initia custom

core/bloombits/docs.go

@@ -0,0 +1,18 @@
+// Copyright 2017 The go-ethereum Authors
+// This file is part of the go-ethereum library.
+//
+// The go-ethereum library is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Lesser General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// The go-ethereum library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU Lesser General Public License for more details.
+//
+// You should have received a copy of the GNU Lesser General Public License
+// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
+
+// Package bloombits implements bloom filtering on batches of data.
+package bloombits

core/bloombits/generator.go

@@ -0,0 +1,98 @@
+// Copyright 2017 The go-ethereum Authors
+// This file is part of the go-ethereum library.
+//
+// The go-ethereum library is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Lesser General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// The go-ethereum library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU Lesser General Public License for more details.
+//
+// You should have received a copy of the GNU Lesser General Public License
+// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
+
+package bloombits
+
+import (
+	"errors"
+
+	"github.com/ethereum/go-ethereum/core/types"
+)
+
+var (
+	// errSectionOutOfBounds is returned if the user tried to add more bloom filters
+	// to the batch than available space, or if tries to retrieve above the capacity.
+	errSectionOutOfBounds = errors.New("section out of bounds")
+
+	// errBloomBitOutOfBounds is returned if the user tried to retrieve specified
+	// bit bloom above the capacity.
+	errBloomBitOutOfBounds = errors.New("bloom bit out of bounds")
+)
+
+// Generator takes a number of bloom filters and generates the rotated bloom bits
+// to be used for batched filtering.
+type Generator struct {
+	blooms   [types.BloomBitLength][]byte // Rotated blooms for per-bit matching
+	sections uint                         // Number of sections to batch together
+	nextSec  uint                         // Next section to set when adding a bloom
+}
+
+// NewGenerator creates a rotated bloom generator that can iteratively fill a
+// batched bloom filter's bits.
+func NewGenerator(sections uint) (*Generator, error) {
+	if sections%8 != 0 {
+		return nil, errors.New("section count not multiple of 8")
+	}
+	b := &Generator{sections: sections}
+	for i := 0; i < types.BloomBitLength; i++ {
+		b.blooms[i] = make([]byte, sections/8)
+	}
+	return b, nil
+}
+
+// AddBloom takes a single bloom filter and sets the corresponding bit column
+// in memory accordingly.
+func (b *Generator) AddBloom(index uint, bloom types.Bloom) error {
+	// Make sure we're not adding more bloom filters than our capacity
+	if b.nextSec >= b.sections {
+		return errSectionOutOfBounds
+	}
+	if b.nextSec != index {
+		return errors.New("bloom filter with unexpected index")
+	}
+	// Rotate the bloom and insert into our collection
+	byteIndex := b.nextSec / 8
+	bitIndex := byte(7 - b.nextSec%8)
+	for byt := 0; byt < types.BloomByteLength; byt++ {
+		bloomByte := bloom[types.BloomByteLength-1-byt]
+		if bloomByte == 0 {
+			continue
+		}
+		base := 8 * byt
+		b.blooms[base+7][byteIndex] |= ((bloomByte >> 7) & 1) << bitIndex
+		b.blooms[base+6][byteIndex] |= ((bloomByte >> 6) & 1) << bitIndex
+		b.blooms[base+5][byteIndex] |= ((bloomByte >> 5) & 1) << bitIndex
+		b.blooms[base+4][byteIndex] |= ((bloomByte >> 4) & 1) << bitIndex
+		b.blooms[base+3][byteIndex] |= ((bloomByte >> 3) & 1) << bitIndex
+		b.blooms[base+2][byteIndex] |= ((bloomByte >> 2) & 1) << bitIndex
+		b.blooms[base+1][byteIndex] |= ((bloomByte >> 1) & 1) << bitIndex
+		b.blooms[base][byteIndex] |= (bloomByte & 1) << bitIndex
+	}
+	b.nextSec++
+	return nil
+}
+
+// Bitset returns the bit vector belonging to the given bit index after all
+// blooms have been added.
+func (b *Generator) Bitset(idx uint) ([]byte, error) {
+	if b.nextSec != b.sections {
+		return nil, errors.New("bloom not fully generated yet")
+	}
+	if idx >= types.BloomBitLength {
+		return nil, errBloomBitOutOfBounds
+	}
+	return b.blooms[idx], nil
+}

core/bloombits/generator_test.go

@@ -0,0 +1,100 @@
+// Copyright 2017 The go-ethereum Authors
+// This file is part of the go-ethereum library.
+//
+// The go-ethereum library is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Lesser General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// The go-ethereum library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU Lesser General Public License for more details.
+//
+// You should have received a copy of the GNU Lesser General Public License
+// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
+
+package bloombits
+
+import (
+	"bytes"
+	crand "crypto/rand"
+	"math/rand"
+	"testing"
+
+	"github.com/ethereum/go-ethereum/core/types"
+)
+
+// Tests that batched bloom bits are correctly rotated from the input bloom
+// filters.
+func TestGenerator(t *testing.T) {
+	// Generate the input and the rotated output
+	var input, output [types.BloomBitLength][types.BloomByteLength]byte
+
+	for i := 0; i < types.BloomBitLength; i++ {
+		for j := 0; j < types.BloomBitLength; j++ {
+			bit := byte(rand.Int() % 2)
+
+			input[i][j/8] |= bit << byte(7-j%8)
+			output[types.BloomBitLength-1-j][i/8] |= bit << byte(7-i%8)
+		}
+	}
+	// Crunch the input through the generator and verify the result
+	gen, err := NewGenerator(types.BloomBitLength)
+	if err != nil {
+		t.Fatalf("failed to create bloombit generator: %v", err)
+	}
+	for i, bloom := range input {
+		if err := gen.AddBloom(uint(i), bloom); err != nil {
+			t.Fatalf("bloom %d: failed to add: %v", i, err)
+		}
+	}
+	for i, want := range output {
+		have, err := gen.Bitset(uint(i))
+		if err != nil {
+			t.Fatalf("output %d: failed to retrieve bits: %v", i, err)
+		}
+		if !bytes.Equal(have, want[:]) {
+			t.Errorf("output %d: bit vector mismatch have %x, want %x", i, have, want)
+		}
+	}
+}
+
+func BenchmarkGenerator(b *testing.B) {
+	var input [types.BloomBitLength][types.BloomByteLength]byte
+	b.Run("empty", func(b *testing.B) {
+		b.ReportAllocs()
+		b.ResetTimer()
+		for i := 0; i < b.N; i++ {
+			// Crunch the input through the generator and verify the result
+			gen, err := NewGenerator(types.BloomBitLength)
+			if err != nil {
+				b.Fatalf("failed to create bloombit generator: %v", err)
+			}
+			for j, bloom := range &input {
+				if err := gen.AddBloom(uint(j), bloom); err != nil {
+					b.Fatalf("bloom %d: failed to add: %v", i, err)
+				}
+			}
+		}
+	})
+	for i := 0; i < types.BloomBitLength; i++ {
+		crand.Read(input[i][:])
+	}
+	b.Run("random", func(b *testing.B) {
+		b.ReportAllocs()
+		b.ResetTimer()
+		for i := 0; i < b.N; i++ {
+			// Crunch the input through the generator and verify the result
+			gen, err := NewGenerator(types.BloomBitLength)
+			if err != nil {
+				b.Fatalf("failed to create bloombit generator: %v", err)
+			}
+			for j, bloom := range &input {
+				if err := gen.AddBloom(uint(j), bloom); err != nil {
+					b.Fatalf("bloom %d: failed to add: %v", i, err)
+				}
+			}
+		}
+	})
+}

