
    virtual void Updated(TxGraphImpl& graph, int level, bool rename) noexcept = 0;
190
    /** Remove all chunk index entries for this cluster (level=0 only). */
191
    virtual void RemoveChunkData(TxGraphImpl& graph) noexcept = 0;
192
    /** Create a copy of this Cluster in staging, returning a pointer to it (used by PullIn). */
193
    virtual Cluster* CopyToStaging(TxGraphImpl& graph) const noexcept = 0;
194
    /** Get the list of Clusters in main that conflict with this one (which is assumed to be in staging). */
195
    virtual void GetConflicts(const TxGraphImpl& graph, std::vector<Cluster*>& out) const noexcept = 0;
196
    /** Mark all the Entry objects belonging to this staging Cluster as missing. The Cluster must be
197
     *  deleted immediately after. */
198
    virtual void MakeStagingTransactionsMissing(TxGraphImpl& graph) noexcept = 0;
199
    /** Remove all transactions from a (non-empty) Cluster. */
200
    virtual void Clear(TxGraphImpl& graph, int level) noexcept = 0;
201
    /** Change a Cluster's level from 1 (staging) to 0 (main). */
202
    virtual void MoveToMain(TxGraphImpl& graph) noexcept = 0;
203
    /** Minimize this Cluster's memory usage. */
204
    virtual void Compact() noexcept = 0;
205
206
    // Functions that implement the Cluster-specific side of internal TxGraphImpl mutations.
207
208
    /** Apply all removals from the front of to_remove that apply to this Cluster, popping them
209
     *  off. There must be at least one such entry. */
210
    virtual void ApplyRemovals(TxGraphImpl& graph, int level, std::span<GraphIndex>& to_remove) noexcept = 0;
211
    /** Split this cluster (must have a NEEDS_SPLIT* quality). Returns whether to delete this
212
     *  Cluster afterwards. */
213
    [[nodiscard]] virtual bool Split(TxGraphImpl& graph, int level) noexcept = 0;
214
    /** Move all transactions from cluster to *this (as separate components). */
215
    virtual void Merge(TxGraphImpl& graph, int level, Cluster& cluster) noexcept = 0;
216
    /** Given a span of (parent, child) pairs that all belong to this Cluster, apply them. */
217
    virtual void ApplyDependencies(TxGraphImpl& graph, int level, std::span<std::pair<GraphIndex, GraphIndex>> to_apply) noexcept = 0;
218
    /** Improve the linearization of this Cluster. Returns how much work was performed and whether
219
     *  the Cluster's QualityLevel improved as a result. */
220
    virtual std::pair<uint64_t, bool> Relinearize(TxGraphImpl& graph, int level, uint64_t max_cost) noexcept = 0;
221
    /** For every chunk in the cluster, append its FeeFrac to ret. */
222
    virtual void AppendChunkFeerates(std::vector<FeeFrac>& ret) const noexcept = 0;
223
    /** Add a TrimTxData entry (filling m_chunk_feerate, m_index, m_tx_size) for every
224
     *  transaction in the Cluster to ret. Implicit dependencies between consecutive transactions
225
     *  in the linearization are added to deps. Return the Cluster's total transaction size. */
226
    virtual uint64_t AppendTrimData(std::vector<TrimTxData>& ret, std::vector<std::pair<GraphIndex, GraphIndex>>& deps) const noexcept = 0;
227
228
    // Functions that implement the Cluster-specific side of public TxGraph functions.
229
230
    /** Process elements from the front of args that apply to this cluster, and append Refs for the
231
     *  union of their ancestors to output. */
232
    virtual void GetAncestorRefs(const TxGraphImpl& graph, std::span<std::pair<Cluster*, DepGraphIndex>>& args, std::vector<TxGraph::Ref*>& output) noexcept = 0;
233
    /** Process elements from the front of args that apply to this cluster, and append Refs for the
234
     *  union of their descendants to output. */
235
    virtual void GetDescendantRefs(const TxGraphImpl& graph, std::span<std::pair<Cluster*, DepGraphIndex>>& args, std::vector<TxGraph::Ref*>& output) noexcept = 0;
236
    /** Populate range with refs for the transactions in this Cluster's linearization, from
237
     *  position start_pos until start_pos+range.size()-1, inclusive. Returns whether that
238
     *  range includes the last transaction in the linearization. */
239
    virtual bool GetClusterRefs(TxGraphImpl& graph, std::span<TxGraph::Ref*> range, LinearizationIndex start_pos) noexcept = 0;
240
    /** Get the individual transaction feerate of a Cluster element. */
241
    virtual FeePerWeight GetIndividualFeerate(DepGraphIndex idx) noexcept = 0;
242
    /** Modify the fee of a Cluster element. */
243
    virtual void SetFee(TxGraphImpl& graph, int level, DepGraphIndex idx, int64_t fee) noexcept = 0;
244
245
    // Debugging functions.
246
247
    virtual void SanityCheck(const TxGraphImpl& graph, int level) const = 0;
248
};
249
250
/** An implementation of Cluster that uses a DepGraph and vectors, to support arbitrary numbers of
251
 *  transactions up to MAX_CLUSTER_COUNT_LIMIT. */
252
class GenericClusterImpl final : public Cluster
253
{
254
    friend class TxGraphImpl;
255
    /** The DepGraph for this cluster, holding all feerates, and ancestors/descendants. */
256
    DepGraph<SetType> m_depgraph;
257
    /** m_mapping[i] gives the GraphIndex for the position i transaction in m_depgraph. Values for
258
     *  positions i that do not exist in m_depgraph shouldn't ever be accessed and thus don't
259
     *  matter. m_mapping.size() equals m_depgraph.PositionRange(). */
260
    std::vector<GraphIndex> m_mapping;
261
    /** The current linearization of the cluster. m_linearization.size() equals
262
     *  m_depgraph.TxCount(). This is always kept topological. */
263
    std::vector<DepGraphIndex> m_linearization;
264
265
public:
266
    /** The smallest number of transactions this Cluster implementation is intended for. */
267
    static constexpr DepGraphIndex MIN_INTENDED_TX_COUNT{2};
268
    /** The largest number of transactions this Cluster implementation supports. */
269
    static constexpr DepGraphIndex MAX_TX_COUNT{SetType::Size()};
270
271
    GenericClusterImpl() noexcept = delete;
272
    /** Construct an empty GenericClusterImpl. */
273
    explicit GenericClusterImpl(uint64_t sequence) noexcept;
274
275
    size_t TotalMemoryUsage() const noexcept final;
276
54.6k
    constexpr DepGraphIndex GetMinIntendedTxCount() const noexcept final { return MIN_INTENDED_TX_COUNT; }
277
58.5k
    constexpr DepGraphIndex GetMaxTxCount() const noexcept final { return MAX_TX_COUNT; }
278
11
    DepGraphIndex GetDepGraphIndexRange() const noexcept final { return m_depgraph.PositionRange(); }
279
305k
    LinearizationIndex GetTxCount() const noexcept final { return m_linearization.size(); }
280
    uint64_t GetTotalTxSize() const noexcept final;
281
230k
    GraphIndex GetClusterEntry(DepGraphIndex index) const noexcept final { return m_mapping[index]; }
282
    DepGraphIndex AppendTransaction(GraphIndex graph_idx, FeePerWeight feerate) noexcept final;
283
    void AddDependencies(SetType parents, DepGraphIndex child) noexcept final;
284
    void ExtractTransactions(const std::function<void (DepGraphIndex, GraphIndex, FeePerWeight)>& visit1_fn, const std::function<void (DepGraphIndex, GraphIndex, SetType)>& visit2_fn) noexcept final;
285
    int GetLevel(const TxGraphImpl& graph) const noexcept final;
286
560
    void UpdateMapping(DepGraphIndex cluster_idx, GraphIndex graph_idx) noexcept final { m_mapping[cluster_idx] = graph_idx; }
287
    void Updated(TxGraphImpl& graph, int level, bool rename) noexcept final;
288
    void RemoveChunkData(TxGraphImpl& graph) noexcept final;
289
    Cluster* CopyToStaging(TxGraphImpl& graph) const noexcept final;
290
    void GetConflicts(const TxGraphImpl& graph, std::vector<Cluster*>& out) const noexcept final;
291
    void MakeStagingTransactionsMissing(TxGraphImpl& graph) noexcept final;
292
    void Clear(TxGraphImpl& graph, int level) noexcept final;
293
    void MoveToMain(TxGraphImpl& graph) noexcept final;
294
    void Compact() noexcept final;
295
    void ApplyRemovals(TxGraphImpl& graph, int level, std::span<GraphIndex>& to_remove) noexcept final;
296
    [[nodiscard]] bool Split(TxGraphImpl& graph, int level) noexcept final;
297
    void Merge(TxGraphImpl& graph, int level, Cluster& cluster) noexcept final;
298
    void ApplyDependencies(TxGraphImpl& graph, int level, std::span<std::pair<GraphIndex, GraphIndex>> to_apply) noexcept final;
299
    std::pair<uint64_t, bool> Relinearize(TxGraphImpl& graph, int level, uint64_t max_cost) noexcept final;
300
    void AppendChunkFeerates(std::vector<FeeFrac>& ret) const noexcept final;
301
    uint64_t AppendTrimData(std::vector<TrimTxData>& ret, std::vector<std::pair<GraphIndex, GraphIndex>>& deps) const noexcept final;
302
    void GetAncestorRefs(const TxGraphImpl& graph, std::span<std::pair<Cluster*, DepGraphIndex>>& args, std::vector<TxGraph::Ref*>& output) noexcept final;
303
    void GetDescendantRefs(const TxGraphImpl& graph, std::span<std::pair<Cluster*, DepGraphIndex>>& args, std::vector<TxGraph::Ref*>& output) noexcept final;
304
    bool GetClusterRefs(TxGraphImpl& graph, std::span<TxGraph::Ref*> range, LinearizationIndex start_pos) noexcept final;
305
    FeePerWeight GetIndividualFeerate(DepGraphIndex idx) noexcept final;
306
    void SetFee(TxGraphImpl& graph, int level, DepGraphIndex idx, int64_t fee) noexcept final;
307
    void SanityCheck(const TxGraphImpl& graph, int level) const final;
308
};
309
310
/** An implementation of Cluster that only supports 1 transaction. */
311
class SingletonClusterImpl final : public Cluster
312
{
313
    friend class TxGraphImpl;
314
315
    /** The feerate of the (singular) transaction in this Cluster. */
316
    FeePerWeight m_feerate;
317
    /** Constant to indicate that this Cluster is empty. */
318
    static constexpr auto NO_GRAPH_INDEX = GraphIndex(-1);
319
    /** The GraphIndex of the transaction. NO_GRAPH_INDEX if this Cluster is empty. */
320
    GraphIndex m_graph_index = NO_GRAPH_INDEX;
321
322
public:
323
    /** The smallest number of transactions this Cluster implementation is intended for. */
324
    static constexpr DepGraphIndex MIN_INTENDED_TX_COUNT{1};
325
    /** The largest number of transactions this Cluster implementation supports. */
326
    static constexpr DepGraphIndex MAX_TX_COUNT{1};
327
328
    SingletonClusterImpl() noexcept = delete;
329
    /** Construct an empty SingletonClusterImpl. */
330
129k
    explicit SingletonClusterImpl(uint64_t sequence) noexcept : Cluster(sequence) {}
331
332
    size_t TotalMemoryUsage() const noexcept final;
333
10.6M
    constexpr DepGraphIndex GetMinIntendedTxCount() const noexcept final { return MIN_INTENDED_TX_COUNT; }
334
10.6M
    constexpr DepGraphIndex GetMaxTxCount() const noexcept final { return MAX_TX_COUNT; }
335
97.3M
    LinearizationIndex GetTxCount() const noexcept final { return m_graph_index != NO_GRAPH_INDEX; }
336
6.98k
    DepGraphIndex GetDepGraphIndexRange() const noexcept final { return GetTxCount(); }
337
10.8M
    uint64_t GetTotalTxSize() const noexcept final { return GetTxCount() ? m_feerate.size : 0; }
338
10.6M
    GraphIndex GetClusterEntry(DepGraphIndex index) const noexcept final { Assume(index == 0); Assume(GetTxCount()); return m_graph_index; }
339
    DepGraphIndex AppendTransaction(GraphIndex graph_idx, FeePerWeight feerate) noexcept final;
340
    void AddDependencies(SetType parents, DepGraphIndex child) noexcept final;
341
    void ExtractTransactions(const std::function<void (DepGraphIndex, GraphIndex, FeePerWeight)>& visit1_fn, const std::function<void (DepGraphIndex, GraphIndex, SetType)>& visit2_fn) noexcept final;
342
    int GetLevel(const TxGraphImpl& graph) const noexcept final;
343
16.9k
    void UpdateMapping(DepGraphIndex cluster_idx, GraphIndex graph_idx) noexcept final { Assume(cluster_idx == 0); m_graph_index = graph_idx; }
344
    void Updated(TxGraphImpl& graph, int level, bool rename) noexcept final;
345
    void RemoveChunkData(TxGraphImpl& graph) noexcept final;
346
    Cluster* CopyToStaging(TxGraphImpl& graph) const noexcept final;
347
    void GetConflicts(const TxGraphImpl& graph, std::vector<Cluster*>& out) const noexcept final;
348
    void MakeStagingTransactionsMissing(TxGraphImpl& graph) noexcept final;
349
    void Clear(TxGraphImpl& graph, int level) noexcept final;
350
    void MoveToMain(TxGraphImpl& graph) noexcept final;
351
    void Compact() noexcept final;
352
    void ApplyRemovals(TxGraphImpl& graph, int level, std::span<GraphIndex>& to_remove) noexcept final;
353
    [[nodiscard]] bool Split(TxGraphImpl& graph, int level) noexcept final;
354
    void Merge(TxGraphImpl& graph, int level, Cluster& cluster) noexcept final;
355
    void ApplyDependencies(TxGraphImpl& graph, int level, std::span<std::pair<GraphIndex, GraphIndex>> to_apply) noexcept final;
356
    std::pair<uint64_t, bool> Relinearize(TxGraphImpl& graph, int level, uint64_t max_cost) noexcept final;
357
    void AppendChunkFeerates(std::vector<FeeFrac>& ret) const noexcept final;
358
    uint64_t AppendTrimData(std::vector<TrimTxData>& ret, std::vector<std::pair<GraphIndex, GraphIndex>>& deps) const noexcept final;
359
    void GetAncestorRefs(const TxGraphImpl& graph, std::span<std::pair<Cluster*, DepGraphIndex>>& args, std::vector<TxGraph::Ref*>& output) noexcept final;
360
    void GetDescendantRefs(const TxGraphImpl& graph, std::span<std::pair<Cluster*, DepGraphIndex>>& args, std::vector<TxGraph::Ref*>& output) noexcept final;
361
    bool GetClusterRefs(TxGraphImpl& graph, std::span<TxGraph::Ref*> range, LinearizationIndex start_pos) noexcept final;
362
    FeePerWeight GetIndividualFeerate(DepGraphIndex idx) noexcept final;
363
    void SetFee(TxGraphImpl& graph, int level, DepGraphIndex idx, int64_t fee) noexcept final;
364
    void SanityCheck(const TxGraphImpl& graph, int level) const final;
365
};
366
367
/** The transaction graph, including staged changes.
368
 *
369
 * The overall design of the data structure consists of 3 interlinked representations:
370
 * - The transactions (held as a vector of TxGraphImpl::Entry inside TxGraphImpl).
371
 * - The clusters (Cluster objects in per-quality vectors inside TxGraphImpl::ClusterSet).
372
 * - The Refs (TxGraph::Ref objects, held externally by users of the TxGraph class)
373
 *
374
 * The Clusters are kept in one or two ClusterSet objects, one for the "main" graph, and one for
375
 * the proposed changes ("staging"). If a transaction occurs in both, they share the same Entry,
376
 * but there will be a separate Cluster per graph.
377
 *
378
 * Clusters and Refs contain the index of the Entry objects they refer to, and the Entry objects
379
 * refer back to the Clusters and Refs the corresponding transaction is contained in.
380
 *
381
 * While redundant, this permits moving all of them independently, without invalidating things
382
 * or costly iteration to fix up everything:
383
 * - Entry objects can be moved to fill holes left by removed transactions in the Entry vector
384
 *   (see TxGraphImpl::Compact).
385
 * - Clusters can be rewritten continuously (removals can cause them to split, new dependencies
386
 *   can cause them to be merged).
387
 * - Ref objects can be held outside the class, while permitting them to be moved around, and
388
 *   inherited from.
389
 */
390
class TxGraphImpl final : public TxGraph
391
{
392
    friend class Cluster;
393
    friend class SingletonClusterImpl;
394
    friend class GenericClusterImpl;
395
    friend class BlockBuilderImpl;
396
private:
397
    /** Internal RNG. */
398
    FastRandomContext m_rng;
399
    /** This TxGraphImpl's maximum cluster count limit. */
400
    const DepGraphIndex m_max_cluster_count;
401
    /** This TxGraphImpl's maximum cluster size limit. */
402
    const uint64_t m_max_cluster_size;
403
    /** The amount of linearization work needed per cluster to be considered acceptable. */
404
    const uint64_t m_acceptable_cost;
405
    /** Fallback ordering for transactions. */
406
    const std::function<std::strong_ordering(const TxGraph::Ref&, const TxGraph::Ref&)> m_fallback_order;
407
408
    /** Information about one group of Clusters to be merged. */
409
    struct GroupEntry
410
    {
411
        /** Where the clusters to be merged start in m_group_clusters. */
412
        uint32_t m_cluster_offset;
413
        /** How many clusters to merge. */
414
        uint32_t m_cluster_count;
415
        /** Where the dependencies for this cluster group in m_deps_to_add start. */
416
        uint32_t m_deps_offset;
417
        /** How many dependencies to add. */
418
        uint32_t m_deps_count;
419
    };
420
421
    /** Information about all groups of Clusters to be merged. */
422
    struct GroupData
423
    {
424
        /** The groups of Clusters to be merged. */
425
        std::vector<GroupEntry> m_groups;
426
        /** Which clusters are to be merged. GroupEntry::m_cluster_offset indexes into this. */
427
        std::vector<Cluster*> m_group_clusters;
428
    };
429
430
    /** The collection of all Clusters in main or staged. */
431
    struct ClusterSet
432
    {
433
        /** The vectors of clusters, one vector per quality level. ClusterSetIndex indexes into each. */
434
        std::array<std::vector<std::unique_ptr<Cluster>>, int(QualityLevel::NONE)> m_clusters;
435
        /** Which removals have yet to be applied. */
436
        std::vector<GraphIndex> m_to_remove;
437
        /** Which dependencies are to be added ((parent,child) pairs). GroupData::m_deps_offset indexes
438
         *  into this. */
439
        std::vector<std::pair<GraphIndex, GraphIndex>> m_deps_to_add;
440
        /** Information about the merges to be performed, if known. */
441
        std::optional<GroupData> m_group_data = GroupData{};
442
        /** Which entries were removed in this ClusterSet (so they can be wiped on abort). This
443
         *  includes all entries which have an (R) removed locator at this level (staging only),
444
         *  plus optionally any transaction in m_unlinked. */
445
        std::vector<GraphIndex> m_removed;
446
        /** Total number of transactions in this graph (sum of all transaction counts in all
447
         *  Clusters, and for staging also those inherited from the main ClusterSet). */
448
        GraphIndex m_txcount{0};
449
        /** Total number of individually oversized transactions in the graph. */
450
        GraphIndex m_txcount_oversized{0};
451
        /** Whether this graph is oversized (if known). */
452
        std::optional<bool> m_oversized{false};
453
        /** The combined TotalMemoryUsage of all clusters in this level (only Clusters that
454
         *  are materialized; in staging, implicit Clusters from main are not counted), */
455
        size_t m_cluster_usage{0};
456
457
63.8k
        ClusterSet() noexcept = default;
458
    };
459
460
    /** The main ClusterSet. */
461
    ClusterSet m_main_clusterset;
462
    /** The staging ClusterSet, if any. */
463
    std::optional<ClusterSet> m_staging_clusterset;
464
    /** Next sequence number to assign to created Clusters. */
465
    uint64_t m_next_sequence_counter{0};
466
467
    /** Information about a chunk in the main graph. */
468
    struct ChunkData
469
    {
470
        /** The Entry which is the last transaction of the chunk. */
471
        mutable GraphIndex m_graph_index;
472
        /** How many transactions the chunk contains (-1 = singleton tail of cluster). */
473
        LinearizationIndex m_chunk_count;
474
475
        ChunkData(GraphIndex graph_index, LinearizationIndex chunk_count) noexcept :
476
176k
            m_graph_index{graph_index}, m_chunk_count{chunk_count} {}
477
    };
478
479
    /** Compare two Cluster* by their m_sequence value (while supporting nullptr). */
480
    static std::strong_ordering CompareClusters(Cluster* a, Cluster* b) noexcept
481
286k
    {
482
        // The nullptr pointer compares before everything else.
483
286k
        if (a == nullptr || b == nullptr) {
484
0
            return (a != nullptr) <=> (b != nullptr);
485
0
        }
486
        // If neither pointer is nullptr, compare the Clusters' sequence numbers.
487
286k
        Assume(a == b || a->m_sequence != b->m_sequence);
488
286k
        return a->m_sequence <=> b->m_sequence;
489
286k
    }
490
491
    /** Compare two entries (which must both exist within the main graph). */
492
    std::strong_ordering CompareMainTransactions(GraphIndex a, GraphIndex b) const noexcept
493
139M
    {
494
139M
        if (a == b) return std::strong_ordering::equal;
495
139M
        Assume(a < m_entries.size() && b < m_entries.size());
496
139M
        const auto& entry_a = m_entries[a];
497
139M
        const auto& entry_b = m_entries[b];
498
        // Compare chunk feerates, and return result if it differs.
499
139M
        auto feerate_cmp = FeeRateCompare(entry_b.m_main_chunk_feerate, entry_a.m_main_chunk_feerate);
500
139M
        if (feerate_cmp < 0) return std::strong_ordering::less;
501
99.1M
        if (feerate_cmp > 0) return std::strong_ordering::greater;
502
        // Compare equal-feerate chunk prefix size for comparing equal chunk feerates. This does two
503
        // things: it distinguishes equal-feerate chunks within the same cluster (because later
504
        // ones will always have a higher prefix size), and it may distinguish equal-feerate chunks
505
        // from distinct clusters.
506
77.1M
        if (entry_a.m_main_equal_feerate_chunk_prefix_size != entry_b.m_main_equal_feerate_chunk_prefix_size) {
507
1.57M
            return entry_a.m_main_equal_feerate_chunk_prefix_size <=> entry_b.m_main_equal_feerate_chunk_prefix_size;
508
1.57M
        }
509
        // Compare by maximum m_fallback_order element to order equal-feerate chunks in distinct
510
        // clusters, when the equal-feerate-prefix size is also the same.
511
75.5M
        const auto& locator_a = entry_a.m_locator[0];
512
75.5M
        const auto& locator_b = entry_b.m_locator[0];
513
75.5M
        Assume(locator_a.IsPresent() && locator_b.IsPresent());
514
75.5M
        if (locator_a.cluster != locator_b.cluster) {
515
75.4M
            auto fallback_cmp = m_fallback_order(*m_entries[entry_a.m_main_max_chunk_fallback].m_ref,
516
75.4M
                                                 *m_entries[entry_b.m_main_max_chunk_fallback].m_ref);
517
75.4M
            if (fallback_cmp != 0) return fallback_cmp;
518
            // This shouldn't be reachable as m_fallback_order defines a strong ordering.
519
0
            Assume(false);
520
0
            return CompareClusters(locator_a.cluster, locator_b.cluster);
521
75.4M
        }
522
        // Within a single chunk, sort by position within cluster linearization.
523
87.8k
        return entry_a.m_main_lin_index <=> entry_b.m_main_lin_index;
524
75.5M
    }
525
526
    /** Comparator for ChunkData objects in mining order. */
527
    class ChunkOrder
528
    {
529
        const TxGraphImpl* const m_graph;
530
    public:
531
1.20k
        explicit ChunkOrder(const TxGraphImpl* graph) : m_graph(graph) {}
532
533
        bool operator()(const ChunkData& a, const ChunkData& b) const noexcept
534
2.02M
        {
535
2.02M
            return m_graph->CompareMainTransactions(a.m_graph_index, b.m_graph_index) < 0;
536
2.02M
        }
537
    };
538
539
    /** Definition for the mining index type. */
540
    using ChunkIndex = std::set<ChunkData, ChunkOrder>;
541
542
    /** Index of ChunkData objects, indexing the last transaction in each chunk in the main
543
     *  graph. */
544
    ChunkIndex m_main_chunkindex;
545
    /** Number of index-observing objects in existence (BlockBuilderImpls). */
546
    size_t m_main_chunkindex_observers{0};
547
    /** Cache of discarded ChunkIndex node handles to reuse, avoiding additional allocation. */
548
    std::vector<ChunkIndex::node_type> m_main_chunkindex_discarded;
549
550
    /** A Locator that describes whether, where, and in which Cluster an Entry appears.
551
     *  Every Entry has MAX_LEVELS locators, as it may appear in one Cluster per level.
552
     *
553
     *  Each level of a Locator is in one of three states:
554
     *
555
     *  - (P)resent: actually occurs in a Cluster at that level.
556
     *
557
     *  - (M)issing:
558
     *    - In the main graph:    the transaction does not exist in main.
