Integrate BACKBEAT SDK and resolve KACHING license validation

Major integrations and fixes:
- Added BACKBEAT SDK integration for P2P operation timing
- Implemented beat-aware status tracking for distributed operations
- Added Docker secrets support for secure license management
- Resolved KACHING license validation via HTTPS/TLS
- Updated docker-compose configuration for clean stack deployment
- Disabled rollback policies to prevent deployment failures
- Added license credential storage (CHORUS-DEV-MULTI-001)

Technical improvements:
- BACKBEAT P2P operation tracking with phase management
- Enhanced configuration system with file-based secrets
- Improved error handling for license validation
- Clean separation of KACHING and CHORUS deployment stacks

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

vendor/go.etcd.io/bbolt/.gitignore

@@ -0,0 +1,12 @@
+*.prof
+*.test
+*.swp
+/bin/
+cover.out
+cover-*.out
+/.idea
+*.iml
+/bbolt
+/cmd/bbolt/bbolt
+.DS_Store
+

vendor/go.etcd.io/bbolt/.go-version

@@ -0,0 +1 @@
+1.23.6

vendor/go.etcd.io/bbolt/LICENSE

@@ -0,0 +1,20 @@
+The MIT License (MIT)
+
+Copyright (c) 2013 Ben Johnson
+
+Permission is hereby granted, free of charge, to any person obtaining a copy of
+this software and associated documentation files (the "Software"), to deal in
+the Software without restriction, including without limitation the rights to
+use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of
+the Software, and to permit persons to whom the Software is furnished to do so,
+subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS
+FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
+COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
+IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

vendor/go.etcd.io/bbolt/Makefile

@@ -0,0 +1,108 @@
+BRANCH=`git rev-parse --abbrev-ref HEAD`
+COMMIT=`git rev-parse --short HEAD`
+GOLDFLAGS="-X main.branch $(BRANCH) -X main.commit $(COMMIT)"
+GOFILES = $(shell find . -name \*.go)
+
+TESTFLAGS_RACE=-race=false
+ifdef ENABLE_RACE
+	TESTFLAGS_RACE=-race=true
+endif
+
+TESTFLAGS_CPU=
+ifdef CPU
+	TESTFLAGS_CPU=-cpu=$(CPU)
+endif
+TESTFLAGS = $(TESTFLAGS_RACE) $(TESTFLAGS_CPU) $(EXTRA_TESTFLAGS)
+
+TESTFLAGS_TIMEOUT=30m
+ifdef TIMEOUT
+	TESTFLAGS_TIMEOUT=$(TIMEOUT)
+endif
+
+TESTFLAGS_ENABLE_STRICT_MODE=false
+ifdef ENABLE_STRICT_MODE
+	TESTFLAGS_ENABLE_STRICT_MODE=$(ENABLE_STRICT_MODE)
+endif
+
+.EXPORT_ALL_VARIABLES:
+TEST_ENABLE_STRICT_MODE=${TESTFLAGS_ENABLE_STRICT_MODE}
+
+.PHONY: fmt
+fmt:
+	@echo "Verifying gofmt, failures can be fixed with ./scripts/fix.sh"
+	@!(gofmt -l -s -d ${GOFILES} | grep '[a-z]')
+
+	@echo "Verifying goimports, failures can be fixed with ./scripts/fix.sh"
+	@!(go run golang.org/x/tools/cmd/goimports@latest -l -d ${GOFILES} | grep '[a-z]')
+
+.PHONY: lint
+lint:
+	golangci-lint run ./...
+
+.PHONY: test
+test:
+	@echo "hashmap freelist test"
+	BBOLT_VERIFY=all TEST_FREELIST_TYPE=hashmap go test -v ${TESTFLAGS} -timeout ${TESTFLAGS_TIMEOUT}
+	BBOLT_VERIFY=all TEST_FREELIST_TYPE=hashmap go test -v ${TESTFLAGS} ./internal/...
+	BBOLT_VERIFY=all TEST_FREELIST_TYPE=hashmap go test -v ${TESTFLAGS} ./cmd/bbolt
+
+	@echo "array freelist test"
+	BBOLT_VERIFY=all TEST_FREELIST_TYPE=array go test -v ${TESTFLAGS} -timeout ${TESTFLAGS_TIMEOUT}
+	BBOLT_VERIFY=all TEST_FREELIST_TYPE=array go test -v ${TESTFLAGS} ./internal/...
+	BBOLT_VERIFY=all TEST_FREELIST_TYPE=array go test -v ${TESTFLAGS} ./cmd/bbolt
+
+.PHONY: coverage
+coverage:
+	@echo "hashmap freelist test"
+	TEST_FREELIST_TYPE=hashmap go test -v -timeout ${TESTFLAGS_TIMEOUT} \
+		-coverprofile cover-freelist-hashmap.out -covermode atomic
+
+	@echo "array freelist test"
+	TEST_FREELIST_TYPE=array go test -v -timeout ${TESTFLAGS_TIMEOUT} \
+		-coverprofile cover-freelist-array.out -covermode atomic
+
+BOLT_CMD=bbolt
+
+build:
+	go build -o bin/${BOLT_CMD} ./cmd/${BOLT_CMD}
+
+.PHONY: clean
+clean: # Clean binaries
+	rm -f ./bin/${BOLT_CMD}
+
+.PHONY: gofail-enable
+gofail-enable: install-gofail
+	gofail enable .
+
+.PHONY: gofail-disable
+gofail-disable: install-gofail
+	gofail disable .
+
+.PHONY: install-gofail
+install-gofail:
+	go install go.etcd.io/gofail
+
+.PHONY: test-failpoint
+test-failpoint:
+	@echo "[failpoint] hashmap freelist test"
+	BBOLT_VERIFY=all TEST_FREELIST_TYPE=hashmap go test -v ${TESTFLAGS} -timeout 30m ./tests/failpoint
+
+	@echo "[failpoint] array freelist test"
+	BBOLT_VERIFY=all TEST_FREELIST_TYPE=array go test -v ${TESTFLAGS} -timeout 30m ./tests/failpoint
+
+.PHONY: test-robustness # Running robustness tests requires root permission for now
+# TODO: Remove sudo once we fully migrate to the prow infrastructure
+test-robustness: gofail-enable build
+	sudo env PATH=$$PATH go test -v ${TESTFLAGS} ./tests/dmflakey -test.root
+	sudo env PATH=$(PWD)/bin:$$PATH go test -v ${TESTFLAGS} ${ROBUSTNESS_TESTFLAGS} ./tests/robustness -test.root
+
+.PHONY: test-benchmark-compare
+# Runs benchmark tests on the current git ref and the given REF, and compares
+# the two.
+test-benchmark-compare: install-benchstat
+	@git fetch
+	./scripts/compare_benchmarks.sh $(REF)
+
+.PHONY: install-benchstat
+install-benchstat:
+	go install golang.org/x/perf/cmd/benchstat@latest

vendor/go.etcd.io/bbolt/OWNERS

@@ -0,0 +1,10 @@
+# See the OWNERS docs at https://go.k8s.io/owners
+
+approvers:
+  - ahrtr           # Benjamin Wang <benjamin.ahrtr@gmail.com> <benjamin.wang@broadcom.com>
+  - serathius       # Marek Siarkowicz <siarkowicz@google.com> <marek.siarkowicz@gmail.com>
+  - ptabor          # Piotr Tabor <piotr.tabor@gmail.com>
+  - spzala          # Sahdev Zala <spzala@us.ibm.com>
+reviewers:
+  - fuweid          # Wei Fu <fuweid89@gmail.com>
+  - tjungblu        # Thomas Jungblut <tjungblu@redhat.com>

vendor/go.etcd.io/bbolt/bolt_386.go

@@ -0,0 +1,7 @@
+package bbolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0x7FFFFFFF // 2GB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0xFFFFFFF

vendor/go.etcd.io/bbolt/bolt_aix.go

@@ -0,0 +1,90 @@
+//go:build aix
+
+package bbolt
+
+import (
+	"fmt"
+	"syscall"
+	"time"
+	"unsafe"
+
+	"golang.org/x/sys/unix"
+)
+
+// flock acquires an advisory lock on a file descriptor.
+func flock(db *DB, exclusive bool, timeout time.Duration) error {
+	var t time.Time
+	if timeout != 0 {
+		t = time.Now()
+	}
+	fd := db.file.Fd()
+	var lockType int16
+	if exclusive {
+		lockType = syscall.F_WRLCK
+	} else {
+		lockType = syscall.F_RDLCK
+	}
+	for {
+		// Attempt to obtain an exclusive lock.
+		lock := syscall.Flock_t{Type: lockType}
+		err := syscall.FcntlFlock(fd, syscall.F_SETLK, &lock)
+		if err == nil {
+			return nil
+		} else if err != syscall.EAGAIN {
+			return err
+		}
+
+		// If we timed out then return an error.
+		if timeout != 0 && time.Since(t) > timeout-flockRetryTimeout {
+			return ErrTimeout
+		}
+
+		// Wait for a bit and try again.
+		time.Sleep(flockRetryTimeout)
+	}
+}
+
+// funlock releases an advisory lock on a file descriptor.
+func funlock(db *DB) error {
+	var lock syscall.Flock_t
+	lock.Start = 0
+	lock.Len = 0
+	lock.Type = syscall.F_UNLCK
+	lock.Whence = 0
+	return syscall.FcntlFlock(uintptr(db.file.Fd()), syscall.F_SETLK, &lock)
+}
+
+// mmap memory maps a DB's data file.
+func mmap(db *DB, sz int) error {
+	// Map the data file to memory.
+	b, err := unix.Mmap(int(db.file.Fd()), 0, sz, syscall.PROT_READ, syscall.MAP_SHARED|db.MmapFlags)
+	if err != nil {
+		return err
+	}
+
+	// Advise the kernel that the mmap is accessed randomly.
+	if err := unix.Madvise(b, syscall.MADV_RANDOM); err != nil {
+		return fmt.Errorf("madvise: %s", err)
+	}
+
+	// Save the original byte slice and convert to a byte array pointer.
+	db.dataref = b
+	db.data = (*[maxMapSize]byte)(unsafe.Pointer(&b[0]))
+	db.datasz = sz
+	return nil
+}
+
+// munmap unmaps a DB's data file from memory.
+func munmap(db *DB) error {
+	// Ignore the unmap if we have no mapped data.
+	if db.dataref == nil {
+		return nil
+	}
+
+	// Unmap using the original byte slice.
+	err := unix.Munmap(db.dataref)
+	db.dataref = nil
+	db.data = nil
+	db.datasz = 0
+	return err
+}

vendor/go.etcd.io/bbolt/bolt_amd64.go

@@ -0,0 +1,7 @@
+package bbolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0xFFFFFFFFFFFF // 256TB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0x7FFFFFFF

vendor/go.etcd.io/bbolt/bolt_android.go

@@ -0,0 +1,90 @@
+package bbolt
+
+import (
+	"fmt"
+	"syscall"
+	"time"
+	"unsafe"
+
+	"golang.org/x/sys/unix"
+)
+
+// flock acquires an advisory lock on a file descriptor.
+func flock(db *DB, exclusive bool, timeout time.Duration) error {
+	var t time.Time
+	if timeout != 0 {
+		t = time.Now()
+	}
+	fd := db.file.Fd()
+	var lockType int16
+	if exclusive {
+		lockType = syscall.F_WRLCK
+	} else {
+		lockType = syscall.F_RDLCK
+	}
+	for {
+		// Attempt to obtain an exclusive lock.
+		lock := syscall.Flock_t{Type: lockType}
+		err := syscall.FcntlFlock(fd, syscall.F_SETLK, &lock)
+		if err == nil {
+			return nil
+		} else if err != syscall.EAGAIN {
+			return err
+		}
+
+		// If we timed out then return an error.
+		if timeout != 0 && time.Since(t) > timeout-flockRetryTimeout {
+			return ErrTimeout
+		}
+
+		// Wait for a bit and try again.
+		time.Sleep(flockRetryTimeout)
+	}
+}
+
+// funlock releases an advisory lock on a file descriptor.
+func funlock(db *DB) error {
+	var lock syscall.Flock_t
+	lock.Start = 0
+	lock.Len = 0
+	lock.Type = syscall.F_UNLCK
+	lock.Whence = 0
+	return syscall.FcntlFlock(uintptr(db.file.Fd()), syscall.F_SETLK, &lock)
+}
+
+// mmap memory maps a DB's data file.
+func mmap(db *DB, sz int) error {
+	// Map the data file to memory.
+	b, err := unix.Mmap(int(db.file.Fd()), 0, sz, syscall.PROT_READ, syscall.MAP_SHARED|db.MmapFlags)
+	if err != nil {
+		return err
+	}
+
+	// Advise the kernel that the mmap is accessed randomly.
+	err = unix.Madvise(b, syscall.MADV_RANDOM)
+	if err != nil && err != syscall.ENOSYS {
+		// Ignore not implemented error in kernel because it still works.
+		return fmt.Errorf("madvise: %s", err)
+	}
+
+	// Save the original byte slice and convert to a byte array pointer.
+	db.dataref = b
+	db.data = (*[maxMapSize]byte)(unsafe.Pointer(&b[0]))
+	db.datasz = sz
+	return nil
+}
+
+// munmap unmaps a DB's data file from memory.
+func munmap(db *DB) error {
+	// Ignore the unmap if we have no mapped data.
+	if db.dataref == nil {
+		return nil
+	}
+
+	// Unmap using the original byte slice.
+	err := unix.Munmap(db.dataref)
+	db.dataref = nil
+	db.data = nil
+	db.datasz = 0
+	return err
+}

vendor/go.etcd.io/bbolt/bolt_arm.go

@@ -0,0 +1,7 @@
+package bbolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0x7FFFFFFF // 2GB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0xFFFFFFF

vendor/go.etcd.io/bbolt/bolt_arm64.go

@@ -0,0 +1,9 @@
+//go:build arm64
+
+package bbolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0xFFFFFFFFFFFF // 256TB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0x7FFFFFFF

vendor/go.etcd.io/bbolt/bolt_linux.go

@@ -0,0 +1,10 @@
+package bbolt
+
+import (
+	"syscall"
+)
+
+// fdatasync flushes written data to a file descriptor.
+func fdatasync(db *DB) error {
+	return syscall.Fdatasync(int(db.file.Fd()))
+}

vendor/go.etcd.io/bbolt/bolt_loong64.go

@@ -0,0 +1,9 @@
+//go:build loong64
+
+package bbolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0xFFFFFFFFFFFF // 256TB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0x7FFFFFFF

vendor/go.etcd.io/bbolt/bolt_mips64x.go

@@ -0,0 +1,9 @@
+//go:build mips64 || mips64le
+
+package bbolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0x8000000000 // 512GB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0x7FFFFFFF

vendor/go.etcd.io/bbolt/bolt_mipsx.go

@@ -0,0 +1,9 @@
+//go:build mips || mipsle
+
+package bbolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0x40000000 // 1GB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0xFFFFFFF

vendor/go.etcd.io/bbolt/bolt_openbsd.go

@@ -0,0 +1,16 @@
+package bbolt
+
+import (
+	"golang.org/x/sys/unix"
+)
+
+func msync(db *DB) error {
+	return unix.Msync(db.data[:db.datasz], unix.MS_INVALIDATE)
+}
+
+func fdatasync(db *DB) error {
+	if db.data != nil {
+		return msync(db)
+	}
+	return db.file.Sync()
+}

vendor/go.etcd.io/bbolt/bolt_ppc.go

@@ -0,0 +1,9 @@
+//go:build ppc
+
+package bbolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0x7FFFFFFF // 2GB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0xFFFFFFF