core/bloombits/matcher.go

@@ -0,0 +1,649 @@
+// Copyright 2017 The go-ethereum Authors
+// This file is part of the go-ethereum library.
+//
+// The go-ethereum library is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Lesser General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// The go-ethereum library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU Lesser General Public License for more details.
+//
+// You should have received a copy of the GNU Lesser General Public License
+// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
+
+package bloombits
+
+import (
+	"bytes"
+	"context"
+	"errors"
+	"math"
+	"sort"
+	"sync"
+	"sync/atomic"
+	"time"
+
+	"github.com/ethereum/go-ethereum/common/bitutil"
+	"github.com/ethereum/go-ethereum/crypto"
+)
+
+// bloomIndexes represents the bit indexes inside the bloom filter that belong
+// to some key.
+type bloomIndexes [3]uint
+
+// calcBloomIndexes returns the bloom filter bit indexes belonging to the given key.
+func calcBloomIndexes(b []byte) bloomIndexes {
+	b = crypto.Keccak256(b)
+
+	var idxs bloomIndexes
+	for i := 0; i < len(idxs); i++ {
+		idxs[i] = (uint(b[2*i])<<8)&2047 + uint(b[2*i+1])
+	}
+	return idxs
+}
+
+// partialMatches with a non-nil vector represents a section in which some sub-
+// matchers have already found potential matches. Subsequent sub-matchers will
+// binary AND their matches with this vector. If vector is nil, it represents a
+// section to be processed by the first sub-matcher.
+type partialMatches struct {
+	section uint64
+	bitset  []byte
+}
+
+// Retrieval represents a request for retrieval task assignments for a given
+// bit with the given number of fetch elements, or a response for such a request.
+// It can also have the actual results set to be used as a delivery data struct.
+//
+// The context and error fields are used by the light client to terminate matching
+// early if an error is encountered on some path of the pipeline.
+type Retrieval struct {
+	Bit      uint
+	Sections []uint64
+	Bitsets  [][]byte
+
+	Context context.Context
+	Error   error
+}
+
+// Matcher is a pipelined system of schedulers and logic matchers which perform
+// binary AND/OR operations on the bit-streams, creating a stream of potential
+// blocks to inspect for data content.
+type Matcher struct {
+	sectionSize uint64 // Size of the data batches to filter on
+
+	filters    [][]bloomIndexes    // Filter the system is matching for
+	schedulers map[uint]*scheduler // Retrieval schedulers for loading bloom bits
+
+	retrievers chan chan uint       // Retriever processes waiting for bit allocations
+	counters   chan chan uint       // Retriever processes waiting for task count reports
+	retrievals chan chan *Retrieval // Retriever processes waiting for task allocations
+	deliveries chan *Retrieval      // Retriever processes waiting for task response deliveries
+
+	running atomic.Bool // Atomic flag whether a session is live or not
+}
+
+// NewMatcher creates a new pipeline for retrieving bloom bit streams and doing
+// address and topic filtering on them. Setting a filter component to `nil` is
+// allowed and will result in that filter rule being skipped (OR 0x11...1).
+func NewMatcher(sectionSize uint64, filters [][][]byte) *Matcher {
+	// Create the matcher instance
+	m := &Matcher{
+		sectionSize: sectionSize,
+		schedulers:  make(map[uint]*scheduler),
+		retrievers:  make(chan chan uint),
+		counters:    make(chan chan uint),
+		retrievals:  make(chan chan *Retrieval),
+		deliveries:  make(chan *Retrieval),
+	}
+	// Calculate the bloom bit indexes for the groups we're interested in
+	m.filters = nil
+
+	for _, filter := range filters {
+		// Gather the bit indexes of the filter rule, special casing the nil filter
+		if len(filter) == 0 {
+			continue
+		}
+		bloomBits := make([]bloomIndexes, len(filter))
+		for i, clause := range filter {
+			if clause == nil {
+				bloomBits = nil
+				break
+			}
+			bloomBits[i] = calcBloomIndexes(clause)
+		}
+		// Accumulate the filter rules if no nil rule was within
+		if bloomBits != nil {
+			m.filters = append(m.filters, bloomBits)
+		}
+	}
+	// For every bit, create a scheduler to load/download the bit vectors
+	for _, bloomIndexLists := range m.filters {
+		for _, bloomIndexList := range bloomIndexLists {
+			for _, bloomIndex := range bloomIndexList {
+				m.addScheduler(bloomIndex)
+			}
+		}
+	}
+	return m
+}
+
+// addScheduler adds a bit stream retrieval scheduler for the given bit index if
+// it has not existed before. If the bit is already selected for filtering, the
+// existing scheduler can be used.
+func (m *Matcher) addScheduler(idx uint) {
+	if _, ok := m.schedulers[idx]; ok {
+		return
+	}
+	m.schedulers[idx] = newScheduler(idx)
+}
+
+// Start starts the matching process and returns a stream of bloom matches in
+// a given range of blocks. If there are no more matches in the range, the result
+// channel is closed.
+func (m *Matcher) Start(ctx context.Context, begin, end uint64, results chan uint64) (*MatcherSession, error) {
+	// Make sure we're not creating concurrent sessions
+	if m.running.Swap(true) {
+		return nil, errors.New("matcher already running")
+	}
+	defer m.running.Store(false)
+
+	// Initiate a new matching round
+	session := &MatcherSession{
+		matcher: m,
+		quit:    make(chan struct{}),
+		ctx:     ctx,
+	}
+	for _, scheduler := range m.schedulers {
+		scheduler.reset()
+	}
+	sink := m.run(begin, end, cap(results), session)
+
+	// Read the output from the result sink and deliver to the user
+	session.pend.Add(1)
+	go func() {
+		defer session.pend.Done()
+		defer close(results)
+
+		for {
+			select {
+			case <-session.quit:
+				return
+
+			case res, ok := <-sink:
+				// New match result found
+				if !ok {
+					return
+				}
+				// Calculate the first and last blocks of the section
+				sectionStart := res.section * m.sectionSize
+
+				first := sectionStart
+				if begin > first {
+					first = begin
+				}
+				last := sectionStart + m.sectionSize - 1
+				if end < last {
+					last = end
+				}
+				// Iterate over all the blocks in the section and return the matching ones
+				for i := first; i <= last; i++ {
+					// Skip the entire byte if no matches are found inside (and we're processing an entire byte!)
+					next := res.bitset[(i-sectionStart)/8]
+					if next == 0 {
+						if i%8 == 0 {
+							i += 7
+						}
+						continue
+					}
+					// Some bit it set, do the actual submatching
+					if bit := 7 - i%8; next&(1<<bit) != 0 {
+						select {
+						case <-session.quit:
+							return
+						case results <- i:
+						}
+					}
+				}
+			}
+		}
+	}()
+	return session, nil
+}
+
+// run creates a daisy-chain of sub-matchers, one for the address set and one
+// for each topic set, each sub-matcher receiving a section only if the previous