559
     *    - In the staging graph: the transaction's existence is the same as in main. If it doesn't
560
     *                            exist in main, (M) in staging means it does not exist there
561
     *                            either. If it does exist in main, (M) in staging means the
562
     *                            cluster it is in has not been modified in staging, and thus the
563
     *                            transaction implicitly exists in staging too (without explicit
564
     *                            Cluster object; see PullIn() to create it in staging too).
565
     *
566
     *  - (R)emoved: only possible in staging; it means the transaction exists in main, but is
567
     *               removed in staging.
568
     *
569
     * The following combinations are possible:
570
     * - (M,M): the transaction doesn't exist in either graph.
571
     * - (P,M): the transaction exists in both, but only exists explicitly in a Cluster object in
572
     *          main. Its existence in staging is inherited from main.
573
     * - (P,P): the transaction exists in both, and is materialized in both. Thus, the clusters
574
     *          and/or their linearizations may be different in main and staging.
575
     * - (M,P): the transaction is added in staging, and does not exist in main.
576
     * - (P,R): the transaction exists in main, but is removed in staging.
577
     *
578
     * When staging does not exist, only (M,M) and (P,M) are possible.
579
     */
580
    struct Locator
581
    {
582
        /** Which Cluster the Entry appears in (nullptr = missing). */
583
        Cluster* cluster{nullptr};
584
        /** Where in the Cluster it appears (if cluster == nullptr: 0 = missing, -1 = removed). */
585
        DepGraphIndex index{0};
586
587
        /** Mark this Locator as missing (= same as lower level, or non-existing if level 0). */
588
116k
        void SetMissing() noexcept { cluster = nullptr; index = 0; }
589
        /** Mark this Locator as removed (not allowed in level 0). */
590
2.17k
        void SetRemoved() noexcept { cluster = nullptr; index = DepGraphIndex(-1); }
591
        /** Mark this Locator as present, in the specified Cluster. */
592
345k
        void SetPresent(Cluster* c, DepGraphIndex i) noexcept { cluster = c; index = i; }
593
        /** Check if this Locator is missing. */
594
23.2M
        bool IsMissing() const noexcept { return cluster == nullptr && index == 0; }
595
        /** Check if this Locator is removed. */
596
34.0M
        bool IsRemoved() const noexcept { return cluster == nullptr && index == DepGraphIndex(-1); }
597
        /** Check if this Locator is present (in some Cluster). */
598
480M
        bool IsPresent() const noexcept { return cluster != nullptr; }
599
    };
600
601
    /** Internal information about each transaction in a TxGraphImpl. */
602
    struct Entry
603
    {
604
        /** Pointer to the corresponding Ref object if any, or nullptr if unlinked. */
605
        Ref* m_ref{nullptr};
606
        /** Iterator to the corresponding ChunkData, if any, and m_main_chunkindex.end() otherwise.
607
         *  This is initialized on construction of the Entry, in AddTransaction. */
608
        ChunkIndex::iterator m_main_chunkindex_iterator;
609
        /** Which Cluster and position therein this Entry appears in. ([0] = main, [1] = staged). */
610
        Locator m_locator[MAX_LEVELS];
611
        /** The chunk feerate of this transaction in main (if present in m_locator[0]). */
612
        FeePerWeight m_main_chunk_feerate;
613
        /** The equal-feerate chunk prefix size of this transaction in main. If the transaction is
614
         *  part of chunk C in main, then this gives the sum of the sizes of all chunks in C's
615
         *  cluster, whose feerate is equal to that of C, which do not appear after C itself in
616
         *  the cluster's linearization.
617
         *  This provides a way to sort equal-feerate chunks across clusters, in a way that agrees
618
         *  with the within-cluster chunk ordering. */
619
        int32_t m_main_equal_feerate_chunk_prefix_size;
620
        /** The position this transaction has in the main linearization (if present). */
621
        LinearizationIndex m_main_lin_index;
622
        /** Of all transactions within this transaction's chunk in main (if present there), the
623
         *  maximal one according to m_fallback_order. */
624
        GraphIndex m_main_max_chunk_fallback = GraphIndex(-1);
625
    };
626
627
    /** The set of all transactions (in all levels combined). GraphIndex values index into this. */
628
    std::vector<Entry> m_entries;
629
630
    /** Set of Entries which have no linked Ref anymore. */
631
    std::vector<GraphIndex> m_unlinked;
632
633
public:
634
    /** Construct a new TxGraphImpl with the specified limits and fallback order. */
635
    explicit TxGraphImpl(
636
        DepGraphIndex max_cluster_count,
637
        uint64_t max_cluster_size,
638
        uint64_t acceptable_cost,
639
        const std::function<std::strong_ordering(const TxGraph::Ref&, const TxGraph::Ref&)>& fallback_order
640
    ) noexcept :
641
1.20k
        m_max_cluster_count(max_cluster_count),
642
1.20k
        m_max_cluster_size(max_cluster_size),
643
1.20k
        m_acceptable_cost(acceptable_cost),
644
1.20k
        m_fallback_order(fallback_order),
645
1.20k
        m_main_chunkindex(ChunkOrder(this))
646
1.20k
    {
647
1.20k
        Assume(max_cluster_count >= 1);
648
1.20k
        Assume(max_cluster_count <= MAX_CLUSTER_COUNT_LIMIT);
649
1.20k
    }
650
651
    /** Destructor. */
652
    ~TxGraphImpl() noexcept;
653
654
    // Cannot move or copy (would invalidate TxGraphImpl* in Ref, MiningOrder, EvictionOrder).
655
    TxGraphImpl(const TxGraphImpl&) = delete;
656
    TxGraphImpl& operator=(const TxGraphImpl&) = delete;
657
    TxGraphImpl(TxGraphImpl&&) = delete;
658
    TxGraphImpl& operator=(TxGraphImpl&&) = delete;
659
660
    // Simple helper functions.
661
662
    /** Swap the Entry referred to by a and the one referred to by b. Gather main clusters to
663
     *  update afterwards. */
664
    void SwapIndexes(GraphIndex a, GraphIndex b, std::vector<Cluster*>& affected_main) noexcept;
665
    /** If idx exists in the specified level ClusterSet (explicitly, or in the level below and not
666
    *   removed), return the Cluster it is in. Otherwise, return nullptr. */
667
993k
    Cluster* FindCluster(GraphIndex idx, int level) const noexcept { return FindClusterAndLevel(idx, level).first; }
668
    /** Like FindCluster, but also return what level the match was found in (-1 if not found). */
669
    std::pair<Cluster*, int> FindClusterAndLevel(GraphIndex idx, int level) const noexcept;
670
    /** Extract a Cluster from its ClusterSet, and set its quality to QualityLevel::NONE. */
671
    std::unique_ptr<Cluster> ExtractCluster(int level, QualityLevel quality, ClusterSetIndex setindex) noexcept;
672
    /** Delete a Cluster. */
673
    void DeleteCluster(Cluster& cluster, int level) noexcept;
674
    /** Insert a Cluster into its ClusterSet. */
675
    ClusterSetIndex InsertCluster(int level, std::unique_ptr<Cluster>&& cluster, QualityLevel quality) noexcept;
676
    /** Change the QualityLevel of a Cluster (identified by old_quality and old_index). */
677
    void SetClusterQuality(int level, QualityLevel old_quality, ClusterSetIndex old_index, QualityLevel new_quality) noexcept;
678
    /** Get the index of the top level ClusterSet (staging if it exists, main otherwise). */
679
3.90M
    int GetTopLevel() const noexcept { return m_staging_clusterset.has_value(); }
680
    /** Get the specified level (staging if it exists and level is TOP, main otherwise). */
681
532k
    int GetSpecifiedLevel(Level level) const noexcept { return level == Level::TOP && m_staging_clusterset.has_value(); }
682
    /** Get a reference to the ClusterSet at the specified level (which must exist). */
683
    ClusterSet& GetClusterSet(int level) noexcept;
684
    const ClusterSet& GetClusterSet(int level) const noexcept;
685
    /** Make a transaction not exist at a specified level. It must currently exist there.
686
     *  oversized_tx indicates whether the transaction is an individually-oversized one
687
     *  (OVERSIZED_SINGLETON). */
688
    void ClearLocator(int level, GraphIndex index, bool oversized_tx) noexcept;
689
    /** Find which Clusters in main conflict with ones in staging. */
690
    std::vector<Cluster*> GetConflicts() const noexcept;
691
    /** Clear an Entry's ChunkData. */
692
    void ClearChunkData(Entry& entry) noexcept;
693
    /** Give an Entry a ChunkData object. */
694
    void CreateChunkData(GraphIndex idx, LinearizationIndex chunk_count) noexcept;
695
    /** Create an empty GenericClusterImpl object. */
696
    std::unique_ptr<GenericClusterImpl> CreateEmptyGenericCluster() noexcept
697
1.63k
    {
698
1.63k
        return std::make_unique<GenericClusterImpl>(m_next_sequence_counter++);
699
1.63k
    }
700
    /** Create an empty SingletonClusterImpl object. */
701
    std::unique_ptr<SingletonClusterImpl> CreateEmptySingletonCluster() noexcept
702
129k
    {
703
129k
        return std::make_unique<SingletonClusterImpl>(m_next_sequence_counter++);
704
129k
    }
705
    /** Create an empty Cluster of the appropriate implementation for the specified (maximum) tx
706
     *  count. */
707
    std::unique_ptr<Cluster> CreateEmptyCluster(DepGraphIndex tx_count) noexcept
708
129k
    {
709
129k
        if (tx_count >= SingletonClusterImpl::MIN_INTENDED_TX_COUNT && tx_count <= SingletonClusterImpl::MAX_TX_COUNT) {
710
127k
            return CreateEmptySingletonCluster();
711
127k
        }
712
1.29k
        if (tx_count >= GenericClusterImpl::MIN_INTENDED_TX_COUNT && tx_count <= GenericClusterImpl::MAX_TX_COUNT) {
713
1.29k
            return CreateEmptyGenericCluster();
714
1.29k
        }
715
1.29k
        assert(false);
716
0
        return {};
717
0
    }
718
719
    // Functions for handling Refs.
720
721
    /** Only called by Ref's move constructor/assignment to update Ref locations. */
722
    void UpdateRef(GraphIndex idx, Ref& new_location) noexcept final
723
65.5k
    {
724
65.5k
        auto& entry = m_entries[idx];
725
65.5k
        Assume(entry.m_ref != nullptr);
726
65.5k
        entry.m_ref = &new_location;
727
65.5k
    }
728
729
    /** Only called by Ref::~Ref to unlink Refs, and Ref's move assignment. */
730
    void UnlinkRef(GraphIndex idx) noexcept final
731
63.5k
    {
732
63.5k
        auto& entry = m_entries[idx];
733
63.5k
        Assume(entry.m_ref != nullptr);
734
63.5k
        Assume(m_main_chunkindex_observers == 0 || !entry.m_locator[0].IsPresent());
735
        // Remove all chunk index entries for the affected cluster, to avoid any chunk indexes
736
        // referencing unlinked/destroyed Refs.
737
63.5k
        if (entry.m_locator[0].IsPresent()) {
738
50.6k
            entry.m_locator[0].cluster->RemoveChunkData(*this);
739
50.6k
        }
740
63.5k
        entry.m_ref = nullptr;
741
        // Mark the transaction as to be removed in all levels where it explicitly or implicitly
742
        // exists.
743
63.5k
        bool exists_anywhere{false};
744
63.5k
        bool exists{false};
745
127k
        for (int level = 0; level <= GetTopLevel(); ++level) {
746
63.5k
            if (entry.m_locator[level].IsPresent()) {
747
50.6k
                exists_anywhere = true;
748
50.6k
                exists = true;
749
50.6k
            } else if (entry.m_locator[level].IsRemoved()) {
750
0
                exists = false;
751
0
            }
752
63.5k
            if (exists) {
753
50.6k
                auto& clusterset = GetClusterSet(level);
754
50.6k
                clusterset.m_to_remove.push_back(idx);
755
                // Force recomputation of grouping data.
756
50.6k
                clusterset.m_group_data = std::nullopt;
757
                // Do not wipe the oversized state of main if staging exists. The reason for this
758
                // is that the alternative would mean that cluster merges may need to be applied to
759
                // a formerly-oversized main graph while staging exists (to satisfy chunk feerate
760
                // queries into main, for example), and such merges could conflict with pulls of
761
                // some of their constituents into staging.
762
50.6k
                if (level == GetTopLevel() && clusterset.m_oversized == true) {
763
0
                    clusterset.m_oversized = std::nullopt;
764
0
                }
765
50.6k
            }
766
63.5k
        }
767
63.5k
        m_unlinked.push_back(idx);
768
63.5k
        if (!exists_anywhere) Compact();
769
63.5k
    }
770
771
    // Functions related to various normalization/application steps.
772
    /** Get rid of unlinked Entry objects in m_entries, if possible (this changes the GraphIndex
773
     *  values for remaining Entry objects, so this only does something when no to-be-applied
774
     *  operations or staged removals referring to GraphIndexes remain). */
775
    void Compact() noexcept;
776
    /** If cluster is not in staging, copy it there, and return a pointer to it.
777
    *   Staging must exist, and this modifies the locators of its
778
    *   transactions from inherited (P,M) to explicit (P,P). */
779
    Cluster* PullIn(Cluster* cluster, int level) noexcept;
780
    /** Apply all removals queued up in m_to_remove to the relevant Clusters (which get a
781
     *  NEEDS_SPLIT* QualityLevel) up to the specified level. */
782
    void ApplyRemovals(int up_to_level) noexcept;
783
    /** Split an individual cluster. */
784
    void Split(Cluster& cluster, int level) noexcept;
785
    /** Split all clusters that need splitting up to the specified level. */
786
    void SplitAll(int up_to_level) noexcept;
787
    /** Populate m_group_data based on m_deps_to_add in the specified level. */
788
    void GroupClusters(int level) noexcept;
789
    /** Merge the specified clusters. */
790
    void Merge(std::span<Cluster*> to_merge, int level) noexcept;
791
    /** Apply all m_deps_to_add to the relevant Clusters in the specified level. */
792
    void ApplyDependencies(int level) noexcept;
793
    /** Make a specified Cluster have quality ACCEPTABLE or OPTIMAL. */
794
    void MakeAcceptable(Cluster& cluster, int level) noexcept;
795
    /** Make all Clusters at the specified level have quality ACCEPTABLE or OPTIMAL. */
796
    void MakeAllAcceptable(int level) noexcept;
797
798
    // Implementations for the public TxGraph interface.
799
800
    void AddTransaction(Ref& arg, const FeePerWeight& feerate) noexcept final;
801
    void RemoveTransaction(const Ref& arg) noexcept final;
802
    void AddDependency(const Ref& parent, const Ref& child) noexcept final;
803
    void SetTransactionFee(const Ref&, int64_t fee) noexcept final;
804
805
    bool DoWork(uint64_t max_cost) noexcept final;
806
807
    void StartStaging() noexcept final;
808
    void CommitStaging() noexcept final;
809
    void AbortStaging() noexcept final;
810
62.6k
    bool HaveStaging() const noexcept final { return m_staging_clusterset.has_value(); }
811
812
    bool Exists(const Ref& arg, Level level) noexcept final;
813
    FeePerWeight GetMainChunkFeerate(const Ref& arg) noexcept final;
814
    FeePerWeight GetIndividualFeerate(const Ref& arg) noexcept final;
815
    std::vector<Ref*> GetCluster(const Ref& arg, Level level) noexcept final;
816
    std::vector<Ref*> GetAncestors(const Ref& arg, Level level) noexcept final;
817
    std::vector<Ref*> GetDescendants(const Ref& arg, Level level) noexcept final;
818
    std::vector<Ref*> GetAncestorsUnion(std::span<const Ref* const> args, Level level) noexcept final;
819
    std::vector<Ref*> GetDescendantsUnion(std::span<const Ref* const> args, Level level) noexcept final;
820
    GraphIndex GetTransactionCount(Level level) noexcept final;
821
    bool IsOversized(Level level) noexcept final;
822
    std::strong_ordering CompareMainOrder(const Ref& a, const Ref& b) noexcept final;
823
    GraphIndex CountDistinctClusters(std::span<const Ref* const> refs, Level level) noexcept final;
824
    std::pair<std::vector<FeeFrac>, std::vector<FeeFrac>> GetMainStagingDiagrams() noexcept final;
825
    std::vector<Ref*> Trim() noexcept final;
826
827
    std::unique_ptr<BlockBuilder> GetBlockBuilder() noexcept final;
828
    std::pair<std::vector<Ref*>, FeePerWeight> GetWorstMainChunk() noexcept final;
829
830
    size_t GetMainMemoryUsage() noexcept final;
831
832
    void SanityCheck() const final;
833
};
834
835
TxGraphImpl::ClusterSet& TxGraphImpl::GetClusterSet(int level) noexcept
836
281M
{
837
281M
    if (level == 0) return m_main_clusterset;
838
425k
    Assume(level == 1);
839
425k
    Assume(m_staging_clusterset.has_value());
840
425k
    return *m_staging_clusterset;
841
281M
}
842
843
const TxGraphImpl::ClusterSet& TxGraphImpl::GetClusterSet(int level) const noexcept
844
181k
{
845
181k
    if (level == 0) return m_main_clusterset;
846
53.2k
    Assume(level == 1);
847
53.2k
    Assume(m_staging_clusterset.has_value());
848
53.2k
    return *m_staging_clusterset;
849
181k
}
850
851
/** Implementation of the TxGraph::BlockBuilder interface. */
852
class BlockBuilderImpl final : public TxGraph::BlockBuilder
853
{
854
    /** Which TxGraphImpl this object is doing block building for. It will have its
855
     *  m_main_chunkindex_observers incremented as long as this BlockBuilderImpl exists. */
856
    TxGraphImpl* const m_graph;
857
    /** Cluster sequence numbers which we're not including further transactions from. */
858
    std::unordered_set<uint64_t> m_excluded_clusters;
859
    /** Iterator to the current chunk in the chunk index. end() if nothing further remains. */
860
    TxGraphImpl::ChunkIndex::const_iterator m_cur_iter;
861
    /** Which cluster the current chunk belongs to, so we can exclude further transactions from it
862
     *  when that chunk is skipped. */
863
    Cluster* m_cur_cluster;
864
    /** Whether we know that m_cur_iter points to the last chunk of m_cur_cluster. */
865
    bool m_known_end_of_cluster;
866
867
    // Move m_cur_iter / m_cur_cluster to the next acceptable chunk.
868
    void Next() noexcept;
869
870
public:
871
    /** Construct a new BlockBuilderImpl to build blocks for the provided graph. */
872
    BlockBuilderImpl(TxGraphImpl& graph) noexcept;
873
874
    // Implement the public interface.
875
    ~BlockBuilderImpl() final;
876
    std::optional<std::pair<std::vector<TxGraph::Ref*>, FeePerWeight>> GetCurrentChunk() noexcept final;
877
    void Include() noexcept final;
878
    void Skip() noexcept final;
879
};
880
881
void TxGraphImpl::ClearChunkData(Entry& entry) noexcept
882
489k
{
883
489k
    if (entry.m_main_chunkindex_iterator != m_main_chunkindex.end()) {
884
113k
        Assume(m_main_chunkindex_observers == 0);
885
        // If the Entry has a non-empty m_main_chunkindex_iterator, extract it, and move the handle
886
        // to the cache of discarded chunkindex entries.
887
113k
        m_main_chunkindex_discarded.emplace_back(m_main_chunkindex.extract(entry.m_main_chunkindex_iterator));
888
113k
        entry.m_main_chunkindex_iterator = m_main_chunkindex.end();
889
113k
    }
890
489k
}
891
892
void TxGraphImpl::CreateChunkData(GraphIndex idx, LinearizationIndex chunk_count) noexcept
893
177k
{
894
177k
    auto& entry = m_entries[idx];
895
    // Make sure to not create chunk data for unlinked entries, which would make invoking
896
    // m_fallback_order on them impossible.
897
177k
    Assume(entry.m_ref != nullptr);
898
177k
    if (!m_main_chunkindex_discarded.empty()) {
899
        // Reuse an discarded node handle.
900
1.38k
        auto& node = m_main_chunkindex_discarded.back().value();
901
1.38k
        node.m_graph_index = idx;
902
1.38k
        node.m_chunk_count = chunk_count;
903
1.38k
        auto insert_result = m_main_chunkindex.insert(std::move(m_main_chunkindex_discarded.back()));
904
1.38k
        Assume(insert_result.inserted);
905
1.38k
        entry.m_main_chunkindex_iterator = insert_result.position;
906
1.38k
        m_main_chunkindex_discarded.pop_back();
907
176k
    } else {
908
        // Construct a new entry.
909
176k
        auto emplace_result = m_main_chunkindex.emplace(idx, chunk_count);
910
176k
        Assume(emplace_result.second);
911
176k
        entry.m_main_chunkindex_iterator = emplace_result.first;
912
176k
    }
913
177k
}
914
915
size_t GenericClusterImpl::TotalMemoryUsage() const noexcept
916
67.7k
{
917
67.7k
    return // Dynamic memory allocated in this Cluster.
918
67.7k
           memusage::DynamicUsage(m_mapping) + memusage::DynamicUsage(m_linearization) +
919
           // Dynamic memory usage inside m_depgraph.
920
67.7k
           m_depgraph.DynamicMemoryUsage() +
921
           // Memory usage of the allocated Cluster itself.
922
67.7k
           memusage::MallocUsage(sizeof(GenericClusterImpl)) +
923
           // Memory usage of the ClusterSet::m_clusters entry.
924
67.7k
           sizeof(std::unique_ptr<Cluster>);
925
67.7k
}
926
927
size_t SingletonClusterImpl::TotalMemoryUsage() const noexcept
928
10.9M
{
929
10.9M
    return // Memory usage of the allocated SingletonClusterImpl itself.
930
10.9M
           memusage::MallocUsage(sizeof(SingletonClusterImpl)) +
931
           // Memory usage of the ClusterSet::m_clusters entry.
932
10.9M
           sizeof(std::unique_ptr<Cluster>);
933
10.9M
}
934
935
uint64_t GenericClusterImpl::GetTotalTxSize() const noexcept
936
59.4k
{
937
59.4k
    uint64_t ret{0};
938
300k
    for (auto i : m_linearization) {
939
300k
        ret += m_depgraph.FeeRate(i).size;
940
300k
    }
941
59.4k
    return ret;
942
59.4k
}
943
944
DepGraphIndex GenericClusterImpl::AppendTransaction(GraphIndex graph_idx, FeePerWeight feerate) noexcept
945
15
{
946
15
    Assume(graph_idx != GraphIndex(-1));
947
15
    auto ret = m_depgraph.AddTransaction(feerate);
948
15
    m_mapping.push_back(graph_idx);
949
15
    m_linearization.push_back(ret);
950
15
    return ret;
951
15
}
952
953
DepGraphIndex SingletonClusterImpl::AppendTransaction(GraphIndex graph_idx, FeePerWeight feerate) noexcept
954
127k
{
955
127k
    Assume(!GetTxCount());
956
127k
    m_graph_index = graph_idx;
957
127k
    m_feerate = feerate;
958
127k
    return 0;
959
127k
}
960
961
void GenericClusterImpl::AddDependencies(SetType parents, DepGraphIndex child) noexcept
962
15
{
963
15
    m_depgraph.AddDependencies(parents, child);
964
15
}
965
966
void SingletonClusterImpl::AddDependencies(SetType parents, DepGraphIndex child) noexcept
967
386
{
968
    // Singletons cannot have any dependencies.
969
386
    Assume(child == 0);
970
386
    Assume(parents == SetType{} || parents == SetType::Fill(0));
971
386
}
972
973
void GenericClusterImpl::ExtractTransactions(const std::function<void (DepGraphIndex, GraphIndex, FeePerWeight)>& visit1_fn, const std::function<void (DepGraphIndex, GraphIndex, SetType)>& visit2_fn) noexcept
974
11
{
975
36
    for (auto pos : m_linearization) {
976
36
        visit1_fn(pos, m_mapping[pos], FeePerWeight::FromFeeFrac(m_depgraph.FeeRate(pos)));
977
36
    }
978
36
    for (auto pos : m_linearization) {
979
36
        visit2_fn(pos, m_mapping[pos], m_depgraph.GetReducedParents(pos));
980
36
    }
981
    // Purge this Cluster, now that everything has been moved.
982
11
    m_depgraph = DepGraph<SetType>{};
983
11
    m_linearization.clear();
984
11
    m_mapping.clear();
985
11
}
986
987
void SingletonClusterImpl::ExtractTransactions(const std::function<void (DepGraphIndex, GraphIndex, FeePerWeight)>& visit1_fn, const std::function<void (DepGraphIndex, GraphIndex, SetType)>& visit2_fn) noexcept
988
6.98k
{
989
6.98k
    if (GetTxCount()) {
990
6.98k
        visit1_fn(0, m_graph_index, m_feerate);
991
6.98k
        visit2_fn(0, m_graph_index, SetType{});
992
6.98k
        m_graph_index = NO_GRAPH_INDEX;
993
6.98k
    }
994
6.98k
}
995
996
int GenericClusterImpl::GetLevel(const TxGraphImpl& graph) const noexcept
997
54.6k
{
998
    // GetLevel() does not work for empty Clusters.
999
54.6k
    if (!Assume(!m_linearization.empty())) return -1;
1000
1001
    // Pick an arbitrary Entry that occurs in this Cluster.
1002
54.6k
    const auto& entry = graph.m_entries[m_mapping[m_linearization.front()]];
1003
    // See if there is a level whose Locator matches this Cluster, if so return that level.