vendor/go.etcd.io/bbolt/bolt_ppc64.go

@@ -0,0 +1,9 @@
+//go:build ppc64
+
+package bbolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0xFFFFFFFFFFFF // 256TB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0x7FFFFFFF

vendor/go.etcd.io/bbolt/bolt_ppc64le.go

@@ -0,0 +1,9 @@
+//go:build ppc64le
+
+package bbolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0xFFFFFFFFFFFF // 256TB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0x7FFFFFFF

vendor/go.etcd.io/bbolt/bolt_riscv64.go

@@ -0,0 +1,9 @@
+//go:build riscv64
+
+package bbolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0xFFFFFFFFFFFF // 256TB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0x7FFFFFFF

vendor/go.etcd.io/bbolt/bolt_s390x.go

@@ -0,0 +1,9 @@
+//go:build s390x
+
+package bbolt
+
+// maxMapSize represents the largest mmap size supported by Bolt.
+const maxMapSize = 0xFFFFFFFFFFFF // 256TB
+
+// maxAllocSize is the size used when creating array pointers.
+const maxAllocSize = 0x7FFFFFFF

vendor/go.etcd.io/bbolt/bolt_solaris.go

@@ -0,0 +1,88 @@
+package bbolt
+
+import (
+	"fmt"
+	"syscall"
+	"time"
+	"unsafe"
+
+	"golang.org/x/sys/unix"
+)
+
+// flock acquires an advisory lock on a file descriptor.
+func flock(db *DB, exclusive bool, timeout time.Duration) error {
+	var t time.Time
+	if timeout != 0 {
+		t = time.Now()
+	}
+	fd := db.file.Fd()
+	var lockType int16
+	if exclusive {
+		lockType = syscall.F_WRLCK
+	} else {
+		lockType = syscall.F_RDLCK
+	}
+	for {
+		// Attempt to obtain an exclusive lock.
+		lock := syscall.Flock_t{Type: lockType}
+		err := syscall.FcntlFlock(fd, syscall.F_SETLK, &lock)
+		if err == nil {
+			return nil
+		} else if err != syscall.EAGAIN {
+			return err
+		}
+
+		// If we timed out then return an error.
+		if timeout != 0 && time.Since(t) > timeout-flockRetryTimeout {
+			return ErrTimeout
+		}
+
+		// Wait for a bit and try again.
+		time.Sleep(flockRetryTimeout)
+	}
+}
+
+// funlock releases an advisory lock on a file descriptor.
+func funlock(db *DB) error {
+	var lock syscall.Flock_t
+	lock.Start = 0
+	lock.Len = 0
+	lock.Type = syscall.F_UNLCK
+	lock.Whence = 0
+	return syscall.FcntlFlock(uintptr(db.file.Fd()), syscall.F_SETLK, &lock)
+}
+
+// mmap memory maps a DB's data file.
+func mmap(db *DB, sz int) error {
+	// Map the data file to memory.
+	b, err := unix.Mmap(int(db.file.Fd()), 0, sz, syscall.PROT_READ, syscall.MAP_SHARED|db.MmapFlags)
+	if err != nil {
+		return err
+	}
+
+	// Advise the kernel that the mmap is accessed randomly.
+	if err := unix.Madvise(b, syscall.MADV_RANDOM); err != nil {
+		return fmt.Errorf("madvise: %s", err)
+	}
+
+	// Save the original byte slice and convert to a byte array pointer.
+	db.dataref = b
+	db.data = (*[maxMapSize]byte)(unsafe.Pointer(&b[0]))
+	db.datasz = sz
+	return nil
+}
+
+// munmap unmaps a DB's data file from memory.
+func munmap(db *DB) error {
+	// Ignore the unmap if we have no mapped data.
+	if db.dataref == nil {
+		return nil
+	}
+
+	// Unmap using the original byte slice.
+	err := unix.Munmap(db.dataref)
+	db.dataref = nil
+	db.data = nil
+	db.datasz = 0
+	return err
+}

vendor/go.etcd.io/bbolt/bolt_unix.go

@@ -0,0 +1,88 @@
+//go:build !windows && !plan9 && !solaris && !aix && !android
+
+package bbolt
+
+import (
+	"fmt"
+	"syscall"
+	"time"
+	"unsafe"
+
+	"golang.org/x/sys/unix"
+
+	"go.etcd.io/bbolt/errors"
+)
+
+// flock acquires an advisory lock on a file descriptor.
+func flock(db *DB, exclusive bool, timeout time.Duration) error {
+	var t time.Time
+	if timeout != 0 {
+		t = time.Now()
+	}
+	fd := db.file.Fd()
+	flag := syscall.LOCK_NB
+	if exclusive {
+		flag |= syscall.LOCK_EX
+	} else {
+		flag |= syscall.LOCK_SH
+	}
+	for {
+		// Attempt to obtain an exclusive lock.
+		err := syscall.Flock(int(fd), flag)
+		if err == nil {
+			return nil
+		} else if err != syscall.EWOULDBLOCK {
+			return err
+		}
+
+		// If we timed out then return an error.
+		if timeout != 0 && time.Since(t) > timeout-flockRetryTimeout {
+			return errors.ErrTimeout
+		}
+
+		// Wait for a bit and try again.
+		time.Sleep(flockRetryTimeout)
+	}
+}
+
+// funlock releases an advisory lock on a file descriptor.
+func funlock(db *DB) error {
+	return syscall.Flock(int(db.file.Fd()), syscall.LOCK_UN)
+}
+
+// mmap memory maps a DB's data file.
+func mmap(db *DB, sz int) error {
+	// Map the data file to memory.
+	b, err := unix.Mmap(int(db.file.Fd()), 0, sz, syscall.PROT_READ, syscall.MAP_SHARED|db.MmapFlags)
+	if err != nil {
+		return err
+	}
+
+	// Advise the kernel that the mmap is accessed randomly.
+	err = unix.Madvise(b, syscall.MADV_RANDOM)
+	if err != nil && err != syscall.ENOSYS {
+		// Ignore not implemented error in kernel because it still works.
+		return fmt.Errorf("madvise: %s", err)
+	}
+
+	// Save the original byte slice and convert to a byte array pointer.
+	db.dataref = b
+	db.data = (*[maxMapSize]byte)(unsafe.Pointer(&b[0]))
+	db.datasz = sz
+	return nil
+}
+
+// munmap unmaps a DB's data file from memory.
+func munmap(db *DB) error {
+	// Ignore the unmap if we have no mapped data.
+	if db.dataref == nil {
+		return nil
+	}
+
+	// Unmap using the original byte slice.
+	err := unix.Munmap(db.dataref)
+	db.dataref = nil
+	db.data = nil
+	db.datasz = 0
+	return err
+}

vendor/go.etcd.io/bbolt/bolt_windows.go

@@ -0,0 +1,119 @@
+package bbolt
+
+import (
+	"fmt"
+	"os"
+	"syscall"
+	"time"
+	"unsafe"
+
+	"golang.org/x/sys/windows"
+
+	"go.etcd.io/bbolt/errors"
+)
+
+// fdatasync flushes written data to a file descriptor.
+func fdatasync(db *DB) error {
+	return db.file.Sync()
+}
+
+// flock acquires an advisory lock on a file descriptor.
+func flock(db *DB, exclusive bool, timeout time.Duration) error {
+	var t time.Time
+	if timeout != 0 {
+		t = time.Now()
+	}
+	var flags uint32 = windows.LOCKFILE_FAIL_IMMEDIATELY
+	if exclusive {
+		flags |= windows.LOCKFILE_EXCLUSIVE_LOCK
+	}
+	for {
+		// Fix for https://github.com/etcd-io/bbolt/issues/121. Use byte-range
+		// -1..0 as the lock on the database file.
+		var m1 uint32 = (1 << 32) - 1 // -1 in a uint32
+		err := windows.LockFileEx(windows.Handle(db.file.Fd()), flags, 0, 1, 0, &windows.Overlapped{
+			Offset:     m1,
+			OffsetHigh: m1,
+		})
+
+		if err == nil {
+			return nil
+		} else if err != windows.ERROR_LOCK_VIOLATION {
+			return err
+		}
+
+		// If we timed oumercit then return an error.
+		if timeout != 0 && time.Since(t) > timeout-flockRetryTimeout {
+			return errors.ErrTimeout
+		}
+
+		// Wait for a bit and try again.
+		time.Sleep(flockRetryTimeout)
+	}
+}
+
+// funlock releases an advisory lock on a file descriptor.
+func funlock(db *DB) error {
+	var m1 uint32 = (1 << 32) - 1 // -1 in a uint32
+	return windows.UnlockFileEx(windows.Handle(db.file.Fd()), 0, 1, 0, &windows.Overlapped{
+		Offset:     m1,
+		OffsetHigh: m1,
+	})
+}
+
+// mmap memory maps a DB's data file.
+// Based on: https://github.com/edsrzf/mmap-go
+func mmap(db *DB, sz int) error {
+	var sizelo, sizehi uint32
+
+	if !db.readOnly {
+		// Truncate the database to the size of the mmap.
+		if err := db.file.Truncate(int64(sz)); err != nil {
+			return fmt.Errorf("truncate: %s", err)
+		}
+		sizehi = uint32(sz >> 32)
+		sizelo = uint32(sz)
+	}
+
+	// Open a file mapping handle.
+	h, errno := syscall.CreateFileMapping(syscall.Handle(db.file.Fd()), nil, syscall.PAGE_READONLY, sizehi, sizelo, nil)
+	if h == 0 {
+		return os.NewSyscallError("CreateFileMapping", errno)
+	}
+
+	// Create the memory map.
+	addr, errno := syscall.MapViewOfFile(h, syscall.FILE_MAP_READ, 0, 0, 0)
+	if addr == 0 {
+		// Do our best and report error returned from MapViewOfFile.
+		_ = syscall.CloseHandle(h)
+		return os.NewSyscallError("MapViewOfFile", errno)
+	}
+
+	// Close mapping handle.
+	if err := syscall.CloseHandle(syscall.Handle(h)); err != nil {
+		return os.NewSyscallError("CloseHandle", err)
+	}
+
+	// Convert to a byte array.
+	db.data = (*[maxMapSize]byte)(unsafe.Pointer(addr))
+	db.datasz = sz
+
+	return nil
+}
+
+// munmap unmaps a pointer from a file.
+// Based on: https://github.com/edsrzf/mmap-go
+func munmap(db *DB) error {
+	if db.data == nil {
+		return nil
+	}
+
+	addr := (uintptr)(unsafe.Pointer(&db.data[0]))
+	var err1 error
+	if err := syscall.UnmapViewOfFile(addr); err != nil {
+		err1 = os.NewSyscallError("UnmapViewOfFile", err)
+	}
+	db.data = nil
+	db.datasz = 0
+	return err1
+}

vendor/go.etcd.io/bbolt/boltsync_unix.go

@@ -0,0 +1,8 @@
+//go:build !windows && !plan9 && !linux && !openbsd
+
+package bbolt
+
+// fdatasync flushes written data to a file descriptor.
+func fdatasync(db *DB) error {
+	return db.file.Sync()
+}

vendor/go.etcd.io/bbolt/compact.go

@@ -0,0 +1,119 @@
+package bbolt
+
+// Compact will create a copy of the source DB and in the destination DB. This may
+// reclaim space that the source database no longer has use for. txMaxSize can be
+// used to limit the transactions size of this process and may trigger intermittent
+// commits. A value of zero will ignore transaction sizes.
+// TODO: merge with: https://github.com/etcd-io/etcd/blob/b7f0f52a16dbf83f18ca1d803f7892d750366a94/mvcc/backend/backend.go#L349
+func Compact(dst, src *DB, txMaxSize int64) error {
+	// commit regularly, or we'll run out of memory for large datasets if using one transaction.
+	var size int64
+	tx, err := dst.Begin(true)
+	if err != nil {
+		return err
+	}
+	defer func() {
+		if tempErr := tx.Rollback(); tempErr != nil {
+			err = tempErr
+		}
+	}()
+
+	if err := walk(src, func(keys [][]byte, k, v []byte, seq uint64) error {
+		// On each key/value, check if we have exceeded tx size.
+		sz := int64(len(k) + len(v))
+		if size+sz > txMaxSize && txMaxSize != 0 {
+			// Commit previous transaction.
+			if err := tx.Commit(); err != nil {
+				return err
+			}
+
+			// Start new transaction.
+			tx, err = dst.Begin(true)
+			if err != nil {
+				return err
+			}
+			size = 0
+		}
+		size += sz
+
+		// Create bucket on the root transaction if this is the first level.
+		nk := len(keys)
+		if nk == 0 {
+			bkt, err := tx.CreateBucket(k)
+			if err != nil {
+				return err
+			}
+			if err := bkt.SetSequence(seq); err != nil {
+				return err
+			}
+			return nil
+		}
+
+		// Create buckets on subsequent levels, if necessary.
+		b := tx.Bucket(keys[0])
+		if nk > 1 {
+			for _, k := range keys[1:] {
+				b = b.Bucket(k)
+			}
+		}
+
+		// Fill the entire page for best compaction.
+		b.FillPercent = 1.0
+
+		// If there is no value then this is a bucket call.
+		if v == nil {
+			bkt, err := b.CreateBucket(k)
+			if err != nil {
+				return err
+			}
+			if err := bkt.SetSequence(seq); err != nil {
+				return err
+			}
+			return nil
+		}
+
+		// Otherwise treat it as a key/value pair.
+		return b.Put(k, v)
+	}); err != nil {
+		return err
+	}
+	err = tx.Commit()
+
+	return err
+}
+
+// walkFunc is the type of the function called for keys (buckets and "normal"
+// values) discovered by Walk. keys is the list of keys to descend to the bucket
+// owning the discovered key/value pair k/v.
+type walkFunc func(keys [][]byte, k, v []byte, seq uint64) error
+
+// walk walks recursively the bolt database db, calling walkFn for each key it finds.
+func walk(db *DB, walkFn walkFunc) error {
+	return db.View(func(tx *Tx) error {
+		return tx.ForEach(func(name []byte, b *Bucket) error {
+			return walkBucket(b, nil, name, nil, b.Sequence(), walkFn)
+		})
+	})
+}
+
+func walkBucket(b *Bucket, keypath [][]byte, k, v []byte, seq uint64, fn walkFunc) error {
+	// Execute callback.
+	if err := fn(keypath, k, v, seq); err != nil {
+		return err
+	}
+
+	// If this is not a bucket then stop.
+	if v != nil {
+		return nil
+	}
+
+	// Iterate over each child key/value.
+	keypath = append(keypath, k)
+	return b.ForEach(func(k, v []byte) error {
+		if v == nil {
+			bkt := b.Bucket(k)
+			return walkBucket(bkt, keypath, k, nil, bkt.Sequence(), fn)
+		}
+		return walkBucket(b, keypath, k, v, b.Sequence(), fn)
+	})
+}