+// ones have all found a potential match in one of the blocks of the section,
+// then binary AND-ing its own matches and forwarding the result to the next one.
+//
+// The method starts feeding the section indexes into the first sub-matcher on a
+// new goroutine and returns a sink channel receiving the results.
+func (m *Matcher) run(begin, end uint64, buffer int, session *MatcherSession) chan *partialMatches {
+	// Create the source channel and feed section indexes into
+	source := make(chan *partialMatches, buffer)
+
+	session.pend.Add(1)
+	go func() {
+		defer session.pend.Done()
+		defer close(source)
+
+		for i := begin / m.sectionSize; i <= end/m.sectionSize; i++ {
+			select {
+			case <-session.quit:
+				return
+			case source <- &partialMatches{i, bytes.Repeat([]byte{0xff}, int(m.sectionSize/8))}:
+			}
+		}
+	}()
+	// Assemble the daisy-chained filtering pipeline
+	next := source
+	dist := make(chan *request, buffer)
+
+	for _, bloom := range m.filters {
+		next = m.subMatch(next, dist, bloom, session)
+	}
+	// Start the request distribution
+	session.pend.Add(1)
+	go m.distributor(dist, session)
+
+	return next
+}
+
+// subMatch creates a sub-matcher that filters for a set of addresses or topics, binary OR-s those matches, then
+// binary AND-s the result to the daisy-chain input (source) and forwards it to the daisy-chain output.
+// The matches of each address/topic are calculated by fetching the given sections of the three bloom bit indexes belonging to
+// that address/topic, and binary AND-ing those vectors together.
+func (m *Matcher) subMatch(source chan *partialMatches, dist chan *request, bloom []bloomIndexes, session *MatcherSession) chan *partialMatches {
+	// Start the concurrent schedulers for each bit required by the bloom filter
+	sectionSources := make([][3]chan uint64, len(bloom))
+	sectionSinks := make([][3]chan []byte, len(bloom))
+	for i, bits := range bloom {
+		for j, bit := range bits {
+			sectionSources[i][j] = make(chan uint64, cap(source))
+			sectionSinks[i][j] = make(chan []byte, cap(source))
+
+			m.schedulers[bit].run(sectionSources[i][j], dist, sectionSinks[i][j], session.quit, &session.pend)
+		}
+	}
+
+	process := make(chan *partialMatches, cap(source)) // entries from source are forwarded here after fetches have been initiated
+	results := make(chan *partialMatches, cap(source))
+
+	session.pend.Add(2)
+	go func() {
+		// Tear down the goroutine and terminate all source channels
+		defer session.pend.Done()
+		defer close(process)
+
+		defer func() {
+			for _, bloomSources := range sectionSources {
+				for _, bitSource := range bloomSources {
+					close(bitSource)
+				}
+			}
+		}()
+		// Read sections from the source channel and multiplex into all bit-schedulers
+		for {
+			select {
+			case <-session.quit:
+				return
+
+			case subres, ok := <-source:
+				// New subresult from previous link
+				if !ok {
+					return
+				}
+				// Multiplex the section index to all bit-schedulers
+				for _, bloomSources := range sectionSources {
+					for _, bitSource := range bloomSources {
+						select {
+						case <-session.quit:
+							return
+						case bitSource <- subres.section:
+						}
+					}
+				}
+				// Notify the processor that this section will become available
+				select {
+				case <-session.quit:
+					return
+				case process <- subres:
+				}
+			}
+		}
+	}()
+
+	go func() {
+		// Tear down the goroutine and terminate the final sink channel
+		defer session.pend.Done()
+		defer close(results)
+
+		// Read the source notifications and collect the delivered results
+		for {
+			select {
+			case <-session.quit:
+				return
+
+			case subres, ok := <-process:
+				// Notified of a section being retrieved
+				if !ok {
+					return
+				}
+				// Gather all the sub-results and merge them together
+				var orVector []byte
+				for _, bloomSinks := range sectionSinks {
+					var andVector []byte
+					for _, bitSink := range bloomSinks {
+						var data []byte
+						select {
+						case <-session.quit:
+							return
+						case data = <-bitSink:
+						}
+						if andVector == nil {
+							andVector = make([]byte, int(m.sectionSize/8))
+							copy(andVector, data)
+						} else {
+							bitutil.ANDBytes(andVector, andVector, data)
+						}
+					}
+					if orVector == nil {
+						orVector = andVector
+					} else {
+						bitutil.ORBytes(orVector, orVector, andVector)
+					}
+				}
+
+				if orVector == nil {
+					orVector = make([]byte, int(m.sectionSize/8))
+				}
+				if subres.bitset != nil {
+					bitutil.ANDBytes(orVector, orVector, subres.bitset)
+				}
+				if bitutil.TestBytes(orVector) {
+					select {
+					case <-session.quit:
+						return
+					case results <- &partialMatches{subres.section, orVector}:
+					}
+				}
+			}
+		}
+	}()
+	return results
+}
+
+// distributor receives requests from the schedulers and queues them into a set
+// of pending requests, which are assigned to retrievers wanting to fulfil them.
+func (m *Matcher) distributor(dist chan *request, session *MatcherSession) {
+	defer session.pend.Done()
+
+	var (
+		requests   = make(map[uint][]uint64) // Per-bit list of section requests, ordered by section number
+		unallocs   = make(map[uint]struct{}) // Bits with pending requests but not allocated to any retriever
+		retrievers chan chan uint            // Waiting retrievers (toggled to nil if unallocs is empty)
+		allocs     int                       // Number of active allocations to handle graceful shutdown requests
+		shutdown   = session.quit            // Shutdown request channel, will gracefully wait for pending requests
+	)
+
+	// assign is a helper method to try to assign a pending bit an actively
+	// listening servicer, or schedule it up for later when one arrives.
+	assign := func(bit uint) {
+		select {
+		case fetcher := <-m.retrievers:
+			allocs++
+			fetcher <- bit
+		default:
+			// No retrievers active, start listening for new ones
+			retrievers = m.retrievers
+			unallocs[bit] = struct{}{}
+		}
+	}
+
+	for {
+		select {
+		case <-shutdown:
+			// Shutdown requested. No more retrievers can be allocated,
+			// but we still need to wait until all pending requests have returned.
+			shutdown = nil
+			if allocs == 0 {
+				return
+			}
+
+		case req := <-dist:
+			// New retrieval request arrived to be distributed to some fetcher process
+			queue := requests[req.bit]
+			index := sort.Search(len(queue), func(i int) bool { return queue[i] >= req.section })
+			requests[req.bit] = append(queue[:index], append([]uint64{req.section}, queue[index:]...)...)
+
+			// If it's a new bit and we have waiting fetchers, allocate to them
+			if len(queue) == 0 {
+				assign(req.bit)
+			}
+
+		case fetcher := <-retrievers:
+			// New retriever arrived, find the lowest section-ed bit to assign
+			bit, best := uint(0), uint64(math.MaxUint64)