1004
54.7k
    for (int level = 0; level < MAX_LEVELS; ++level) {
1005
54.7k
        if (entry.m_locator[level].cluster == this) return level;
1006
54.7k
    }
1007
    // Given that we started with an Entry that occurs in this Cluster, one of its Locators must
1008
    // point back to it.
1009
54.6k
    assert(false);
1010
0
    return -1;
1011
0
}
1012
1013
int SingletonClusterImpl::GetLevel(const TxGraphImpl& graph) const noexcept
1014
10.7M
{
1015
    // GetLevel() does not work for empty Clusters.
1016
10.7M
    if (!Assume(GetTxCount())) return -1;
1017
1018
    // Get the Entry in this Cluster.
1019
10.7M
    const auto& entry = graph.m_entries[m_graph_index];
1020
    // See if there is a level whose Locator matches this Cluster, if so return that level.
1021
10.7M
    for (int level = 0; level < MAX_LEVELS; ++level) {
1022
10.7M
        if (entry.m_locator[level].cluster == this) return level;
1023
10.7M
    }
1024
    // Given that we started with an Entry that occurs in this Cluster, one of its Locators must
1025
    // point back to it.
1026
10.7M
    assert(false);
1027
0
    return -1;
1028
0
}
1029
1030
void TxGraphImpl::ClearLocator(int level, GraphIndex idx, bool oversized_tx) noexcept
1031
51.1k
{
1032
51.1k
    auto& entry = m_entries[idx];
1033
51.1k
    auto& clusterset = GetClusterSet(level);
1034
51.1k
    Assume(entry.m_locator[level].IsPresent());
1035
    // Change the locator from Present to Missing or Removed.
1036
51.1k
    if (level == 0 || !entry.m_locator[level - 1].IsPresent()) {
1037
49.0k
        entry.m_locator[level].SetMissing();
1038
49.0k
    } else {
1039
2.17k
        entry.m_locator[level].SetRemoved();
1040
2.17k
        clusterset.m_removed.push_back(idx);
1041
2.17k
    }
1042
    // Update the transaction count.
1043
51.1k
    --clusterset.m_txcount;
1044
51.1k
    clusterset.m_txcount_oversized -= oversized_tx;
1045
    // If clearing main, adjust the status of Locators of this transaction in staging, if it exists.
1046
51.1k
    if (level == 0 && GetTopLevel() == 1) {
1047
2.14k
        if (entry.m_locator[1].IsRemoved()) {
1048
1.59k
            entry.m_locator[1].SetMissing();
1049
1.59k
        } else if (!entry.m_locator[1].IsPresent()) {
1050
0
            --m_staging_clusterset->m_txcount;
1051
0
            m_staging_clusterset->m_txcount_oversized -= oversized_tx;
1052
0
        }
1053
2.14k
    }
1054
51.1k
    if (level == 0) ClearChunkData(entry);
1055
51.1k
}
1056
1057
void GenericClusterImpl::RemoveChunkData(TxGraphImpl& graph) noexcept
1058
6.03k
{
1059
134k
    for (DepGraphIndex idx : m_linearization) {
1060
134k
        auto& entry = graph.m_entries[m_mapping[idx]];
1061
134k
        graph.ClearChunkData(entry);
1062
134k
    }
1063
6.03k
}
1064
1065
void SingletonClusterImpl::RemoveChunkData(TxGraphImpl& graph) noexcept
1066
44.6k
{
1067
44.6k
    if (GetTxCount() == 0) return;
1068
44.6k
    auto& entry = graph.m_entries[m_graph_index];
1069
44.6k
    graph.ClearChunkData(entry);
1070
44.6k
}
1071
1072
void GenericClusterImpl::Updated(TxGraphImpl& graph, int level, bool rename) noexcept
1073
12.4k
{
1074
    // Update all the Locators for this Cluster's Entry objects.
1075
145k
    for (DepGraphIndex idx : m_linearization) {
1076
145k
        auto& entry = graph.m_entries[m_mapping[idx]];
1077
        // Discard any potential ChunkData prior to modifying the Cluster (as that could
1078
        // invalidate its ordering).
1079
145k
        if (level == 0 && !rename) graph.ClearChunkData(entry);
1080
145k
        entry.m_locator[level].SetPresent(this, idx);
1081
145k
    }
1082
    // If this is for the main graph (level = 0), and the Cluster's quality is ACCEPTABLE or
1083
    // OPTIMAL, compute its chunking and store its information in the Entry's m_main_lin_index
1084
    // and m_main_chunk_feerate. These fields are only accessed after making the entire graph
1085
    // ACCEPTABLE, so it is pointless to compute these if we haven't reached that quality level
1086
    // yet.
1087
    // When rename=true, this is always performed for level 0, to make sure the values inside the
1088
    // entries remain consistent with the chunk index (otherwise unrelated chunk index operations
1089
    // could cause the index to become corrupted, by inserting elements in the wrong place).
1090
12.4k
    if (level == 0 && (rename || IsAcceptable())) {
1091
5.68k
        auto chunking = ChunkLinearizationInfo(m_depgraph, m_linearization);
1092
5.68k
        LinearizationIndex lin_idx{0};
1093
        /** The sum of all chunk feerate FeeFracs with the same feerate as the current chunk,
1094
         *  up to and including the current chunk. */
1095
5.68k
        FeeFrac equal_feerate_chunk_feerate;
1096
        // Iterate over the chunks.
1097
70.3k
        for (unsigned chunk_idx = 0; chunk_idx < chunking.size(); ++chunk_idx) {
1098
64.7k
            auto& chunk = chunking[chunk_idx];
1099
64.7k
            auto chunk_count = chunk.transactions.Count();
1100
64.7k
            Assume(chunk_count > 0);
1101
            // Update equal_feerate_chunk_feerate to include this chunk, starting over when the
1102
            // feerate changed.
1103
64.7k
            if (chunk.feerate << equal_feerate_chunk_feerate) {
1104
3.24k
                equal_feerate_chunk_feerate = chunk.feerate;
1105
61.4k
            } else {
1106
                // Note that this is adding fees to fees, and sizes to sizes, so the overall
1107
                // ratio remains the same; it's just accounting for the size of the added chunk.
1108
61.4k
                equal_feerate_chunk_feerate += chunk.feerate;
1109
61.4k
            }
1110
            // Determine the m_fallback_order maximum transaction in the chunk.
1111
64.7k
            auto it = chunk.transactions.begin();
1112
64.7k
            GraphIndex max_element = m_mapping[*it];
1113
64.7k
            ++it;
1114
72.9k
            while (it != chunk.transactions.end()) {
1115
8.23k
                GraphIndex this_element = m_mapping[*it];
1116
8.23k
                if (graph.m_fallback_order(*graph.m_entries[this_element].m_ref, *graph.m_entries[max_element].m_ref) > 0) {
1117
1.69k
                    max_element = this_element;
1118
1.69k
                }
1119
8.23k
                ++it;
1120
8.23k
            }
1121
            // Iterate over the transactions in the linearization, which must match those in chunk.
1122
72.9k
            while (true) {
1123
72.9k
                DepGraphIndex idx = m_linearization[lin_idx];
1124
72.9k
                GraphIndex graph_idx = m_mapping[idx];
1125
72.9k
                auto& entry = graph.m_entries[graph_idx];
1126
72.9k
                entry.m_main_lin_index = lin_idx++;
1127
72.9k
                entry.m_main_chunk_feerate = FeePerWeight::FromFeeFrac(chunk.feerate);
1128
72.9k
                entry.m_main_equal_feerate_chunk_prefix_size = equal_feerate_chunk_feerate.size;
1129
72.9k
                entry.m_main_max_chunk_fallback = max_element;
1130
72.9k
                Assume(chunk.transactions[idx]);
1131
72.9k
                chunk.transactions.Reset(idx);
1132
72.9k
                if (chunk.transactions.None()) {
1133
                    // Last transaction in the chunk.
1134
64.7k
                    if (chunk_count == 1 && chunk_idx + 1 == chunking.size()) {
1135
                        // If this is the final chunk of the cluster, and it contains just a single
1136
                        // transaction (which will always be true for the very common singleton
1137
                        // clusters), store the special value -1 as chunk count.
1138
4.25k
                        chunk_count = LinearizationIndex(-1);
1139
4.25k
                    }
1140
64.7k
                    if (!rename) graph.CreateChunkData(graph_idx, chunk_count);
1141
64.7k
                    break;
1142
64.7k
                }
1143
72.9k
            }
1144
64.7k
        }
1145
5.68k
    }
1146
12.4k
}
1147
1148
void SingletonClusterImpl::Updated(TxGraphImpl& graph, int level, bool rename) noexcept
1149
193k
{
1150
    // Don't do anything if this is empty.
1151
193k
    if (GetTxCount() == 0) return;
1152
1153
193k
    auto& entry = graph.m_entries[m_graph_index];
1154
    // Discard any potential ChunkData prior to modifying the Cluster (as that could
1155
    // invalidate its ordering).
1156
193k
    if (level == 0 && !rename) graph.ClearChunkData(entry);
1157
193k
    entry.m_locator[level].SetPresent(this, 0);
1158
    // If this is for the main graph (level = 0), compute its chunking and store its information in
1159
    // the Entry's m_main_lin_index and m_main_chunk_feerate.
1160
193k
    if (level == 0 && (rename || IsAcceptable())) {
1161
128k
        entry.m_main_lin_index = 0;
1162
128k
        entry.m_main_chunk_feerate = m_feerate;
1163
128k
        entry.m_main_equal_feerate_chunk_prefix_size = m_feerate.size;
1164
128k
        entry.m_main_max_chunk_fallback = m_graph_index;
1165
        // Always use the special LinearizationIndex(-1), indicating singleton chunk at end of
1166
        // Cluster, here.
1167
128k
        if (!rename) graph.CreateChunkData(m_graph_index, LinearizationIndex(-1));
1168
128k
    }
1169
193k
}
1170
1171
void GenericClusterImpl::GetConflicts(const TxGraphImpl& graph, std::vector<Cluster*>& out) const noexcept
1172
141
{
1173
1.26k
    for (auto i : m_linearization) {
1174
1.26k
        auto& entry = graph.m_entries[m_mapping[i]];
1175
        // For every transaction Entry in this Cluster, if it also exists in a lower-level Cluster,
1176
        // then that Cluster conflicts.
1177
1.26k
        if (entry.m_locator[0].IsPresent()) {
1178
1.10k
            out.push_back(entry.m_locator[0].cluster);
1179
1.10k
        }
1180
1.26k
    }
1181
141
}
1182
1183
void SingletonClusterImpl::GetConflicts(const TxGraphImpl& graph, std::vector<Cluster*>& out) const noexcept
1184
53.2k
{
1185
    // Empty clusters have no conflicts.
1186
53.2k
    if (GetTxCount() == 0) return;
1187
1188
53.2k
    auto& entry = graph.m_entries[m_graph_index];
1189
    // If the transaction in this Cluster also exists in a lower-level Cluster, then that Cluster
1190
    // conflicts.
1191
53.2k
    if (entry.m_locator[0].IsPresent()) {
1192
130
        out.push_back(entry.m_locator[0].cluster);
1193
130
    }
1194
53.2k
}
1195
1196
std::vector<Cluster*> TxGraphImpl::GetConflicts() const noexcept
1197
53.2k
{
1198
53.2k
    Assume(GetTopLevel() == 1);
1199
53.2k
    auto& clusterset = GetClusterSet(1);
1200
53.2k
    std::vector<Cluster*> ret;
1201
    // All main Clusters containing transactions in m_removed (so (P,R) ones) are conflicts.
1202
53.2k
    for (auto i : clusterset.m_removed) {
1203
3.75k
        auto& entry = m_entries[i];
1204
3.75k
        if (entry.m_locator[0].IsPresent()) {
1205
3.75k
            ret.push_back(entry.m_locator[0].cluster);
1206
3.75k
        }
1207
3.75k
    }
1208
    // Then go over all Clusters at this level, and find their conflicts (the (P,P) ones).
1209
425k
    for (int quality = 0; quality < int(QualityLevel::NONE); ++quality) {
1210
372k
        auto& clusters = clusterset.m_clusters[quality];
1211
372k
        for (const auto& cluster : clusters) {
1212
53.4k
            cluster->GetConflicts(*this, ret);
1213
53.4k
        }
1214
372k
    }
1215
    // Deduplicate the result (the same Cluster may appear multiple times).
1216
53.2k
    std::sort(ret.begin(), ret.end(), [](Cluster* a, Cluster* b) noexcept { return CompareClusters(a, b) < 0; });
1217
53.2k
    ret.erase(std::unique(ret.begin(), ret.end()), ret.end());
1218
53.2k
    return ret;
1219
53.2k
}
1220
1221
Cluster* GenericClusterImpl::CopyToStaging(TxGraphImpl& graph) const noexcept
1222
336
{
1223
    // Construct an empty Cluster.
1224
336
    auto ret = graph.CreateEmptyGenericCluster();
1225
336
    auto ptr = ret.get();
1226
    // Copy depgraph, mapping, and linearization.
1227
336
    ptr->m_depgraph = m_depgraph;
1228
336
    ptr->m_mapping = m_mapping;
1229
336
    ptr->m_linearization = m_linearization;
1230
    // Insert the new Cluster into the graph.
1231
336
    graph.InsertCluster(/*level=*/1, std::move(ret), m_quality);
1232
    // Update its Locators.
1233
336
    ptr->Updated(graph, /*level=*/1, /*rename=*/false);
1234
    // Update memory usage.
1235
336
    graph.GetClusterSet(/*level=*/1).m_cluster_usage += ptr->TotalMemoryUsage();
1236
336
    return ptr;
1237
336
}
1238
1239
Cluster* SingletonClusterImpl::CopyToStaging(TxGraphImpl& graph) const noexcept
1240
1.40k
{
1241
    // Construct an empty Cluster.
1242
1.40k
    auto ret = graph.CreateEmptySingletonCluster();
1243
1.40k
    auto ptr = ret.get();
1244
    // Copy data.
1245
1.40k
    ptr->m_graph_index = m_graph_index;
1246
1.40k
    ptr->m_feerate = m_feerate;
1247
    // Insert the new Cluster into the graph.
1248
1.40k
    graph.InsertCluster(/*level=*/1, std::move(ret), m_quality);
1249
    // Update its Locators.
1250
1.40k
    ptr->Updated(graph, /*level=*/1, /*rename=*/false);
1251
    // Update memory usage.
1252
1.40k
    graph.GetClusterSet(/*level=*/1).m_cluster_usage += ptr->TotalMemoryUsage();
1253
1.40k
    return ptr;
1254
1.40k
}
1255
1256
void GenericClusterImpl::ApplyRemovals(TxGraphImpl& graph, int level, std::span<GraphIndex>& to_remove) noexcept
1257
1.13k
{
1258
    // Iterate over the prefix of to_remove that applies to this cluster.
1259
1.13k
    Assume(!to_remove.empty());
1260
1.13k
    SetType todo;
1261
1.13k
    graph.GetClusterSet(level).m_cluster_usage -= TotalMemoryUsage();
1262
5.17k
    do {
1263
5.17k
        GraphIndex idx = to_remove.front();
1264
5.17k
        Assume(idx < graph.m_entries.size());
1265
5.17k
        auto& entry = graph.m_entries[idx];
1266
5.17k
        auto& locator = entry.m_locator[level];
1267
        // Stop once we hit an entry that applies to another Cluster.
1268
5.17k
        if (locator.cluster != this) break;
1269
        // - Remember it in a set of to-remove DepGraphIndexes.
1270
4.40k
        todo.Set(locator.index);
1271
        // - Remove from m_mapping. This isn't strictly necessary as unused positions in m_mapping
1272
        //   are just never accessed, but set it to -1 here to increase the ability to detect a bug
1273
        //   that causes it to be accessed regardless.
1274
4.40k
        m_mapping[locator.index] = GraphIndex(-1);
1275
        // - Remove its linearization index from the Entry (if in main).
1276
4.40k
        if (level == 0) {
1277
3.63k
            entry.m_main_lin_index = LinearizationIndex(-1);
1278
3.63k
        }
1279
        // - Mark it as missing/removed in the Entry's locator.
1280
4.40k
        graph.ClearLocator(level, idx, m_quality == QualityLevel::OVERSIZED_SINGLETON);
1281
4.40k
        to_remove = to_remove.subspan(1);
1282
4.40k
    } while(!to_remove.empty());
1283
1284
1.13k
    Assume(todo.Any());
1285
    // Wipe from the Cluster's DepGraph (this is O(n) regardless of the number of entries
1286
    // removed, so we benefit from batching all the removals).
1287
1.13k
    m_depgraph.RemoveTransactions(todo);
1288
1.13k
    m_mapping.resize(m_depgraph.PositionRange());
1289
1290
    // Filter removed transactions out of m_linearization.
1291
1.13k
    m_linearization.erase(std::remove_if(m_linearization.begin(), m_linearization.end(),
1292
5.22k
                                         [&](auto pos) { return todo[pos]; }),
1293
1.13k
                          m_linearization.end());
1294
1295
1.13k
    Compact();
1296
1.13k
    graph.GetClusterSet(level).m_cluster_usage += TotalMemoryUsage();
1297
1.13k
    auto new_quality = IsTopological() ? QualityLevel::NEEDS_SPLIT : QualityLevel::NEEDS_SPLIT_FIX;
1298
1.13k
    graph.SetClusterQuality(level, m_quality, m_setindex, new_quality);
1299
1.13k
    Updated(graph, /*level=*/level, /*rename=*/false);
1300
1.13k
}
1301
1302
void SingletonClusterImpl::ApplyRemovals(TxGraphImpl& graph, int level, std::span<GraphIndex>& to_remove) noexcept
1303
44.6k
{
1304
    // We can only remove the one transaction this Cluster has.
1305
44.6k
    Assume(!to_remove.empty());
1306
44.6k
    Assume(GetTxCount());
1307
44.6k
    Assume(to_remove.front() == m_graph_index);
1308
    // Pop all copies of m_graph_index from the front of to_remove (at least one, but there may be
1309
    // multiple).
1310
44.6k
    do {
1311
44.6k
        to_remove = to_remove.subspan(1);
1312
44.6k
    } while (!to_remove.empty() && to_remove.front() == m_graph_index);
1313
    // Clear this cluster.
1314
44.6k
    graph.ClearLocator(level, m_graph_index, m_quality == QualityLevel::OVERSIZED_SINGLETON);
1315
44.6k
    m_graph_index = NO_GRAPH_INDEX;
1316
44.6k
    graph.SetClusterQuality(level, m_quality, m_setindex, QualityLevel::NEEDS_SPLIT);
1317
    // No need to account for m_cluster_usage changes here, as SingletonClusterImpl has constant
1318
    // memory usage.
1319
44.6k
}
1320
1321
void GenericClusterImpl::Clear(TxGraphImpl& graph, int level) noexcept
1322
232
{
1323
232
    Assume(GetTxCount());
1324
232
    graph.GetClusterSet(level).m_cluster_usage -= TotalMemoryUsage();
1325
1.10k
    for (auto i : m_linearization) {
1326
1.10k
        graph.ClearLocator(level, m_mapping[i], m_quality == QualityLevel::OVERSIZED_SINGLETON);
1327
1.10k
    }
1328
232
    m_depgraph = {};
1329
232
    m_linearization.clear();
1330
232
    m_mapping.clear();
1331
232
}
1332
1333
void SingletonClusterImpl::Clear(TxGraphImpl& graph, int level) noexcept
1334
1.03k
{
1335
1.03k
    Assume(GetTxCount());
1336
1.03k
    graph.GetClusterSet(level).m_cluster_usage -= TotalMemoryUsage();
1337
1.03k
    graph.ClearLocator(level, m_graph_index, m_quality == QualityLevel::OVERSIZED_SINGLETON);
1338
1.03k
    m_graph_index = NO_GRAPH_INDEX;
1339
1.03k
}
1340
1341
void GenericClusterImpl::MoveToMain(TxGraphImpl& graph) noexcept
1342
58
{
1343
565
    for (auto i : m_linearization) {
1344
565
        GraphIndex idx = m_mapping[i];
1345
565
        auto& entry = graph.m_entries[idx];
1346
565
        entry.m_locator[1].SetMissing();
1347
565
    }
1348
58
    auto quality = m_quality;
1349
    // Subtract memory usage from staging and add it to main.
1350
58
    graph.GetClusterSet(/*level=*/1).m_cluster_usage -= TotalMemoryUsage();
1351
58
    graph.GetClusterSet(/*level=*/0).m_cluster_usage += TotalMemoryUsage();
1352
    // Remove cluster itself from staging and add it to main.
1353
58
    auto cluster = graph.ExtractCluster(1, quality, m_setindex);
1354
58
    graph.InsertCluster(/*level=*/0, std::move(cluster), quality);
1355
58
    Updated(graph, /*level=*/0, /*rename=*/false);
1356
58
}
1357
1358
void SingletonClusterImpl::MoveToMain(TxGraphImpl& graph) noexcept
1359
51.9k
{
1360
51.9k
    if (GetTxCount()) {
1361
51.9k
        auto& entry = graph.m_entries[m_graph_index];
1362
51.9k
        entry.m_locator[1].SetMissing();
1363
51.9k
    }
1364
51.9k
    auto quality = m_quality;
1365
51.9k
    graph.GetClusterSet(/*level=*/1).m_cluster_usage -= TotalMemoryUsage();
1366
51.9k
    auto cluster = graph.ExtractCluster(/*level=*/1, quality, m_setindex);
1367
51.9k
    graph.InsertCluster(/*level=*/0, std::move(cluster), quality);
1368
51.9k
    graph.GetClusterSet(/*level=*/0).m_cluster_usage += TotalMemoryUsage();
1369
51.9k
    Updated(graph, /*level=*/0, /*rename=*/false);
1370
51.9k
}
1371
1372
void GenericClusterImpl::Compact() noexcept
1373
6.30k
{
1374
6.30k
    m_linearization.shrink_to_fit();
1375
6.30k
    m_mapping.shrink_to_fit();
1376
6.30k
    m_depgraph.Compact();
1377
6.30k
}
1378
1379
void SingletonClusterImpl::Compact() noexcept
1380
386
{
1381
    // Nothing to compact; SingletonClusterImpl is constant size.
1382
386
}
1383
1384
void GenericClusterImpl::AppendChunkFeerates(std::vector<FeeFrac>& ret) const noexcept
1385
418
{
1386
418
    auto chunk_feerates = ChunkLinearization(m_depgraph, m_linearization);
1387
418
    ret.reserve(ret.size() + chunk_feerates.size());
1388
418
    ret.insert(ret.end(), chunk_feerates.begin(), chunk_feerates.end());
1389
418
}
1390
1391
void SingletonClusterImpl::AppendChunkFeerates(std::vector<FeeFrac>& ret) const noexcept
1392
2.73k
{
1393
2.73k
    if (GetTxCount()) {
1394
2.73k
        ret.push_back(m_feerate);
1395
2.73k
    }
1396
2.73k
}
1397
1398
uint64_t GenericClusterImpl::AppendTrimData(std::vector<TrimTxData>& ret, std::vector<std::pair<GraphIndex, GraphIndex>>& deps) const noexcept
1399
0
{
1400
0
    Assume(IsAcceptable());
1401
0
    auto linchunking = ChunkLinearizationInfo(m_depgraph, m_linearization);
1402
0
    LinearizationIndex pos{0};
1403
0
    uint64_t size{0};
1404
0
    auto prev_index = GraphIndex(-1);
1405
    // Iterate over the chunks of this cluster's linearization.
1406
0
    for (const auto& [chunk, chunk_feerate] : linchunking) {
1407
        // Iterate over the transactions of that chunk, in linearization order.
1408
0
        auto chunk_tx_count = chunk.Count();
1409
0
        for (unsigned j = 0; j < chunk_tx_count; ++j) {
1410
0
            auto cluster_idx = m_linearization[pos];
1411
            // The transaction must appear in the chunk.
1412
0
            Assume(chunk[cluster_idx]);
1413
            // Construct a new element in ret.
1414
0
            auto& entry = ret.emplace_back();
1415
0
            entry.m_chunk_feerate = FeePerWeight::FromFeeFrac(chunk_feerate);
1416
0
            entry.m_index = m_mapping[cluster_idx];
1417
            // If this is not the first transaction of the cluster linearization, it has an
1418
            // implicit dependency on its predecessor.
1419
0
            if (pos != 0) deps.emplace_back(prev_index, entry.m_index);
1420
0
            prev_index = entry.m_index;
1421
0
            entry.m_tx_size = m_depgraph.FeeRate(cluster_idx).size;
1422
0
            size += entry.m_tx_size;
1423
0
            ++pos;
1424
0
        }
1425
0
    }
1426
0
    return size;
1427
0
}
1428
1429
uint64_t SingletonClusterImpl::AppendTrimData(std::vector<TrimTxData>& ret, std::vector<std::pair<GraphIndex, GraphIndex>>& deps) const noexcept
1430
64.2k
{
1431
64.2k
    if (!GetTxCount()) return 0;
1432
64.2k
    auto& entry = ret.emplace_back();
1433
64.2k
    entry.m_chunk_feerate = m_feerate;
1434
64.2k
    entry.m_index = m_graph_index;
1435
64.2k
    entry.m_tx_size = m_feerate.size;
1436
64.2k
    return m_feerate.size;
1437
64.2k
}
1438
1439
bool GenericClusterImpl::Split(TxGraphImpl& graph, int level) noexcept
1440
1.13k
{
1441
    // This function can only be called when the Cluster needs splitting.
1442
1.13k
    Assume(NeedsSplitting());
1443
    // Determine the new quality the split-off Clusters will have.