vendor/go.etcd.io/bbolt/cursor.go

@@ -0,0 +1,432 @@
+package bbolt
+
+import (
+	"bytes"
+	"fmt"
+	"sort"
+
+	"go.etcd.io/bbolt/errors"
+	"go.etcd.io/bbolt/internal/common"
+)
+
+// Cursor represents an iterator that can traverse over all key/value pairs in a bucket
+// in lexicographical order.
+// Cursors see nested buckets with value == nil.
+// Cursors can be obtained from a transaction and are valid as long as the transaction is open.
+//
+// Keys and values returned from the cursor are only valid for the life of the transaction.
+//
+// Changing data while traversing with a cursor may cause it to be invalidated
+// and return unexpected keys and/or values. You must reposition your cursor
+// after mutating data.
+type Cursor struct {
+	bucket *Bucket
+	stack  []elemRef
+}
+
+// Bucket returns the bucket that this cursor was created from.
+func (c *Cursor) Bucket() *Bucket {
+	return c.bucket
+}
+
+// First moves the cursor to the first item in the bucket and returns its key and value.
+// If the bucket is empty then a nil key and value are returned.
+// The returned key and value are only valid for the life of the transaction.
+func (c *Cursor) First() (key []byte, value []byte) {
+	common.Assert(c.bucket.tx.db != nil, "tx closed")
+	k, v, flags := c.first()
+	if (flags & uint32(common.BucketLeafFlag)) != 0 {
+		return k, nil
+	}
+	return k, v
+}
+
+func (c *Cursor) first() (key []byte, value []byte, flags uint32) {
+	c.stack = c.stack[:0]
+	p, n := c.bucket.pageNode(c.bucket.RootPage())
+	c.stack = append(c.stack, elemRef{page: p, node: n, index: 0})
+	c.goToFirstElementOnTheStack()
+
+	// If we land on an empty page then move to the next value.
+	// https://github.com/boltdb/bolt/issues/450
+	if c.stack[len(c.stack)-1].count() == 0 {
+		c.next()
+	}
+
+	k, v, flags := c.keyValue()
+	if (flags & uint32(common.BucketLeafFlag)) != 0 {
+		return k, nil, flags
+	}
+	return k, v, flags
+}
+
+// Last moves the cursor to the last item in the bucket and returns its key and value.
+// If the bucket is empty then a nil key and value are returned.
+// The returned key and value are only valid for the life of the transaction.
+func (c *Cursor) Last() (key []byte, value []byte) {
+	common.Assert(c.bucket.tx.db != nil, "tx closed")
+	c.stack = c.stack[:0]
+	p, n := c.bucket.pageNode(c.bucket.RootPage())
+	ref := elemRef{page: p, node: n}
+	ref.index = ref.count() - 1
+	c.stack = append(c.stack, ref)
+	c.last()
+
+	// If this is an empty page (calling Delete may result in empty pages)
+	// we call prev to find the last page that is not empty
+	for len(c.stack) > 1 && c.stack[len(c.stack)-1].count() == 0 {
+		c.prev()
+	}
+
+	if len(c.stack) == 0 {
+		return nil, nil
+	}
+
+	k, v, flags := c.keyValue()
+	if (flags & uint32(common.BucketLeafFlag)) != 0 {
+		return k, nil
+	}
+	return k, v
+}
+
+// Next moves the cursor to the next item in the bucket and returns its key and value.
+// If the cursor is at the end of the bucket then a nil key and value are returned.
+// The returned key and value are only valid for the life of the transaction.
+func (c *Cursor) Next() (key []byte, value []byte) {
+	common.Assert(c.bucket.tx.db != nil, "tx closed")
+	k, v, flags := c.next()
+	if (flags & uint32(common.BucketLeafFlag)) != 0 {
+		return k, nil
+	}
+	return k, v
+}
+
+// Prev moves the cursor to the previous item in the bucket and returns its key and value.
+// If the cursor is at the beginning of the bucket then a nil key and value are returned.
+// The returned key and value are only valid for the life of the transaction.
+func (c *Cursor) Prev() (key []byte, value []byte) {
+	common.Assert(c.bucket.tx.db != nil, "tx closed")
+	k, v, flags := c.prev()
+	if (flags & uint32(common.BucketLeafFlag)) != 0 {
+		return k, nil
+	}
+	return k, v
+}
+
+// Seek moves the cursor to a given key using a b-tree search and returns it.
+// If the key does not exist then the next key is used. If no keys
+// follow, a nil key is returned.
+// The returned key and value are only valid for the life of the transaction.
+func (c *Cursor) Seek(seek []byte) (key []byte, value []byte) {
+	common.Assert(c.bucket.tx.db != nil, "tx closed")
+
+	k, v, flags := c.seek(seek)
+
+	// If we ended up after the last element of a page then move to the next one.
+	if ref := &c.stack[len(c.stack)-1]; ref.index >= ref.count() {
+		k, v, flags = c.next()
+	}
+
+	if k == nil {
+		return nil, nil
+	} else if (flags & uint32(common.BucketLeafFlag)) != 0 {
+		return k, nil
+	}
+	return k, v
+}
+
+// Delete removes the current key/value under the cursor from the bucket.
+// Delete fails if current key/value is a bucket or if the transaction is not writable.
+func (c *Cursor) Delete() error {
+	if c.bucket.tx.db == nil {
+		return errors.ErrTxClosed
+	} else if !c.bucket.Writable() {
+		return errors.ErrTxNotWritable
+	}
+
+	key, _, flags := c.keyValue()
+	// Return an error if current value is a bucket.
+	if (flags & common.BucketLeafFlag) != 0 {
+		return errors.ErrIncompatibleValue
+	}
+	c.node().del(key)
+
+	return nil
+}
+
+// seek moves the cursor to a given key and returns it.
+// If the key does not exist then the next key is used.
+func (c *Cursor) seek(seek []byte) (key []byte, value []byte, flags uint32) {
+	// Start from root page/node and traverse to correct page.
+	c.stack = c.stack[:0]
+	c.search(seek, c.bucket.RootPage())
+
+	// If this is a bucket then return a nil value.
+	return c.keyValue()
+}
+
+// first moves the cursor to the first leaf element under the last page in the stack.
+func (c *Cursor) goToFirstElementOnTheStack() {
+	for {
+		// Exit when we hit a leaf page.
+		var ref = &c.stack[len(c.stack)-1]
+		if ref.isLeaf() {
+			break
+		}
+
+		// Keep adding pages pointing to the first element to the stack.
+		var pgId common.Pgid
+		if ref.node != nil {
+			pgId = ref.node.inodes[ref.index].Pgid()
+		} else {
+			pgId = ref.page.BranchPageElement(uint16(ref.index)).Pgid()
+		}
+		p, n := c.bucket.pageNode(pgId)
+		c.stack = append(c.stack, elemRef{page: p, node: n, index: 0})
+	}
+}
+
+// last moves the cursor to the last leaf element under the last page in the stack.
+func (c *Cursor) last() {
+	for {
+		// Exit when we hit a leaf page.
+		ref := &c.stack[len(c.stack)-1]
+		if ref.isLeaf() {
+			break
+		}
+
+		// Keep adding pages pointing to the last element in the stack.
+		var pgId common.Pgid
+		if ref.node != nil {
+			pgId = ref.node.inodes[ref.index].Pgid()
+		} else {
+			pgId = ref.page.BranchPageElement(uint16(ref.index)).Pgid()
+		}
+		p, n := c.bucket.pageNode(pgId)
+
+		var nextRef = elemRef{page: p, node: n}
+		nextRef.index = nextRef.count() - 1
+		c.stack = append(c.stack, nextRef)
+	}
+}
+
+// next moves to the next leaf element and returns the key and value.
+// If the cursor is at the last leaf element then it stays there and returns nil.
+func (c *Cursor) next() (key []byte, value []byte, flags uint32) {
+	for {
+		// Attempt to move over one element until we're successful.
+		// Move up the stack as we hit the end of each page in our stack.
+		var i int
+		for i = len(c.stack) - 1; i >= 0; i-- {
+			elem := &c.stack[i]
+			if elem.index < elem.count()-1 {
+				elem.index++
+				break
+			}
+		}
+
+		// If we've hit the root page then stop and return. This will leave the
+		// cursor on the last element of the last page.
+		if i == -1 {
+			return nil, nil, 0
+		}
+
+		// Otherwise start from where we left off in the stack and find the
+		// first element of the first leaf page.
+		c.stack = c.stack[:i+1]
+		c.goToFirstElementOnTheStack()
+
+		// If this is an empty page then restart and move back up the stack.
+		// https://github.com/boltdb/bolt/issues/450
+		if c.stack[len(c.stack)-1].count() == 0 {
+			continue
+		}
+
+		return c.keyValue()
+	}
+}
+
+// prev moves the cursor to the previous item in the bucket and returns its key and value.
+// If the cursor is at the beginning of the bucket then a nil key and value are returned.
+func (c *Cursor) prev() (key []byte, value []byte, flags uint32) {
+	// Attempt to move back one element until we're successful.
+	// Move up the stack as we hit the beginning of each page in our stack.
+	for i := len(c.stack) - 1; i >= 0; i-- {
+		elem := &c.stack[i]
+		if elem.index > 0 {
+			elem.index--
+			break
+		}
+		// If we've hit the beginning, we should stop moving the cursor,
+		// and stay at the first element, so that users can continue to
+		// iterate over the elements in reverse direction by calling `Next`.
+		// We should return nil in such case.
+		// Refer to https://github.com/etcd-io/bbolt/issues/733
+		if len(c.stack) == 1 {
+			c.first()
+			return nil, nil, 0
+		}
+		c.stack = c.stack[:i]
+	}
+
+	// If we've hit the end then return nil.
+	if len(c.stack) == 0 {
+		return nil, nil, 0
+	}
+
+	// Move down the stack to find the last element of the last leaf under this branch.
+	c.last()
+	return c.keyValue()
+}
+
+// search recursively performs a binary search against a given page/node until it finds a given key.
+func (c *Cursor) search(key []byte, pgId common.Pgid) {
+	p, n := c.bucket.pageNode(pgId)
+	if p != nil && !p.IsBranchPage() && !p.IsLeafPage() {
+		panic(fmt.Sprintf("invalid page type: %d: %x", p.Id(), p.Flags()))
+	}
+	e := elemRef{page: p, node: n}
+	c.stack = append(c.stack, e)
+
+	// If we're on a leaf page/node then find the specific node.
+	if e.isLeaf() {
+		c.nsearch(key)
+		return
+	}
+
+	if n != nil {
+		c.searchNode(key, n)
+		return
+	}
+	c.searchPage(key, p)
+}
+
+func (c *Cursor) searchNode(key []byte, n *node) {
+	var exact bool
+	index := sort.Search(len(n.inodes), func(i int) bool {
+		// TODO(benbjohnson): Optimize this range search. It's a bit hacky right now.
+		// sort.Search() finds the lowest index where f() != -1 but we need the highest index.
+		ret := bytes.Compare(n.inodes[i].Key(), key)
+		if ret == 0 {
+			exact = true
+		}
+		return ret != -1
+	})
+	if !exact && index > 0 {
+		index--
+	}
+	c.stack[len(c.stack)-1].index = index
+
+	// Recursively search to the next page.
+	c.search(key, n.inodes[index].Pgid())
+}
+
+func (c *Cursor) searchPage(key []byte, p *common.Page) {
+	// Binary search for the correct range.
+	inodes := p.BranchPageElements()
+
+	var exact bool
+	index := sort.Search(int(p.Count()), func(i int) bool {
+		// TODO(benbjohnson): Optimize this range search. It's a bit hacky right now.
+		// sort.Search() finds the lowest index where f() != -1 but we need the highest index.
+		ret := bytes.Compare(inodes[i].Key(), key)
+		if ret == 0 {
+			exact = true
+		}
+		return ret != -1
+	})
+	if !exact && index > 0 {
+		index--
+	}
+	c.stack[len(c.stack)-1].index = index
+
+	// Recursively search to the next page.
+	c.search(key, inodes[index].Pgid())
+}
+
+// nsearch searches the leaf node on the top of the stack for a key.
+func (c *Cursor) nsearch(key []byte) {
+	e := &c.stack[len(c.stack)-1]
+	p, n := e.page, e.node
+
+	// If we have a node then search its inodes.
+	if n != nil {
+		index := sort.Search(len(n.inodes), func(i int) bool {
+			return bytes.Compare(n.inodes[i].Key(), key) != -1
+		})
+		e.index = index
+		return
+	}
+
+	// If we have a page then search its leaf elements.
+	inodes := p.LeafPageElements()
+	index := sort.Search(int(p.Count()), func(i int) bool {
+		return bytes.Compare(inodes[i].Key(), key) != -1
+	})
+	e.index = index
+}
+
+// keyValue returns the key and value of the current leaf element.
+func (c *Cursor) keyValue() ([]byte, []byte, uint32) {
+	ref := &c.stack[len(c.stack)-1]
+
+	// If the cursor is pointing to the end of page/node then return nil.
+	if ref.count() == 0 || ref.index >= ref.count() {
+		return nil, nil, 0
+	}
+
+	// Retrieve value from node.
+	if ref.node != nil {
+		inode := &ref.node.inodes[ref.index]
+		return inode.Key(), inode.Value(), inode.Flags()
+	}
+
+	// Or retrieve value from page.
+	elem := ref.page.LeafPageElement(uint16(ref.index))
+	return elem.Key(), elem.Value(), elem.Flags()
+}
+
+// node returns the node that the cursor is currently positioned on.
+func (c *Cursor) node() *node {
+	common.Assert(len(c.stack) > 0, "accessing a node with a zero-length cursor stack")
+
+	// If the top of the stack is a leaf node then just return it.
+	if ref := &c.stack[len(c.stack)-1]; ref.node != nil && ref.isLeaf() {
+		return ref.node
+	}
+
+	// Start from root and traverse down the hierarchy.
+	var n = c.stack[0].node
+	if n == nil {
+		n = c.bucket.node(c.stack[0].page.Id(), nil)
+	}
+	for _, ref := range c.stack[:len(c.stack)-1] {
+		common.Assert(!n.isLeaf, "expected branch node")
+		n = n.childAt(ref.index)
+	}
+	common.Assert(n.isLeaf, "expected leaf node")
+	return n
+}
+
+// elemRef represents a reference to an element on a given page/node.
+type elemRef struct {
+	page  *common.Page
+	node  *node
+	index int
+}
+
+// isLeaf returns whether the ref is pointing at a leaf page/node.
+func (r *elemRef) isLeaf() bool {
+	if r.node != nil {
+		return r.node.isLeaf
+	}
+	return r.page.IsLeafPage()
+}
+
+// count returns the number of inodes or page elements.
+func (r *elemRef) count() int {
+	if r.node != nil {
+		return len(r.node.inodes)
+	}
+	return int(r.page.Count())
+}

vendor/go.etcd.io/bbolt/doc.go

@@ -0,0 +1,40 @@
+/*
+package bbolt implements a low-level key/value store in pure Go. It supports
+fully serializable transactions, ACID semantics, and lock-free MVCC with
+multiple readers and a single writer. Bolt can be used for projects that
+want a simple data store without the need to add large dependencies such as
+Postgres or MySQL.
+
+Bolt is a single-level, zero-copy, B+tree data store. This means that Bolt is
+optimized for fast read access and does not require recovery in the event of a
+system crash. Transactions which have not finished committing will simply be
+rolled back in the event of a crash.
+
+The design of Bolt is based on Howard Chu's LMDB database project.
+
+Bolt currently works on Windows, Mac OS X, and Linux.
+
+# Basics
+
+There are only a few types in Bolt: DB, Bucket, Tx, and Cursor. The DB is
+a collection of buckets and is represented by a single file on disk. A bucket is
+a collection of unique keys that are associated with values.
+
+Transactions provide either read-only or read-write access to the database.
+Read-only transactions can retrieve key/value pairs and can use Cursors to
+iterate over the dataset sequentially. Read-write transactions can create and
+delete buckets and can insert and remove keys. Only one read-write transaction
+is allowed at a time.
+
+# Caveats
+
+The database uses a read-only, memory-mapped data file to ensure that
+applications cannot corrupt the database, however, this means that keys and
+values returned from Bolt cannot be changed. Writing to a read-only byte slice
+will cause Go to panic.
+
+Keys and values retrieved from the database are only valid for the life of
+the transaction. When used outside the transaction, these byte slices can
+point to different data or can point to invalid memory which will cause a panic.
+*/
+package bbolt