+			for idx := range unallocs {
+				if requests[idx][0] < best {
+					bit, best = idx, requests[idx][0]
+				}
+			}
+			// Stop tracking this bit (and alloc notifications if no more work is available)
+			delete(unallocs, bit)
+			if len(unallocs) == 0 {
+				retrievers = nil
+			}
+			allocs++
+			fetcher <- bit
+
+		case fetcher := <-m.counters:
+			// New task count request arrives, return number of items
+			fetcher <- uint(len(requests[<-fetcher]))
+
+		case fetcher := <-m.retrievals:
+			// New fetcher waiting for tasks to retrieve, assign
+			task := <-fetcher
+			if want := len(task.Sections); want >= len(requests[task.Bit]) {
+				task.Sections = requests[task.Bit]
+				delete(requests, task.Bit)
+			} else {
+				task.Sections = append(task.Sections[:0], requests[task.Bit][:want]...)
+				requests[task.Bit] = append(requests[task.Bit][:0], requests[task.Bit][want:]...)
+			}
+			fetcher <- task
+
+			// If anything was left unallocated, try to assign to someone else
+			if len(requests[task.Bit]) > 0 {
+				assign(task.Bit)
+			}
+
+		case result := <-m.deliveries:
+			// New retrieval task response from fetcher, split out missing sections and
+			// deliver complete ones
+			var (
+				sections = make([]uint64, 0, len(result.Sections))
+				bitsets  = make([][]byte, 0, len(result.Bitsets))
+				missing  = make([]uint64, 0, len(result.Sections))
+			)
+			for i, bitset := range result.Bitsets {
+				if len(bitset) == 0 {
+					missing = append(missing, result.Sections[i])
+					continue
+				}
+				sections = append(sections, result.Sections[i])
+				bitsets = append(bitsets, bitset)
+			}
+			m.schedulers[result.Bit].deliver(sections, bitsets)
+			allocs--
+
+			// Reschedule missing sections and allocate bit if newly available
+			if len(missing) > 0 {
+				queue := requests[result.Bit]
+				for _, section := range missing {
+					index := sort.Search(len(queue), func(i int) bool { return queue[i] >= section })
+					queue = append(queue[:index], append([]uint64{section}, queue[index:]...)...)
+				}
+				requests[result.Bit] = queue
+
+				if len(queue) == len(missing) {
+					assign(result.Bit)
+				}
+			}
+
+			// End the session when all pending deliveries have arrived.
+			if shutdown == nil && allocs == 0 {
+				return
+			}
+		}
+	}
+}
+
+// MatcherSession is returned by a started matcher to be used as a terminator
+// for the actively running matching operation.
+type MatcherSession struct {
+	matcher *Matcher
+
+	closer sync.Once     // Sync object to ensure we only ever close once
+	quit   chan struct{} // Quit channel to request pipeline termination
+
+	ctx     context.Context // Context used by the light client to abort filtering
+	err     error           // Global error to track retrieval failures deep in the chain
+	errLock sync.Mutex
+
+	pend sync.WaitGroup
+}
+
+// Close stops the matching process and waits for all subprocesses to terminate
+// before returning. The timeout may be used for graceful shutdown, allowing the
+// currently running retrievals to complete before this time.
+func (s *MatcherSession) Close() {
+	s.closer.Do(func() {
+		// Signal termination and wait for all goroutines to tear down
+		close(s.quit)
+		s.pend.Wait()
+	})
+}
+
+// Error returns any failure encountered during the matching session.
+func (s *MatcherSession) Error() error {
+	s.errLock.Lock()
+	defer s.errLock.Unlock()
+
+	return s.err
+}
+
+// allocateRetrieval assigns a bloom bit index to a client process that can either
+// immediately request and fetch the section contents assigned to this bit or wait
+// a little while for more sections to be requested.
+func (s *MatcherSession) allocateRetrieval() (uint, bool) {
+	fetcher := make(chan uint)
+
+	select {
+	case <-s.quit:
+		return 0, false
+	case s.matcher.retrievers <- fetcher:
+		bit, ok := <-fetcher
+		return bit, ok
+	}
+}
+
+// pendingSections returns the number of pending section retrievals belonging to
+// the given bloom bit index.
+func (s *MatcherSession) pendingSections(bit uint) int {
+	fetcher := make(chan uint)
+
+	select {
+	case <-s.quit:
+		return 0
+	case s.matcher.counters <- fetcher:
+		fetcher <- bit
+		return int(<-fetcher)
+	}
+}
+
+// allocateSections assigns all or part of an already allocated bit-task queue
+// to the requesting process.
+func (s *MatcherSession) allocateSections(bit uint, count int) []uint64 {
+	fetcher := make(chan *Retrieval)
+
+	select {
+	case <-s.quit:
+		return nil
+	case s.matcher.retrievals <- fetcher:
+		task := &Retrieval{
+			Bit:      bit,
+			Sections: make([]uint64, count),
+		}
+		fetcher <- task
+		return (<-fetcher).Sections
+	}
+}
+
+// deliverSections delivers a batch of section bit-vectors for a specific bloom
+// bit index to be injected into the processing pipeline.
+func (s *MatcherSession) deliverSections(bit uint, sections []uint64, bitsets [][]byte) {
+	s.matcher.deliveries <- &Retrieval{Bit: bit, Sections: sections, Bitsets: bitsets}
+}
+
+// Multiplex polls the matcher session for retrieval tasks and multiplexes it into
+// the requested retrieval queue to be serviced together with other sessions.
+//
+// This method will block for the lifetime of the session. Even after termination
+// of the session, any request in-flight need to be responded to! Empty responses
+// are fine though in that case.
+func (s *MatcherSession) Multiplex(batch int, wait time.Duration, mux chan chan *Retrieval) {
+	waitTimer := time.NewTimer(wait)
+	defer waitTimer.Stop()
+
+	for {
+		// Allocate a new bloom bit index to retrieve data for, stopping when done
+		bit, ok := s.allocateRetrieval()
+		if !ok {
+			return
+		}
+		// Bit allocated, throttle a bit if we're below our batch limit
+		if s.pendingSections(bit) < batch {
+			waitTimer.Reset(wait)
+			select {
+			case <-s.quit:
+				// Session terminating, we can't meaningfully service, abort
+				s.allocateSections(bit, 0)
+				s.deliverSections(bit, []uint64{}, [][]byte{})
+				return
+
+			case <-waitTimer.C:
+				// Throttling up, fetch whatever is available
+			}
+		}
+		// Allocate as much as we can handle and request servicing
+		sections := s.allocateSections(bit, batch)
+		request := make(chan *Retrieval)
+
+		select {
+		case <-s.quit:
+			// Session terminating, we can't meaningfully service, abort
+			s.deliverSections(bit, sections, make([][]byte, len(sections)))
+			return
+
+		case mux <- request:
+			// Retrieval accepted, something must arrive before we're aborting
+			request <- &Retrieval{Bit: bit, Sections: sections, Context: s.ctx}
+
+			result := <-request
+
+			// Deliver a result before s.Close() to avoid a deadlock
+			s.deliverSections(result.Bit, result.Sections, result.Bitsets)
+
+			if result.Error != nil {
+				s.errLock.Lock()
+				s.err = result.Error
+				s.errLock.Unlock()
+				s.Close()
+			}
+		}
+	}
+}