1444
1.13k
    QualityLevel new_quality = IsTopological() ? QualityLevel::NEEDS_RELINEARIZE : QualityLevel::NEEDS_FIX;
1445
    /** Which positions are still left in this Cluster. */
1446
1.13k
    auto todo = m_depgraph.Positions();
1447
    /** Mapping from transaction positions in this Cluster to the Cluster where it ends up, and
1448
     *  its position therein. */
1449
1.13k
    std::vector<std::pair<Cluster*, DepGraphIndex>> remap(m_depgraph.PositionRange());
1450
1.13k
    std::vector<Cluster*> new_clusters;
1451
1.13k
    bool first{true};
1452
    // Iterate over the connected components of this Cluster's m_depgraph.
1453
1.52k
    while (todo.Any()) {
1454
417
        auto component = m_depgraph.FindConnectedComponent(todo);
1455
417
        auto component_size = component.Count();
1456
417
        auto split_quality = component_size <= 1 ? QualityLevel::OPTIMAL : new_quality;
1457
417
        if (first && component == todo && SetType::Fill(component_size) == component && component_size >= MIN_INTENDED_TX_COUNT) {
1458
            // The existing Cluster is an entire component, without holes. Leave it be, but update
1459
            // its quality. If there are holes, we continue, so that the Cluster is reconstructed
1460
            // without holes, reducing memory usage. If the component's size is below the intended
1461
            // transaction count for this Cluster implementation, continue so that it can get
1462
            // converted.
1463
25
            Assume(todo == m_depgraph.Positions());
1464
25
            graph.SetClusterQuality(level, m_quality, m_setindex, split_quality);
1465
            // If this made the quality ACCEPTABLE or OPTIMAL, we need to compute and cache its
1466
            // chunking.
1467
25
            Updated(graph, /*level=*/level, /*rename=*/false);
1468
25
            return false;
1469
25
        }
1470
392
        first = false;
1471
        // Construct a new Cluster to hold the found component.
1472
392
        auto new_cluster = graph.CreateEmptyCluster(component_size);
1473
392
        new_clusters.push_back(new_cluster.get());
1474
        // Remember that all the component's transactions go to this new Cluster. The positions
1475
        // will be determined below, so use -1 for now.
1476
401
        for (auto i : component) {
1477
401
            remap[i] = {new_cluster.get(), DepGraphIndex(-1)};
1478
401
        }
1479
392
        graph.InsertCluster(level, std::move(new_cluster), split_quality);
1480
392
        todo -= component;
1481
392
    }
1482
    // We have to split the Cluster up. Remove accounting for the existing one first.
1483
1.10k
    graph.GetClusterSet(level).m_cluster_usage -= TotalMemoryUsage();
1484
    // Redistribute the transactions.
1485
1.10k
    for (auto i : m_linearization) {
1486
        /** The cluster which transaction originally in position i is moved to. */
1487
401
        Cluster* new_cluster = remap[i].first;
1488
        // Copy the transaction to the new cluster's depgraph, and remember the position.
1489
401
        remap[i].second = new_cluster->AppendTransaction(m_mapping[i], FeePerWeight::FromFeeFrac(m_depgraph.FeeRate(i)));
1490
401
    }
1491
    // Redistribute the dependencies.
1492
1.10k
    for (auto i : m_linearization) {
1493
        /** The cluster transaction in position i is moved to. */
1494
401
        Cluster* new_cluster = remap[i].first;
1495
        // Copy its parents, translating positions.
1496
401
        SetType new_parents;
1497
401
        for (auto par : m_depgraph.GetReducedParents(i)) new_parents.Set(remap[par].second);
1498
401
        new_cluster->AddDependencies(new_parents, remap[i].second);
1499
401
    }
1500
    // Update all the Locators of moved transactions, and memory usage.
1501
1.10k
    for (Cluster* new_cluster : new_clusters) {
1502
392
        new_cluster->Updated(graph, /*level=*/level, /*rename=*/false);
1503
392
        new_cluster->Compact();
1504
392
        graph.GetClusterSet(level).m_cluster_usage += new_cluster->TotalMemoryUsage();
1505
392
    }
1506
    // Wipe this Cluster, and return that it needs to be deleted.
1507
1.10k
    m_depgraph = DepGraph<SetType>{};
1508
1.10k
    m_mapping.clear();
1509
1.10k
    m_linearization.clear();
1510
1.10k
    return true;
1511
1.13k
}
1512
1513
bool SingletonClusterImpl::Split(TxGraphImpl& graph, int level) noexcept
1514
44.6k
{
1515
44.6k
    Assume(NeedsSplitting());
1516
44.6k
    Assume(!GetTxCount());
1517
44.6k
    graph.GetClusterSet(level).m_cluster_usage -= TotalMemoryUsage();
1518
44.6k
    return true;
1519
44.6k
}
1520
1521
void GenericClusterImpl::Merge(TxGraphImpl& graph, int level, Cluster& other) noexcept
1522
7.00k
{
1523
    /** Vector to store the positions in this Cluster for each position in other. */
1524
7.00k
    std::vector<DepGraphIndex> remap(other.GetDepGraphIndexRange());
1525
    // Iterate over all transactions in the other Cluster (the one being absorbed).
1526
7.02k
    other.ExtractTransactions([&](DepGraphIndex pos, GraphIndex idx, FeePerWeight feerate) noexcept {
1527
        // Copy the transaction into this Cluster, and remember its position.
1528
7.02k
        auto new_pos = m_depgraph.AddTransaction(feerate);
1529
        // Since this cluster must have been made hole-free before being merged into, all added
1530
        // transactions should appear at the end.
1531
7.02k
        Assume(new_pos == m_mapping.size());
1532
7.02k
        remap[pos] = new_pos;
1533
7.02k
        m_mapping.push_back(idx);
1534
7.02k
        m_linearization.push_back(new_pos);
1535
7.02k
    }, [&](DepGraphIndex pos, GraphIndex idx, SetType other_parents) noexcept {
1536
        // Copy the transaction's dependencies, translating them using remap.
1537
7.02k
        SetType parents;
1538
7.02k
        for (auto par : other_parents) {
1539
25
            parents.Set(remap[par]);
1540
25
        }
1541
7.02k
        m_depgraph.AddDependencies(parents, remap[pos]);
1542
        // Update the transaction's Locator. There is no need to call Updated() to update chunk
1543
        // feerates, as Updated() will be invoked by Cluster::ApplyDependencies on the resulting
1544
        // merged Cluster later anyway.
1545
7.02k
        auto& entry = graph.m_entries[idx];
1546
        // Discard any potential ChunkData prior to modifying the Cluster (as that could
1547
        // invalidate its ordering).
1548
7.02k
        if (level == 0) graph.ClearChunkData(entry);
1549
7.02k
        entry.m_locator[level].SetPresent(this, remap[pos]);
1550
7.02k
    });
1551
7.00k
}
1552
1553
void SingletonClusterImpl::Merge(TxGraphImpl&, int, Cluster&) noexcept
1554
0
{
1555
    // Nothing can be merged into a singleton; it should have been converted to GenericClusterImpl first.
1556
0
    Assume(false);
1557
0
}
1558
1559
void GenericClusterImpl::ApplyDependencies(TxGraphImpl& graph, int level, std::span<std::pair<GraphIndex, GraphIndex>> to_apply) noexcept
1560
5.19k
{
1561
    // Sort the list of dependencies to apply by child, so those can be applied in batch.
1562
29.2k
    std::sort(to_apply.begin(), to_apply.end(), [](auto& a, auto& b) { return a.second < b.second; });
1563
    // Iterate over groups of to-be-added dependencies with the same child.
1564
5.19k
    auto it = to_apply.begin();
1565
10.7k
    while (it != to_apply.end()) {
1566
5.55k
        auto& first_child = graph.m_entries[it->second].m_locator[level];
1567
5.55k
        const auto child_idx = first_child.index;
1568
        // Iterate over all to-be-added dependencies within that same child, gather the relevant
1569
        // parents.
1570
5.55k
        SetType parents;
1571
15.4k
        while (it != to_apply.end()) {
1572
10.2k
            auto& child = graph.m_entries[it->second].m_locator[level];
1573
10.2k
            auto& parent = graph.m_entries[it->first].m_locator[level];
1574
10.2k
            Assume(child.cluster == this && parent.cluster == this);
1575
10.2k
            if (child.index != child_idx) break;
1576
9.92k
            parents.Set(parent.index);
1577
9.92k
            ++it;
1578
9.92k
        }
1579
        // Push all dependencies to the underlying DepGraph. Note that this is O(N) in the size of
1580
        // the cluster, regardless of the number of parents being added, so batching them together
1581
        // has a performance benefit.
1582
5.55k
        m_depgraph.AddDependencies(parents, child_idx);
1583
5.55k
    }
1584
1585
    // Finally set the cluster to NEEDS_FIX, so its linearization is fixed the next time it is
1586
    // attempted to be made ACCEPTABLE.
1587
5.19k
    Assume(!NeedsSplitting());
1588
5.19k
    Assume(!IsOversized());
1589
5.19k
    graph.SetClusterQuality(level, m_quality, m_setindex, QualityLevel::NEEDS_FIX);
1590
1591
    // Finally push the changes to graph.m_entries.
1592
5.19k
    Updated(graph, /*level=*/level, /*rename=*/false);
1593
5.19k
}
1594
1595
void SingletonClusterImpl::ApplyDependencies(TxGraphImpl&, int, std::span<std::pair<GraphIndex, GraphIndex>>) noexcept
1596
0
{
1597
    // Nothing can actually be applied.
1598
0
    Assume(false);
1599
0
}
1600
1601
TxGraphImpl::~TxGraphImpl() noexcept
1602
1.20k
{
1603
    // If Refs outlive the TxGraphImpl they refer to, unlink them, so that their destructor does not
1604
    // try to reach into a non-existing TxGraphImpl anymore.
1605
67.8k
    for (auto& entry : m_entries) {
1606
67.8k
        if (entry.m_ref != nullptr) {
1607
64.0k
            GetRefGraph(*entry.m_ref) = nullptr;
1608
64.0k
        }
1609
67.8k
    }
1610
1.20k
}
1611
1612
std::unique_ptr<Cluster> TxGraphImpl::ExtractCluster(int level, QualityLevel quality, ClusterSetIndex setindex) noexcept
1613
162k
{
1614
162k
    Assume(quality != QualityLevel::NONE);
1615
1616
162k
    auto& clusterset = GetClusterSet(level);
1617
162k
    auto& quality_clusters = clusterset.m_clusters[int(quality)];
1618
162k
    Assume(setindex < quality_clusters.size());
1619
1620
    // Extract the Cluster-owning unique_ptr.
1621
162k
    std::unique_ptr<Cluster> ret = std::move(quality_clusters[setindex]);
1622
162k
    ret->m_quality = QualityLevel::NONE;
1623
162k
    ret->m_setindex = ClusterSetIndex(-1);
1624
1625
    // Clean up space in quality_cluster.
1626
162k
    auto max_setindex = quality_clusters.size() - 1;
1627
162k
    if (setindex != max_setindex) {
1628
        // If the cluster was not the last element of quality_clusters, move that to take its place.
1629
34.7k
        quality_clusters.back()->m_setindex = setindex;
1630
34.7k
        quality_clusters[setindex] = std::move(quality_clusters.back());
1631
34.7k
    }
1632
    // The last element of quality_clusters is now unused; drop it.
1633
162k
    quality_clusters.pop_back();
1634
1635
162k
    return ret;
1636
162k
}
1637
1638
ClusterSetIndex TxGraphImpl::InsertCluster(int level, std::unique_ptr<Cluster>&& cluster, QualityLevel quality) noexcept
1639
239k
{
1640
    // Cannot insert with quality level NONE (as that would mean not inserted).
1641
239k
    Assume(quality != QualityLevel::NONE);
1642
    // The passed-in Cluster must not currently be in the TxGraphImpl.
1643
239k
    Assume(cluster->m_quality == QualityLevel::NONE);
1644
1645
    // Append it at the end of the relevant TxGraphImpl::m_cluster.
1646
239k
    auto& clusterset = GetClusterSet(level);
1647
239k
    auto& quality_clusters = clusterset.m_clusters[int(quality)];
1648
239k
    ClusterSetIndex ret = quality_clusters.size();
1649
239k
    cluster->m_quality = quality;
1650
239k
    cluster->m_setindex = ret;
1651
239k
    quality_clusters.push_back(std::move(cluster));
1652
239k
    return ret;
1653
239k
}
1654
1655
void TxGraphImpl::SetClusterQuality(int level, QualityLevel old_quality, ClusterSetIndex old_index, QualityLevel new_quality) noexcept
1656
56.2k
{
1657
56.2k
    Assume(new_quality != QualityLevel::NONE);
1658
1659
    // Don't do anything if the quality did not change.
1660
56.2k
    if (old_quality == new_quality) return;
1661
    // Extract the cluster from where it currently resides.
1662
56.2k
    auto cluster_ptr = ExtractCluster(level, old_quality, old_index);
1663
    // And re-insert it where it belongs.
1664
56.2k
    InsertCluster(level, std::move(cluster_ptr), new_quality);
1665
56.2k
}
1666
1667
void TxGraphImpl::DeleteCluster(Cluster& cluster, int level) noexcept
1668
54.0k
{
1669
    // Extract the cluster from where it currently resides.
1670
54.0k
    auto cluster_ptr = ExtractCluster(level, cluster.m_quality, cluster.m_setindex);
1671
    // And throw it away.
1672
54.0k
    cluster_ptr.reset();
1673
54.0k
}
1674
1675
std::pair<Cluster*, int> TxGraphImpl::FindClusterAndLevel(GraphIndex idx, int level) const noexcept
1676
1.36M
{
1677
1.36M
    Assume(level >= 0 && level <= GetTopLevel());
1678
1.36M
    auto& entry = m_entries[idx];
1679
    // Search the entry's locators from top to bottom.
1680
1.40M
    for (int l = level; l >= 0; --l) {
1681
        // If the locator is missing, dig deeper; it may exist at a lower level and therefore be
1682
        // implicitly existing at this level too.
1683
1.40M
        if (entry.m_locator[l].IsMissing()) continue;
1684
        // If the locator has the entry marked as explicitly removed, stop.
1685
1.36M
        if (entry.m_locator[l].IsRemoved()) break;
1686
        // Otherwise, we have found the topmost ClusterSet that contains this entry.
1687
1.36M
        return {entry.m_locator[l].cluster, l};
1688
1.36M
    }
1689
    // If no non-empty locator was found, or an explicitly removed was hit, return nothing.
1690
1.32k
    return {nullptr, -1};
1691
1.36M
}
1692
1693
Cluster* TxGraphImpl::PullIn(Cluster* cluster, int level) noexcept
1694
2.49k
{
1695
2.49k
    int to_level = GetTopLevel();
1696
2.49k
    Assume(to_level == 1);
1697
2.49k
    Assume(level <= to_level);
1698
    // Copy the Cluster from main to staging, if it's not already there.
1699
2.49k
    if (level == 0) {
1700
        // Make the Cluster Acceptable before copying. This isn't strictly necessary, but doing it
1701
        // now avoids doing double work later.
1702
1.74k
        MakeAcceptable(*cluster, level);
1703
1.74k
        cluster = cluster->CopyToStaging(*this);
1704
1.74k
    }
1705
2.49k
    return cluster;
1706
2.49k
}
1707
1708
void TxGraphImpl::ApplyRemovals(int up_to_level) noexcept
1709
639k
{
1710
639k
    Assume(up_to_level >= 0 && up_to_level <= GetTopLevel());
1711
1.34M
    for (int level = 0; level <= up_to_level; ++level) {
1712
707k
        auto& clusterset = GetClusterSet(level);
1713
707k
        auto& to_remove = clusterset.m_to_remove;
1714
        // Skip if there is nothing to remove in this level.
1715
707k
        if (to_remove.empty()) continue;
1716
        // Pull in all Clusters that are not in staging.
1717
4.06k
        if (level == 1) {
1718
2.17k
            for (GraphIndex index : to_remove) {
1719
2.17k
                auto [cluster, cluster_level] = FindClusterAndLevel(index, level);
1720
2.17k
                if (cluster != nullptr) PullIn(cluster, cluster_level);
1721
2.17k
            }
1722
1.34k
        }
1723
        // Group the set of to-be-removed entries by Cluster::m_sequence.
1724
261k
        std::sort(to_remove.begin(), to_remove.end(), [&](GraphIndex a, GraphIndex b) noexcept {
1725
261k
            Cluster* cluster_a = m_entries[a].m_locator[level].cluster;
1726
261k
            Cluster* cluster_b = m_entries[b].m_locator[level].cluster;
1727
261k
            return CompareClusters(cluster_a, cluster_b) < 0;
1728
261k
        });
1729
        // Process per Cluster.
1730
4.06k
        std::span to_remove_span{to_remove};
1731
49.8k
        while (!to_remove_span.empty()) {
1732
45.7k
            Cluster* cluster = m_entries[to_remove_span.front()].m_locator[level].cluster;
1733
45.7k
            if (cluster != nullptr) {
1734
                // If the first to_remove_span entry's Cluster exists, hand to_remove_span to it, so it
1735
                // can pop off whatever applies to it.
1736
45.7k
                cluster->ApplyRemovals(*this, level, to_remove_span);
1737
45.7k
            } else {
1738
                // Otherwise, skip this already-removed entry. This may happen when
1739
                // RemoveTransaction was called twice on the same Ref, for example.
1740
0
                to_remove_span = to_remove_span.subspan(1);
1741
0
            }
1742
45.7k
        }
1743
4.06k
        to_remove.clear();
1744
4.06k
    }
1745
639k
    Compact();
1746
639k
}
1747
1748
void TxGraphImpl::SwapIndexes(GraphIndex a, GraphIndex b, std::vector<Cluster*>& affected_main) noexcept
1749
18.2k
{
1750
18.2k
    Assume(a < m_entries.size());
1751
18.2k
    Assume(b < m_entries.size());
1752
    // Swap the Entry objects.
1753
18.2k
    std::swap(m_entries[a], m_entries[b]);
1754
    // Iterate over both objects.
1755
54.8k
    for (int i = 0; i < 2; ++i) {
1756
36.5k
        GraphIndex idx = i ? b : a;
1757
36.5k
        Entry& entry = m_entries[idx];
1758
        // Update linked Ref, if any exists.
1759
36.5k
        if (entry.m_ref) GetRefIndex(*entry.m_ref) = idx;
1760
        // Update linked chunk index entries, if any exist.
1761
36.5k
        if (entry.m_main_chunkindex_iterator != m_main_chunkindex.end()) {
1762
17.1k
            entry.m_main_chunkindex_iterator->m_graph_index = idx;
1763
17.1k
        }
1764
        // Update the locators for both levels.
1765
109k
        for (int level = 0; level < MAX_LEVELS; ++level) {
1766
73.1k
            Locator& locator = entry.m_locator[level];
1767
73.1k
            if (locator.IsPresent()) {
1768
17.4k
                locator.cluster->UpdateMapping(locator.index, idx);
1769
17.4k
                if (level == 0) affected_main.push_back(locator.cluster);
1770
17.4k
            }
1771
73.1k
        }
1772
36.5k
    }
1773
18.2k
}
1774
1775
void TxGraphImpl::Compact() noexcept
1776
730k
{
1777
    // We cannot compact while any to-be-applied operations or staged removals remain as we'd need
1778
    // to rewrite them. It is easier to delay the compaction until they have been applied.
1779
730k
    if (!m_main_clusterset.m_deps_to_add.empty()) return;
1780
725k
    if (!m_main_clusterset.m_to_remove.empty()) return;
1781
725k
    Assume(m_main_clusterset.m_removed.empty()); // non-staging m_removed is always empty
1782
725k
    if (m_staging_clusterset.has_value()) {
1783
76.1k
        if (!m_staging_clusterset->m_deps_to_add.empty()) return;
1784
50.9k
        if (!m_staging_clusterset->m_to_remove.empty()) return;
1785
50.9k
        if (!m_staging_clusterset->m_removed.empty()) return;
1786
50.9k
    }
1787
1788
    // Release memory used by discarded ChunkData index entries.
1789
695k
    ClearShrink(m_main_chunkindex_discarded);
1790
1791
    // Sort the GraphIndexes that need to be cleaned up. They are sorted in reverse, so the last
1792
    // ones get processed first. This means earlier-processed GraphIndexes will not cause moving of
1793
    // later-processed ones during the "swap with end of m_entries" step below (which might
1794
    // invalidate them).
1795
695k
    std::sort(m_unlinked.begin(), m_unlinked.end(), std::greater{});
1796
1797
695k
    std::vector<Cluster*> affected_main;
1798
695k
    auto last = GraphIndex(-1);
1799
695k
    for (GraphIndex idx : m_unlinked) {
1800
        // m_unlinked should never contain the same GraphIndex twice (the code below would fail
1801
        // if so, because GraphIndexes get invalidated by removing them).
1802
59.6k
        Assume(idx != last);
1803
59.6k
        last = idx;
1804
1805
        // Make sure the entry is unlinked.
1806
59.6k
        Entry& entry = m_entries[idx];
1807
59.6k
        Assume(entry.m_ref == nullptr);
1808
        // Make sure the entry does not occur in the graph.
1809
179k
        for (int level = 0; level < MAX_LEVELS; ++level) {
1810
119k
            Assume(!entry.m_locator[level].IsPresent());
1811
119k
        }
1812
1813
        // Move the entry to the end.
1814
59.6k
        if (idx != m_entries.size() - 1) SwapIndexes(idx, m_entries.size() - 1, affected_main);
1815
        // Drop the entry for idx, now that it is at the end.
1816
59.6k
        m_entries.pop_back();
1817
59.6k
    }
1818
1819
    // Update the affected clusters, to fixup Entry::m_main_max_chunk_fallback values which may
1820
    // have become outdated due to the compaction above.
1821
695k
    std::sort(affected_main.begin(), affected_main.end());
1822
695k
    affected_main.erase(std::unique(affected_main.begin(), affected_main.end()), affected_main.end());
1823
695k
    for (Cluster* cluster : affected_main) {
1824
12.1k
        cluster->Updated(*this, /*level=*/0, /*rename=*/true);
1825
12.1k
    }
1826
695k
    m_unlinked.clear();
1827
695k
}
1828
1829
void TxGraphImpl::Split(Cluster& cluster, int level) noexcept
1830
45.7k
{
1831
    // To split a Cluster, first make sure all removals are applied (as we might need to split
1832
    // again afterwards otherwise).
1833
45.7k
    ApplyRemovals(level);
1834
45.7k
    bool del = cluster.Split(*this, level);
1835
45.7k
    if (del) {
1836
        // Cluster::Split reports whether the Cluster is to be deleted.
1837
45.7k
        DeleteCluster(cluster, level);
1838
45.7k
    }
1839
45.7k
}
1840
1841
void TxGraphImpl::SplitAll(int up_to_level) noexcept
1842
534k
{
1843
534k
    Assume(up_to_level >= 0 && up_to_level <= GetTopLevel());
1844
    // Before splitting all Cluster, first make sure all removals are applied.
1845
534k
    ApplyRemovals(up_to_level);
1846
1.07M
    for (int level = 0; level <= up_to_level; ++level) {
1847
1.08M
        for (auto quality : {QualityLevel::NEEDS_SPLIT_FIX, QualityLevel::NEEDS_SPLIT}) {
1848
1.08M
            auto& queue = GetClusterSet(level).m_clusters[int(quality)];
1849
1.13M
            while (!queue.empty()) {
1850
45.7k
                Split(*queue.back().get(), level);
1851
45.7k
            }
1852
1.08M
        }
1853
542k
    }
1854
534k
}
1855
1856
void TxGraphImpl::GroupClusters(int level) noexcept
1857
138M
{
1858
138M
    auto& clusterset = GetClusterSet(level);
1859
    // If the groupings have been computed already, nothing is left to be done.
1860
138M
    if (clusterset.m_group_data.has_value()) return;
1861
1862
    // Before computing which Clusters need to be merged together, first apply all removals and
1863
    // split the Clusters into connected components. If we would group first, we might end up
1864
    // with inefficient and/or oversized Clusters which just end up being split again anyway.
1865
10.6k
    SplitAll(level);
1866
1867
    /** Annotated clusters: an entry for each Cluster, together with the sequence number for the
1868
     *  representative for the partition it is in (initially its own, later that of the
1869
     *  to-be-merged group). */
1870
10.6k
    std::vector<std::pair<Cluster*, uint64_t>> an_clusters;
1871
    /** Annotated dependencies: an entry for each m_deps_to_add entry (excluding ones that apply
1872
     *  to removed transactions), together with the sequence number of the representative root of
1873
     *  Clusters it applies to (initially that of the child Cluster, later that of the
1874
     *  to-be-merged group). */
1875
10.6k
    std::vector<std::pair<std::pair<GraphIndex, GraphIndex>, uint64_t>> an_deps;
1876
1877
    // Construct an an_clusters entry for every oversized cluster, including ones from levels below,
1878
    // as they may be inherited in this one.
1879
29.1k
    for (int level_iter = 0; level_iter <= level; ++level_iter) {
1880
18.5k
        for (auto& cluster : GetClusterSet(level_iter).m_clusters[int(QualityLevel::OVERSIZED_SINGLETON)]) {
1881
6
            auto graph_idx = cluster->GetClusterEntry(0);
1882
6
            auto cur_cluster = FindCluster(graph_idx, level);
1883
6
            if (cur_cluster == nullptr) continue;
1884
6
            an_clusters.emplace_back(cur_cluster, cur_cluster->m_sequence);
1885
6
        }
1886
18.5k
    }
1887
1888
    // Construct a an_clusters entry for every parent and child in the to-be-applied dependencies,
1889
    // and an an_deps entry for each dependency to be applied.