vendor/go.etcd.io/bbolt/errors.go

@@ -0,0 +1,108 @@
+package bbolt
+
+import "go.etcd.io/bbolt/errors"
+
+// These errors can be returned when opening or calling methods on a DB.
+var (
+	// ErrDatabaseNotOpen is returned when a DB instance is accessed before it
+	// is opened or after it is closed.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrDatabaseNotOpen = errors.ErrDatabaseNotOpen
+
+	// ErrInvalid is returned when both meta pages on a database are invalid.
+	// This typically occurs when a file is not a bolt database.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrInvalid = errors.ErrInvalid
+
+	// ErrInvalidMapping is returned when the database file fails to get mapped.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrInvalidMapping = errors.ErrInvalidMapping
+
+	// ErrVersionMismatch is returned when the data file was created with a
+	// different version of Bolt.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrVersionMismatch = errors.ErrVersionMismatch
+
+	// ErrChecksum is returned when a checksum mismatch occurs on either of the two meta pages.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrChecksum = errors.ErrChecksum
+
+	// ErrTimeout is returned when a database cannot obtain an exclusive lock
+	// on the data file after the timeout passed to Open().
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrTimeout = errors.ErrTimeout
+)
+
+// These errors can occur when beginning or committing a Tx.
+var (
+	// ErrTxNotWritable is returned when performing a write operation on a
+	// read-only transaction.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrTxNotWritable = errors.ErrTxNotWritable
+
+	// ErrTxClosed is returned when committing or rolling back a transaction
+	// that has already been committed or rolled back.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrTxClosed = errors.ErrTxClosed
+
+	// ErrDatabaseReadOnly is returned when a mutating transaction is started on a
+	// read-only database.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrDatabaseReadOnly = errors.ErrDatabaseReadOnly
+
+	// ErrFreePagesNotLoaded is returned when a readonly transaction without
+	// preloading the free pages is trying to access the free pages.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrFreePagesNotLoaded = errors.ErrFreePagesNotLoaded
+)
+
+// These errors can occur when putting or deleting a value or a bucket.
+var (
+	// ErrBucketNotFound is returned when trying to access a bucket that has
+	// not been created yet.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrBucketNotFound = errors.ErrBucketNotFound
+
+	// ErrBucketExists is returned when creating a bucket that already exists.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrBucketExists = errors.ErrBucketExists
+
+	// ErrBucketNameRequired is returned when creating a bucket with a blank name.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrBucketNameRequired = errors.ErrBucketNameRequired
+
+	// ErrKeyRequired is returned when inserting a zero-length key.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrKeyRequired = errors.ErrKeyRequired
+
+	// ErrKeyTooLarge is returned when inserting a key that is larger than MaxKeySize.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrKeyTooLarge = errors.ErrKeyTooLarge
+
+	// ErrValueTooLarge is returned when inserting a value that is larger than MaxValueSize.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrValueTooLarge = errors.ErrValueTooLarge
+
+	// ErrIncompatibleValue is returned when trying create or delete a bucket
+	// on an existing non-bucket key or when trying to create or delete a
+	// non-bucket key on an existing bucket key.
+	//
+	// Deprecated: Use the error variables defined in the bbolt/errors package.
+	ErrIncompatibleValue = errors.ErrIncompatibleValue
+)

vendor/go.etcd.io/bbolt/errors/errors.go

@@ -0,0 +1,84 @@
+// Package errors defines the error variables that may be returned
+// during bbolt operations.
+package errors
+
+import "errors"
+
+// These errors can be returned when opening or calling methods on a DB.
+var (
+	// ErrDatabaseNotOpen is returned when a DB instance is accessed before it
+	// is opened or after it is closed.
+	ErrDatabaseNotOpen = errors.New("database not open")
+
+	// ErrInvalid is returned when both meta pages on a database are invalid.
+	// This typically occurs when a file is not a bolt database.
+	ErrInvalid = errors.New("invalid database")
+
+	// ErrInvalidMapping is returned when the database file fails to get mapped.
+	ErrInvalidMapping = errors.New("database isn't correctly mapped")
+
+	// ErrVersionMismatch is returned when the data file was created with a
+	// different version of Bolt.
+	ErrVersionMismatch = errors.New("version mismatch")
+
+	// ErrChecksum is returned when a checksum mismatch occurs on either of the two meta pages.
+	ErrChecksum = errors.New("checksum error")
+
+	// ErrTimeout is returned when a database cannot obtain an exclusive lock
+	// on the data file after the timeout passed to Open().
+	ErrTimeout = errors.New("timeout")
+)
+
+// These errors can occur when beginning or committing a Tx.
+var (
+	// ErrTxNotWritable is returned when performing a write operation on a
+	// read-only transaction.
+	ErrTxNotWritable = errors.New("tx not writable")
+
+	// ErrTxClosed is returned when committing or rolling back a transaction
+	// that has already been committed or rolled back.
+	ErrTxClosed = errors.New("tx closed")
+
+	// ErrDatabaseReadOnly is returned when a mutating transaction is started on a
+	// read-only database.
+	ErrDatabaseReadOnly = errors.New("database is in read-only mode")
+
+	// ErrFreePagesNotLoaded is returned when a readonly transaction without
+	// preloading the free pages is trying to access the free pages.
+	ErrFreePagesNotLoaded = errors.New("free pages are not pre-loaded")
+)
+
+// These errors can occur when putting or deleting a value or a bucket.
+var (
+	// ErrBucketNotFound is returned when trying to access a bucket that has
+	// not been created yet.
+	ErrBucketNotFound = errors.New("bucket not found")
+
+	// ErrBucketExists is returned when creating a bucket that already exists.
+	ErrBucketExists = errors.New("bucket already exists")
+
+	// ErrBucketNameRequired is returned when creating a bucket with a blank name.
+	ErrBucketNameRequired = errors.New("bucket name required")
+
+	// ErrKeyRequired is returned when inserting a zero-length key.
+	ErrKeyRequired = errors.New("key required")
+
+	// ErrKeyTooLarge is returned when inserting a key that is larger than MaxKeySize.
+	ErrKeyTooLarge = errors.New("key too large")
+
+	// ErrValueTooLarge is returned when inserting a value that is larger than MaxValueSize.
+	ErrValueTooLarge = errors.New("value too large")
+
+	// ErrIncompatibleValue is returned when trying to create or delete a bucket
+	// on an existing non-bucket key or when trying to create or delete a
+	// non-bucket key on an existing bucket key.
+	ErrIncompatibleValue = errors.New("incompatible value")
+
+	// ErrSameBuckets is returned when trying to move a sub-bucket between
+	// source and target buckets, while source and target buckets are the same.
+	ErrSameBuckets = errors.New("the source and target are the same bucket")
+
+	// ErrDifferentDB is returned when trying to move a sub-bucket between
+	// source and target buckets, while source and target buckets are in different database files.
+	ErrDifferentDB = errors.New("the source and target buckets are in different database files")
+)

vendor/go.etcd.io/bbolt/internal/common/bucket.go

@@ -0,0 +1,54 @@
+package common
+
+import (
+	"fmt"
+	"unsafe"
+)
+
+const BucketHeaderSize = int(unsafe.Sizeof(InBucket{}))
+
+// InBucket represents the on-file representation of a bucket.
+// This is stored as the "value" of a bucket key. If the bucket is small enough,
+// then its root page can be stored inline in the "value", after the bucket
+// header. In the case of inline buckets, the "root" will be 0.
+type InBucket struct {
+	root     Pgid   // page id of the bucket's root-level page
+	sequence uint64 // monotonically incrementing, used by NextSequence()
+}
+
+func NewInBucket(root Pgid, seq uint64) InBucket {
+	return InBucket{
+		root:     root,
+		sequence: seq,
+	}
+}
+
+func (b *InBucket) RootPage() Pgid {
+	return b.root
+}
+
+func (b *InBucket) SetRootPage(id Pgid) {
+	b.root = id
+}
+
+// InSequence returns the sequence. The reason why not naming it `Sequence`
+// is to avoid duplicated name as `(*Bucket) Sequence()`
+func (b *InBucket) InSequence() uint64 {
+	return b.sequence
+}
+
+func (b *InBucket) SetInSequence(v uint64) {
+	b.sequence = v
+}
+
+func (b *InBucket) IncSequence() {
+	b.sequence++
+}
+
+func (b *InBucket) InlinePage(v []byte) *Page {
+	return (*Page)(unsafe.Pointer(&v[BucketHeaderSize]))
+}
+
+func (b *InBucket) String() string {
+	return fmt.Sprintf("<pgid=%d,seq=%d>", b.root, b.sequence)
+}

vendor/go.etcd.io/bbolt/internal/common/inode.go

@@ -0,0 +1,115 @@
+package common
+
+import "unsafe"
+
+// Inode represents an internal node inside of a node.
+// It can be used to point to elements in a page or point
+// to an element which hasn't been added to a page yet.
+type Inode struct {
+	flags uint32
+	pgid  Pgid
+	key   []byte
+	value []byte
+}
+
+type Inodes []Inode
+
+func (in *Inode) Flags() uint32 {
+	return in.flags
+}
+
+func (in *Inode) SetFlags(flags uint32) {
+	in.flags = flags
+}
+
+func (in *Inode) Pgid() Pgid {
+	return in.pgid
+}
+
+func (in *Inode) SetPgid(id Pgid) {
+	in.pgid = id
+}
+
+func (in *Inode) Key() []byte {
+	return in.key
+}
+
+func (in *Inode) SetKey(key []byte) {
+	in.key = key
+}
+
+func (in *Inode) Value() []byte {
+	return in.value
+}
+
+func (in *Inode) SetValue(value []byte) {
+	in.value = value
+}
+
+func ReadInodeFromPage(p *Page) Inodes {
+	inodes := make(Inodes, int(p.Count()))
+	isLeaf := p.IsLeafPage()
+	for i := 0; i < int(p.Count()); i++ {
+		inode := &inodes[i]
+		if isLeaf {
+			elem := p.LeafPageElement(uint16(i))
+			inode.SetFlags(elem.Flags())
+			inode.SetKey(elem.Key())
+			inode.SetValue(elem.Value())
+		} else {
+			elem := p.BranchPageElement(uint16(i))
+			inode.SetPgid(elem.Pgid())
+			inode.SetKey(elem.Key())
+		}
+		Assert(len(inode.Key()) > 0, "read: zero-length inode key")
+	}
+
+	return inodes
+}
+
+func WriteInodeToPage(inodes Inodes, p *Page) uint32 {
+	// Loop over each item and write it to the page.
+	// off tracks the offset into p of the start of the next data.
+	off := unsafe.Sizeof(*p) + p.PageElementSize()*uintptr(len(inodes))
+	isLeaf := p.IsLeafPage()
+	for i, item := range inodes {
+		Assert(len(item.Key()) > 0, "write: zero-length inode key")
+
+		// Create a slice to write into of needed size and advance
+		// byte pointer for next iteration.
+		sz := len(item.Key()) + len(item.Value())
+		b := UnsafeByteSlice(unsafe.Pointer(p), off, 0, sz)
+		off += uintptr(sz)
+
+		// Write the page element.
+		if isLeaf {
+			elem := p.LeafPageElement(uint16(i))
+			elem.SetPos(uint32(uintptr(unsafe.Pointer(&b[0])) - uintptr(unsafe.Pointer(elem))))
+			elem.SetFlags(item.Flags())
+			elem.SetKsize(uint32(len(item.Key())))
+			elem.SetVsize(uint32(len(item.Value())))
+		} else {
+			elem := p.BranchPageElement(uint16(i))
+			elem.SetPos(uint32(uintptr(unsafe.Pointer(&b[0])) - uintptr(unsafe.Pointer(elem))))
+			elem.SetKsize(uint32(len(item.Key())))
+			elem.SetPgid(item.Pgid())
+			Assert(elem.Pgid() != p.Id(), "write: circular dependency occurred")
+		}
+
+		// Write data for the element to the end of the page.
+		l := copy(b, item.Key())
+		copy(b[l:], item.Value())
+	}
+
+	return uint32(off)
+}
+
+func UsedSpaceInPage(inodes Inodes, p *Page) uint32 {
+	off := unsafe.Sizeof(*p) + p.PageElementSize()*uintptr(len(inodes))
+	for _, item := range inodes {
+		sz := len(item.Key()) + len(item.Value())
+		off += uintptr(sz)
+	}
+
+	return uint32(off)
+}

vendor/go.etcd.io/bbolt/internal/common/meta.go

@@ -0,0 +1,161 @@
+package common
+
+import (
+	"fmt"
+	"hash/fnv"
+	"io"
+	"unsafe"
+
+	"go.etcd.io/bbolt/errors"
+)
+
+type Meta struct {
+	magic    uint32
+	version  uint32
+	pageSize uint32
+	flags    uint32
+	root     InBucket
+	freelist Pgid
+	pgid     Pgid
+	txid     Txid
+	checksum uint64
+}
+
+// Validate checks the marker bytes and version of the meta page to ensure it matches this binary.
+func (m *Meta) Validate() error {
+	if m.magic != Magic {
+		return errors.ErrInvalid
+	} else if m.version != Version {
+		return errors.ErrVersionMismatch
+	} else if m.checksum != m.Sum64() {
+		return errors.ErrChecksum
+	}
+	return nil
+}
+
+// Copy copies one meta object to another.
+func (m *Meta) Copy(dest *Meta) {
+	*dest = *m
+}
+
+// Write writes the meta onto a page.
+func (m *Meta) Write(p *Page) {
+	if m.root.root >= m.pgid {
+		panic(fmt.Sprintf("root bucket pgid (%d) above high water mark (%d)", m.root.root, m.pgid))
+	} else if m.freelist >= m.pgid && m.freelist != PgidNoFreelist {
+		// TODO: reject pgidNoFreeList if !NoFreelistSync
+		panic(fmt.Sprintf("freelist pgid (%d) above high water mark (%d)", m.freelist, m.pgid))
+	}
+
+	// Page id is either going to be 0 or 1 which we can determine by the transaction ID.
+	p.id = Pgid(m.txid % 2)
+	p.SetFlags(MetaPageFlag)
+
+	// Calculate the checksum.
+	m.checksum = m.Sum64()
+
+	m.Copy(p.Meta())
+}
+
+// Sum64 generates the checksum for the meta.
+func (m *Meta) Sum64() uint64 {
+	var h = fnv.New64a()
+	_, _ = h.Write((*[unsafe.Offsetof(Meta{}.checksum)]byte)(unsafe.Pointer(m))[:])
+	return h.Sum64()
+}
+
+func (m *Meta) Magic() uint32 {
+	return m.magic
+}
+
+func (m *Meta) SetMagic(v uint32) {
+	m.magic = v
+}
+
+func (m *Meta) Version() uint32 {
+	return m.version
+}
+
+func (m *Meta) SetVersion(v uint32) {
+	m.version = v
+}
+
+func (m *Meta) PageSize() uint32 {
+	return m.pageSize
+}
+
+func (m *Meta) SetPageSize(v uint32) {
+	m.pageSize = v
+}
+
+func (m *Meta) Flags() uint32 {
+	return m.flags
+}
+
+func (m *Meta) SetFlags(v uint32) {
+	m.flags = v
+}
+
+func (m *Meta) SetRootBucket(b InBucket) {
+	m.root = b
+}
+
+func (m *Meta) RootBucket() *InBucket {
+	return &m.root
+}
+
+func (m *Meta) Freelist() Pgid {
+	return m.freelist
+}
+
+func (m *Meta) SetFreelist(v Pgid) {
+	m.freelist = v
+}
+
+func (m *Meta) IsFreelistPersisted() bool {
+	return m.freelist != PgidNoFreelist
+}
+
+func (m *Meta) Pgid() Pgid {
+	return m.pgid
+}
+
+func (m *Meta) SetPgid(id Pgid) {
+	m.pgid = id
+}
+
+func (m *Meta) Txid() Txid {
+	return m.txid
+}
+
+func (m *Meta) SetTxid(id Txid) {
+	m.txid = id
+}
+
+func (m *Meta) IncTxid() {
+	m.txid += 1
+}
+
+func (m *Meta) DecTxid() {
+	m.txid -= 1
+}
+
+func (m *Meta) Checksum() uint64 {
+	return m.checksum
+}
+
+func (m *Meta) SetChecksum(v uint64) {
+	m.checksum = v
+}
+
+func (m *Meta) Print(w io.Writer) {
+	fmt.Fprintf(w, "Version:    %d\n", m.version)
+	fmt.Fprintf(w, "Page Size:  %d bytes\n", m.pageSize)
+	fmt.Fprintf(w, "Flags:      %08x\n", m.flags)
+	fmt.Fprintf(w, "Root:       <pgid=%d>\n", m.root.root)
+	fmt.Fprintf(w, "Freelist:   <pgid=%d>\n", m.freelist)
+	fmt.Fprintf(w, "HWM:        <pgid=%d>\n", m.pgid)
+	fmt.Fprintf(w, "Txn ID:     %d\n", m.txid)
+	fmt.Fprintf(w, "Checksum:   %016x\n", m.checksum)
+	fmt.Fprintf(w, "\n")
+}