core/bloombits/matcher_test.go

@@ -0,0 +1,292 @@
+// Copyright 2017 The go-ethereum Authors
+// This file is part of the go-ethereum library.
+//
+// The go-ethereum library is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Lesser General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// The go-ethereum library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU Lesser General Public License for more details.
+//
+// You should have received a copy of the GNU Lesser General Public License
+// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
+
+package bloombits
+
+import (
+	"context"
+	"math/rand"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"github.com/ethereum/go-ethereum/common"
+)
+
+const testSectionSize = 4096
+
+// Tests that wildcard filter rules (nil) can be specified and are handled well.
+func TestMatcherWildcards(t *testing.T) {
+	t.Parallel()
+	matcher := NewMatcher(testSectionSize, [][][]byte{
+		{common.Address{}.Bytes(), common.Address{0x01}.Bytes()}, // Default address is not a wildcard
+		{common.Hash{}.Bytes(), common.Hash{0x01}.Bytes()},       // Default hash is not a wildcard
+		{common.Hash{0x01}.Bytes()},                              // Plain rule, sanity check
+		{common.Hash{0x01}.Bytes(), nil},                         // Wildcard suffix, drop rule
+		{nil, common.Hash{0x01}.Bytes()},                         // Wildcard prefix, drop rule
+		{nil, nil},                                               // Wildcard combo, drop rule
+		{},                                                       // Inited wildcard rule, drop rule
+		nil,                                                      // Proper wildcard rule, drop rule
+	})
+	if len(matcher.filters) != 3 {
+		t.Fatalf("filter system size mismatch: have %d, want %d", len(matcher.filters), 3)
+	}
+	if len(matcher.filters[0]) != 2 {
+		t.Fatalf("address clause size mismatch: have %d, want %d", len(matcher.filters[0]), 2)
+	}
+	if len(matcher.filters[1]) != 2 {
+		t.Fatalf("combo topic clause size mismatch: have %d, want %d", len(matcher.filters[1]), 2)
+	}
+	if len(matcher.filters[2]) != 1 {
+		t.Fatalf("singletone topic clause size mismatch: have %d, want %d", len(matcher.filters[2]), 1)
+	}
+}
+
+// Tests the matcher pipeline on a single continuous workflow without interrupts.
+func TestMatcherContinuous(t *testing.T) {
+	t.Parallel()
+	testMatcherDiffBatches(t, [][]bloomIndexes{{{10, 20, 30}}}, 0, 100000, false, 75)
+	testMatcherDiffBatches(t, [][]bloomIndexes{{{32, 3125, 100}}, {{40, 50, 10}}}, 0, 100000, false, 81)
+	testMatcherDiffBatches(t, [][]bloomIndexes{{{4, 8, 11}, {7, 8, 17}}, {{9, 9, 12}, {15, 20, 13}}, {{18, 15, 15}, {12, 10, 4}}}, 0, 10000, false, 36)
+}
+
+// Tests the matcher pipeline on a constantly interrupted and resumed work pattern
+// with the aim of ensuring data items are requested only once.
+func TestMatcherIntermittent(t *testing.T) {
+	t.Parallel()
+	testMatcherDiffBatches(t, [][]bloomIndexes{{{10, 20, 30}}}, 0, 100000, true, 75)
+	testMatcherDiffBatches(t, [][]bloomIndexes{{{32, 3125, 100}}, {{40, 50, 10}}}, 0, 100000, true, 81)
+	testMatcherDiffBatches(t, [][]bloomIndexes{{{4, 8, 11}, {7, 8, 17}}, {{9, 9, 12}, {15, 20, 13}}, {{18, 15, 15}, {12, 10, 4}}}, 0, 10000, true, 36)
+}
+
+// Tests the matcher pipeline on random input to hopefully catch anomalies.
+func TestMatcherRandom(t *testing.T) {
+	t.Parallel()
+	for i := 0; i < 10; i++ {
+		testMatcherBothModes(t, makeRandomIndexes([]int{1}, 50), 0, 10000, 0)
+		testMatcherBothModes(t, makeRandomIndexes([]int{3}, 50), 0, 10000, 0)
+		testMatcherBothModes(t, makeRandomIndexes([]int{2, 2, 2}, 20), 0, 10000, 0)
+		testMatcherBothModes(t, makeRandomIndexes([]int{5, 5, 5}, 50), 0, 10000, 0)
+		testMatcherBothModes(t, makeRandomIndexes([]int{4, 4, 4}, 20), 0, 10000, 0)
+	}
+}
+
+// Tests that the matcher can properly find matches if the starting block is
+// shifted from a multiple of 8. This is needed to cover an optimisation with
+// bitset matching https://github.com/ethereum/go-ethereum/issues/15309.
+func TestMatcherShifted(t *testing.T) {
+	t.Parallel()
+	// Block 0 always matches in the tests, skip ahead of first 8 blocks with the
+	// start to get a potential zero byte in the matcher bitset.
+
+	// To keep the second bitset byte zero, the filter must only match for the first
+	// time in block 16, so doing an all-16 bit filter should suffice.
+
+	// To keep the starting block non divisible by 8, block number 9 is the first
+	// that would introduce a shift and not match block 0.
+	testMatcherBothModes(t, [][]bloomIndexes{{{16, 16, 16}}}, 9, 64, 0)
+}
+
+// Tests that matching on everything doesn't crash (special case internally).
+func TestWildcardMatcher(t *testing.T) {
+	t.Parallel()
+	testMatcherBothModes(t, nil, 0, 10000, 0)
+}
+
+// makeRandomIndexes generates a random filter system, composed of multiple filter
+// criteria, each having one bloom list component for the address and arbitrarily
+// many topic bloom list components.
+func makeRandomIndexes(lengths []int, max int) [][]bloomIndexes {
+	res := make([][]bloomIndexes, len(lengths))
+	for i, topics := range lengths {
+		res[i] = make([]bloomIndexes, topics)
+		for j := 0; j < topics; j++ {
+			for k := 0; k < len(res[i][j]); k++ {
+				res[i][j][k] = uint(rand.Intn(max-1) + 2)
+			}
+		}
+	}
+	return res
+}
+
+// testMatcherDiffBatches runs the given matches test in single-delivery and also
+// in batches delivery mode, verifying that all kinds of deliveries are handled
+// correctly within.
+func testMatcherDiffBatches(t *testing.T, filter [][]bloomIndexes, start, blocks uint64, intermittent bool, retrievals uint32) {
+	singleton := testMatcher(t, filter, start, blocks, intermittent, retrievals, 1)
+	batched := testMatcher(t, filter, start, blocks, intermittent, retrievals, 16)
+
+	if singleton != batched {
+		t.Errorf("filter = %v blocks = %v intermittent = %v: request count mismatch, %v in singleton vs. %v in batched mode", filter, blocks, intermittent, singleton, batched)
+	}
+}
+
+// testMatcherBothModes runs the given matcher test in both continuous as well as
+// in intermittent mode, verifying that the request counts match each other.
+func testMatcherBothModes(t *testing.T, filter [][]bloomIndexes, start, blocks uint64, retrievals uint32) {
+	continuous := testMatcher(t, filter, start, blocks, false, retrievals, 16)
+	intermittent := testMatcher(t, filter, start, blocks, true, retrievals, 16)
+
+	if continuous != intermittent {
+		t.Errorf("filter = %v blocks = %v: request count mismatch, %v in continuous vs. %v in intermittent mode", filter, blocks, continuous, intermittent)
+	}
+}
+
+// testMatcher is a generic tester to run the given matcher test and return the
+// number of requests made for cross validation between different modes.
+func testMatcher(t *testing.T, filter [][]bloomIndexes, start, blocks uint64, intermittent bool, retrievals uint32, maxReqCount int) uint32 {
+	// Create a new matcher an simulate our explicit random bitsets
+	matcher := NewMatcher(testSectionSize, nil)
+	matcher.filters = filter
+
+	for _, rule := range filter {
+		for _, topic := range rule {
+			for _, bit := range topic {
+				matcher.addScheduler(bit)
+			}
+		}
+	}
+	// Track the number of retrieval requests made
+	var requested atomic.Uint32
+
+	// Start the matching session for the filter and the retriever goroutines
+	quit := make(chan struct{})
+	matches := make(chan uint64, 16)
+
+	session, err := matcher.Start(context.Background(), start, blocks-1, matches)
+	if err != nil {
+		t.Fatalf("failed to stat matcher session: %v", err)
+	}
+	startRetrievers(session, quit, &requested, maxReqCount)
+
+	// Iterate over all the blocks and verify that the pipeline produces the correct matches
+	for i := start; i < blocks; i++ {
+		if expMatch3(filter, i) {
+			match, ok := <-matches
+			if !ok {
+				t.Errorf("filter = %v  blocks = %v  intermittent = %v: expected #%v, results channel closed", filter, blocks, intermittent, i)
+				return 0
+			}
+			if match != i {
+				t.Errorf("filter = %v  blocks = %v  intermittent = %v: expected #%v, got #%v", filter, blocks, intermittent, i, match)
+			}
+			// If we're testing intermittent mode, abort and restart the pipeline
+			if intermittent {
+				session.Close()
+				close(quit)
+
+				quit = make(chan struct{})
+				matches = make(chan uint64, 16)
+
+				session, err = matcher.Start(context.Background(), i+1, blocks-1, matches)
+				if err != nil {
+					t.Fatalf("failed to stat matcher session: %v", err)
+				}
+				startRetrievers(session, quit, &requested, maxReqCount)
+			}
+		}
+	}
+	// Ensure the result channel is torn down after the last block
+	match, ok := <-matches
+	if ok {
+		t.Errorf("filter = %v  blocks = %v  intermittent = %v: expected closed channel, got #%v", filter, blocks, intermittent, match)
+	}
+	// Clean up the session and ensure we match the expected retrieval count
+	session.Close()
+	close(quit)
+
+	if retrievals != 0 && requested.Load() != retrievals {
+		t.Errorf("filter = %v  blocks = %v  intermittent = %v: request count mismatch, have #%v, want #%v", filter, blocks, intermittent, requested.Load(), retrievals)
+	}
+	return requested.Load()
+}
+
+// startRetrievers starts a batch of goroutines listening for section requests
+// and serving them.
+func startRetrievers(session *MatcherSession, quit chan struct{}, retrievals *atomic.Uint32, batch int) {
+	requests := make(chan chan *Retrieval)
+
+	for i := 0; i < 10; i++ {
+		// Start a multiplexer to test multiple threaded execution
+		go session.Multiplex(batch, 100*time.Microsecond, requests)
+
+		// Start a services to match the above multiplexer
+		go func() {
+			for {
+				// Wait for a service request or a shutdown
+				select {
+				case <-quit:
+					return
+
+				case request := <-requests:
+					task := <-request
+
+					task.Bitsets = make([][]byte, len(task.Sections))
+					for i, section := range task.Sections {
+						if rand.Int()%4 != 0 { // Handle occasional missing deliveries
+							task.Bitsets[i] = generateBitset(task.Bit, section)
+							retrievals.Add(1)
+						}
+					}
+					request <- task
+				}
+			}
+		}()
+	}
+}
+
+// generateBitset generates the rotated bitset for the given bloom bit and section
+// numbers.
+func generateBitset(bit uint, section uint64) []byte {
+	bitset := make([]byte, testSectionSize/8)
+	for i := 0; i < len(bitset); i++ {
+		for b := 0; b < 8; b++ {
+			blockIdx := section*testSectionSize + uint64(i*8+b)
+			bitset[i] += bitset[i]
+			if (blockIdx % uint64(bit)) == 0 {
+				bitset[i]++
+			}
+		}
+	}
+	return bitset
+}
+
+func expMatch1(filter bloomIndexes, i uint64) bool {
+	for _, ii := range filter {
+		if (i % uint64(ii)) != 0 {
+			return false
+		}
+	}
+	return true
+}
+
+func expMatch2(filter []bloomIndexes, i uint64) bool {
+	for _, ii := range filter {
+		if expMatch1(ii, i) {
+			return true
+		}
+	}
+	return false
+}
+
+func expMatch3(filter [][]bloomIndexes, i uint64) bool {
+	for _, ii := range filter {
+		if !expMatch2(ii, i) {
+			return false
+		}
+	}
+	return true
+}