1890
10.6k
    an_deps.reserve(clusterset.m_deps_to_add.size());
1891
139k
    for (const auto& [par, chl] : clusterset.m_deps_to_add) {
1892
139k
        auto par_cluster = FindCluster(par, level);
1893
139k
        auto chl_cluster = FindCluster(chl, level);
1894
        // Skip dependencies for which the parent or child transaction is removed.
1895
139k
        if (par_cluster == nullptr || chl_cluster == nullptr) continue;
1896
139k
        an_clusters.emplace_back(par_cluster, par_cluster->m_sequence);
1897
        // Do not include a duplicate when parent and child are identical, as it'll be removed
1898
        // below anyway.
1899
139k
        if (chl_cluster != par_cluster) an_clusters.emplace_back(chl_cluster, chl_cluster->m_sequence);
1900
        // Add entry to an_deps, using the child sequence number.
1901
139k
        an_deps.emplace_back(std::pair{par, chl}, chl_cluster->m_sequence);
1902
139k
    }
1903
    // Sort and deduplicate an_clusters, so we end up with a sorted list of all involved Clusters
1904
    // to which dependencies apply, or which are oversized.
1905
4.92M
    std::sort(an_clusters.begin(), an_clusters.end(), [](auto& a, auto& b) noexcept { return a.second < b.second; });
1906
10.6k
    an_clusters.erase(std::unique(an_clusters.begin(), an_clusters.end()), an_clusters.end());
1907
    // Sort an_deps by applying the same order to the involved child cluster.
1908
2.35M
    std::sort(an_deps.begin(), an_deps.end(), [&](auto& a, auto& b) noexcept { return a.second < b.second; });
1909
1910
    // Run the union-find algorithm to find partitions of the input Clusters which need to be
1911
    // grouped together. See https://en.wikipedia.org/wiki/Disjoint-set_data_structure.
1912
10.6k
    {
1913
        /** Each PartitionData entry contains information about a single input Cluster. */
1914
10.6k
        struct PartitionData
1915
10.6k
        {
1916
            /** The sequence number of the cluster this holds information for. */
1917
10.6k
            uint64_t sequence;
1918
            /** All PartitionData entries belonging to the same partition are organized in a tree.
1919
             *  Each element points to its parent, or to itself if it is the root. The root is then
1920
             *  a representative for the entire tree, and can be found by walking upwards from any
1921
             *  element. */
1922
10.6k
            PartitionData* parent;
1923
            /** (only if this is a root, so when parent == this) An upper bound on the height of
1924
             *  tree for this partition. */
1925
10.6k
            unsigned rank;
1926
10.6k
        };
1927
        /** Information about each input Cluster. Sorted by Cluster::m_sequence. */
1928
10.6k
        std::vector<PartitionData> partition_data;
1929
1930
        /** Given a Cluster, find its corresponding PartitionData. */
1931
271k
        auto locate_fn = [&](uint64_t sequence) noexcept -> PartitionData* {
1932
271k
            auto it = std::lower_bound(partition_data.begin(), partition_data.end(), sequence,
1933
4.08M
                                       [](auto& a, uint64_t seq) noexcept { return a.sequence < seq; });
1934
271k
            Assume(it != partition_data.end());
1935
271k
            Assume(it->sequence == sequence);
1936
271k
            return &*it;
1937
271k
        };
1938
1939
        /** Given a PartitionData, find the root of the tree it is in (its representative). */
1940
419k
        static constexpr auto find_root_fn = [](PartitionData* data) noexcept -> PartitionData* {
1941
742k
            while (data->parent != data) {
1942
                // Replace pointers to parents with pointers to grandparents.
1943
                // See https://en.wikipedia.org/wiki/Disjoint-set_data_structure#Finding_set_representatives.
1944
322k
                auto par = data->parent;
1945
322k
                data->parent = par->parent;
1946
322k
                data = par;
1947
322k
            }
1948
419k
            return data;
1949
419k
        };
1950
1951
        /** Given two PartitionDatas, union the partitions they are in, and return their
1952
         *  representative. */
1953
138k
        static constexpr auto union_fn = [](PartitionData* arg1, PartitionData* arg2) noexcept {
1954
            // Find the roots of the trees, and bail out if they are already equal (which would
1955
            // mean they are in the same partition already).
1956
138k
            auto rep1 = find_root_fn(arg1);
1957
138k
            auto rep2 = find_root_fn(arg2);
1958
138k
            if (rep1 == rep2) return rep1;
1959
            // Pick the lower-rank root to become a child of the higher-rank one.
1960
            // See https://en.wikipedia.org/wiki/Disjoint-set_data_structure#Union_by_rank.
1961
135k
            if (rep1->rank < rep2->rank) std::swap(rep1, rep2);
1962
135k
            rep2->parent = rep1;
1963
135k
            rep1->rank += (rep1->rank == rep2->rank);
1964
135k
            return rep1;
1965
138k
        };
1966
1967
        // Start by initializing every Cluster as its own singleton partition.
1968
10.6k
        partition_data.resize(an_clusters.size());
1969
154k
        for (size_t i = 0; i < an_clusters.size(); ++i) {
1970
143k
            partition_data[i].sequence = an_clusters[i].first->m_sequence;
1971
143k
            partition_data[i].parent = &partition_data[i];
1972
143k
            partition_data[i].rank = 0;
1973
143k
        }
1974
1975
        // Run through all parent/child pairs in an_deps, and union the partitions their Clusters
1976
        // are in.
1977
10.6k
        Cluster* last_chl_cluster{nullptr};
1978
10.6k
        PartitionData* last_partition{nullptr};
1979
139k
        for (const auto& [dep, _] : an_deps) {
1980
139k
            auto [par, chl] = dep;
1981
139k
            auto par_cluster = FindCluster(par, level);
1982
139k
            auto chl_cluster = FindCluster(chl, level);
1983
139k
            Assume(chl_cluster != nullptr && par_cluster != nullptr);
1984
            // Nothing to do if parent and child are in the same Cluster.
1985
139k
            if (par_cluster == chl_cluster) continue;
1986
138k
            Assume(par != chl);
1987
138k
            if (chl_cluster == last_chl_cluster) {
1988
                // If the child Clusters is the same as the previous iteration, union with the
1989
                // tree they were in, avoiding the need for another lookup. Note that an_deps
1990
                // is sorted by child Cluster, so batches with the same child are expected.
1991
5.20k
                last_partition = union_fn(locate_fn(par_cluster->m_sequence), last_partition);
1992
133k
            } else {
1993
133k
                last_chl_cluster = chl_cluster;
1994
133k
                last_partition = union_fn(locate_fn(par_cluster->m_sequence), locate_fn(chl_cluster->m_sequence));
1995
133k
            }
1996
138k
        }
1997
1998
        // Update the sequence numbers in an_clusters and an_deps to be those of the partition
1999
        // representative.
2000
10.6k
        auto deps_it = an_deps.begin();
2001
154k
        for (size_t i = 0; i < partition_data.size(); ++i) {
2002
143k
            auto& data = partition_data[i];
2003
            // Find the sequence of the representative of the partition Cluster i is in, and store
2004
            // it with the Cluster.
2005
143k
            auto rep_seq = find_root_fn(&data)->sequence;
2006
143k
            an_clusters[i].second = rep_seq;
2007
            // Find all dependencies whose child Cluster is Cluster i, and annotate them with rep.
2008
283k
            while (deps_it != an_deps.end()) {
2009
275k
                auto [par, chl] = deps_it->first;
2010
275k
                auto chl_cluster = FindCluster(chl, level);
2011
275k
                Assume(chl_cluster != nullptr);
2012
275k
                if (chl_cluster->m_sequence > data.sequence) break;
2013
139k
                deps_it->second = rep_seq;
2014
139k
                ++deps_it;
2015
139k
            }
2016
143k
        }
2017
10.6k
    }
2018
2019
    // Sort both an_clusters and an_deps by sequence number of the representative of the
2020
    // partition they are in, grouping all those applying to the same partition together.
2021
1.81M
    std::sort(an_deps.begin(), an_deps.end(), [](auto& a, auto& b) noexcept { return a.second < b.second; });
2022
1.80M
    std::sort(an_clusters.begin(), an_clusters.end(), [](auto& a, auto& b) noexcept { return a.second < b.second; });
2023
2024
    // Translate the resulting cluster groups to the m_group_data structure, and the dependencies
2025
    // back to m_deps_to_add.
2026
10.6k
    clusterset.m_group_data = GroupData{};
2027
10.6k
    clusterset.m_group_data->m_group_clusters.reserve(an_clusters.size());
2028
10.6k
    clusterset.m_deps_to_add.clear();
2029
10.6k
    clusterset.m_deps_to_add.reserve(an_deps.size());
2030
10.6k
    clusterset.m_oversized = false;
2031
10.6k
    auto an_deps_it = an_deps.begin();
2032
10.6k
    auto an_clusters_it = an_clusters.begin();
2033
18.5k
    while (an_clusters_it != an_clusters.end()) {
2034
        // Process all clusters/dependencies belonging to the partition with representative rep.
2035
7.89k
        auto rep = an_clusters_it->second;
2036
        // Create and initialize a new GroupData entry for the partition.
2037
7.89k
        auto& new_entry = clusterset.m_group_data->m_groups.emplace_back();
2038
7.89k
        new_entry.m_cluster_offset = clusterset.m_group_data->m_group_clusters.size();
2039
7.89k
        new_entry.m_cluster_count = 0;
2040
7.89k
        new_entry.m_deps_offset = clusterset.m_deps_to_add.size();
2041
7.89k
        new_entry.m_deps_count = 0;
2042
7.89k
        uint32_t total_count{0};
2043
7.89k
        uint64_t total_size{0};
2044
        // Add all its clusters to it (copying those from an_clusters to m_group_clusters).
2045
151k
        while (an_clusters_it != an_clusters.end() && an_clusters_it->second == rep) {
2046
143k
            clusterset.m_group_data->m_group_clusters.push_back(an_clusters_it->first);
2047
143k
            total_count += an_clusters_it->first->GetTxCount();
2048
143k
            total_size += an_clusters_it->first->GetTotalTxSize();
2049
143k
            ++an_clusters_it;
2050
143k
            ++new_entry.m_cluster_count;
2051
143k
        }
2052
        // Add all its dependencies to it (copying those back from an_deps to m_deps_to_add).
2053
147k
        while (an_deps_it != an_deps.end() && an_deps_it->second == rep) {
2054
139k
            clusterset.m_deps_to_add.push_back(an_deps_it->first);
2055
139k
            ++an_deps_it;
2056
139k
            ++new_entry.m_deps_count;
2057
139k
        }
2058
        // Detect oversizedness.
2059
7.89k
        if (total_count > m_max_cluster_count || total_size > m_max_cluster_size) {
2060
136
            clusterset.m_oversized = true;
2061
136
        }
2062
7.89k
    }
2063
10.6k
    Assume(an_deps_it == an_deps.end());
2064
10.6k
    Assume(an_clusters_it == an_clusters.end());
2065
10.6k
    Compact();
2066
10.6k
}
2067
2068
void TxGraphImpl::Merge(std::span<Cluster*> to_merge, int level) noexcept
2069
5.19k
{
2070
5.19k
    Assume(!to_merge.empty());
2071
    // Nothing to do if a group consists of just a single Cluster.
2072
5.19k
    if (to_merge.size() == 1) return;
2073
2074
    // Move the largest Cluster to the front of to_merge. As all transactions in other to-be-merged
2075
    // Clusters will be moved to that one, putting the largest one first minimizes the number of
2076
    // moves.
2077
5.16k
    size_t max_size_pos{0};
2078
5.16k
    DepGraphIndex max_size = to_merge[max_size_pos]->GetTxCount();
2079
5.16k
    GetClusterSet(level).m_cluster_usage -= to_merge[max_size_pos]->TotalMemoryUsage();
2080
5.16k
    DepGraphIndex total_size = max_size;
2081
10.8k
    for (size_t i = 1; i < to_merge.size(); ++i) {
2082
5.70k
        GetClusterSet(level).m_cluster_usage -= to_merge[i]->TotalMemoryUsage();
2083
5.70k
        DepGraphIndex size = to_merge[i]->GetTxCount();
2084
5.70k
        total_size += size;
2085
5.70k
        if (size > max_size) {
2086
67
            max_size_pos = i;
2087
67
            max_size = size;
2088
67
        }
2089
5.70k
    }
2090
5.16k
    if (max_size_pos != 0) std::swap(to_merge[0], to_merge[max_size_pos]);
2091
2092
5.16k
    size_t start_idx = 1;
2093
5.16k
    Cluster* into_cluster = to_merge[0];
2094
5.16k
    if (total_size > into_cluster->GetMaxTxCount()) {
2095
        // The into_merge cluster is too small to fit all transactions being merged. Construct a
2096
        // a new Cluster using an implementation that matches the total size, and merge everything
2097
        // in there.
2098
1.29k
        auto new_cluster = CreateEmptyCluster(total_size);
2099
1.29k
        into_cluster = new_cluster.get();
2100
1.29k
        InsertCluster(level, std::move(new_cluster), QualityLevel::OPTIMAL);
2101
1.29k
        start_idx = 0;
2102
1.29k
    }
2103
2104
    // Merge all further Clusters in the group into the result (first one, or new one), and delete
2105
    // them.
2106
12.1k
    for (size_t i = start_idx; i < to_merge.size(); ++i) {
2107
7.00k
        into_cluster->Merge(*this, level, *to_merge[i]);
2108
7.00k
        DeleteCluster(*to_merge[i], level);
2109
7.00k
    }
2110
5.16k
    into_cluster->Compact();
2111
5.16k
    GetClusterSet(level).m_cluster_usage += into_cluster->TotalMemoryUsage();
2112
5.16k
}
2113
2114
void TxGraphImpl::ApplyDependencies(int level) noexcept
2115
138M
{
2116
138M
    auto& clusterset = GetClusterSet(level);
2117
    // Do not bother computing groups if we already know the result will be oversized.
2118
138M
    if (clusterset.m_oversized == true) return;
2119
    // Compute the groups of to-be-merged Clusters (which also applies all removals, and splits).
2120
138M
    GroupClusters(level);
2121
138M
    Assume(clusterset.m_group_data.has_value());
2122
    // Nothing to do if there are no dependencies to be added.
2123
138M
    if (clusterset.m_deps_to_add.empty()) return;
2124
    // Dependencies cannot be applied if it would result in oversized clusters.
2125
4.98k
    if (clusterset.m_oversized == true) return;
2126
2127
    // For each group of to-be-merged Clusters.
2128
5.19k
    for (const auto& group_entry : clusterset.m_group_data->m_groups) {
2129
5.19k
        auto cluster_span = std::span{clusterset.m_group_data->m_group_clusters}
2130
5.19k
                                .subspan(group_entry.m_cluster_offset, group_entry.m_cluster_count);
2131
        // Pull in all the Clusters that contain dependencies.
2132
5.19k
        if (level == 1) {
2133
313
            for (Cluster*& cluster : cluster_span) {
2134
313
                cluster = PullIn(cluster, cluster->GetLevel(*this));
2135
313
            }
2136
83
        }
2137
        // Invoke Merge() to merge them into a single Cluster.
2138
5.19k
        Merge(cluster_span, level);
2139
        // Actually apply all to-be-added dependencies (all parents and children from this grouping
2140
        // belong to the same Cluster at this point because of the merging above).
2141
5.19k
        auto deps_span = std::span{clusterset.m_deps_to_add}
2142
5.19k
                             .subspan(group_entry.m_deps_offset, group_entry.m_deps_count);
2143
5.19k
        Assume(!deps_span.empty());
2144
5.19k
        const auto& loc = m_entries[deps_span[0].second].m_locator[level];
2145
5.19k
        Assume(loc.IsPresent());
2146
5.19k
        loc.cluster->ApplyDependencies(*this, level, deps_span);
2147
5.19k
    }
2148
2149
    // Wipe the list of to-be-added dependencies now that they are applied.
2150
4.98k
    clusterset.m_deps_to_add.clear();
2151
4.98k
    Compact();
2152
    // Also no further Cluster mergings are needed (note that we clear, but don't set to
2153
    // std::nullopt, as that would imply the groupings are unknown).
2154
4.98k
    clusterset.m_group_data = GroupData{};
2155
4.98k
}
2156
2157
std::pair<uint64_t, bool> GenericClusterImpl::Relinearize(TxGraphImpl& graph, int level, uint64_t max_cost) noexcept
2158
5.20k
{
2159
    // We can only relinearize Clusters that do not need splitting.
2160
5.20k
    Assume(!NeedsSplitting());
2161
    // No work is required for Clusters which are already optimally linearized.
2162
5.20k
    if (IsOptimal()) return {0, false};
2163
    // Invoke the actual linearization algorithm (passing in the existing one).
2164
5.20k
    uint64_t rng_seed = graph.m_rng.rand64();
2165
72.1k
    const auto fallback_order = [&](DepGraphIndex a, DepGraphIndex b) noexcept {
2166
72.1k
        const auto ref_a = graph.m_entries[m_mapping[a]].m_ref;
2167
72.1k
        const auto ref_b = graph.m_entries[m_mapping[b]].m_ref;
2168
72.1k
        return graph.m_fallback_order(*ref_a, *ref_b);
2169
72.1k
    };
2170
5.20k
    auto [linearization, optimal, cost] = Linearize(
2171
5.20k
        /*depgraph=*/m_depgraph,
2172
5.20k
        /*max_cost=*/max_cost,
2173
5.20k
        /*rng_seed=*/rng_seed,
2174
5.20k
        /*fallback_order=*/fallback_order,
2175
5.20k
        /*old_linearization=*/m_linearization,
2176
5.20k
        /*is_topological=*/IsTopological());
2177
    // Postlinearize to improve the linearization (if optimal, only the sub-chunk order).
2178
    // This also guarantees that all chunks are connected (even when non-optimal).
2179
5.20k
    PostLinearize(m_depgraph, linearization);
2180
    // Update the linearization.
2181
5.20k
    m_linearization = std::move(linearization);
2182
    // Update the Cluster's quality.
2183
5.20k
    bool improved = false;
2184
5.20k
    if (optimal) {
2185
5.20k
        graph.SetClusterQuality(level, m_quality, m_setindex, QualityLevel::OPTIMAL);
2186
5.20k
        improved = true;
2187
5.20k
    } else if (max_cost >= graph.m_acceptable_cost && !IsAcceptable()) {
2188
1
        graph.SetClusterQuality(level, m_quality, m_setindex, QualityLevel::ACCEPTABLE);
2189
1
        improved = true;
2190
1
    } else if (!IsTopological()) {
2191
0
        graph.SetClusterQuality(level, m_quality, m_setindex, QualityLevel::NEEDS_RELINEARIZE);
2192
0
        improved = true;
2193
0
    }
2194
    // Update the Entry objects.
2195
5.20k
    Updated(graph, /*level=*/level, /*rename=*/false);
2196
5.20k
    return {cost, improved};
2197
5.20k
}
2198
2199
std::pair<uint64_t, bool> SingletonClusterImpl::Relinearize(TxGraphImpl& graph, int level, uint64_t max_cost) noexcept
2200
0
{
2201
    // All singletons are optimal, oversized, or need splitting. Each of these precludes
2202
    // Relinearize from being called.
2203
0
    assert(false);
2204
0
    return {0, false};
2205
0
}
2206
2207
void TxGraphImpl::MakeAcceptable(Cluster& cluster, int level) noexcept
2208
274M
{
2209
    // Relinearize the Cluster if needed.
2210
274M
    if (!cluster.NeedsSplitting() && !cluster.IsAcceptable() && !cluster.IsOversized()) {
2211
96
        cluster.Relinearize(*this, level, m_acceptable_cost);
2212
96
    }
2213
274M
}
2214
2215
void TxGraphImpl::MakeAllAcceptable(int level) noexcept
2216
168k
{
2217
168k
    ApplyDependencies(level);
2218
168k
    auto& clusterset = GetClusterSet(level);
2219
168k
    if (clusterset.m_oversized == true) return;
2220
337k
    for (auto quality : {QualityLevel::NEEDS_FIX, QualityLevel::NEEDS_RELINEARIZE}) {
2221
337k
        auto& queue = clusterset.m_clusters[int(quality)];
2222
337k
        while (!queue.empty()) {
2223
87
            MakeAcceptable(*queue.back().get(), level);
2224
87
        }
2225
337k
    }
2226
168k
}
2227
2228
1.63k
GenericClusterImpl::GenericClusterImpl(uint64_t sequence) noexcept : Cluster{sequence} {}
2229
2230
void TxGraphImpl::AddTransaction(Ref& arg, const FeePerWeight& feerate) noexcept
2231
127k
{
2232
127k
    Assume(m_main_chunkindex_observers == 0 || GetTopLevel() != 0);
2233
127k
    Assume(feerate.size > 0);
2234
127k
    Assume(GetRefGraph(arg) == nullptr);
2235
    // Construct a new Entry, and link it with the Ref.
2236
127k
    auto idx = m_entries.size();
2237
127k
    m_entries.emplace_back();
2238
127k
    auto& entry = m_entries.back();
2239
127k
    entry.m_main_chunkindex_iterator = m_main_chunkindex.end();
2240
127k
    entry.m_ref = &arg;
2241
127k
    GetRefGraph(arg) = this;
2242
127k
    GetRefIndex(arg) = idx;
2243
    // Construct a new singleton Cluster (which is necessarily optimally linearized).
2244
127k
    bool oversized = uint64_t(feerate.size) > m_max_cluster_size;
2245
127k
    auto cluster = CreateEmptyCluster(1);
2246
127k
    cluster->AppendTransaction(idx, feerate);
2247
127k
    auto cluster_ptr = cluster.get();
2248
127k
    int level = GetTopLevel();
2249
127k
    auto& clusterset = GetClusterSet(level);
2250
127k
    InsertCluster(level, std::move(cluster), oversized ? QualityLevel::OVERSIZED_SINGLETON : QualityLevel::OPTIMAL);
2251
127k
    cluster_ptr->Updated(*this, /*level=*/level, /*rename=*/false);
2252
127k
    clusterset.m_cluster_usage += cluster_ptr->TotalMemoryUsage();
2253
127k
    ++clusterset.m_txcount;
2254
    // Deal with individually oversized transactions.
2255
127k
    if (oversized) {
2256
9
        ++clusterset.m_txcount_oversized;
2257
9
        clusterset.m_oversized = true;
2258
9
        clusterset.m_group_data = std::nullopt;
2259
9
    }
2260
127k
}
2261
2262
void TxGraphImpl::RemoveTransaction(const Ref& arg) noexcept
2263
2.18k
{
2264
    // Don't do anything if the Ref is empty (which may be indicative of the transaction already
2265
    // having been removed).
2266
2.18k
    if (GetRefGraph(arg) == nullptr) return;
2267
2.18k
    Assume(GetRefGraph(arg) == this);
2268
2.18k
    Assume(m_main_chunkindex_observers == 0 || GetTopLevel() != 0);
2269
    // Find the Cluster the transaction is in, and stop if it isn't in any.
2270
2.18k
    int level = GetTopLevel();
2271
2.18k
    auto cluster = FindCluster(GetRefIndex(arg), level);
2272
2.18k
    if (cluster == nullptr) return;
2273
    // Remember that the transaction is to be removed.
2274
2.18k
    auto& clusterset = GetClusterSet(level);
2275
2.18k
    clusterset.m_to_remove.push_back(GetRefIndex(arg));
2276
    // Wipe m_group_data (as it will need to be recomputed).
2277
2.18k
    clusterset.m_group_data.reset();
2278
2.18k
    if (clusterset.m_oversized == true) clusterset.m_oversized = std::nullopt;
2279
2.18k
}
2280
2281
void TxGraphImpl::AddDependency(const Ref& parent, const Ref& child) noexcept
2282
76.7k
{
2283
    // Don't do anything if either Ref is empty (which may be indicative of it having already been
2284
    // removed).
2285
76.7k
    if (GetRefGraph(parent) == nullptr || GetRefGraph(child) == nullptr) return;
2286
76.7k
    Assume(GetRefGraph(parent) == this && GetRefGraph(child) == this);
2287
76.7k
    Assume(m_main_chunkindex_observers == 0 || GetTopLevel() != 0);
2288
    // Don't do anything if this is a dependency on self.
2289
76.7k
    if (GetRefIndex(parent) == GetRefIndex(child)) return;
2290
    // Find the Cluster the parent and child transaction are in, and stop if either appears to be
2291
    // already removed.
2292
76.7k
    int level = GetTopLevel();
2293
76.7k
    auto par_cluster = FindCluster(GetRefIndex(parent), level);
2294
76.7k
    if (par_cluster == nullptr) return;
2295
76.7k
    auto chl_cluster = FindCluster(GetRefIndex(child), level);
2296
76.7k
    if (chl_cluster == nullptr) return;
2297
    // Remember that this dependency is to be applied.
2298
76.7k
    auto& clusterset = GetClusterSet(level);
2299
76.7k
    clusterset.m_deps_to_add.emplace_back(GetRefIndex(parent), GetRefIndex(child));
2300
    // Wipe m_group_data (as it will need to be recomputed).