vendor/go.etcd.io/bbolt/internal/common/page.go

@@ -0,0 +1,391 @@
+package common
+
+import (
+	"fmt"
+	"os"
+	"sort"
+	"unsafe"
+)
+
+const PageHeaderSize = unsafe.Sizeof(Page{})
+
+const MinKeysPerPage = 2
+
+const BranchPageElementSize = unsafe.Sizeof(branchPageElement{})
+const LeafPageElementSize = unsafe.Sizeof(leafPageElement{})
+const pgidSize = unsafe.Sizeof(Pgid(0))
+
+const (
+	BranchPageFlag   = 0x01
+	LeafPageFlag     = 0x02
+	MetaPageFlag     = 0x04
+	FreelistPageFlag = 0x10
+)
+
+const (
+	BucketLeafFlag = 0x01
+)
+
+type Pgid uint64
+
+type Page struct {
+	id       Pgid
+	flags    uint16
+	count    uint16
+	overflow uint32
+}
+
+func NewPage(id Pgid, flags, count uint16, overflow uint32) *Page {
+	return &Page{
+		id:       id,
+		flags:    flags,
+		count:    count,
+		overflow: overflow,
+	}
+}
+
+// Typ returns a human-readable page type string used for debugging.
+func (p *Page) Typ() string {
+	if p.IsBranchPage() {
+		return "branch"
+	} else if p.IsLeafPage() {
+		return "leaf"
+	} else if p.IsMetaPage() {
+		return "meta"
+	} else if p.IsFreelistPage() {
+		return "freelist"
+	}
+	return fmt.Sprintf("unknown<%02x>", p.flags)
+}
+
+func (p *Page) IsBranchPage() bool {
+	return p.flags == BranchPageFlag
+}
+
+func (p *Page) IsLeafPage() bool {
+	return p.flags == LeafPageFlag
+}
+
+func (p *Page) IsMetaPage() bool {
+	return p.flags == MetaPageFlag
+}
+
+func (p *Page) IsFreelistPage() bool {
+	return p.flags == FreelistPageFlag
+}
+
+// Meta returns a pointer to the metadata section of the page.
+func (p *Page) Meta() *Meta {
+	return (*Meta)(UnsafeAdd(unsafe.Pointer(p), unsafe.Sizeof(*p)))
+}
+
+func (p *Page) FastCheck(id Pgid) {
+	Assert(p.id == id, "Page expected to be: %v, but self identifies as %v", id, p.id)
+	// Only one flag of page-type can be set.
+	Assert(p.IsBranchPage() ||
+		p.IsLeafPage() ||
+		p.IsMetaPage() ||
+		p.IsFreelistPage(),
+		"page %v: has unexpected type/flags: %x", p.id, p.flags)
+}
+
+// LeafPageElement retrieves the leaf node by index
+func (p *Page) LeafPageElement(index uint16) *leafPageElement {
+	return (*leafPageElement)(UnsafeIndex(unsafe.Pointer(p), unsafe.Sizeof(*p),
+		LeafPageElementSize, int(index)))
+}
+
+// LeafPageElements retrieves a list of leaf nodes.
+func (p *Page) LeafPageElements() []leafPageElement {
+	if p.count == 0 {
+		return nil
+	}
+	data := UnsafeAdd(unsafe.Pointer(p), unsafe.Sizeof(*p))
+	elems := unsafe.Slice((*leafPageElement)(data), int(p.count))
+	return elems
+}
+
+// BranchPageElement retrieves the branch node by index
+func (p *Page) BranchPageElement(index uint16) *branchPageElement {
+	return (*branchPageElement)(UnsafeIndex(unsafe.Pointer(p), unsafe.Sizeof(*p),
+		unsafe.Sizeof(branchPageElement{}), int(index)))
+}
+
+// BranchPageElements retrieves a list of branch nodes.
+func (p *Page) BranchPageElements() []branchPageElement {
+	if p.count == 0 {
+		return nil
+	}
+	data := UnsafeAdd(unsafe.Pointer(p), unsafe.Sizeof(*p))
+	elems := unsafe.Slice((*branchPageElement)(data), int(p.count))
+	return elems
+}
+
+func (p *Page) FreelistPageCount() (int, int) {
+	Assert(p.IsFreelistPage(), fmt.Sprintf("can't get freelist page count from a non-freelist page: %2x", p.flags))
+
+	// If the page.count is at the max uint16 value (64k) then it's considered
+	// an overflow and the size of the freelist is stored as the first element.
+	var idx, count = 0, int(p.count)
+	if count == 0xFFFF {
+		idx = 1
+		c := *(*Pgid)(UnsafeAdd(unsafe.Pointer(p), unsafe.Sizeof(*p)))
+		count = int(c)
+		if count < 0 {
+			panic(fmt.Sprintf("leading element count %d overflows int", c))
+		}
+	}
+
+	return idx, count
+}
+
+func (p *Page) FreelistPageIds() []Pgid {
+	Assert(p.IsFreelistPage(), fmt.Sprintf("can't get freelist page IDs from a non-freelist page: %2x", p.flags))
+
+	idx, count := p.FreelistPageCount()
+
+	if count == 0 {
+		return nil
+	}
+
+	data := UnsafeIndex(unsafe.Pointer(p), unsafe.Sizeof(*p), pgidSize, idx)
+	ids := unsafe.Slice((*Pgid)(data), count)
+
+	return ids
+}
+
+// dump writes n bytes of the page to STDERR as hex output.
+func (p *Page) hexdump(n int) {
+	buf := UnsafeByteSlice(unsafe.Pointer(p), 0, 0, n)
+	fmt.Fprintf(os.Stderr, "%x\n", buf)
+}
+
+func (p *Page) PageElementSize() uintptr {
+	if p.IsLeafPage() {
+		return LeafPageElementSize
+	}
+	return BranchPageElementSize
+}
+
+func (p *Page) Id() Pgid {
+	return p.id
+}
+
+func (p *Page) SetId(target Pgid) {
+	p.id = target
+}
+
+func (p *Page) Flags() uint16 {
+	return p.flags
+}
+
+func (p *Page) SetFlags(v uint16) {
+	p.flags = v
+}
+
+func (p *Page) Count() uint16 {
+	return p.count
+}
+
+func (p *Page) SetCount(target uint16) {
+	p.count = target
+}
+
+func (p *Page) Overflow() uint32 {
+	return p.overflow
+}
+
+func (p *Page) SetOverflow(target uint32) {
+	p.overflow = target
+}
+
+func (p *Page) String() string {
+	return fmt.Sprintf("ID: %d, Type: %s, count: %d, overflow: %d", p.id, p.Typ(), p.count, p.overflow)
+}
+
+type Pages []*Page
+
+func (s Pages) Len() int           { return len(s) }
+func (s Pages) Swap(i, j int)      { s[i], s[j] = s[j], s[i] }
+func (s Pages) Less(i, j int) bool { return s[i].id < s[j].id }
+
+// branchPageElement represents a node on a branch page.
+type branchPageElement struct {
+	pos   uint32
+	ksize uint32
+	pgid  Pgid
+}
+
+func (n *branchPageElement) Pos() uint32 {
+	return n.pos
+}
+
+func (n *branchPageElement) SetPos(v uint32) {
+	n.pos = v
+}
+
+func (n *branchPageElement) Ksize() uint32 {
+	return n.ksize
+}
+
+func (n *branchPageElement) SetKsize(v uint32) {
+	n.ksize = v
+}
+
+func (n *branchPageElement) Pgid() Pgid {
+	return n.pgid
+}
+
+func (n *branchPageElement) SetPgid(v Pgid) {
+	n.pgid = v
+}
+
+// Key returns a byte slice of the node key.
+func (n *branchPageElement) Key() []byte {
+	return UnsafeByteSlice(unsafe.Pointer(n), 0, int(n.pos), int(n.pos)+int(n.ksize))
+}
+
+// leafPageElement represents a node on a leaf page.
+type leafPageElement struct {
+	flags uint32
+	pos   uint32
+	ksize uint32
+	vsize uint32
+}
+
+func NewLeafPageElement(flags, pos, ksize, vsize uint32) *leafPageElement {
+	return &leafPageElement{
+		flags: flags,
+		pos:   pos,
+		ksize: ksize,
+		vsize: vsize,
+	}
+}
+
+func (n *leafPageElement) Flags() uint32 {
+	return n.flags
+}
+
+func (n *leafPageElement) SetFlags(v uint32) {
+	n.flags = v
+}
+
+func (n *leafPageElement) Pos() uint32 {
+	return n.pos
+}
+
+func (n *leafPageElement) SetPos(v uint32) {
+	n.pos = v
+}
+
+func (n *leafPageElement) Ksize() uint32 {
+	return n.ksize
+}
+
+func (n *leafPageElement) SetKsize(v uint32) {
+	n.ksize = v
+}
+
+func (n *leafPageElement) Vsize() uint32 {
+	return n.vsize
+}
+
+func (n *leafPageElement) SetVsize(v uint32) {
+	n.vsize = v
+}
+
+// Key returns a byte slice of the node key.
+func (n *leafPageElement) Key() []byte {
+	i := int(n.pos)
+	j := i + int(n.ksize)
+	return UnsafeByteSlice(unsafe.Pointer(n), 0, i, j)
+}
+
+// Value returns a byte slice of the node value.
+func (n *leafPageElement) Value() []byte {
+	i := int(n.pos) + int(n.ksize)
+	j := i + int(n.vsize)
+	return UnsafeByteSlice(unsafe.Pointer(n), 0, i, j)
+}
+
+func (n *leafPageElement) IsBucketEntry() bool {
+	return n.flags&uint32(BucketLeafFlag) != 0
+}
+
+func (n *leafPageElement) Bucket() *InBucket {
+	if n.IsBucketEntry() {
+		return LoadBucket(n.Value())
+	} else {
+		return nil
+	}
+}
+
+// PageInfo represents human readable information about a page.
+type PageInfo struct {
+	ID            int
+	Type          string
+	Count         int
+	OverflowCount int
+}
+
+type Pgids []Pgid
+
+func (s Pgids) Len() int           { return len(s) }
+func (s Pgids) Swap(i, j int)      { s[i], s[j] = s[j], s[i] }
+func (s Pgids) Less(i, j int) bool { return s[i] < s[j] }
+
+// Merge returns the sorted union of a and b.
+func (a Pgids) Merge(b Pgids) Pgids {
+	// Return the opposite slice if one is nil.
+	if len(a) == 0 {
+		return b
+	}
+	if len(b) == 0 {
+		return a
+	}
+	merged := make(Pgids, len(a)+len(b))
+	Mergepgids(merged, a, b)
+	return merged
+}
+
+// Mergepgids copies the sorted union of a and b into dst.
+// If dst is too small, it panics.
+func Mergepgids(dst, a, b Pgids) {
+	if len(dst) < len(a)+len(b) {
+		panic(fmt.Errorf("mergepgids bad len %d < %d + %d", len(dst), len(a), len(b)))
+	}
+	// Copy in the opposite slice if one is nil.
+	if len(a) == 0 {
+		copy(dst, b)
+		return
+	}
+	if len(b) == 0 {
+		copy(dst, a)
+		return
+	}
+
+	// Merged will hold all elements from both lists.
+	merged := dst[:0]
+
+	// Assign lead to the slice with a lower starting value, follow to the higher value.
+	lead, follow := a, b
+	if b[0] < a[0] {
+		lead, follow = b, a
+	}
+
+	// Continue while there are elements in the lead.
+	for len(lead) > 0 {
+		// Merge largest prefix of lead that is ahead of follow[0].
+		n := sort.Search(len(lead), func(i int) bool { return lead[i] > follow[0] })
+		merged = append(merged, lead[:n]...)
+		if n >= len(lead) {
+			break
+		}
+
+		// Swap lead and follow.
+		lead, follow = follow, lead[n:]
+	}
+
+	// Append what's left in follow.
+	_ = append(merged, follow...)
+}

vendor/go.etcd.io/bbolt/internal/common/types.go

@@ -0,0 +1,40 @@
+package common
+
+import (
+	"os"
+	"runtime"
+	"time"
+)
+
+// MaxMmapStep is the largest step that can be taken when remapping the mmap.
+const MaxMmapStep = 1 << 30 // 1GB
+
+// Version represents the data file format version.
+const Version uint32 = 2
+
+// Magic represents a marker value to indicate that a file is a Bolt DB.
+const Magic uint32 = 0xED0CDAED
+
+const PgidNoFreelist Pgid = 0xffffffffffffffff
+
+// DO NOT EDIT. Copied from the "bolt" package.
+const pageMaxAllocSize = 0xFFFFFFF
+
+// IgnoreNoSync specifies whether the NoSync field of a DB is ignored when
+// syncing changes to a file.  This is required as some operating systems,
+// such as OpenBSD, do not have a unified buffer cache (UBC) and writes
+// must be synchronized using the msync(2) syscall.
+const IgnoreNoSync = runtime.GOOS == "openbsd"
+
+// Default values if not set in a DB instance.
+const (
+	DefaultMaxBatchSize  int = 1000
+	DefaultMaxBatchDelay     = 10 * time.Millisecond
+	DefaultAllocSize         = 16 * 1024 * 1024
+)
+
+// DefaultPageSize is the default page size for db which is set to the OS page size.
+var DefaultPageSize = os.Getpagesize()
+
+// Txid represents the internal transaction identifier.
+type Txid uint64

vendor/go.etcd.io/bbolt/internal/common/unsafe.go

@@ -0,0 +1,27 @@
+package common
+
+import (
+	"unsafe"
+)
+
+func UnsafeAdd(base unsafe.Pointer, offset uintptr) unsafe.Pointer {
+	return unsafe.Pointer(uintptr(base) + offset)
+}
+
+func UnsafeIndex(base unsafe.Pointer, offset uintptr, elemsz uintptr, n int) unsafe.Pointer {
+	return unsafe.Pointer(uintptr(base) + offset + uintptr(n)*elemsz)
+}
+
+func UnsafeByteSlice(base unsafe.Pointer, offset uintptr, i, j int) []byte {
+	// See: https://github.com/golang/go/wiki/cgo#turning-c-arrays-into-go-slices
+	//
+	// This memory is not allocated from C, but it is unmanaged by Go's
+	// garbage collector and should behave similarly, and the compiler
+	// should produce similar code.  Note that this conversion allows a
+	// subslice to begin after the base address, with an optional offset,
+	// while the URL above does not cover this case and only slices from
+	// index 0.  However, the wiki never says that the address must be to
+	// the beginning of a C allocation (or even that malloc was used at
+	// all), so this is believed to be correct.
+	return (*[pageMaxAllocSize]byte)(UnsafeAdd(base, offset))[i:j:j]
+}