core/bloombits/scheduler.go

@@ -0,0 +1,181 @@
+// Copyright 2017 The go-ethereum Authors
+// This file is part of the go-ethereum library.
+//
+// The go-ethereum library is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Lesser General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// The go-ethereum library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU Lesser General Public License for more details.
+//
+// You should have received a copy of the GNU Lesser General Public License
+// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
+
+package bloombits
+
+import (
+	"sync"
+)
+
+// request represents a bloom retrieval task to prioritize and pull from the local
+// database or remotely from the network.
+type request struct {
+	section uint64 // Section index to retrieve the bit-vector from
+	bit     uint   // Bit index within the section to retrieve the vector of
+}
+
+// response represents the state of a requested bit-vector through a scheduler.
+type response struct {
+	cached []byte        // Cached bits to dedup multiple requests
+	done   chan struct{} // Channel to allow waiting for completion
+}
+
+// scheduler handles the scheduling of bloom-filter retrieval operations for
+// entire section-batches belonging to a single bloom bit. Beside scheduling the
+// retrieval operations, this struct also deduplicates the requests and caches
+// the results to minimize network/database overhead even in complex filtering
+// scenarios.
+type scheduler struct {
+	bit       uint                 // Index of the bit in the bloom filter this scheduler is responsible for
+	responses map[uint64]*response // Currently pending retrieval requests or already cached responses
+	lock      sync.Mutex           // Lock protecting the responses from concurrent access
+}
+
+// newScheduler creates a new bloom-filter retrieval scheduler for a specific
+// bit index.
+func newScheduler(idx uint) *scheduler {
+	return &scheduler{
+		bit:       idx,
+		responses: make(map[uint64]*response),
+	}
+}
+
+// run creates a retrieval pipeline, receiving section indexes from sections and
+// returning the results in the same order through the done channel. Concurrent
+// runs of the same scheduler are allowed, leading to retrieval task deduplication.
+func (s *scheduler) run(sections chan uint64, dist chan *request, done chan []byte, quit chan struct{}, wg *sync.WaitGroup) {
+	// Create a forwarder channel between requests and responses of the same size as
+	// the distribution channel (since that will block the pipeline anyway).
+	pend := make(chan uint64, cap(dist))
+
+	// Start the pipeline schedulers to forward between user -> distributor -> user
+	wg.Add(2)
+	go s.scheduleRequests(sections, dist, pend, quit, wg)
+	go s.scheduleDeliveries(pend, done, quit, wg)
+}
+
+// reset cleans up any leftovers from previous runs. This is required before a
+// restart to ensure the no previously requested but never delivered state will
+// cause a lockup.
+func (s *scheduler) reset() {
+	s.lock.Lock()
+	defer s.lock.Unlock()
+
+	for section, res := range s.responses {
+		if res.cached == nil {
+			delete(s.responses, section)
+		}
+	}
+}
+
+// scheduleRequests reads section retrieval requests from the input channel,
+// deduplicates the stream and pushes unique retrieval tasks into the distribution
+// channel for a database or network layer to honour.
+func (s *scheduler) scheduleRequests(reqs chan uint64, dist chan *request, pend chan uint64, quit chan struct{}, wg *sync.WaitGroup) {
+	// Clean up the goroutine and pipeline when done
+	defer wg.Done()
+	defer close(pend)
+
+	// Keep reading and scheduling section requests
+	for {
+		select {
+		case <-quit:
+			return
+
+		case section, ok := <-reqs:
+			// New section retrieval requested
+			if !ok {
+				return
+			}
+			// Deduplicate retrieval requests
+			unique := false
+
+			s.lock.Lock()
+			if s.responses[section] == nil {
+				s.responses[section] = &response{
+					done: make(chan struct{}),
+				}
+				unique = true
+			}
+			s.lock.Unlock()
+
+			// Schedule the section for retrieval and notify the deliverer to expect this section
+			if unique {
+				select {
+				case <-quit:
+					return
+				case dist <- &request{bit: s.bit, section: section}:
+				}
+			}
+			select {
+			case <-quit:
+				return
+			case pend <- section:
+			}
+		}
+	}
+}
+
+// scheduleDeliveries reads section acceptance notifications and waits for them
+// to be delivered, pushing them into the output data buffer.
+func (s *scheduler) scheduleDeliveries(pend chan uint64, done chan []byte, quit chan struct{}, wg *sync.WaitGroup) {
+	// Clean up the goroutine and pipeline when done
+	defer wg.Done()
+	defer close(done)
+
+	// Keep reading notifications and scheduling deliveries
+	for {
+		select {
+		case <-quit:
+			return
+
+		case idx, ok := <-pend:
+			// New section retrieval pending
+			if !ok {
+				return
+			}
+			// Wait until the request is honoured
+			s.lock.Lock()
+			res := s.responses[idx]
+			s.lock.Unlock()
+
+			select {
+			case <-quit:
+				return
+			case <-res.done:
+			}
+			// Deliver the result
+			select {
+			case <-quit:
+				return
+			case done <- res.cached:
+			}
+		}
+	}
+}
+
+// deliver is called by the request distributor when a reply to a request arrives.
+func (s *scheduler) deliver(sections []uint64, data [][]byte) {
+	s.lock.Lock()
+	defer s.lock.Unlock()
+
+	for i, section := range sections {
+		if res := s.responses[section]; res != nil && res.cached == nil { // Avoid non-requests and double deliveries
+			res.cached = data[i]
+			close(res.done)
+		}
+	}
+}