2301
76.7k
    clusterset.m_group_data.reset();
2302
76.7k
    if (clusterset.m_oversized == false) clusterset.m_oversized = std::nullopt;
2303
76.7k
}
2304
2305
bool TxGraphImpl::Exists(const Ref& arg, Level level_select) noexcept
2306
302
{
2307
302
    if (GetRefGraph(arg) == nullptr) return false;
2308
302
    Assume(GetRefGraph(arg) == this);
2309
302
    size_t level = GetSpecifiedLevel(level_select);
2310
    // Make sure the transaction isn't scheduled for removal.
2311
302
    ApplyRemovals(level);
2312
302
    auto cluster = FindCluster(GetRefIndex(arg), level);
2313
302
    return cluster != nullptr;
2314
302
}
2315
2316
void GenericClusterImpl::GetAncestorRefs(const TxGraphImpl& graph, std::span<std::pair<Cluster*, DepGraphIndex>>& args, std::vector<TxGraph::Ref*>& output) noexcept
2317
47.7k
{
2318
    /** The union of all ancestors to be returned. */
2319
47.7k
    SetType ancestors_union;
2320
    // Process elements from the front of args, as long as they apply.
2321
95.5k
    while (!args.empty()) {
2322
47.7k
        if (args.front().first != this) break;
2323
47.7k
        ancestors_union |= m_depgraph.Ancestors(args.front().second);
2324
47.7k
        args = args.subspan(1);
2325
47.7k
    }
2326
47.7k
    Assume(ancestors_union.Any());
2327
    // Translate all ancestors (in arbitrary order) to Refs (if they have any), and return them.
2328
257k
    for (auto idx : ancestors_union) {
2329
257k
        const auto& entry = graph.m_entries[m_mapping[idx]];
2330
257k
        Assume(entry.m_ref != nullptr);
2331
257k
        output.push_back(entry.m_ref);
2332
257k
    }
2333
47.7k
}
2334
2335
void SingletonClusterImpl::GetAncestorRefs(const TxGraphImpl& graph, std::span<std::pair<Cluster*, DepGraphIndex>>& args, std::vector<TxGraph::Ref*>& output) noexcept
2336
119k
{
2337
119k
    Assume(GetTxCount());
2338
239k
    while (!args.empty()) {
2339
119k
        if (args.front().first != this) break;
2340
119k
        args = args.subspan(1);
2341
119k
    }
2342
119k
    const auto& entry = graph.m_entries[m_graph_index];
2343
119k
    Assume(entry.m_ref != nullptr);
2344
119k
    output.push_back(entry.m_ref);
2345
119k
}
2346
2347
void GenericClusterImpl::GetDescendantRefs(const TxGraphImpl& graph, std::span<std::pair<Cluster*, DepGraphIndex>>& args, std::vector<TxGraph::Ref*>& output) noexcept
2348
46.5k
{
2349
    /** The union of all descendants to be returned. */
2350
46.5k
    SetType descendants_union;
2351
    // Process elements from the front of args, as long as they apply.
2352
93.1k
    while (!args.empty()) {
2353
46.5k
        if (args.front().first != this) break;
2354
46.5k
        descendants_union |= m_depgraph.Descendants(args.front().second);
2355
46.5k
        args = args.subspan(1);
2356
46.5k
    }
2357
46.5k
    Assume(descendants_union.Any());
2358
    // Translate all descendants (in arbitrary order) to Refs (if they have any), and return them.
2359
395k
    for (auto idx : descendants_union) {
2360
395k
        const auto& entry = graph.m_entries[m_mapping[idx]];
2361
395k
        Assume(entry.m_ref != nullptr);
2362
395k
        output.push_back(entry.m_ref);
2363
395k
    }
2364
46.5k
}
2365
2366
void SingletonClusterImpl::GetDescendantRefs(const TxGraphImpl& graph, std::span<std::pair<Cluster*, DepGraphIndex>>& args, std::vector<TxGraph::Ref*>& output) noexcept
2367
42.2k
{
2368
    // In a singleton cluster, the ancestors or descendants are always just the entire cluster.
2369
42.2k
    GetAncestorRefs(graph, args, output);
2370
42.2k
}
2371
2372
bool GenericClusterImpl::GetClusterRefs(TxGraphImpl& graph, std::span<TxGraph::Ref*> range, LinearizationIndex start_pos) noexcept
2373
193k
{
2374
    // Translate the transactions in the Cluster (in linearization order, starting at start_pos in
2375
    // the linearization) to Refs, and fill them in range.
2376
279k
    for (auto& ref : range) {
2377
279k
        Assume(start_pos < m_linearization.size());
2378
279k
        const auto& entry = graph.m_entries[m_mapping[m_linearization[start_pos++]]];
2379
279k
        Assume(entry.m_ref != nullptr);
2380
279k
        ref = entry.m_ref;
2381
279k
    }
2382
    // Return whether start_pos has advanced to the end of the Cluster.
2383
193k
    return start_pos == m_linearization.size();
2384
193k
}
2385
2386
bool SingletonClusterImpl::GetClusterRefs(TxGraphImpl& graph, std::span<TxGraph::Ref*> range, LinearizationIndex start_pos) noexcept
2387
74.9k
{
2388
74.9k
    Assume(!range.empty());
2389
74.9k
    Assume(GetTxCount());
2390
74.9k
    Assume(start_pos == 0);
2391
74.9k
    const auto& entry = graph.m_entries[m_graph_index];
2392
74.9k
    Assume(entry.m_ref != nullptr);
2393
74.9k
    range[0] = entry.m_ref;
2394
74.9k
    return true;
2395
74.9k
}
2396
2397
FeePerWeight GenericClusterImpl::GetIndividualFeerate(DepGraphIndex idx) noexcept
2398
9
{
2399
9
    return FeePerWeight::FromFeeFrac(m_depgraph.FeeRate(idx));
2400
9
}
2401
2402
FeePerWeight SingletonClusterImpl::GetIndividualFeerate(DepGraphIndex idx) noexcept
2403
2
{
2404
2
    Assume(GetTxCount());
2405
2
    Assume(idx == 0);
2406
2
    return m_feerate;
2407
2
}
2408
2409
void GenericClusterImpl::MakeStagingTransactionsMissing(TxGraphImpl& graph) noexcept
2410
26
{
2411
    // Mark all transactions of a Cluster missing, needed when aborting staging, so that the
2412
    // corresponding Locators don't retain references into aborted Clusters.
2413
137
    for (auto ci : m_linearization) {
2414
137
        GraphIndex idx = m_mapping[ci];
2415
137
        auto& entry = graph.m_entries[idx];
2416
137
        entry.m_locator[1].SetMissing();
2417
137
    }
2418
26
}
2419
2420
void SingletonClusterImpl::MakeStagingTransactionsMissing(TxGraphImpl& graph) noexcept
2421
11.2k
{
2422
11.2k
    if (GetTxCount()) {
2423
11.2k
        auto& entry = graph.m_entries[m_graph_index];
2424
11.2k
        entry.m_locator[1].SetMissing();
2425
11.2k
    }
2426
11.2k
}
2427
2428
std::vector<TxGraph::Ref*> TxGraphImpl::GetAncestors(const Ref& arg, Level level_select) noexcept
2429
126k
{
2430
    // Return the empty vector if the Ref is empty.
2431
126k
    if (GetRefGraph(arg) == nullptr) return {};
2432
126k
    Assume(GetRefGraph(arg) == this);
2433
    // Apply all removals and dependencies, as the result might be incorrect otherwise.
2434
126k
    size_t level = GetSpecifiedLevel(level_select);
2435
126k
    ApplyDependencies(level);
2436
    // Ancestry cannot be known if unapplied dependencies remain.
2437
126k
    Assume(GetClusterSet(level).m_deps_to_add.empty());
2438
    // Find the Cluster the argument is in, and return the empty vector if it isn't in any.
2439
126k
    auto [cluster, cluster_level] = FindClusterAndLevel(GetRefIndex(arg), level);
2440
126k
    if (cluster == nullptr) return {};
2441
    // Dispatch to the Cluster.
2442
125k
    std::pair<Cluster*, DepGraphIndex> match = {cluster, m_entries[GetRefIndex(arg)].m_locator[cluster_level].index};
2443
125k
    auto matches = std::span(&match, 1);
2444
125k
    std::vector<TxGraph::Ref*> ret;
2445
125k
    cluster->GetAncestorRefs(*this, matches, ret);
2446
125k
    return ret;
2447
126k
}
2448
2449
std::vector<TxGraph::Ref*> TxGraphImpl::GetDescendants(const Ref& arg, Level level_select) noexcept
2450
88.6k
{
2451
    // Return the empty vector if the Ref is empty.
2452
88.6k
    if (GetRefGraph(arg) == nullptr) return {};
2453
88.6k
    Assume(GetRefGraph(arg) == this);
2454
    // Apply all removals and dependencies, as the result might be incorrect otherwise.
2455
88.6k
    size_t level = GetSpecifiedLevel(level_select);
2456
88.6k
    ApplyDependencies(level);
2457
    // Ancestry cannot be known if unapplied dependencies remain.
2458
88.6k
    Assume(GetClusterSet(level).m_deps_to_add.empty());
2459
    // Find the Cluster the argument is in, and return the empty vector if it isn't in any.
2460
88.6k
    auto [cluster, cluster_level] = FindClusterAndLevel(GetRefIndex(arg), level);
2461
88.6k
    if (cluster == nullptr) return {};
2462
    // Dispatch to the Cluster.
2463
88.6k
    std::pair<Cluster*, DepGraphIndex> match = {cluster, m_entries[GetRefIndex(arg)].m_locator[cluster_level].index};
2464
88.6k
    auto matches = std::span(&match, 1);
2465
88.6k
    std::vector<TxGraph::Ref*> ret;
2466
88.6k
    cluster->GetDescendantRefs(*this, matches, ret);
2467
88.6k
    return ret;
2468
88.6k
}
2469
2470
std::vector<TxGraph::Ref*> TxGraphImpl::GetAncestorsUnion(std::span<const Ref* const> args, Level level_select) noexcept
2471
0
{
2472
    // Apply all dependencies, as the result might be incorrect otherwise.
2473
0
    size_t level = GetSpecifiedLevel(level_select);
2474
0
    ApplyDependencies(level);
2475
    // Ancestry cannot be known if unapplied dependencies remain.
2476
0
    Assume(GetClusterSet(level).m_deps_to_add.empty());
2477
2478
    // Translate args to matches.
2479
0
    std::vector<std::pair<Cluster*, DepGraphIndex>> matches;
2480
0
    matches.reserve(args.size());
2481
0
    for (auto arg : args) {
2482
0
        Assume(arg);
2483
        // Skip empty Refs.
2484
0
        if (GetRefGraph(*arg) == nullptr) continue;
2485
0
        Assume(GetRefGraph(*arg) == this);
2486
        // Find the Cluster the argument is in, and skip if none is found.
2487
0
        auto [cluster, cluster_level] = FindClusterAndLevel(GetRefIndex(*arg), level);
2488
0
        if (cluster == nullptr) continue;
2489
        // Append to matches.
2490
0
        matches.emplace_back(cluster, m_entries[GetRefIndex(*arg)].m_locator[cluster_level].index);
2491
0
    }
2492
    // Group by Cluster.
2493
0
    std::sort(matches.begin(), matches.end(), [](auto& a, auto& b) noexcept { return CompareClusters(a.first, b.first) < 0; });
2494
    // Dispatch to the Clusters.
2495
0
    std::span match_span(matches);
2496
0
    std::vector<TxGraph::Ref*> ret;
2497
0
    while (!match_span.empty()) {
2498
0
        match_span.front().first->GetAncestorRefs(*this, match_span, ret);
2499
0
    }
2500
0
    return ret;
2501
0
}
2502
2503
std::vector<TxGraph::Ref*> TxGraphImpl::GetDescendantsUnion(std::span<const Ref* const> args, Level level_select) noexcept
2504
21.8k
{
2505
    // Apply all dependencies, as the result might be incorrect otherwise.
2506
21.8k
    size_t level = GetSpecifiedLevel(level_select);
2507
21.8k
    ApplyDependencies(level);
2508
    // Ancestry cannot be known if unapplied dependencies remain.
2509
21.8k
    Assume(GetClusterSet(level).m_deps_to_add.empty());
2510
2511
    // Translate args to matches.
2512
21.8k
    std::vector<std::pair<Cluster*, DepGraphIndex>> matches;
2513
21.8k
    matches.reserve(args.size());
2514
21.8k
    for (auto arg : args) {
2515
89
        Assume(arg);
2516
        // Skip empty Refs.
2517
89
        if (GetRefGraph(*arg) == nullptr) continue;
2518
89
        Assume(GetRefGraph(*arg) == this);
2519
        // Find the Cluster the argument is in, and skip if none is found.
2520
89
        auto [cluster, cluster_level] = FindClusterAndLevel(GetRefIndex(*arg), level);
2521
89
        if (cluster == nullptr) continue;
2522
        // Append to matches.
2523
89
        matches.emplace_back(cluster, m_entries[GetRefIndex(*arg)].m_locator[cluster_level].index);
2524
89
    }
2525
    // Group by Cluster.
2526
21.8k
    std::sort(matches.begin(), matches.end(), [](auto& a, auto& b) noexcept { return CompareClusters(a.first, b.first) < 0; });
2527
    // Dispatch to the Clusters.
2528
21.8k
    std::span match_span(matches);
2529
21.8k
    std::vector<TxGraph::Ref*> ret;
2530
21.8k
    while (!match_span.empty()) {
2531
89
        match_span.front().first->GetDescendantRefs(*this, match_span, ret);
2532
89
    }
2533
21.8k
    return ret;
2534
21.8k
}
2535
2536
std::vector<TxGraph::Ref*> TxGraphImpl::GetCluster(const Ref& arg, Level level_select) noexcept
2537
98.5k
{
2538
    // Return the empty vector if the Ref is empty (which may be indicative of the transaction
2539
    // having been removed already.
2540
98.5k
    if (GetRefGraph(arg) == nullptr) return {};
2541
98.5k
    Assume(GetRefGraph(arg) == this);
2542
    // Apply all removals and dependencies, as the result might be incorrect otherwise.
2543
98.5k
    size_t level = GetSpecifiedLevel(level_select);
2544
98.5k
    ApplyDependencies(level);
2545
    // Cluster linearization cannot be known if unapplied dependencies remain.
2546
98.5k
    Assume(GetClusterSet(level).m_deps_to_add.empty());
2547
    // Find the Cluster the argument is in, and return the empty vector if it isn't in any.
2548
98.5k
    auto [cluster, cluster_level] = FindClusterAndLevel(GetRefIndex(arg), level);
2549
98.5k
    if (cluster == nullptr) return {};
2550
    // Make sure the Cluster has an acceptable quality level, and then dispatch to it.
2551
98.5k
    MakeAcceptable(*cluster, cluster_level);
2552
98.5k
    std::vector<TxGraph::Ref*> ret(cluster->GetTxCount());
2553
98.5k
    cluster->GetClusterRefs(*this, ret, 0);
2554
98.5k
    return ret;
2555
98.5k
}
2556
2557
TxGraph::GraphIndex TxGraphImpl::GetTransactionCount(Level level_select) noexcept
2558
23
{
2559
23
    size_t level = GetSpecifiedLevel(level_select);
2560
23
    ApplyRemovals(level);
2561
23
    return GetClusterSet(level).m_txcount;
2562
23
}
2563
2564
FeePerWeight TxGraphImpl::GetIndividualFeerate(const Ref& arg) noexcept
2565
11
{
2566
    // Return the empty FeePerWeight if the passed Ref is empty.
2567
11
    if (GetRefGraph(arg) == nullptr) return {};
2568
11
    Assume(GetRefGraph(arg) == this);
2569
    // Find the cluster the argument is in (the level does not matter as individual feerates will
2570
    // be identical if it occurs in both), and return the empty FeePerWeight if it isn't in any.
2571
11
    Cluster* cluster{nullptr};
2572
11
    int level;
2573
11
    for (level = 0; level <= GetTopLevel(); ++level) {
2574
        // Apply removals, so that we can correctly report FeePerWeight{} for non-existing
2575
        // transactions.
2576
11
        ApplyRemovals(level);
2577
11
        if (m_entries[GetRefIndex(arg)].m_locator[level].IsPresent()) {
2578
11
            cluster = m_entries[GetRefIndex(arg)].m_locator[level].cluster;
2579
11
            break;
2580
11
        }
2581
11
    }
2582
11
    if (cluster == nullptr) return {};
2583
    // Dispatch to the Cluster.
2584
11
    return cluster->GetIndividualFeerate(m_entries[GetRefIndex(arg)].m_locator[level].index);
2585
11
}
2586
2587
FeePerWeight TxGraphImpl::GetMainChunkFeerate(const Ref& arg) noexcept
2588
59.8k
{
2589
    // Return the empty FeePerWeight if the passed Ref is empty.
2590
59.8k
    if (GetRefGraph(arg) == nullptr) return {};
2591
59.8k
    Assume(GetRefGraph(arg) == this);
2592
    // Apply all removals and dependencies, as the result might be inaccurate otherwise.
2593
59.8k
    ApplyDependencies(/*level=*/0);
2594
    // Chunk feerates cannot be accurately known if unapplied dependencies remain.
2595
59.8k
    Assume(m_main_clusterset.m_deps_to_add.empty());
2596
    // Find the cluster the argument is in, and return the empty FeePerWeight if it isn't in any.
2597
59.8k
    auto [cluster, cluster_level] = FindClusterAndLevel(GetRefIndex(arg), /*level=*/0);
2598
59.8k
    if (cluster == nullptr) return {};
2599
    // Make sure the Cluster has an acceptable quality level, and then return the transaction's
2600
    // chunk feerate.
2601
59.8k
    MakeAcceptable(*cluster, cluster_level);
2602
59.8k
    const auto& entry = m_entries[GetRefIndex(arg)];
2603
59.8k
    return entry.m_main_chunk_feerate;
2604
59.8k
}
2605
2606
bool TxGraphImpl::IsOversized(Level level_select) noexcept
2607
195k
{
2608
195k
    size_t level = GetSpecifiedLevel(level_select);
2609
195k
    auto& clusterset = GetClusterSet(level);
2610
195k
    if (clusterset.m_oversized.has_value()) {
2611
        // Return cached value if known.
2612
188k
        return *clusterset.m_oversized;
2613
188k
    }
2614
6.64k
    ApplyRemovals(level);
2615
6.64k
    if (clusterset.m_txcount_oversized > 0) {
2616
0
        clusterset.m_oversized = true;
2617
6.64k
    } else {
2618
        // Find which Clusters will need to be merged together, as that is where the oversize
2619
        // property is assessed.
2620
6.64k
        GroupClusters(level);
2621
6.64k
    }
2622
6.64k
    Assume(clusterset.m_oversized.has_value());
2623
6.64k
    return *clusterset.m_oversized;
2624
195k
}
2625
2626
void TxGraphImpl::StartStaging() noexcept
2627
62.5k
{
2628
    // Staging cannot already exist.
2629
62.5k
    Assume(!m_staging_clusterset.has_value());
2630
    // Apply all remaining dependencies in main before creating a staging graph. Once staging
2631
    // exists, we cannot merge Clusters anymore (because of interference with Clusters being
2632
    // pulled into staging), so to make sure all inspectors are available (if not oversized), do
2633
    // all merging work now. Call SplitAll() first, so that even if ApplyDependencies does not do
2634
    // any thing due to knowing the result is oversized, splitting is still performed.
2635
62.5k
    SplitAll(0);
2636
62.5k
    ApplyDependencies(0);
2637
    // Construct the staging ClusterSet.
2638
62.5k
    m_staging_clusterset.emplace();
2639
    // Copy statistics, precomputed data, and to-be-applied dependencies (only if oversized) to
2640
    // the new graph. To-be-applied removals will always be empty at this point.
2641
62.5k
    m_staging_clusterset->m_txcount = m_main_clusterset.m_txcount;
2642
62.5k
    m_staging_clusterset->m_txcount_oversized = m_main_clusterset.m_txcount_oversized;
2643
62.5k
    m_staging_clusterset->m_deps_to_add = m_main_clusterset.m_deps_to_add;
2644
62.5k
    m_staging_clusterset->m_group_data = m_main_clusterset.m_group_data;
2645
62.5k
    m_staging_clusterset->m_oversized = m_main_clusterset.m_oversized;
2646
62.5k
    Assume(m_staging_clusterset->m_oversized.has_value());
2647
62.5k
    m_staging_clusterset->m_cluster_usage = 0;
2648
62.5k
}
2649
2650
void TxGraphImpl::AbortStaging() noexcept
2651
10.6k
{
2652
    // Staging must exist.
2653
10.6k
    Assume(m_staging_clusterset.has_value());
2654
    // Mark all removed transactions as Missing (so the staging locator for these transactions
2655
    // can be reused if another staging is created).
2656
10.6k
    for (auto idx : m_staging_clusterset->m_removed) {
2657
586
        m_entries[idx].m_locator[1].SetMissing();
2658
586
    }
2659
    // Do the same with the non-removed transactions in staging Clusters.
2660
85.5k
    for (int quality = 0; quality < int(QualityLevel::NONE); ++quality) {
2661
74.8k
        for (auto& cluster : m_staging_clusterset->m_clusters[quality]) {
2662
11.2k
            cluster->MakeStagingTransactionsMissing(*this);
2663
11.2k
        }
2664
74.8k
    }
2665
    // Destroy the staging ClusterSet.
2666
10.6k
    m_staging_clusterset.reset();
2667
10.6k
    Compact();
2668
10.6k
    if (!m_main_clusterset.m_group_data.has_value()) {
2669
        // In case m_oversized in main was kept after a Ref destruction while staging exists, we
2670
        // need to re-evaluate m_oversized now.
2671
0
        if (m_main_clusterset.m_to_remove.empty() && m_main_clusterset.m_txcount_oversized > 0) {
2672
            // It is possible that a Ref destruction caused a removal in main while staging existed.
2673
            // In this case, m_txcount_oversized may be inaccurate.
2674
0
            m_main_clusterset.m_oversized = true;
2675
0
        } else {
2676
0
            m_main_clusterset.m_oversized = std::nullopt;
2677
0
        }
2678
0
    }
2679
10.6k
}
2680
2681
void TxGraphImpl::CommitStaging() noexcept
2682
51.9k
{
2683
    // Staging must exist.
2684
51.9k
    Assume(m_staging_clusterset.has_value());
2685
51.9k
    Assume(m_main_chunkindex_observers == 0);
2686
    // Get rid of removed transactions in staging before moving to main, so they do not need to be
2687
    // added to the chunk index there. Doing so is impossible if they were unlinked, and thus have
2688
    // no Ref anymore to pass to the fallback comparator.
2689
51.9k
    ApplyRemovals(/*up_to_level=*/1);
2690
    // Delete all conflicting Clusters in main, to make place for moving the staging ones
2691
    // there. All of these have been copied to staging in PullIn().
2692
51.9k
    auto conflicts = GetConflicts();
2693
51.9k
    for (Cluster* conflict : conflicts) {
2694
1.26k
        conflict->Clear(*this, /*level=*/0);
2695
1.26k
        DeleteCluster(*conflict, /*level=*/0);
2696
1.26k
    }
2697
    // Mark the removed transactions as Missing (so the staging locator for these transactions
2698
    // can be reused if another staging is created).
2699
51.9k
    for (auto idx : m_staging_clusterset->m_removed) {
2700
1.59k
        m_entries[idx].m_locator[1].SetMissing();
2701
1.59k
    }
2702
    // Then move all Clusters in staging to main.
2703
415k
    for (int quality = 0; quality < int(QualityLevel::NONE); ++quality) {
2704
363k
        auto& stage_sets = m_staging_clusterset->m_clusters[quality];
2705
415k
        while (!stage_sets.empty()) {
2706
52.0k
            stage_sets.back()->MoveToMain(*this);
2707
52.0k
        }
2708
363k
    }
2709
    // Move all statistics, precomputed data, and to-be-applied removals and dependencies.
2710
51.9k
    m_main_clusterset.m_deps_to_add = std::move(m_staging_clusterset->m_deps_to_add);
2711
51.9k
    m_main_clusterset.m_to_remove = std::move(m_staging_clusterset->m_to_remove);
2712
51.9k
    m_main_clusterset.m_group_data = std::move(m_staging_clusterset->m_group_data);
2713
51.9k
    m_main_clusterset.m_oversized = std::move(m_staging_clusterset->m_oversized);
2714
51.9k
    m_main_clusterset.m_txcount = std::move(m_staging_clusterset->m_txcount);
2715
51.9k
    m_main_clusterset.m_txcount_oversized = std::move(m_staging_clusterset->m_txcount_oversized);
2716
    // Delete the old staging graph, after all its information was moved to main.
2717
51.9k
    m_staging_clusterset.reset();
2718
51.9k
    Compact();
2719
51.9k
}
2720
2721
void GenericClusterImpl::SetFee(TxGraphImpl& graph, int level, DepGraphIndex idx, int64_t fee) noexcept
2722
16
{
2723
    // Make sure the specified DepGraphIndex exists in this Cluster.
2724
16
    Assume(m_depgraph.Positions()[idx]);
2725
    // Bail out if the fee isn't actually being changed.
2726
16
    if (m_depgraph.FeeRate(idx).fee == fee) return;
2727
    // Update the fee, remember that relinearization will be necessary, and update the Entries
2728
    // in the same Cluster.
2729
16
    m_depgraph.FeeRate(idx).fee = fee;
2730
16
    if (m_quality == QualityLevel::OVERSIZED_SINGLETON) {
2731
        // Nothing to do.