vendor/go.etcd.io/bbolt/internal/common/utils.go

@@ -0,0 +1,64 @@
+package common
+
+import (
+	"fmt"
+	"io"
+	"os"
+	"unsafe"
+)
+
+func LoadBucket(buf []byte) *InBucket {
+	return (*InBucket)(unsafe.Pointer(&buf[0]))
+}
+
+func LoadPage(buf []byte) *Page {
+	return (*Page)(unsafe.Pointer(&buf[0]))
+}
+
+func LoadPageMeta(buf []byte) *Meta {
+	return (*Meta)(unsafe.Pointer(&buf[PageHeaderSize]))
+}
+
+func CopyFile(srcPath, dstPath string) error {
+	// Ensure source file exists.
+	_, err := os.Stat(srcPath)
+	if os.IsNotExist(err) {
+		return fmt.Errorf("source file %q not found", srcPath)
+	} else if err != nil {
+		return err
+	}
+
+	// Ensure output file not exist.
+	_, err = os.Stat(dstPath)
+	if err == nil {
+		return fmt.Errorf("output file %q already exists", dstPath)
+	} else if !os.IsNotExist(err) {
+		return err
+	}
+
+	srcDB, err := os.Open(srcPath)
+	if err != nil {
+		return fmt.Errorf("failed to open source file %q: %w", srcPath, err)
+	}
+	defer srcDB.Close()
+	dstDB, err := os.Create(dstPath)
+	if err != nil {
+		return fmt.Errorf("failed to create output file %q: %w", dstPath, err)
+	}
+	defer dstDB.Close()
+	written, err := io.Copy(dstDB, srcDB)
+	if err != nil {
+		return fmt.Errorf("failed to copy database file from %q to %q: %w", srcPath, dstPath, err)
+	}
+
+	srcFi, err := srcDB.Stat()
+	if err != nil {
+		return fmt.Errorf("failed to get source file info %q: %w", srcPath, err)
+	}
+	initialSize := srcFi.Size()
+	if initialSize != written {
+		return fmt.Errorf("the byte copied (%q: %d) isn't equal to the initial db size (%q: %d)", dstPath, written, srcPath, initialSize)
+	}
+
+	return nil
+}

vendor/go.etcd.io/bbolt/internal/common/verify.go

@@ -0,0 +1,67 @@
+// Copied from https://github.com/etcd-io/etcd/blob/main/client/pkg/verify/verify.go
+package common
+
+import (
+	"fmt"
+	"os"
+	"strings"
+)
+
+const ENV_VERIFY = "BBOLT_VERIFY"
+
+type VerificationType string
+
+const (
+	ENV_VERIFY_VALUE_ALL    VerificationType = "all"
+	ENV_VERIFY_VALUE_ASSERT VerificationType = "assert"
+)
+
+func getEnvVerify() string {
+	return strings.ToLower(os.Getenv(ENV_VERIFY))
+}
+
+func IsVerificationEnabled(verification VerificationType) bool {
+	env := getEnvVerify()
+	return env == string(ENV_VERIFY_VALUE_ALL) || env == strings.ToLower(string(verification))
+}
+
+// EnableVerifications sets `ENV_VERIFY` and returns a function that
+// can be used to bring the original settings.
+func EnableVerifications(verification VerificationType) func() {
+	previousEnv := getEnvVerify()
+	os.Setenv(ENV_VERIFY, string(verification))
+	return func() {
+		os.Setenv(ENV_VERIFY, previousEnv)
+	}
+}
+
+// EnableAllVerifications enables verification and returns a function
+// that can be used to bring the original settings.
+func EnableAllVerifications() func() {
+	return EnableVerifications(ENV_VERIFY_VALUE_ALL)
+}
+
+// DisableVerifications unsets `ENV_VERIFY` and returns a function that
+// can be used to bring the original settings.
+func DisableVerifications() func() {
+	previousEnv := getEnvVerify()
+	os.Unsetenv(ENV_VERIFY)
+	return func() {
+		os.Setenv(ENV_VERIFY, previousEnv)
+	}
+}
+
+// Verify performs verification if the assertions are enabled.
+// In the default setup running in tests and skipped in the production code.
+func Verify(f func()) {
+	if IsVerificationEnabled(ENV_VERIFY_VALUE_ASSERT) {
+		f()
+	}
+}
+
+// Assert will panic with a given formatted message if the given condition is false.
+func Assert(condition bool, msg string, v ...any) {
+	if !condition {
+		panic(fmt.Sprintf("assertion failed: "+msg, v...))
+	}
+}

vendor/go.etcd.io/bbolt/internal/freelist/array.go

@@ -0,0 +1,108 @@
+package freelist
+
+import (
+	"fmt"
+	"sort"
+
+	"go.etcd.io/bbolt/internal/common"
+)
+
+type array struct {
+	*shared
+
+	ids []common.Pgid // all free and available free page ids.
+}
+
+func (f *array) Init(ids common.Pgids) {
+	f.ids = ids
+	f.reindex()
+}
+
+func (f *array) Allocate(txid common.Txid, n int) common.Pgid {
+	if len(f.ids) == 0 {
+		return 0
+	}
+
+	var initial, previd common.Pgid
+	for i, id := range f.ids {
+		if id <= 1 {
+			panic(fmt.Sprintf("invalid page allocation: %d", id))
+		}
+
+		// Reset initial page if this is not contiguous.
+		if previd == 0 || id-previd != 1 {
+			initial = id
+		}
+
+		// If we found a contiguous block then remove it and return it.
+		if (id-initial)+1 == common.Pgid(n) {
+			// If we're allocating off the beginning then take the fast path
+			// and just adjust the existing slice. This will use extra memory
+			// temporarily but the append() in free() will realloc the slice
+			// as is necessary.
+			if (i + 1) == n {
+				f.ids = f.ids[i+1:]
+			} else {
+				copy(f.ids[i-n+1:], f.ids[i+1:])
+				f.ids = f.ids[:len(f.ids)-n]
+			}
+
+			// Remove from the free cache.
+			for i := common.Pgid(0); i < common.Pgid(n); i++ {
+				delete(f.cache, initial+i)
+			}
+			f.allocs[initial] = txid
+			return initial
+		}
+
+		previd = id
+	}
+	return 0
+}
+
+func (f *array) FreeCount() int {
+	return len(f.ids)
+}
+
+func (f *array) freePageIds() common.Pgids {
+	return f.ids
+}
+
+func (f *array) mergeSpans(ids common.Pgids) {
+	sort.Sort(ids)
+	common.Verify(func() {
+		idsIdx := make(map[common.Pgid]struct{})
+		for _, id := range f.ids {
+			// The existing f.ids shouldn't have duplicated free ID.
+			if _, ok := idsIdx[id]; ok {
+				panic(fmt.Sprintf("detected duplicated free page ID: %d in existing f.ids: %v", id, f.ids))
+			}
+			idsIdx[id] = struct{}{}
+		}
+
+		prev := common.Pgid(0)
+		for _, id := range ids {
+			// The ids shouldn't have duplicated free ID. Note page 0 and 1
+			// are reserved for meta pages, so they can never be free page IDs.
+			if prev == id {
+				panic(fmt.Sprintf("detected duplicated free ID: %d in ids: %v", id, ids))
+			}
+			prev = id
+
+			// The ids shouldn't have any overlap with the existing f.ids.
+			if _, ok := idsIdx[id]; ok {
+				panic(fmt.Sprintf("detected overlapped free page ID: %d between ids: %v and existing f.ids: %v", id, ids, f.ids))
+			}
+		}
+	})
+	f.ids = common.Pgids(f.ids).Merge(ids)
+}
+
+func NewArrayFreelist() Interface {
+	a := &array{
+		shared: newShared(),
+		ids:    []common.Pgid{},
+	}
+	a.Interface = a
+	return a
+}

vendor/go.etcd.io/bbolt/internal/freelist/freelist.go

@@ -0,0 +1,82 @@
+package freelist
+
+import (
+	"go.etcd.io/bbolt/internal/common"
+)
+
+type ReadWriter interface {
+	// Read calls Init with the page ids stored in the given page.
+	Read(page *common.Page)
+
+	// Write writes the freelist into the given page.
+	Write(page *common.Page)
+
+	// EstimatedWritePageSize returns the size in bytes of the freelist after serialization in Write.
+	// This should never underestimate the size.
+	EstimatedWritePageSize() int
+}
+
+type Interface interface {
+	ReadWriter
+
+	// Init initializes this freelist with the given list of pages.
+	Init(ids common.Pgids)
+
+	// Allocate tries to allocate the given number of contiguous pages
+	// from the free list pages. It returns the starting page ID if
+	// available; otherwise, it returns 0.
+	Allocate(txid common.Txid, numPages int) common.Pgid
+
+	// Count returns the number of free and pending pages.
+	Count() int
+
+	// FreeCount returns the number of free pages.
+	FreeCount() int
+
+	// PendingCount returns the number of pending pages.
+	PendingCount() int
+
+	// AddReadonlyTXID adds a given read-only transaction id for pending page tracking.
+	AddReadonlyTXID(txid common.Txid)
+
+	// RemoveReadonlyTXID removes a given read-only transaction id for pending page tracking.
+	RemoveReadonlyTXID(txid common.Txid)
+
+	// ReleasePendingPages releases any pages associated with closed read-only transactions.
+	ReleasePendingPages()
+
+	// Free releases a page and its overflow for a given transaction id.
+	// If the page is already free or is one of the meta pages, then a panic will occur.
+	Free(txId common.Txid, p *common.Page)
+
+	// Freed returns whether a given page is in the free list.
+	Freed(pgId common.Pgid) bool
+
+	// Rollback removes the pages from a given pending tx.
+	Rollback(txId common.Txid)
+
+	// Copyall copies a list of all free ids and all pending ids in one sorted list.
+	// f.count returns the minimum length required for dst.
+	Copyall(dst []common.Pgid)
+
+	// Reload reads the freelist from a page and filters out pending items.
+	Reload(p *common.Page)
+
+	// NoSyncReload reads the freelist from Pgids and filters out pending items.
+	NoSyncReload(pgIds common.Pgids)
+
+	// freePageIds returns the IDs of all free pages. Returns an empty slice if no free pages are available.
+	freePageIds() common.Pgids
+
+	// pendingPageIds returns all pending pages by transaction id.
+	pendingPageIds() map[common.Txid]*txPending
+
+	// release moves all page ids for a transaction id (or older) to the freelist.
+	release(txId common.Txid)
+
+	// releaseRange moves pending pages allocated within an extent [begin,end] to the free list.
+	releaseRange(begin, end common.Txid)
+
+	// mergeSpans is merging the given pages into the freelist
+	mergeSpans(ids common.Pgids)
+}

vendor/go.etcd.io/bbolt/internal/freelist/hashmap.go

@@ -0,0 +1,292 @@
+package freelist
+
+import (
+	"fmt"
+	"reflect"
+	"sort"
+
+	"go.etcd.io/bbolt/internal/common"
+)
+
+// pidSet holds the set of starting pgids which have the same span size
+type pidSet map[common.Pgid]struct{}
+
+type hashMap struct {
+	*shared
+
+	freePagesCount uint64                 // count of free pages(hashmap version)
+	freemaps       map[uint64]pidSet      // key is the size of continuous pages(span), value is a set which contains the starting pgids of same size
+	forwardMap     map[common.Pgid]uint64 // key is start pgid, value is its span size
+	backwardMap    map[common.Pgid]uint64 // key is end pgid, value is its span size
+}
+
+func (f *hashMap) Init(pgids common.Pgids) {
+	// reset the counter when freelist init
+	f.freePagesCount = 0
+	f.freemaps = make(map[uint64]pidSet)
+	f.forwardMap = make(map[common.Pgid]uint64)
+	f.backwardMap = make(map[common.Pgid]uint64)
+
+	if len(pgids) == 0 {
+		return
+	}
+
+	if !sort.SliceIsSorted([]common.Pgid(pgids), func(i, j int) bool { return pgids[i] < pgids[j] }) {
+		panic("pgids not sorted")
+	}
+
+	size := uint64(1)
+	start := pgids[0]
+
+	for i := 1; i < len(pgids); i++ {
+		// continuous page
+		if pgids[i] == pgids[i-1]+1 {
+			size++
+		} else {
+			f.addSpan(start, size)
+
+			size = 1
+			start = pgids[i]
+		}
+	}
+
+	// init the tail
+	if size != 0 && start != 0 {
+		f.addSpan(start, size)
+	}
+
+	f.reindex()
+}
+
+func (f *hashMap) Allocate(txid common.Txid, n int) common.Pgid {
+	if n == 0 {
+		return 0
+	}
+
+	// if we have a exact size match just return short path
+	if bm, ok := f.freemaps[uint64(n)]; ok {
+		for pid := range bm {
+			// remove the span
+			f.delSpan(pid, uint64(n))
+
+			f.allocs[pid] = txid
+
+			for i := common.Pgid(0); i < common.Pgid(n); i++ {
+				delete(f.cache, pid+i)
+			}
+			return pid
+		}
+	}
+
+	// lookup the map to find larger span
+	for size, bm := range f.freemaps {
+		if size < uint64(n) {
+			continue
+		}
+
+		for pid := range bm {
+			// remove the initial
+			f.delSpan(pid, size)
+
+			f.allocs[pid] = txid
+
+			remain := size - uint64(n)
+
+			// add remain span
+			f.addSpan(pid+common.Pgid(n), remain)
+
+			for i := common.Pgid(0); i < common.Pgid(n); i++ {
+				delete(f.cache, pid+i)
+			}
+			return pid
+		}
+	}
+
+	return 0
+}
+
+func (f *hashMap) FreeCount() int {
+	common.Verify(func() {
+		expectedFreePageCount := f.hashmapFreeCountSlow()
+		common.Assert(int(f.freePagesCount) == expectedFreePageCount,
+			"freePagesCount (%d) is out of sync with free pages map (%d)", f.freePagesCount, expectedFreePageCount)
+	})
+	return int(f.freePagesCount)
+}
+
+func (f *hashMap) freePageIds() common.Pgids {
+	count := f.FreeCount()
+	if count == 0 {
+		return common.Pgids{}
+	}
+
+	m := make([]common.Pgid, 0, count)
+
+	startPageIds := make([]common.Pgid, 0, len(f.forwardMap))
+	for k := range f.forwardMap {
+		startPageIds = append(startPageIds, k)
+	}
+	sort.Sort(common.Pgids(startPageIds))
+
+	for _, start := range startPageIds {
+		if size, ok := f.forwardMap[start]; ok {
+			for i := 0; i < int(size); i++ {
+				m = append(m, start+common.Pgid(i))
+			}
+		}
+	}
+
+	return m
+}
+
+func (f *hashMap) hashmapFreeCountSlow() int {
+	count := 0
+	for _, size := range f.forwardMap {
+		count += int(size)
+	}
+	return count
+}
+
+func (f *hashMap) addSpan(start common.Pgid, size uint64) {
+	f.backwardMap[start-1+common.Pgid(size)] = size
+	f.forwardMap[start] = size
+	if _, ok := f.freemaps[size]; !ok {
+		f.freemaps[size] = make(map[common.Pgid]struct{})
+	}
+
+	f.freemaps[size][start] = struct{}{}
+	f.freePagesCount += size
+}
+
+func (f *hashMap) delSpan(start common.Pgid, size uint64) {
+	delete(f.forwardMap, start)
+	delete(f.backwardMap, start+common.Pgid(size-1))
+	delete(f.freemaps[size], start)
+	if len(f.freemaps[size]) == 0 {
+		delete(f.freemaps, size)
+	}
+	f.freePagesCount -= size
+}
+
+func (f *hashMap) mergeSpans(ids common.Pgids) {
+	common.Verify(func() {
+		ids1Freemap := f.idsFromFreemaps()
+		ids2Forward := f.idsFromForwardMap()
+		ids3Backward := f.idsFromBackwardMap()
+
+		if !reflect.DeepEqual(ids1Freemap, ids2Forward) {
+			panic(fmt.Sprintf("Detected mismatch, f.freemaps: %v, f.forwardMap: %v", f.freemaps, f.forwardMap))
+		}
+		if !reflect.DeepEqual(ids1Freemap, ids3Backward) {
+			panic(fmt.Sprintf("Detected mismatch, f.freemaps: %v, f.backwardMap: %v", f.freemaps, f.backwardMap))
+		}
+
+		sort.Sort(ids)
+		prev := common.Pgid(0)
+		for _, id := range ids {
+			// The ids shouldn't have duplicated free ID.
+			if prev == id {
+				panic(fmt.Sprintf("detected duplicated free ID: %d in ids: %v", id, ids))
+			}
+			prev = id
+
+			// The ids shouldn't have any overlap with the existing f.freemaps.
+			if _, ok := ids1Freemap[id]; ok {
+				panic(fmt.Sprintf("detected overlapped free page ID: %d between ids: %v and existing f.freemaps: %v", id, ids, f.freemaps))
+			}
+		}
+	})
+	for _, id := range ids {
+		// try to see if we can merge and update
+		f.mergeWithExistingSpan(id)
+	}
+}
+
+// mergeWithExistingSpan merges pid to the existing free spans, try to merge it backward and forward
+func (f *hashMap) mergeWithExistingSpan(pid common.Pgid) {
+	prev := pid - 1
+	next := pid + 1
+
+	preSize, mergeWithPrev := f.backwardMap[prev]
+	nextSize, mergeWithNext := f.forwardMap[next]
+	newStart := pid
+	newSize := uint64(1)
+
+	if mergeWithPrev {
+		//merge with previous span
+		start := prev + 1 - common.Pgid(preSize)
+		f.delSpan(start, preSize)
+
+		newStart -= common.Pgid(preSize)
+		newSize += preSize
+	}
+
+	if mergeWithNext {
+		// merge with next span
+		f.delSpan(next, nextSize)
+		newSize += nextSize
+	}
+
+	f.addSpan(newStart, newSize)
+}
+
+// idsFromFreemaps get all free page IDs from f.freemaps.
+// used by test only.
+func (f *hashMap) idsFromFreemaps() map[common.Pgid]struct{} {
+	ids := make(map[common.Pgid]struct{})
+	for size, idSet := range f.freemaps {
+		for start := range idSet {
+			for i := 0; i < int(size); i++ {
+				id := start + common.Pgid(i)
+				if _, ok := ids[id]; ok {
+					panic(fmt.Sprintf("detected duplicated free page ID: %d in f.freemaps: %v", id, f.freemaps))
+				}
+				ids[id] = struct{}{}
+			}
+		}
+	}
+	return ids
+}
+
+// idsFromForwardMap get all free page IDs from f.forwardMap.
+// used by test only.
+func (f *hashMap) idsFromForwardMap() map[common.Pgid]struct{} {
+	ids := make(map[common.Pgid]struct{})
+	for start, size := range f.forwardMap {
+		for i := 0; i < int(size); i++ {
+			id := start + common.Pgid(i)
+			if _, ok := ids[id]; ok {
+				panic(fmt.Sprintf("detected duplicated free page ID: %d in f.forwardMap: %v", id, f.forwardMap))
+			}
+			ids[id] = struct{}{}
+		}
+	}
+	return ids
+}
+
+// idsFromBackwardMap get all free page IDs from f.backwardMap.
+// used by test only.
+func (f *hashMap) idsFromBackwardMap() map[common.Pgid]struct{} {
+	ids := make(map[common.Pgid]struct{})
+	for end, size := range f.backwardMap {
+		for i := 0; i < int(size); i++ {
+			id := end - common.Pgid(i)
+			if _, ok := ids[id]; ok {
+				panic(fmt.Sprintf("detected duplicated free page ID: %d in f.backwardMap: %v", id, f.backwardMap))
+			}
+			ids[id] = struct{}{}
+		}
+	}
+	return ids
+}
+
+func NewHashMapFreelist() Interface {
+	hm := &hashMap{
+		shared:      newShared(),
+		freemaps:    make(map[uint64]pidSet),
+		forwardMap:  make(map[common.Pgid]uint64),
+		backwardMap: make(map[common.Pgid]uint64),
+	}
+	hm.Interface = hm
+	return hm
+}