core/bloombits/scheduler_test.go

@@ -0,0 +1,103 @@
+// Copyright 2017 The go-ethereum Authors
+// This file is part of the go-ethereum library.
+//
+// The go-ethereum library is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Lesser General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// The go-ethereum library is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU Lesser General Public License for more details.
+//
+// You should have received a copy of the GNU Lesser General Public License
+// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
+
+package bloombits
+
+import (
+	"bytes"
+	"math/big"
+	"sync"
+	"sync/atomic"
+	"testing"
+)
+
+// Tests that the scheduler can deduplicate and forward retrieval requests to
+// underlying fetchers and serve responses back, irrelevant of the concurrency
+// of the requesting clients or serving data fetchers.
+func TestSchedulerSingleClientSingleFetcher(t *testing.T) { testScheduler(t, 1, 1, 5000) }
+func TestSchedulerSingleClientMultiFetcher(t *testing.T)  { testScheduler(t, 1, 10, 5000) }
+func TestSchedulerMultiClientSingleFetcher(t *testing.T)  { testScheduler(t, 10, 1, 5000) }
+func TestSchedulerMultiClientMultiFetcher(t *testing.T)   { testScheduler(t, 10, 10, 5000) }
+
+func testScheduler(t *testing.T, clients int, fetchers int, requests int) {
+	t.Parallel()
+	f := newScheduler(0)
+
+	// Create a batch of handler goroutines that respond to bloom bit requests and
+	// deliver them to the scheduler.
+	var fetchPend sync.WaitGroup
+	fetchPend.Add(fetchers)
+	defer fetchPend.Wait()
+
+	fetch := make(chan *request, 16)
+	defer close(fetch)
+
+	var delivered atomic.Uint32
+	for i := 0; i < fetchers; i++ {
+		go func() {
+			defer fetchPend.Done()
+
+			for req := range fetch {
+				delivered.Add(1)
+
+				f.deliver([]uint64{
+					req.section + uint64(requests), // Non-requested data (ensure it doesn't go out of bounds)
+					req.section,                    // Requested data
+					req.section,                    // Duplicated data (ensure it doesn't double close anything)
+				}, [][]byte{
+					{},
+					new(big.Int).SetUint64(req.section).Bytes(),
+					new(big.Int).SetUint64(req.section).Bytes(),
+				})
+			}
+		}()
+	}
+	// Start a batch of goroutines to concurrently run scheduling tasks
+	quit := make(chan struct{})
+
+	var pend sync.WaitGroup
+	pend.Add(clients)
+
+	for i := 0; i < clients; i++ {
+		go func() {
+			defer pend.Done()
+
+			in := make(chan uint64, 16)
+			out := make(chan []byte, 16)
+
+			f.run(in, fetch, out, quit, &pend)
+
+			go func() {
+				for j := 0; j < requests; j++ {
+					in <- uint64(j)
+				}
+				close(in)
+			}()
+			b := new(big.Int)
+			for j := 0; j < requests; j++ {
+				bits := <-out
+				if want := b.SetUint64(uint64(j)).Bytes(); !bytes.Equal(bits, want) {
+					t.Errorf("vector %d: delivered content mismatch: have %x, want %x", j, bits, want)
+				}
+			}
+		}()
+	}
+	pend.Wait()
+
+	if have := delivered.Load(); int(have) != requests {
+		t.Errorf("request count mismatch: have %v, want %v", have, requests)
+	}
+}

core/types/bloom9.go

@@ -161,3 +161,17 @@ func bloomValues(data []byte, hashbuf []byte) (uint, byte, uint, byte, uint, byt
 func BloomLookup(bin Bloom, topic bytesBacked) bool {
 	return bin.Test(topic.Bytes())
 }
+
+// INITIA CUSTOM
+// LogsBloom returns the bloom bytes for the given logs
+func LogsBloom(logs []*Log) []byte {
+	buf := make([]byte, 6)
+	var bin Bloom
+	for _, log := range logs {
+		bin.add(log.Address.Bytes(), buf)
+		for _, b := range log.Topics {
+			bin.add(b[:], buf)
+		}
+	}
+	return bin[:]
+}

core/vm/contracts.go

@@ -47,6 +47,17 @@ type PrecompiledContract interface {
 	Run(input []byte) ([]byte, error) // Run runs the precompiled contract
 }
 
+// INITIA CUSTOM
+// ExtendedPrecompiledContract is the initia custom specific interface for native Go contracts.
+type ExtendedPrecompiledContract interface {
+	ExtendedRun(
+		caller common.Address,
+		input []byte,
+		suppliedGas uint64,
+		readOnly bool,
+	) ([]byte, uint64 /* gas cost */, error)
+}
+
 // PrecompiledContracts contains the precompiled contracts supported at the given fork.
 type PrecompiledContracts map[common.Address]PrecompiledContract
 