2732
16
    } else if (IsAcceptable()) {
2733
12
        graph.SetClusterQuality(level, m_quality, m_setindex, QualityLevel::NEEDS_RELINEARIZE);
2734
12
    }
2735
16
    Updated(graph, /*level=*/level, /*rename=*/false);
2736
16
}
2737
2738
void SingletonClusterImpl::SetFee(TxGraphImpl& graph, int level, DepGraphIndex idx, int64_t fee) noexcept
2739
287
{
2740
287
    Assume(GetTxCount());
2741
287
    Assume(idx == 0);
2742
287
    m_feerate.fee = fee;
2743
287
    Updated(graph, /*level=*/level, /*rename=*/false);
2744
287
}
2745
2746
void TxGraphImpl::SetTransactionFee(const Ref& ref, int64_t fee) noexcept
2747
303
{
2748
    // Don't do anything if the passed Ref is empty.
2749
303
    if (GetRefGraph(ref) == nullptr) return;
2750
303
    Assume(GetRefGraph(ref) == this);
2751
303
    Assume(m_main_chunkindex_observers == 0);
2752
    // Find the entry, its locator, and inform its Cluster about the new feerate, if any.
2753
303
    auto& entry = m_entries[GetRefIndex(ref)];
2754
909
    for (int level = 0; level < MAX_LEVELS; ++level) {
2755
606
        auto& locator = entry.m_locator[level];
2756
606
        if (locator.IsPresent()) {
2757
303
            locator.cluster->SetFee(*this, level, locator.index, fee);
2758
303
        }
2759
606
    }
2760
303
}
2761
2762
std::strong_ordering TxGraphImpl::CompareMainOrder(const Ref& a, const Ref& b) noexcept
2763
137M
{
2764
    // The references must not be empty.
2765
137M
    Assume(GetRefGraph(a) == this);
2766
137M
    Assume(GetRefGraph(b) == this);
2767
    // Apply dependencies in main.
2768
137M
    ApplyDependencies(0);
2769
137M
    Assume(m_main_clusterset.m_deps_to_add.empty());
2770
    // Make both involved Clusters acceptable, so chunk feerates are relevant.
2771
137M
    const auto& entry_a = m_entries[GetRefIndex(a)];
2772
137M
    const auto& entry_b = m_entries[GetRefIndex(b)];
2773
137M
    const auto& locator_a = entry_a.m_locator[0];
2774
137M
    const auto& locator_b = entry_b.m_locator[0];
2775
137M
    Assume(locator_a.IsPresent());
2776
137M
    Assume(locator_b.IsPresent());
2777
137M
    MakeAcceptable(*locator_a.cluster, /*level=*/0);
2778
137M
    MakeAcceptable(*locator_b.cluster, /*level=*/0);
2779
    // Invoke comparison logic.
2780
137M
    return CompareMainTransactions(GetRefIndex(a), GetRefIndex(b));
2781
137M
}
2782
2783
TxGraph::GraphIndex TxGraphImpl::CountDistinctClusters(std::span<const Ref* const> refs, Level level_select) noexcept
2784
1.38k
{
2785
1.38k
    size_t level = GetSpecifiedLevel(level_select);
2786
1.38k
    ApplyDependencies(level);
2787
1.38k
    auto& clusterset = GetClusterSet(level);
2788
1.38k
    Assume(clusterset.m_deps_to_add.empty());
2789
    // Build a vector of Clusters that the specified Refs occur in.
2790
1.38k
    std::vector<Cluster*> clusters;
2791
1.38k
    clusters.reserve(refs.size());
2792
2.64k
    for (const Ref* ref : refs) {
2793
2.64k
        Assume(ref);
2794
2.64k
        if (GetRefGraph(*ref) == nullptr) continue;
2795
2.64k
        Assume(GetRefGraph(*ref) == this);
2796
2.64k
        auto cluster = FindCluster(GetRefIndex(*ref), level);
2797
2.64k
        if (cluster != nullptr) clusters.push_back(cluster);
2798
2.64k
    }
2799
    // Count the number of distinct elements in clusters.
2800
9.55k
    std::sort(clusters.begin(), clusters.end(), [](Cluster* a, Cluster* b) noexcept { return CompareClusters(a, b) < 0; });
2801
1.38k
    Cluster* last{nullptr};
2802
1.38k
    GraphIndex ret{0};
2803
2.64k
    for (Cluster* cluster : clusters) {
2804
2.64k
        ret += (cluster != last);
2805
2.64k
        last = cluster;
2806
2.64k
    }
2807
1.38k
    return ret;
2808
1.38k
}
2809
2810
std::pair<std::vector<FeeFrac>, std::vector<FeeFrac>> TxGraphImpl::GetMainStagingDiagrams() noexcept
2811
1.33k
{
2812
1.33k
    Assume(m_staging_clusterset.has_value());
2813
1.33k
    MakeAllAcceptable(0);
2814
1.33k
    Assume(m_main_clusterset.m_deps_to_add.empty()); // can only fail if main is oversized
2815
1.33k
    MakeAllAcceptable(1);
2816
1.33k
    Assume(m_staging_clusterset->m_deps_to_add.empty()); // can only fail if staging is oversized
2817
    // For all Clusters in main which conflict with Clusters in staging (i.e., all that are removed
2818
    // by, or replaced in, staging), gather their chunk feerates.
2819
1.33k
    auto main_clusters = GetConflicts();
2820
1.33k
    std::vector<FeeFrac> main_feerates, staging_feerates;
2821
1.73k
    for (Cluster* cluster : main_clusters) {
2822
1.73k
        cluster->AppendChunkFeerates(main_feerates);
2823
1.73k
    }
2824
    // Do the same for the Clusters in staging themselves.
2825
10.6k
    for (int quality = 0; quality < int(QualityLevel::NONE); ++quality) {
2826
9.35k
        for (const auto& cluster : m_staging_clusterset->m_clusters[quality]) {
2827
1.41k
            cluster->AppendChunkFeerates(staging_feerates);
2828
1.41k
        }
2829
9.35k
    }
2830
    // Sort both by decreasing feerate to obtain diagrams, and return them.
2831
5.18k
    std::sort(main_feerates.begin(), main_feerates.end(), [](auto& a, auto& b) { return a > b; });
2832
1.33k
    std::sort(staging_feerates.begin(), staging_feerates.end(), [](auto& a, auto& b) { return a > b; });
2833
1.33k
    return std::make_pair(std::move(main_feerates), std::move(staging_feerates));
2834
1.33k
}
2835
2836
void GenericClusterImpl::SanityCheck(const TxGraphImpl& graph, int level) const
2837
54.6k
{
2838
    // There must be an m_mapping for each m_depgraph position (including holes).
2839
54.6k
    assert(m_depgraph.PositionRange() == m_mapping.size());
2840
    // The linearization for this Cluster must contain every transaction once.
2841
54.6k
    assert(m_depgraph.TxCount() == m_linearization.size());
2842
    // Unless a split is to be applied, the cluster cannot have any holes.
2843
54.6k
    if (!NeedsSplitting()) {
2844
54.6k
        assert(m_depgraph.Positions() == SetType::Fill(m_depgraph.TxCount()));
2845
54.6k
    }
2846
2847
    // Compute the chunking of m_linearization.
2848
54.6k
    auto linchunking = ChunkLinearizationInfo(m_depgraph, m_linearization);
2849
54.6k
    unsigned chunk_num{0};
2850
2851
    // Verify m_linearization.
2852
54.6k
    SetType m_done;
2853
54.6k
    LinearizationIndex linindex{0};
2854
54.6k
    DepGraphIndex chunk_pos{0}; //!< position within the current chunk
2855
54.6k
    assert(m_depgraph.IsAcyclic());
2856
54.6k
    if (m_linearization.empty()) return;
2857
54.6k
    FeeFrac equal_feerate_prefix = linchunking[chunk_num].feerate;
2858
230k
    for (auto lin_pos : m_linearization) {
2859
230k
        assert(lin_pos < m_mapping.size());
2860
230k
        const auto& entry = graph.m_entries[m_mapping[lin_pos]];
2861
        // Check that the linearization is topological.
2862
230k
        m_done.Set(lin_pos);
2863
230k
        if (IsTopological()) {
2864
230k
            assert(m_done.IsSupersetOf(m_depgraph.Ancestors(lin_pos)));
2865
230k
        }
2866
        // Check that the Entry has a locator pointing back to this Cluster & position within it.
2867
230k
        assert(entry.m_locator[level].cluster == this);
2868
230k
        assert(entry.m_locator[level].index == lin_pos);
2869
        // For main-level entries, check linearization position and chunk feerate.
2870
230k
        if (level == 0 && IsAcceptable()) {
2871
230k
            assert(entry.m_main_lin_index == linindex);
2872
230k
            ++linindex;
2873
230k
            if (!linchunking[chunk_num].transactions[lin_pos]) {
2874
                // First transaction of a new chunk.
2875
157k
                ++chunk_num;
2876
157k
                assert(chunk_num < linchunking.size());
2877
157k
                chunk_pos = 0;
2878
157k
                if (linchunking[chunk_num].feerate << equal_feerate_prefix) {
2879
6.94k
                    equal_feerate_prefix = linchunking[chunk_num].feerate;
2880
150k
                } else {
2881
150k
                    assert(!(linchunking[chunk_num].feerate >> equal_feerate_prefix));
2882
150k
                    equal_feerate_prefix += linchunking[chunk_num].feerate;
2883
150k
                }
2884
157k
            }
2885
230k
            assert(entry.m_main_chunk_feerate == linchunking[chunk_num].feerate);
2886
230k
            assert(entry.m_main_equal_feerate_chunk_prefix_size == equal_feerate_prefix.size);
2887
            // Verify that an entry in the chunk index exists for every chunk-ending transaction.
2888
230k
            ++chunk_pos;
2889
230k
            if (graph.m_main_clusterset.m_to_remove.empty()) {
2890
230k
                bool is_chunk_end = (chunk_pos == linchunking[chunk_num].transactions.Count());
2891
230k
                assert((entry.m_main_chunkindex_iterator != graph.m_main_chunkindex.end()) == is_chunk_end);
2892
230k
                if (is_chunk_end) {
2893
212k
                    auto& chunk_data = *entry.m_main_chunkindex_iterator;
2894
212k
                    if (m_done == m_depgraph.Positions() && chunk_pos == 1) {
2895
46.0k
                        assert(chunk_data.m_chunk_count == LinearizationIndex(-1));
2896
166k
                    } else {
2897
166k
                        assert(chunk_data.m_chunk_count == chunk_pos);
2898
166k
                    }
2899
212k
                }
2900
230k
            }
2901
            // If this Cluster has an acceptable quality level, its chunks must be connected.
2902
230k
            assert(m_depgraph.IsConnected(linchunking[chunk_num].transactions));
2903
230k
        }
2904
230k
    }
2905
    // Verify that each element of m_depgraph occurred in m_linearization.
2906
54.6k
    assert(m_done == m_depgraph.Positions());
2907
54.6k
}
2908
2909
void SingletonClusterImpl::SanityCheck(const TxGraphImpl& graph, int level) const
2910
10.6M
{
2911
    // All singletons are optimal, oversized, or need splitting.
2912
10.6M
    Assume(IsOptimal() || IsOversized() || NeedsSplitting());
2913
10.6M
    if (GetTxCount()) {
2914
10.6M
        const auto& entry = graph.m_entries[m_graph_index];
2915
        // Check that the Entry has a locator pointing back to this Cluster & position within it.
2916
10.6M
        assert(entry.m_locator[level].cluster == this);
2917
10.6M
        assert(entry.m_locator[level].index == 0);
2918
        // For main-level entries, check linearization position and chunk feerate.
2919
10.6M
        if (level == 0 && IsAcceptable()) {
2920
10.6M
            assert(entry.m_main_lin_index == 0);
2921
10.6M
            assert(entry.m_main_chunk_feerate == m_feerate);
2922
10.6M
            assert(entry.m_main_equal_feerate_chunk_prefix_size == m_feerate.size);
2923
10.6M
            if (graph.m_main_clusterset.m_to_remove.empty()) {
2924
10.6M
                assert(entry.m_main_chunkindex_iterator != graph.m_main_chunkindex.end());
2925
10.6M
                auto& chunk_data = *entry.m_main_chunkindex_iterator;
2926
10.6M
                assert(chunk_data.m_chunk_count == LinearizationIndex(-1));
2927
10.6M
            }
2928
10.6M
        }
2929
10.6M
    }
2930
10.6M
}
2931
2932
void TxGraphImpl::SanityCheck() const
2933
128k
{
2934
    /** Which GraphIndexes ought to occur in m_unlinked, based on m_entries. */
2935
128k
    std::set<GraphIndex> expected_unlinked;
2936
    /** Which Clusters ought to occur in ClusterSet::m_clusters, based on m_entries. */
2937
128k
    std::set<const Cluster*> expected_clusters[MAX_LEVELS];
2938
    /** Which GraphIndexes ought to occur in ClusterSet::m_removed, based on m_entries. */
2939
128k
    std::set<GraphIndex> expected_removed[MAX_LEVELS];
2940
    /** Which Cluster::m_sequence values have been encountered. */
2941
128k
    std::set<uint64_t> sequences;
2942
    /** Which GraphIndexes ought to occur in m_main_chunkindex, based on m_entries. */
2943
128k
    std::set<GraphIndex> expected_chunkindex;
2944
    /** Whether compaction is possible in the current state. */
2945
128k
    bool compact_possible{true};
2946
2947
    // Go over all Entry objects in m_entries.
2948
11.0M
    for (GraphIndex idx = 0; idx < m_entries.size(); ++idx) {
2949
10.9M
        const auto& entry = m_entries[idx];
2950
10.9M
        if (entry.m_ref == nullptr) {
2951
            // Unlinked Entry must have indexes appear in m_unlinked.
2952
0
            expected_unlinked.insert(idx);
2953
10.9M
        } else {
2954
            // Every non-unlinked Entry must have a Ref that points back to it.
2955
10.9M
            assert(GetRefGraph(*entry.m_ref) == this);
2956
10.9M
            assert(GetRefIndex(*entry.m_ref) == idx);
2957
10.9M
        }
2958
10.9M
        if (entry.m_main_chunkindex_iterator != m_main_chunkindex.end()) {
2959
            // Remember which entries we see a chunkindex entry for.
2960
10.8M
            assert(entry.m_locator[0].IsPresent());
2961
10.8M
            expected_chunkindex.insert(idx);
2962
10.8M
        }
2963
        // Verify the Entry m_locators.
2964
10.9M
        bool was_present{false}, was_removed{false};
2965
32.7M
        for (int level = 0; level < MAX_LEVELS; ++level) {
2966
21.8M
            const auto& locator = entry.m_locator[level];
2967
            // Every Locator must be in exactly one of these 3 states.
2968
21.8M
            assert(locator.IsMissing() + locator.IsRemoved() + locator.IsPresent() == 1);
2969
21.8M
            if (locator.IsPresent()) {
2970
                // Once removed, a transaction cannot be revived.
2971
10.9M
                assert(!was_removed);
2972
                // Verify that the Cluster agrees with where the Locator claims the transaction is.
2973
10.9M
                assert(locator.cluster->GetClusterEntry(locator.index) == idx);
2974
                // Remember that we expect said Cluster to appear in the ClusterSet::m_clusters.
2975
10.9M
                expected_clusters[level].insert(locator.cluster);
2976
10.9M
                was_present = true;
2977
10.9M
            } else if (locator.IsRemoved()) {
2978
                // Level 0 (main) cannot have IsRemoved locators (IsMissing there means non-existing).
2979
0
                assert(level > 0);
2980
                // A Locator can only be IsRemoved if it was IsPresent before, and only once.
2981
0
                assert(was_present && !was_removed);
2982
                // Remember that we expect this GraphIndex to occur in the ClusterSet::m_removed.
2983
0
                expected_removed[level].insert(idx);
2984
0
                was_removed = true;
2985
0
            }
2986
21.8M
        }
2987
10.9M
    }
2988
2989
    // For all levels (0 = main, 1 = staged)...
2990
257k
    for (int level = 0; level <= GetTopLevel(); ++level) {
2991
128k
        assert(level < MAX_LEVELS);
2992
128k
        auto& clusterset = GetClusterSet(level);
2993
128k
        std::set<const Cluster*> actual_clusters;
2994
128k
        size_t recomputed_cluster_usage{0};
2995
2996
        // For all quality levels...
2997
1.02M
        for (int qual = 0; qual < int(QualityLevel::NONE); ++qual) {
2998
900k
            QualityLevel quality{qual};
2999
900k
            const auto& quality_clusters = clusterset.m_clusters[qual];
3000
            // ... for all clusters in them ...
3001
11.6M
            for (ClusterSetIndex setindex = 0; setindex < quality_clusters.size(); ++setindex) {
3002
10.7M
                const auto& cluster = *quality_clusters[setindex];
3003
                // The number of transactions in a Cluster cannot exceed m_max_cluster_count.
3004
10.7M
                assert(cluster.GetTxCount() <= m_max_cluster_count);
3005
                // The level must match the Cluster's own idea of what level it is in (but GetLevel
3006
                // can only be called for non-empty Clusters).
3007
10.7M
                assert(cluster.GetTxCount() == 0 || level == cluster.GetLevel(*this));
3008
                // The sum of their sizes cannot exceed m_max_cluster_size, unless it is an
3009
                // individually oversized transaction singleton. Note that groups of to-be-merged
3010
                // clusters which would exceed this limit are marked oversized, which means they
3011
                // are never applied.
3012
10.7M
                assert(cluster.IsOversized() || cluster.GetTotalTxSize() <= m_max_cluster_size);
3013
                // OVERSIZED clusters are singletons.
3014
10.7M
                assert(!cluster.IsOversized() || cluster.GetTxCount() == 1);
3015
                // Transaction counts cannot exceed the Cluster implementation's maximum
3016
                // supported transactions count.
3017
10.7M
                assert(cluster.GetTxCount() <= cluster.GetMaxTxCount());
3018
                // Unless a Split is yet to be applied, the number of transactions must not be
3019
                // below the Cluster implementation's intended transaction count.
3020
10.7M
                if (!cluster.NeedsSplitting()) {
3021
10.7M
                    assert(cluster.GetTxCount() >= cluster.GetMinIntendedTxCount());
3022
10.7M
                }
3023
3024
                // Check the sequence number.
3025
10.7M
                assert(cluster.m_sequence < m_next_sequence_counter);
3026
10.7M
                assert(!sequences.contains(cluster.m_sequence));
3027
10.7M
                sequences.insert(cluster.m_sequence);
3028
                // Remember we saw this Cluster (only if it is non-empty; empty Clusters aren't
3029
                // expected to be referenced by the Entry vector).
3030
10.7M
                if (cluster.GetTxCount() != 0) {
3031
10.7M
                    actual_clusters.insert(&cluster);
3032
10.7M
                }
3033
                // Sanity check the cluster, according to the Cluster's internal rules.
3034
10.7M
                cluster.SanityCheck(*this, level);
3035
                // Check that the cluster's quality and setindex matches its position in the quality list.
3036
10.7M
                assert(cluster.m_quality == quality);
3037
10.7M
                assert(cluster.m_setindex == setindex);
3038
                // Count memory usage.
3039
10.7M
                recomputed_cluster_usage += cluster.TotalMemoryUsage();
3040
10.7M
            }
3041
900k
        }
3042
3043
        // Verify memory usage.
3044
128k
        assert(clusterset.m_cluster_usage == recomputed_cluster_usage);
3045
3046
        // Verify that all to-be-removed transactions have valid identifiers.
3047
128k
        for (GraphIndex idx : clusterset.m_to_remove) {
3048
8
            assert(idx < m_entries.size());
3049
            // We cannot assert that all m_to_remove transactions are still present: ~Ref on a
3050
            // (P,M) transaction (present in main, inherited in staging) will cause an m_to_remove
3051
            // addition in both main and staging, but a subsequence ApplyRemovals in main will
3052
            // cause it to disappear from staging too, leaving the m_to_remove in place.
3053
8
        }
3054
3055
        // Verify that all to-be-added dependencies have valid identifiers.
3056
191k
        for (auto [par_idx, chl_idx] : clusterset.m_deps_to_add) {
3057
191k
            assert(par_idx != chl_idx);
3058
191k
            assert(par_idx < m_entries.size());
3059
191k
            assert(chl_idx < m_entries.size());
3060
191k
        }
3061
3062
        // Verify that the actually encountered clusters match the ones occurring in Entry vector.
3063
128k
        assert(actual_clusters == expected_clusters[level]);
3064
3065
        // Verify that the contents of m_removed matches what was expected based on the Entry vector.
3066
128k
        std::set<GraphIndex> actual_removed(clusterset.m_removed.begin(), clusterset.m_removed.end());
3067
128k
        for (auto i : expected_unlinked) {
3068
            // If a transaction exists in both main and staging, and is removed from staging (adding
3069
            // it to m_removed there), and consequently destroyed (wiping the locator completely),
3070
            // it can remain in m_removed despite not having an IsRemoved() locator. Exclude those
3071
            // transactions from the comparison here.
3072
0
            actual_removed.erase(i);
3073
0
            expected_removed[level].erase(i);
3074
0
        }
3075
3076
128k
        assert(actual_removed == expected_removed[level]);
3077
3078
        // If any GraphIndex entries remain in this ClusterSet, compact is not possible.
3079
128k
        if (!clusterset.m_deps_to_add.empty()) compact_possible = false;
3080
128k
        if (!clusterset.m_to_remove.empty()) compact_possible = false;
3081
128k
        if (!clusterset.m_removed.empty()) compact_possible = false;
3082
3083
        // For non-top levels, m_oversized must be known (as it cannot change until the level
3084
        // on top is gone).
3085
128k
        if (level < GetTopLevel()) assert(clusterset.m_oversized.has_value());
3086
128k
    }
3087
3088
    // Verify that the contents of m_unlinked matches what was expected based on the Entry vector.
3089
128k
    std::set<GraphIndex> actual_unlinked(m_unlinked.begin(), m_unlinked.end());
3090
128k
    assert(actual_unlinked == expected_unlinked);
3091
3092
    // If compaction was possible, it should have been performed already, and m_unlinked must be
3093
    // empty (to prevent memory leaks due to an ever-growing m_entries vector).
3094
128k
    if (compact_possible) {
3095
128k
        assert(actual_unlinked.empty());
3096
128k
    }
3097
3098
    // Finally, check the chunk index.
3099
128k
    std::set<GraphIndex> actual_chunkindex;
3100
128k
    FeeFrac last_chunk_feerate;
3101
10.8M
    for (const auto& chunk : m_main_chunkindex) {
3102
10.8M
        GraphIndex idx = chunk.m_graph_index;
3103
10.8M
        actual_chunkindex.insert(idx);
3104
10.8M
        auto chunk_feerate = m_entries[idx].m_main_chunk_feerate;
3105
10.8M
        if (!last_chunk_feerate.IsEmpty()) {
3106
10.8M
            assert(FeeRateCompare(last_chunk_feerate, chunk_feerate) >= 0);
3107
10.8M
        }
3108
10.8M
        last_chunk_feerate = chunk_feerate;
3109
10.8M
    }
3110
128k
    assert(actual_chunkindex == expected_chunkindex);
3111
128k
}
3112
3113
bool TxGraphImpl::DoWork(uint64_t max_cost) noexcept
3114
161k
{
3115
161k
    uint64_t cost_done{0};
3116
    // First linearize everything in NEEDS_RELINEARIZE to an acceptable level. If more budget
3117
    // remains after that, try to make everything optimal.
3118
485k
    for (QualityLevel quality : {QualityLevel::NEEDS_FIX, QualityLevel::NEEDS_RELINEARIZE, QualityLevel::ACCEPTABLE}) {
3119
        // First linearize staging, if it exists, then main.
3120
970k
        for (int level = GetTopLevel(); level >= 0; --level) {
3121
            // Do not modify main if it has any observers.
3122
485k
            if (level == 0 && m_main_chunkindex_observers != 0) continue;
3123
485k
            ApplyDependencies(level);
3124
485k
            auto& clusterset = GetClusterSet(level);
3125
            // Do not modify oversized levels.
3126
485k
            if (clusterset.m_oversized == true) continue;
3127
485k
            auto& queue = clusterset.m_clusters[int(quality)];
3128
490k
            while (!queue.empty()) {
3129
5.10k
                if (cost_done >= max_cost) return false;
3130
                // Randomize the order in which we process, so that if the first cluster somehow
3131
                // needs more work than what max_cost allows, we don't keep spending it on the same
3132
                // one.
3133
5.10k
                auto pos = m_rng.randrange<size_t>(queue.size());
3134
5.10k
                auto cost_now = max_cost - cost_done;
3135
5.10k
                if (quality == QualityLevel::NEEDS_FIX || quality == QualityLevel::NEEDS_RELINEARIZE) {
3136
                    // If we're working with clusters that need relinearization still, only perform
3137
                    // up to m_acceptable_cost work. If they become ACCEPTABLE, and we still
3138
                    // have budget after all other clusters are ACCEPTABLE too, we'll spend the
3139
                    // remaining budget on trying to make them OPTIMAL.
3140
5.10k
                    cost_now = std::min(cost_now, m_acceptable_cost);
3141
5.10k
                }
3142
5.10k
                auto [cost, improved] = queue[pos].get()->Relinearize(*this, level, cost_now);
3143
5.10k
                cost_done += cost;
3144
                // If no improvement was made to the Cluster, it means we've essentially run out of
3145
                // budget. Even though it may be the case that cost_done < max_cost still, the
3146
                // linearizer decided there wasn't enough budget left to attempt anything with.
3147
                // To avoid an infinite loop that keeps trying clusters with minuscule budgets,
3148
                // stop here too.