vendor/go.etcd.io/bbolt/internal/freelist/shared.go

@@ -0,0 +1,310 @@
+package freelist
+
+import (
+	"fmt"
+	"math"
+	"sort"
+	"unsafe"
+
+	"go.etcd.io/bbolt/internal/common"
+)
+
+type txPending struct {
+	ids              []common.Pgid
+	alloctx          []common.Txid // txids allocating the ids
+	lastReleaseBegin common.Txid   // beginning txid of last matching releaseRange
+}
+
+type shared struct {
+	Interface
+
+	readonlyTXIDs []common.Txid               // all readonly transaction IDs.
+	allocs        map[common.Pgid]common.Txid // mapping of Txid that allocated a pgid.
+	cache         map[common.Pgid]struct{}    // fast lookup of all free and pending page ids.
+	pending       map[common.Txid]*txPending  // mapping of soon-to-be free page ids by tx.
+}
+
+func newShared() *shared {
+	return &shared{
+		pending: make(map[common.Txid]*txPending),
+		allocs:  make(map[common.Pgid]common.Txid),
+		cache:   make(map[common.Pgid]struct{}),
+	}
+}
+
+func (t *shared) pendingPageIds() map[common.Txid]*txPending {
+	return t.pending
+}
+
+func (t *shared) PendingCount() int {
+	var count int
+	for _, txp := range t.pending {
+		count += len(txp.ids)
+	}
+	return count
+}
+
+func (t *shared) Count() int {
+	return t.FreeCount() + t.PendingCount()
+}
+
+func (t *shared) Freed(pgId common.Pgid) bool {
+	_, ok := t.cache[pgId]
+	return ok
+}
+
+func (t *shared) Free(txid common.Txid, p *common.Page) {
+	if p.Id() <= 1 {
+		panic(fmt.Sprintf("cannot free page 0 or 1: %d", p.Id()))
+	}
+
+	// Free page and all its overflow pages.
+	txp := t.pending[txid]
+	if txp == nil {
+		txp = &txPending{}
+		t.pending[txid] = txp
+	}
+	allocTxid, ok := t.allocs[p.Id()]
+	common.Verify(func() {
+		if allocTxid == txid {
+			panic(fmt.Sprintf("free: freed page (%d) was allocated by the same transaction (%d)", p.Id(), txid))
+		}
+	})
+	if ok {
+		delete(t.allocs, p.Id())
+	}
+
+	for id := p.Id(); id <= p.Id()+common.Pgid(p.Overflow()); id++ {
+		// Verify that page is not already free.
+		if _, ok := t.cache[id]; ok {
+			panic(fmt.Sprintf("page %d already freed", id))
+		}
+		// Add to the freelist and cache.
+		txp.ids = append(txp.ids, id)
+		txp.alloctx = append(txp.alloctx, allocTxid)
+		t.cache[id] = struct{}{}
+	}
+}
+
+func (t *shared) Rollback(txid common.Txid) {
+	// Remove page ids from cache.
+	txp := t.pending[txid]
+	if txp == nil {
+		return
+	}
+	for i, pgid := range txp.ids {
+		delete(t.cache, pgid)
+		tx := txp.alloctx[i]
+		if tx == 0 {
+			continue
+		}
+		if tx != txid {
+			// Pending free aborted; restore page back to alloc list.
+			t.allocs[pgid] = tx
+		} else {
+			// A writing TXN should never free a page which was allocated by itself.
+			panic(fmt.Sprintf("rollback: freed page (%d) was allocated by the same transaction (%d)", pgid, txid))
+		}
+	}
+	// Remove pages from pending list and mark as free if allocated by txid.
+	delete(t.pending, txid)
+
+	// Remove pgids which are allocated by this txid
+	for pgid, tid := range t.allocs {
+		if tid == txid {
+			delete(t.allocs, pgid)
+		}
+	}
+}
+
+func (t *shared) AddReadonlyTXID(tid common.Txid) {
+	t.readonlyTXIDs = append(t.readonlyTXIDs, tid)
+}
+
+func (t *shared) RemoveReadonlyTXID(tid common.Txid) {
+	for i := range t.readonlyTXIDs {
+		if t.readonlyTXIDs[i] == tid {
+			last := len(t.readonlyTXIDs) - 1
+			t.readonlyTXIDs[i] = t.readonlyTXIDs[last]
+			t.readonlyTXIDs = t.readonlyTXIDs[:last]
+			break
+		}
+	}
+}
+
+type txIDx []common.Txid
+
+func (t txIDx) Len() int           { return len(t) }
+func (t txIDx) Swap(i, j int)      { t[i], t[j] = t[j], t[i] }
+func (t txIDx) Less(i, j int) bool { return t[i] < t[j] }
+
+func (t *shared) ReleasePendingPages() {
+	// Free all pending pages prior to the earliest open transaction.
+	sort.Sort(txIDx(t.readonlyTXIDs))
+	minid := common.Txid(math.MaxUint64)
+	if len(t.readonlyTXIDs) > 0 {
+		minid = t.readonlyTXIDs[0]
+	}
+	if minid > 0 {
+		t.release(minid - 1)
+	}
+	// Release unused txid extents.
+	for _, tid := range t.readonlyTXIDs {
+		t.releaseRange(minid, tid-1)
+		minid = tid + 1
+	}
+	t.releaseRange(minid, common.Txid(math.MaxUint64))
+	// Any page both allocated and freed in an extent is safe to release.
+}
+
+func (t *shared) release(txid common.Txid) {
+	m := make(common.Pgids, 0)
+	for tid, txp := range t.pending {
+		if tid <= txid {
+			// Move transaction's pending pages to the available freelist.
+			// Don't remove from the cache since the page is still free.
+			m = append(m, txp.ids...)
+			delete(t.pending, tid)
+		}
+	}
+	t.mergeSpans(m)
+}
+
+func (t *shared) releaseRange(begin, end common.Txid) {
+	if begin > end {
+		return
+	}
+	m := common.Pgids{}
+	for tid, txp := range t.pending {
+		if tid < begin || tid > end {
+			continue
+		}
+		// Don't recompute freed pages if ranges haven't updated.
+		if txp.lastReleaseBegin == begin {
+			continue
+		}
+		for i := 0; i < len(txp.ids); i++ {
+			if atx := txp.alloctx[i]; atx < begin || atx > end {
+				continue
+			}
+			m = append(m, txp.ids[i])
+			txp.ids[i] = txp.ids[len(txp.ids)-1]
+			txp.ids = txp.ids[:len(txp.ids)-1]
+			txp.alloctx[i] = txp.alloctx[len(txp.alloctx)-1]
+			txp.alloctx = txp.alloctx[:len(txp.alloctx)-1]
+			i--
+		}
+		txp.lastReleaseBegin = begin
+		if len(txp.ids) == 0 {
+			delete(t.pending, tid)
+		}
+	}
+	t.mergeSpans(m)
+}
+
+// Copyall copies a list of all free ids and all pending ids in one sorted list.
+// f.count returns the minimum length required for dst.
+func (t *shared) Copyall(dst []common.Pgid) {
+	m := make(common.Pgids, 0, t.PendingCount())
+	for _, txp := range t.pendingPageIds() {
+		m = append(m, txp.ids...)
+	}
+	sort.Sort(m)
+	common.Mergepgids(dst, t.freePageIds(), m)
+}
+
+func (t *shared) Reload(p *common.Page) {
+	t.Read(p)
+	t.NoSyncReload(t.freePageIds())
+}
+
+func (t *shared) NoSyncReload(pgIds common.Pgids) {
+	// Build a cache of only pending pages.
+	pcache := make(map[common.Pgid]bool)
+	for _, txp := range t.pending {
+		for _, pendingID := range txp.ids {
+			pcache[pendingID] = true
+		}
+	}
+
+	// Check each page in the freelist and build a new available freelist
+	// with any pages not in the pending lists.
+	a := []common.Pgid{}
+	for _, id := range pgIds {
+		if !pcache[id] {
+			a = append(a, id)
+		}
+	}
+
+	t.Init(a)
+}
+
+// reindex rebuilds the free cache based on available and pending free lists.
+func (t *shared) reindex() {
+	free := t.freePageIds()
+	pending := t.pendingPageIds()
+	t.cache = make(map[common.Pgid]struct{}, len(free))
+	for _, id := range free {
+		t.cache[id] = struct{}{}
+	}
+	for _, txp := range pending {
+		for _, pendingID := range txp.ids {
+			t.cache[pendingID] = struct{}{}
+		}
+	}
+}
+
+func (t *shared) Read(p *common.Page) {
+	if !p.IsFreelistPage() {
+		panic(fmt.Sprintf("invalid freelist page: %d, page type is %s", p.Id(), p.Typ()))
+	}
+
+	ids := p.FreelistPageIds()
+
+	// Copy the list of page ids from the freelist.
+	if len(ids) == 0 {
+		t.Init([]common.Pgid{})
+	} else {
+		// copy the ids, so we don't modify on the freelist page directly
+		idsCopy := make([]common.Pgid, len(ids))
+		copy(idsCopy, ids)
+		// Make sure they're sorted.
+		sort.Sort(common.Pgids(idsCopy))
+
+		t.Init(idsCopy)
+	}
+}
+
+func (t *shared) EstimatedWritePageSize() int {
+	n := t.Count()
+	if n >= 0xFFFF {
+		// The first element will be used to store the count. See freelist.write.
+		n++
+	}
+	return int(common.PageHeaderSize) + (int(unsafe.Sizeof(common.Pgid(0))) * n)
+}
+
+func (t *shared) Write(p *common.Page) {
+	// Combine the old free pgids and pgids waiting on an open transaction.
+
+	// Update the header flag.
+	p.SetFlags(common.FreelistPageFlag)
+
+	// The page.count can only hold up to 64k elements so if we overflow that
+	// number then we handle it by putting the size in the first element.
+	l := t.Count()
+	if l == 0 {
+		p.SetCount(uint16(l))
+	} else if l < 0xFFFF {
+		p.SetCount(uint16(l))
+		data := common.UnsafeAdd(unsafe.Pointer(p), unsafe.Sizeof(*p))
+		ids := unsafe.Slice((*common.Pgid)(data), l)
+		t.Copyall(ids)
+	} else {
+		p.SetCount(0xFFFF)
+		data := common.UnsafeAdd(unsafe.Pointer(p), unsafe.Sizeof(*p))
+		ids := unsafe.Slice((*common.Pgid)(data), l+1)
+		ids[0] = common.Pgid(l)
+		t.Copyall(ids[1:])
+	}
+}