@@ -231,6 +242,39 @@ func RunPrecompiledContract(p PrecompiledContract, input []byte, suppliedGas uin
 	return output, suppliedGas, err
 }
 
+// ExtendedRunPrecompiledContract runs and evaluates the output of a precompiled contract.
+// It returns
+// - the returned bytes,
+// - the _remaining_ gas,
+// - any error that occurred
+func ExtendedRunPrecompiledContract(
+	p PrecompiledContract,
+	caller common.Address,
+	input []byte,
+	suppliedGas uint64,
+	readOnly bool,
+	logger *tracing.Hooks,
+) (
+	ret []byte,
+	remainingGas uint64,
+	err error,
+) {
+	if p, ok := p.(ExtendedPrecompiledContract); ok {
+		output, gasCost, err := p.ExtendedRun(caller, input, suppliedGas, readOnly)
+		if err == ErrOutOfGas || suppliedGas < gasCost {
+			return nil, 0, ErrOutOfGas
+		}
+		if logger != nil && logger.OnGasChange != nil {
+			logger.OnGasChange(suppliedGas, suppliedGas-gasCost, tracing.GasChangeCallPrecompiledContract)
+		}
+
+		suppliedGas -= gasCost
+		return output, suppliedGas, err
+	}
+
+	return RunPrecompiledContract(p, input, suppliedGas, logger)
+}
+
 // ecrecover implemented as a native contract.
 type ecrecover struct{}
 

core/vm/evm.go

@@ -125,6 +125,34 @@ type EVM struct {
 	// jumpDests is the aggregated result of JUMPDEST analysis made through
 	// the life cycle of EVM.
 	jumpDests map[common.Hash]bitvec
+
+	// INITIA CUSTOM
+	// disallowCosmosDispatch is a flag that is used to determine if the EVM should disallow Cosmos dispatch
+	disallowCosmosDispatch bool
+}
+
+// INITIA CUSTOM
+// SetDisallowCosmosDispatch sets the disallowCosmosDispatch flag
+func (evm *EVM) SetDisallowCosmosDispatch(value bool) {
+	evm.disallowCosmosDispatch = value
+}
+
+// INITIA CUSTOM
+// GetDisallowCosmosDispatch returns the disallowCosmosDispatch flag
+func (evm *EVM) GetDisallowCosmosDispatch() bool {
+	return evm.disallowCosmosDispatch
+}
+
+// INITIA CUSTOM
+// IncreaseDepth increases the depth of the EVM
+func (evm *EVM) IncreaseDepth() {
+	evm.depth++
+}
+
+// INITIA CUSTOM
+// DecreaseDepth decreases the depth of the EVM
+func (evm *EVM) DecreaseDepth() {
+	evm.depth--
 }
 
 // NewEVM constructs an EVM instance with the supplied block context, state
@@ -224,7 +252,7 @@ func (evm *EVM) Call(caller common.Address, addr common.Address, input []byte, g
 	evm.Context.Transfer(evm.StateDB, caller, addr, value)
 
 	if isPrecompile {
-		ret, gas, err = RunPrecompiledContract(p, input, gas, evm.Config.Tracer)
+		ret, gas, err = ExtendedRunPrecompiledContract(p, caller, input, gas, false, evm.Config.Tracer)
 	} else {
 		// Initialise a new contract and set the code that is to be used by the EVM.
 		code := evm.resolveCode(addr)
@@ -288,7 +316,7 @@ func (evm *EVM) CallCode(caller common.Address, addr common.Address, input []byt
 
 	// It is allowed to call precompiles, even via delegatecall
 	if p, isPrecompile := evm.precompile(addr); isPrecompile {
-		ret, gas, err = RunPrecompiledContract(p, input, gas, evm.Config.Tracer)
+		ret, gas, err = ExtendedRunPrecompiledContract(p, caller, input, gas, false, evm.Config.Tracer)
 	} else {
 		// Initialise a new contract and set the code that is to be used by the EVM.
 		// The contract is a scoped environment for this execution context only.
@@ -331,7 +359,7 @@ func (evm *EVM) DelegateCall(originCaller common.Address, caller common.Address,
 
 	// It is allowed to call precompiles, even via delegatecall
 	if p, isPrecompile := evm.precompile(addr); isPrecompile {
-		ret, gas, err = RunPrecompiledContract(p, input, gas, evm.Config.Tracer)
+		ret, gas, err = ExtendedRunPrecompiledContract(p, caller, input, gas, true, evm.Config.Tracer)
 	} else {
 		// Initialise a new contract and make initialise the delegate values
 		//
@@ -383,7 +411,7 @@ func (evm *EVM) StaticCall(caller common.Address, addr common.Address, input []b
 	evm.StateDB.AddBalance(addr, new(uint256.Int), tracing.BalanceChangeTouchAccount)
 
 	if p, isPrecompile := evm.precompile(addr); isPrecompile {
-		ret, gas, err = RunPrecompiledContract(p, input, gas, evm.Config.Tracer)
+		ret, gas, err = ExtendedRunPrecompiledContract(p, caller, input, gas, true, evm.Config.Tracer)
 	} else {
 		// Initialise a new contract and set the code that is to be used by the EVM.
 		// The contract is a scoped environment for this execution context only.

core/vm/instructions.go

@@ -436,12 +436,20 @@ func opBlockhash(pc *uint64, interpreter *EVMInterpreter, scope *ScopeContext) (
 		return nil, nil
 	}
 
+	// INITIA CUSTOM
+	// number of blocks to retain block hashes is parameterized by the chain config.
+	// - `0` means all block hashes are retained
+	numOfBlocks := uint64(256)
+	if interpreter.evm.Config.NumRetainBlockHashes != nil {
+		numOfBlocks = *interpreter.evm.Config.NumRetainBlockHashes
+	}
+
 	var upper, lower uint64
 	upper = interpreter.evm.Context.BlockNumber.Uint64()
-	if upper < 257 {
+	if numOfBlocks == 0 || upper < numOfBlocks+1 {
 		lower = 0
 	} else {
-		lower = upper - 256
+		lower = upper - numOfBlocks
 	}
 	if num64 >= lower && num64 < upper {
 		res := interpreter.evm.Context.GetHash(num64)

core/vm/interpreter.go

@@ -35,6 +35,9 @@ type Config struct {
 	ExtraEips               []int // Additional EIPS that are to be enabled
 
 	StatelessSelfValidation bool // Generate execution witnesses and self-check against them (testing purpose)
+
+	// INITIA CUSTOM
+	NumRetainBlockHashes *uint64 // Number of block hashes to retain for the BLOCKHASH opcode
 }
 
 // ScopeContext contains the things that are per-call, such as stack and memory,
@@ -164,6 +167,12 @@ func (in *EVMInterpreter) Run(contract *Contract, input []byte, readOnly bool) (
 	in.evm.depth++
 	defer func() { in.evm.depth-- }()
 
+	// INITIA CUSTOM
+	originDisableCosmosDispatch := in.evm.GetDisallowCosmosDispatch()
+	defer func() {
+		in.evm.SetDisallowCosmosDispatch(originDisableCosmosDispatch)
+	}()
+
 	// Make sure the readOnly is only set if we aren't in readOnly yet.
 	// This also makes sure that the readOnly flag isn't removed for child calls.
 	if readOnly && !in.readOnly {