3149
5.10k
                if (!improved) return false;
3150
5.10k
            }
3151
485k
        }
3152
485k
    }
3153
    // All possible work has been performed, so we can return true. Note that this does *not* mean
3154
    // that all clusters are optimally linearized now. It may be that there is nothing to do left
3155
    // because all non-optimal clusters are in oversized and/or observer-bearing levels.
3156
161k
    return true;
3157
161k
}
3158
3159
void BlockBuilderImpl::Next() noexcept
3160
10.7M
{
3161
    // Don't do anything if we're already done.
3162
10.7M
    if (m_cur_iter == m_graph->m_main_chunkindex.end()) return;
3163
10.7M
    while (true) {
3164
        // Advance the pointer, and stop if we reach the end.
3165
10.7M
        ++m_cur_iter;
3166
10.7M
        m_cur_cluster = nullptr;
3167
10.7M
        if (m_cur_iter == m_graph->m_main_chunkindex.end()) break;
3168
        // Find the cluster pointed to by m_cur_iter.
3169
10.7M
        const auto& chunk_data = *m_cur_iter;
3170
10.7M
        const auto& chunk_end_entry = m_graph->m_entries[chunk_data.m_graph_index];
3171
10.7M
        m_cur_cluster = chunk_end_entry.m_locator[0].cluster;
3172
10.7M
        m_known_end_of_cluster = false;
3173
        // If we previously skipped a chunk from this cluster we cannot include more from it.
3174
10.7M
        if (!m_excluded_clusters.contains(m_cur_cluster->m_sequence)) break;
3175
10.7M
    }
3176
10.7M
}
3177
3178
std::optional<std::pair<std::vector<TxGraph::Ref*>, FeePerWeight>> BlockBuilderImpl::GetCurrentChunk() noexcept
3179
10.9M
{
3180
10.9M
    std::optional<std::pair<std::vector<TxGraph::Ref*>, FeePerWeight>> ret;
3181
    // Populate the return value if we are not done.
3182
10.9M
    if (m_cur_iter != m_graph->m_main_chunkindex.end()) {
3183
10.7M
        ret.emplace();
3184
10.7M
        const auto& chunk_data = *m_cur_iter;
3185
10.7M
        const auto& chunk_end_entry = m_graph->m_entries[chunk_data.m_graph_index];
3186
10.7M
        if (chunk_data.m_chunk_count == LinearizationIndex(-1)) {
3187
            // Special case in case just a single transaction remains, avoiding the need to
3188
            // dispatch to and dereference Cluster.
3189
10.5M
            ret->first.resize(1);
3190
10.5M
            Assume(chunk_end_entry.m_ref != nullptr);
3191
10.5M
            ret->first[0] = chunk_end_entry.m_ref;
3192
10.5M
            m_known_end_of_cluster = true;
3193
10.5M
        } else {
3194
170k
            Assume(m_cur_cluster);
3195
170k
            ret->first.resize(chunk_data.m_chunk_count);
3196
170k
            auto start_pos = chunk_end_entry.m_main_lin_index + 1 - chunk_data.m_chunk_count;
3197
170k
            m_known_end_of_cluster = m_cur_cluster->GetClusterRefs(*m_graph, ret->first, start_pos);
3198
            // If the chunk size was 1 and at end of cluster, then the special case above should
3199
            // have been used.
3200
170k
            Assume(!m_known_end_of_cluster || chunk_data.m_chunk_count > 1);
3201
170k
        }
3202
10.7M
        ret->second = chunk_end_entry.m_main_chunk_feerate;
3203
10.7M
    }
3204
10.9M
    return ret;
3205
10.9M
}
3206
3207
165k
BlockBuilderImpl::BlockBuilderImpl(TxGraphImpl& graph) noexcept : m_graph(&graph)
3208
165k
{
3209
    // Make sure all clusters in main are up to date, and acceptable.
3210
165k
    m_graph->MakeAllAcceptable(0);
3211
    // There cannot remain any inapplicable dependencies (only possible if main is oversized).
3212
165k
    Assume(m_graph->m_main_clusterset.m_deps_to_add.empty());
3213
    // Remember that this object is observing the graph's index, so that we can detect concurrent
3214
    // modifications.
3215
165k
    ++m_graph->m_main_chunkindex_observers;
3216
    // Find the first chunk.
3217
165k
    m_cur_iter = m_graph->m_main_chunkindex.begin();
3218
165k
    m_cur_cluster = nullptr;
3219
165k
    if (m_cur_iter != m_graph->m_main_chunkindex.end()) {
3220
        // Find the cluster pointed to by m_cur_iter.
3221
32.9k
        const auto& chunk_data = *m_cur_iter;
3222
32.9k
        const auto& chunk_end_entry = m_graph->m_entries[chunk_data.m_graph_index];
3223
32.9k
        m_cur_cluster = chunk_end_entry.m_locator[0].cluster;
3224
32.9k
    }
3225
165k
}
3226
3227
BlockBuilderImpl::~BlockBuilderImpl()
3228
165k
{
3229
165k
    Assume(m_graph->m_main_chunkindex_observers > 0);
3230
    // Permit modifications to the main graph again after destroying the BlockBuilderImpl.
3231
165k
    --m_graph->m_main_chunkindex_observers;
3232
165k
}
3233
3234
void BlockBuilderImpl::Include() noexcept
3235
10.6M
{
3236
    // The actual inclusion of the chunk is done by the calling code. All we have to do is switch
3237
    // to the next chunk.
3238
10.6M
    Next();
3239
10.6M
}
3240
3241
void BlockBuilderImpl::Skip() noexcept
3242
41.1k
{
3243
    // When skipping a chunk we need to not include anything more of the cluster, as that could make
3244
    // the result topologically invalid. However, don't do this if the chunk is known to be the last
3245
    // chunk of the cluster. This may significantly reduce the size of m_excluded_clusters,
3246
    // especially when many singleton clusters are ignored.
3247
41.1k
    if (m_cur_cluster != nullptr && !m_known_end_of_cluster) {
3248
1
        m_excluded_clusters.insert(m_cur_cluster->m_sequence);
3249
1
    }
3250
41.1k
    Next();
3251
41.1k
}
3252
3253
std::unique_ptr<TxGraph::BlockBuilder> TxGraphImpl::GetBlockBuilder() noexcept
3254
165k
{
3255
165k
    return std::make_unique<BlockBuilderImpl>(*this);
3256
165k
}
3257
3258
std::pair<std::vector<TxGraph::Ref*>, FeePerWeight> TxGraphImpl::GetWorstMainChunk() noexcept
3259
48
{
3260
48
    std::pair<std::vector<Ref*>, FeePerWeight> ret;
3261
    // Make sure all clusters in main are up to date, and acceptable.
3262
48
    MakeAllAcceptable(0);
3263
48
    Assume(m_main_clusterset.m_deps_to_add.empty());
3264
    // If the graph is not empty, populate ret.
3265
48
    if (!m_main_chunkindex.empty()) {
3266
48
        const auto& chunk_data = *m_main_chunkindex.rbegin();
3267
48
        const auto& chunk_end_entry = m_entries[chunk_data.m_graph_index];
3268
48
        Cluster* cluster = chunk_end_entry.m_locator[0].cluster;
3269
48
        if (chunk_data.m_chunk_count == LinearizationIndex(-1) || chunk_data.m_chunk_count == 1)  {
3270
            // Special case for singletons.
3271
42
            ret.first.resize(1);
3272
42
            Assume(chunk_end_entry.m_ref != nullptr);
3273
42
            ret.first[0] = chunk_end_entry.m_ref;
3274
42
        } else {
3275
6
            ret.first.resize(chunk_data.m_chunk_count);
3276
6
            auto start_pos = chunk_end_entry.m_main_lin_index + 1 - chunk_data.m_chunk_count;
3277
6
            cluster->GetClusterRefs(*this, ret.first, start_pos);
3278
6
            std::reverse(ret.first.begin(), ret.first.end());
3279
6
        }
3280
48
        ret.second = chunk_end_entry.m_main_chunk_feerate;
3281
48
    }
3282
48
    return ret;
3283
48
}
3284
3285
std::vector<TxGraph::Ref*> TxGraphImpl::Trim() noexcept
3286
2.13k
{
3287
2.13k
    int level = GetTopLevel();
3288
2.13k
    Assume(m_main_chunkindex_observers == 0 || level != 0);
3289
2.13k
    std::vector<TxGraph::Ref*> ret;
3290
3291
    // Compute the groups of to-be-merged Clusters (which also applies all removals, and splits).
3292
2.13k
    auto& clusterset = GetClusterSet(level);
3293
2.13k
    if (clusterset.m_oversized == false) return ret;
3294
33
    GroupClusters(level);
3295
33
    Assume(clusterset.m_group_data.has_value());
3296
    // Nothing to do if not oversized.
3297
33
    Assume(clusterset.m_oversized.has_value());
3298
33
    if (clusterset.m_oversized == false) return ret;
3299
3300
    // In this function, would-be clusters (as precomputed in m_group_data by GroupClusters) are
3301
    // trimmed by removing transactions in them such that the resulting clusters satisfy the size
3302
    // and count limits.
3303
    //
3304
    // It works by defining for each would-be cluster a rudimentary linearization: at every point
3305
    // the highest-chunk-feerate remaining transaction is picked among those with no unmet
3306
    // dependencies. "Dependency" here means either a to-be-added dependency (m_deps_to_add), or
3307
    // an implicit dependency added between any two consecutive transaction in their current
3308
    // cluster linearization. So it can be seen as a "merge sort" of the chunks of the clusters,
3309
    // but respecting the dependencies being added.
3310
    //
3311
    // This rudimentary linearization is computed lazily, by putting all eligible (no unmet
3312
    // dependencies) transactions in a heap, and popping the highest-feerate one from it. Along the
3313
    // way, the counts and sizes of the would-be clusters up to that point are tracked (by
3314
    // partitioning the involved transactions using a union-find structure). Any transaction whose
3315
    // addition would cause a violation is removed, along with all their descendants.
3316
    //
3317
    // A next invocation of GroupClusters (after applying the removals) will compute the new
3318
    // resulting clusters, and none of them will violate the limits.
3319
3320
    /** All dependencies (both to be added ones, and implicit ones between consecutive transactions
3321
     *  in existing cluster linearizations), sorted by parent. */
3322
4
    std::vector<std::pair<GraphIndex, GraphIndex>> deps_by_parent;
3323
    /** Same, but sorted by child. */
3324
4
    std::vector<std::pair<GraphIndex, GraphIndex>> deps_by_child;
3325
    /** Information about all transactions involved in a Cluster group to be trimmed, sorted by
3326
     *  GraphIndex. It contains entries both for transactions that have already been included,
3327
     *  and ones that have not yet been. */
3328
4
    std::vector<TrimTxData> trim_data;
3329
    /** Iterators into trim_data, treated as a max heap according to cmp_fn below. Each entry is
3330
     *  a transaction that has not yet been included yet, but all its ancestors have. */
3331
4
    std::vector<std::vector<TrimTxData>::iterator> trim_heap;
3332
    /** The list of representatives of the partitions a given transaction depends on. */
3333
4
    std::vector<TrimTxData*> current_deps;
3334
3335
    /** Function to define the ordering of trim_heap. */
3336
1.08M
    static constexpr auto cmp_fn = [](auto a, auto b) noexcept {
3337
        // Sort by increasing chunk feerate, and then by decreasing size.
3338
        // We do not need to sort by cluster or within clusters, because due to the implicit
3339
        // dependency between consecutive linearization elements, no two transactions from the
3340
        // same Cluster will ever simultaneously be in the heap.
3341
1.08M
        return a->m_chunk_feerate < b->m_chunk_feerate;
3342
1.08M
    };
3343
3344
    /** Given a TrimTxData entry, find the representative of the partition it is in. */
3345
190k
    static constexpr auto find_fn = [](TrimTxData* arg) noexcept {
3346
252k
        while (arg != arg->m_uf_parent) {
3347
            // Replace pointer to parent with pointer to grandparent (path splitting).
3348
            // See https://en.wikipedia.org/wiki/Disjoint-set_data_structure#Finding_set_representatives.
3349
62.0k
            auto par = arg->m_uf_parent;
3350
62.0k
            arg->m_uf_parent = par->m_uf_parent;
3351
62.0k
            arg = par;
3352
62.0k
        }
3353
190k
        return arg;
3354
190k
    };
3355
3356
    /** Given two TrimTxData entries, union the partitions they are in, and return the
3357
     *  representative. */
3358
63.0k
    static constexpr auto union_fn = [](TrimTxData* arg1, TrimTxData* arg2) noexcept {
3359
        // Replace arg1 and arg2 by their representatives.
3360
63.0k
        auto rep1 = find_fn(arg1);
3361
63.0k
        auto rep2 = find_fn(arg2);
3362
        // Bail out if both representatives are the same, because that means arg1 and arg2 are in
3363
        // the same partition already.
3364
63.0k
        if (rep1 == rep2) return rep1;
3365
        // Pick the lower-count root to become a child of the higher-count one.
3366
        // See https://en.wikipedia.org/wiki/Disjoint-set_data_structure#Union_by_size.
3367
63.0k
        if (rep1->m_uf_count < rep2->m_uf_count) std::swap(rep1, rep2);
3368
63.0k
        rep2->m_uf_parent = rep1;
3369
        // Add the statistics of arg2 (which is no longer a representative) to those of arg1 (which
3370
        // is now the representative for both).
3371
63.0k
        rep1->m_uf_size += rep2->m_uf_size;
3372
63.0k
        rep1->m_uf_count += rep2->m_uf_count;
3373
63.0k
        return rep1;
3374
63.0k
    };
3375
3376
    /** Get iterator to TrimTxData entry for a given index. */
3377
128k
    auto locate_fn = [&](GraphIndex index) noexcept {
3378
2.04M
        auto it = std::lower_bound(trim_data.begin(), trim_data.end(), index, [](TrimTxData& elem, GraphIndex idx) noexcept {
3379
2.04M
            return elem.m_index < idx;
3380
2.04M
        });
3381
128k
        Assume(it != trim_data.end() && it->m_index == index);
3382
128k
        return it;
3383
128k
    };
3384
3385
    // For each group of to-be-merged Clusters.
3386
9
    for (const auto& group_data : clusterset.m_group_data->m_groups) {
3387
9
        trim_data.clear();
3388
9
        trim_heap.clear();
3389
9
        deps_by_child.clear();
3390
9
        deps_by_parent.clear();
3391
3392
        // Gather trim data and implicit dependency data from all involved Clusters.
3393
9
        auto cluster_span = std::span{clusterset.m_group_data->m_group_clusters}
3394
9
                                .subspan(group_data.m_cluster_offset, group_data.m_cluster_count);
3395
9
        uint64_t size{0};
3396
64.2k
        for (Cluster* cluster : cluster_span) {
3397
64.2k
            MakeAcceptable(*cluster, cluster->GetLevel(*this));
3398
64.2k
            size += cluster->AppendTrimData(trim_data, deps_by_child);
3399
64.2k
        }
3400
        // If this group of Clusters does not violate any limits, continue to the next group.
3401
9
        if (trim_data.size() <= m_max_cluster_count && size <= m_max_cluster_size) continue;
3402
        // Sort the trim data by GraphIndex. In what follows, we will treat this sorted vector as
3403
        // a map from GraphIndex to TrimTxData via locate_fn, and its ordering will not change
3404
        // anymore.
3405
1.32M
        std::sort(trim_data.begin(), trim_data.end(), [](auto& a, auto& b) noexcept { return a.m_index < b.m_index; });
3406
3407
        // Add the explicitly added dependencies to deps_by_child.
3408
9
        deps_by_child.insert(deps_by_child.end(),
3409
9
                             clusterset.m_deps_to_add.begin() + group_data.m_deps_offset,
3410
9
                             clusterset.m_deps_to_add.begin() + group_data.m_deps_offset + group_data.m_deps_count);
3411
3412
        // Sort deps_by_child by child transaction GraphIndex. The order will not be changed
3413
        // anymore after this.
3414
1.46M
        std::sort(deps_by_child.begin(), deps_by_child.end(), [](auto& a, auto& b) noexcept { return a.second < b.second; });
3415
        // Fill m_parents_count and m_parents_offset in trim_data, as well as m_deps_left, and
3416
        // initially populate trim_heap. Because of the sort above, all dependencies involving the
3417
        // same child are grouped together, so a single linear scan suffices.
3418
9
        auto deps_it = deps_by_child.begin();
3419
64.2k
        for (auto trim_it = trim_data.begin(); trim_it != trim_data.end(); ++trim_it) {
3420
64.2k
            trim_it->m_parent_offset = deps_it - deps_by_child.begin();
3421
64.2k
            trim_it->m_deps_left = 0;
3422
128k
            while (deps_it != deps_by_child.end() && deps_it->second == trim_it->m_index) {
3423
64.2k
                ++trim_it->m_deps_left;
3424
64.2k
                ++deps_it;
3425
64.2k
            }
3426
64.2k
            trim_it->m_parent_count = trim_it->m_deps_left;
3427
            // If this transaction has no unmet dependencies, and is not oversized, add it to the
3428
            // heap (just append for now, the heapification happens below).
3429
64.2k
            if (trim_it->m_deps_left == 0 && trim_it->m_tx_size <= m_max_cluster_size) {
3430
1.15k
                trim_heap.push_back(trim_it);
3431
1.15k
            }
3432
64.2k
        }
3433
9
        Assume(deps_it == deps_by_child.end());
3434
3435
        // Construct deps_by_parent, sorted by parent transaction GraphIndex. The order will not be
3436
        // changed anymore after this.
3437
9
        deps_by_parent = deps_by_child;
3438
2.33M
        std::sort(deps_by_parent.begin(), deps_by_parent.end(), [](auto& a, auto& b) noexcept { return a.first < b.first; });
3439
        // Fill m_children_offset and m_children_count in trim_data. Because of the sort above, all
3440
        // dependencies involving the same parent are grouped together, so a single linear scan
3441
        // suffices.
3442
9
        deps_it = deps_by_parent.begin();
3443
64.2k
        for (auto& trim_entry : trim_data) {
3444
64.2k
            trim_entry.m_children_count = 0;
3445
64.2k
            trim_entry.m_children_offset = deps_it - deps_by_parent.begin();
3446
128k
            while (deps_it != deps_by_parent.end() && deps_it->first == trim_entry.m_index) {
3447
64.2k
                ++trim_entry.m_children_count;
3448
64.2k
                ++deps_it;
3449
64.2k
            }
3450
64.2k
        }
3451
9
        Assume(deps_it == deps_by_parent.end());
3452
3453
        // Build a heap of all transactions with 0 unmet dependencies.
3454
9
        std::make_heap(trim_heap.begin(), trim_heap.end(), cmp_fn);
3455
3456
        // Iterate over to-be-included transactions, and convert them to included transactions, or
3457
        // skip them if adding them would violate resource limits of the would-be cluster.
3458
        //
3459
        // It is possible that the heap empties without ever hitting either cluster limit, in case
3460
        // the implied graph (to be added dependencies plus implicit dependency between each
3461
        // original transaction and its predecessor in the linearization it came from) contains
3462
        // cycles. Such cycles will be removed entirely, because each of the transactions in the
3463
        // cycle permanently have unmet dependencies. However, this cannot occur in real scenarios
3464
        // where Trim() is called to deal with reorganizations that would violate cluster limits,
3465
        // as all added dependencies are in the same direction (from old mempool transactions to
3466
        // new from-block transactions); cycles require dependencies in both directions to be
3467
        // added.
3468
64.2k
        while (!trim_heap.empty()) {
3469
            // Move the best remaining transaction to the end of trim_heap.
3470
64.2k
            std::pop_heap(trim_heap.begin(), trim_heap.end(), cmp_fn);
3471
            // Pop it, and find its TrimTxData.
3472
64.2k
            auto& entry = *trim_heap.back();
3473
64.2k
            trim_heap.pop_back();
3474
3475
            // Initialize it as a singleton partition.
3476
64.2k
            entry.m_uf_parent = &entry;
3477
64.2k
            entry.m_uf_count = 1;
3478
64.2k
            entry.m_uf_size = entry.m_tx_size;
3479
3480
            // Find the distinct transaction partitions this entry depends on.
3481
64.2k
            current_deps.clear();
3482
64.2k
            for (auto& [par, chl] : std::span{deps_by_child}.subspan(entry.m_parent_offset, entry.m_parent_count)) {
3483
64.2k
                Assume(chl == entry.m_index);
3484
64.2k
                current_deps.push_back(find_fn(&*locate_fn(par)));
3485
64.2k
            }
3486
64.2k
            std::sort(current_deps.begin(), current_deps.end());
3487
64.2k
            current_deps.erase(std::unique(current_deps.begin(), current_deps.end()), current_deps.end());
3488
3489
            // Compute resource counts.
3490
64.2k
            uint32_t new_count = 1;
3491
64.2k
            uint64_t new_size = entry.m_tx_size;
3492
64.2k
            for (TrimTxData* ptr : current_deps) {
3493
64.2k
                new_count += ptr->m_uf_count;
3494
64.2k
                new_size += ptr->m_uf_size;
3495
64.2k
            }
3496
            // Skip the entry if this would violate any limit.
3497
64.2k
            if (new_count > m_max_cluster_count || new_size > m_max_cluster_size) continue;
3498
3499
            // Union the partitions this transaction and all its dependencies are in together.
3500
64.1k
            auto rep = &entry;
3501
64.1k
            for (TrimTxData* ptr : current_deps) rep = union_fn(ptr, rep);
3502
            // Mark the entry as included (so the loop below will not remove the transaction).
3503
64.1k
            entry.m_deps_left = uint32_t(-1);
3504
            // Mark each to-be-added dependency involving this transaction as parent satisfied.
3505
64.2k
            for (auto& [par, chl] : std::span{deps_by_parent}.subspan(entry.m_children_offset, entry.m_children_count)) {
3506
64.2k
                Assume(par == entry.m_index);
3507
64.2k
                auto chl_it = locate_fn(chl);
3508
                // Reduce the number of unmet dependencies of chl_it, and if that brings the number
3509
                // to zero, add it to the heap of includable transactions.
3510
64.2k
                Assume(chl_it->m_deps_left > 0);
3511
64.2k
                if (--chl_it->m_deps_left == 0) {
3512
63.0k
                    trim_heap.push_back(chl_it);
3513
63.0k
                    std::push_heap(trim_heap.begin(), trim_heap.end(), cmp_fn);
3514
63.0k
                }
3515
64.2k
            }
3516
64.1k
        }
3517
3518
        // Remove all the transactions that were not processed above. Because nothing gets
3519
        // processed until/unless all its dependencies are met, this automatically guarantees
3520
        // that if a transaction is removed, all its descendants, or would-be descendants, are
3521
        // removed as well.
3522
64.2k
        for (const auto& trim_entry : trim_data) {
3523
64.2k
            if (trim_entry.m_deps_left != uint32_t(-1)) {
3524
19
                ret.push_back(m_entries[trim_entry.m_index].m_ref);
3525
19
                clusterset.m_to_remove.push_back(trim_entry.m_index);
3526
19
            }
3527
64.2k
        }
3528
9
    }
3529
4
    clusterset.m_group_data.reset();
3530
4
    clusterset.m_oversized = false;
3531
4
    Assume(!ret.empty());
3532
4
    return ret;
3533
33
}
3534
3535
size_t TxGraphImpl::GetMainMemoryUsage() noexcept
3536
461k
{
3537
    // Make sure splits/merges are applied, as memory usage may not be representative otherwise.
3538
461k
    SplitAll(/*up_to_level=*/0);
3539
461k
    ApplyDependencies(/*level=*/0);
3540
    // Compute memory usage
3541
461k
    size_t usage = /* From clusters */
3542
461k
                   m_main_clusterset.m_cluster_usage +
3543
                   /* From Entry objects. */
3544
461k
                   sizeof(Entry) * m_main_clusterset.m_txcount +
3545
                   /* From the chunk index. */
3546
461k
                   memusage::DynamicUsage(m_main_chunkindex);
3547
461k
    return usage;
3548
461k
}
3549
3550
} // namespace
3551
3552
TxGraph::Ref::~Ref()
3553
219k
{
3554
219k
    if (m_graph) {
3555
        // Inform the TxGraph about the Ref being destroyed.
3556
63.5k
        m_graph->UnlinkRef(m_index);
3557
63.5k
        m_graph = nullptr;
3558
63.5k
    }
3559
219k
}
3560
3561
TxGraph::Ref::Ref(Ref&& other) noexcept
3562
65.5k
{
3563
    // Inform the TxGraph of other that its Ref is being moved.
3564
65.5k
    if (other.m_graph) other.m_graph->UpdateRef(other.m_index, *this);
3565
    // Actually move the contents.
3566
65.5k
    std::swap(m_graph, other.m_graph);
3567
65.5k
    std::swap(m_index, other.m_index);
3568
65.5k
}
3569
3570
std::unique_ptr<TxGraph> MakeTxGraph(
3571
    unsigned max_cluster_count,
3572
    uint64_t max_cluster_size,
3573
    uint64_t acceptable_cost,
3574
    const std::function<std::strong_ordering(const TxGraph::Ref&, const TxGraph::Ref&)>& fallback_order) noexcept
3575
1.20k
{
3576
1.20k
    return std::make_unique<TxGraphImpl>(
3577
1.20k
        /*max_cluster_count=*/max_cluster_count,
3578
1.20k
        /*max_cluster_size=*/max_cluster_size,
3579
1.20k
        /*acceptable_cost=*/acceptable_cost,
3580
1.20k
        /*fallback_order=*/fallback_order);
3581
1.20k
}