vendor/go.etcd.io/bbolt/logger.go

@@ -0,0 +1,113 @@
+package bbolt
+
+// See https://github.com/etcd-io/raft/blob/main/logger.go
+import (
+	"fmt"
+	"io"
+	"log"
+	"os"
+)
+
+type Logger interface {
+	Debug(v ...interface{})
+	Debugf(format string, v ...interface{})
+
+	Error(v ...interface{})
+	Errorf(format string, v ...interface{})
+
+	Info(v ...interface{})
+	Infof(format string, v ...interface{})
+
+	Warning(v ...interface{})
+	Warningf(format string, v ...interface{})
+
+	Fatal(v ...interface{})
+	Fatalf(format string, v ...interface{})
+
+	Panic(v ...interface{})
+	Panicf(format string, v ...interface{})
+}
+
+func getDiscardLogger() Logger {
+	return discardLogger
+}
+
+var (
+	discardLogger = &DefaultLogger{Logger: log.New(io.Discard, "", 0)}
+)
+
+const (
+	calldepth = 2
+)
+
+// DefaultLogger is a default implementation of the Logger interface.
+type DefaultLogger struct {
+	*log.Logger
+	debug bool
+}
+
+func (l *DefaultLogger) EnableTimestamps() {
+	l.SetFlags(l.Flags() | log.Ldate | log.Ltime)
+}
+
+func (l *DefaultLogger) EnableDebug() {
+	l.debug = true
+}
+
+func (l *DefaultLogger) Debug(v ...interface{}) {
+	if l.debug {
+		_ = l.Output(calldepth, header("DEBUG", fmt.Sprint(v...)))
+	}
+}
+
+func (l *DefaultLogger) Debugf(format string, v ...interface{}) {
+	if l.debug {
+		_ = l.Output(calldepth, header("DEBUG", fmt.Sprintf(format, v...)))
+	}
+}
+
+func (l *DefaultLogger) Info(v ...interface{}) {
+	_ = l.Output(calldepth, header("INFO", fmt.Sprint(v...)))
+}
+
+func (l *DefaultLogger) Infof(format string, v ...interface{}) {
+	_ = l.Output(calldepth, header("INFO", fmt.Sprintf(format, v...)))
+}
+
+func (l *DefaultLogger) Error(v ...interface{}) {
+	_ = l.Output(calldepth, header("ERROR", fmt.Sprint(v...)))
+}
+
+func (l *DefaultLogger) Errorf(format string, v ...interface{}) {
+	_ = l.Output(calldepth, header("ERROR", fmt.Sprintf(format, v...)))
+}
+
+func (l *DefaultLogger) Warning(v ...interface{}) {
+	_ = l.Output(calldepth, header("WARN", fmt.Sprint(v...)))
+}
+
+func (l *DefaultLogger) Warningf(format string, v ...interface{}) {
+	_ = l.Output(calldepth, header("WARN", fmt.Sprintf(format, v...)))
+}
+
+func (l *DefaultLogger) Fatal(v ...interface{}) {
+	_ = l.Output(calldepth, header("FATAL", fmt.Sprint(v...)))
+	os.Exit(1)
+}
+
+func (l *DefaultLogger) Fatalf(format string, v ...interface{}) {
+	_ = l.Output(calldepth, header("FATAL", fmt.Sprintf(format, v...)))
+	os.Exit(1)
+}
+
+func (l *DefaultLogger) Panic(v ...interface{}) {
+	l.Logger.Panic(v...)
+}
+
+func (l *DefaultLogger) Panicf(format string, v ...interface{}) {
+	l.Logger.Panicf(format, v...)
+}
+
+func header(lvl, msg string) string {
+	return fmt.Sprintf("%s: %s", lvl, msg)
+}

vendor/go.etcd.io/bbolt/mlock_unix.go

@@ -0,0 +1,36 @@
+//go:build !windows
+
+package bbolt
+
+import "golang.org/x/sys/unix"
+
+// mlock locks memory of db file
+func mlock(db *DB, fileSize int) error {
+	sizeToLock := fileSize
+	if sizeToLock > db.datasz {
+		// Can't lock more than mmaped slice
+		sizeToLock = db.datasz
+	}
+	if err := unix.Mlock(db.dataref[:sizeToLock]); err != nil {
+		return err
+	}
+	return nil
+}
+
+// munlock unlocks memory of db file
+func munlock(db *DB, fileSize int) error {
+	if db.dataref == nil {
+		return nil
+	}
+
+	sizeToUnlock := fileSize
+	if sizeToUnlock > db.datasz {
+		// Can't unlock more than mmaped slice
+		sizeToUnlock = db.datasz
+	}
+
+	if err := unix.Munlock(db.dataref[:sizeToUnlock]); err != nil {
+		return err
+	}
+	return nil
+}

vendor/go.etcd.io/bbolt/mlock_windows.go

@@ -0,0 +1,11 @@
+package bbolt
+
+// mlock locks memory of db file
+func mlock(_ *DB, _ int) error {
+	panic("mlock is supported only on UNIX systems")
+}
+
+// munlock unlocks memory of db file
+func munlock(_ *DB, _ int) error {
+	panic("munlock is supported only on UNIX systems")
+}

vendor/go.etcd.io/bbolt/node.go

@@ -0,0 +1,538 @@
+package bbolt
+
+import (
+	"bytes"
+	"fmt"
+	"sort"
+
+	"go.etcd.io/bbolt/internal/common"
+)
+
+// node represents an in-memory, deserialized page.
+type node struct {
+	bucket     *Bucket
+	isLeaf     bool
+	unbalanced bool
+	spilled    bool
+	key        []byte
+	pgid       common.Pgid
+	parent     *node
+	children   nodes
+	inodes     common.Inodes
+}
+
+// root returns the top-level node this node is attached to.
+func (n *node) root() *node {
+	if n.parent == nil {
+		return n
+	}
+	return n.parent.root()
+}
+
+// minKeys returns the minimum number of inodes this node should have.
+func (n *node) minKeys() int {
+	if n.isLeaf {
+		return 1
+	}
+	return 2
+}
+
+// size returns the size of the node after serialization.
+func (n *node) size() int {
+	sz, elsz := common.PageHeaderSize, n.pageElementSize()
+	for i := 0; i < len(n.inodes); i++ {
+		item := &n.inodes[i]
+		sz += elsz + uintptr(len(item.Key())) + uintptr(len(item.Value()))
+	}
+	return int(sz)
+}
+
+// sizeLessThan returns true if the node is less than a given size.
+// This is an optimization to avoid calculating a large node when we only need
+// to know if it fits inside a certain page size.
+func (n *node) sizeLessThan(v uintptr) bool {
+	sz, elsz := common.PageHeaderSize, n.pageElementSize()
+	for i := 0; i < len(n.inodes); i++ {
+		item := &n.inodes[i]
+		sz += elsz + uintptr(len(item.Key())) + uintptr(len(item.Value()))
+		if sz >= v {
+			return false
+		}
+	}
+	return true
+}
+
+// pageElementSize returns the size of each page element based on the type of node.
+func (n *node) pageElementSize() uintptr {
+	if n.isLeaf {
+		return common.LeafPageElementSize
+	}
+	return common.BranchPageElementSize
+}
+
+// childAt returns the child node at a given index.
+func (n *node) childAt(index int) *node {
+	if n.isLeaf {
+		panic(fmt.Sprintf("invalid childAt(%d) on a leaf node", index))
+	}
+	return n.bucket.node(n.inodes[index].Pgid(), n)
+}
+
+// childIndex returns the index of a given child node.
+func (n *node) childIndex(child *node) int {
+	index := sort.Search(len(n.inodes), func(i int) bool { return bytes.Compare(n.inodes[i].Key(), child.key) != -1 })
+	return index
+}
+
+// numChildren returns the number of children.
+func (n *node) numChildren() int {
+	return len(n.inodes)
+}
+
+// nextSibling returns the next node with the same parent.
+func (n *node) nextSibling() *node {
+	if n.parent == nil {
+		return nil
+	}
+	index := n.parent.childIndex(n)
+	if index >= n.parent.numChildren()-1 {
+		return nil
+	}
+	return n.parent.childAt(index + 1)
+}
+
+// prevSibling returns the previous node with the same parent.
+func (n *node) prevSibling() *node {
+	if n.parent == nil {
+		return nil
+	}
+	index := n.parent.childIndex(n)
+	if index == 0 {
+		return nil
+	}
+	return n.parent.childAt(index - 1)
+}
+
+// put inserts a key/value.
+func (n *node) put(oldKey, newKey, value []byte, pgId common.Pgid, flags uint32) {
+	if pgId >= n.bucket.tx.meta.Pgid() {
+		panic(fmt.Sprintf("pgId (%d) above high water mark (%d)", pgId, n.bucket.tx.meta.Pgid()))
+	} else if len(oldKey) <= 0 {
+		panic("put: zero-length old key")
+	} else if len(newKey) <= 0 {
+		panic("put: zero-length new key")
+	}
+
+	// Find insertion index.
+	index := sort.Search(len(n.inodes), func(i int) bool { return bytes.Compare(n.inodes[i].Key(), oldKey) != -1 })
+
+	// Add capacity and shift nodes if we don't have an exact match and need to insert.
+	exact := len(n.inodes) > 0 && index < len(n.inodes) && bytes.Equal(n.inodes[index].Key(), oldKey)
+	if !exact {
+		n.inodes = append(n.inodes, common.Inode{})
+		copy(n.inodes[index+1:], n.inodes[index:])
+	}
+
+	inode := &n.inodes[index]
+	inode.SetFlags(flags)
+	inode.SetKey(newKey)
+	inode.SetValue(value)
+	inode.SetPgid(pgId)
+	common.Assert(len(inode.Key()) > 0, "put: zero-length inode key")
+}
+
+// del removes a key from the node.
+func (n *node) del(key []byte) {
+	// Find index of key.
+	index := sort.Search(len(n.inodes), func(i int) bool { return bytes.Compare(n.inodes[i].Key(), key) != -1 })
+
+	// Exit if the key isn't found.
+	if index >= len(n.inodes) || !bytes.Equal(n.inodes[index].Key(), key) {
+		return
+	}
+
+	// Delete inode from the node.
+	n.inodes = append(n.inodes[:index], n.inodes[index+1:]...)
+
+	// Mark the node as needing rebalancing.
+	n.unbalanced = true
+}
+
+// read initializes the node from a page.
+func (n *node) read(p *common.Page) {
+	n.pgid = p.Id()
+	n.isLeaf = p.IsLeafPage()
+	n.inodes = common.ReadInodeFromPage(p)
+
+	// Save first key, so we can find the node in the parent when we spill.
+	if len(n.inodes) > 0 {
+		n.key = n.inodes[0].Key()
+		common.Assert(len(n.key) > 0, "read: zero-length node key")
+	} else {
+		n.key = nil
+	}
+}
+
+// write writes the items onto one or more pages.
+// The page should have p.id (might be 0 for meta or bucket-inline page) and p.overflow set
+// and the rest should be zeroed.
+func (n *node) write(p *common.Page) {
+	common.Assert(p.Count() == 0 && p.Flags() == 0, "node cannot be written into a not empty page")
+
+	// Initialize page.
+	if n.isLeaf {
+		p.SetFlags(common.LeafPageFlag)
+	} else {
+		p.SetFlags(common.BranchPageFlag)
+	}
+
+	if len(n.inodes) >= 0xFFFF {
+		panic(fmt.Sprintf("inode overflow: %d (pgid=%d)", len(n.inodes), p.Id()))
+	}
+	p.SetCount(uint16(len(n.inodes)))
+
+	// Stop here if there are no items to write.
+	if p.Count() == 0 {
+		return
+	}
+
+	common.WriteInodeToPage(n.inodes, p)
+
+	// DEBUG ONLY: n.dump()
+}
+
+// split breaks up a node into multiple smaller nodes, if appropriate.
+// This should only be called from the spill() function.
+func (n *node) split(pageSize uintptr) []*node {
+	var nodes []*node
+
+	node := n
+	for {
+		// Split node into two.
+		a, b := node.splitTwo(pageSize)
+		nodes = append(nodes, a)
+
+		// If we can't split then exit the loop.
+		if b == nil {
+			break
+		}
+
+		// Set node to b so it gets split on the next iteration.
+		node = b
+	}
+
+	return nodes
+}
+
+// splitTwo breaks up a node into two smaller nodes, if appropriate.
+// This should only be called from the split() function.
+func (n *node) splitTwo(pageSize uintptr) (*node, *node) {
+	// Ignore the split if the page doesn't have at least enough nodes for
+	// two pages or if the nodes can fit in a single page.
+	if len(n.inodes) <= (common.MinKeysPerPage*2) || n.sizeLessThan(pageSize) {
+		return n, nil
+	}
+
+	// Determine the threshold before starting a new node.
+	var fillPercent = n.bucket.FillPercent
+	if fillPercent < minFillPercent {
+		fillPercent = minFillPercent
+	} else if fillPercent > maxFillPercent {
+		fillPercent = maxFillPercent
+	}
+	threshold := int(float64(pageSize) * fillPercent)
+
+	// Determine split position and sizes of the two pages.
+	splitIndex, _ := n.splitIndex(threshold)
+
+	// Split node into two separate nodes.
+	// If there's no parent then we'll need to create one.
+	if n.parent == nil {
+		n.parent = &node{bucket: n.bucket, children: []*node{n}}
+	}
+
+	// Create a new node and add it to the parent.
+	next := &node{bucket: n.bucket, isLeaf: n.isLeaf, parent: n.parent}
+	n.parent.children = append(n.parent.children, next)
+
+	// Split inodes across two nodes.
+	next.inodes = n.inodes[splitIndex:]
+	n.inodes = n.inodes[:splitIndex]
+
+	// Update the statistics.
+	n.bucket.tx.stats.IncSplit(1)
+
+	return n, next
+}
+
+// splitIndex finds the position where a page will fill a given threshold.
+// It returns the index as well as the size of the first page.
+// This is only be called from split().
+func (n *node) splitIndex(threshold int) (index, sz uintptr) {
+	sz = common.PageHeaderSize
+
+	// Loop until we only have the minimum number of keys required for the second page.
+	for i := 0; i < len(n.inodes)-common.MinKeysPerPage; i++ {
+		index = uintptr(i)
+		inode := n.inodes[i]
+		elsize := n.pageElementSize() + uintptr(len(inode.Key())) + uintptr(len(inode.Value()))
+
+		// If we have at least the minimum number of keys and adding another
+		// node would put us over the threshold then exit and return.
+		if index >= common.MinKeysPerPage && sz+elsize > uintptr(threshold) {
+			break
+		}
+
+		// Add the element size to the total size.
+		sz += elsize
+	}
+
+	return
+}
+
+// spill writes the nodes to dirty pages and splits nodes as it goes.
+// Returns an error if dirty pages cannot be allocated.
+func (n *node) spill() error {
+	var tx = n.bucket.tx
+	if n.spilled {
+		return nil
+	}
+
+	// Spill child nodes first. Child nodes can materialize sibling nodes in
+	// the case of split-merge so we cannot use a range loop. We have to check
+	// the children size on every loop iteration.
+	sort.Sort(n.children)
+	for i := 0; i < len(n.children); i++ {
+		if err := n.children[i].spill(); err != nil {
+			return err
+		}
+	}
+
+	// We no longer need the child list because it's only used for spill tracking.
+	n.children = nil
+
+	// Split nodes into appropriate sizes. The first node will always be n.
+	var nodes = n.split(uintptr(tx.db.pageSize))
+	for _, node := range nodes {
+		// Add node's page to the freelist if it's not new.
+		if node.pgid > 0 {
+			tx.db.freelist.Free(tx.meta.Txid(), tx.page(node.pgid))
+			node.pgid = 0
+		}
+
+		// Allocate contiguous space for the node.
+		p, err := tx.allocate((node.size() + tx.db.pageSize - 1) / tx.db.pageSize)
+		if err != nil {
+			return err
+		}
+
+		// Write the node.
+		if p.Id() >= tx.meta.Pgid() {
+			panic(fmt.Sprintf("pgid (%d) above high water mark (%d)", p.Id(), tx.meta.Pgid()))
+		}
+		node.pgid = p.Id()
+		node.write(p)
+		node.spilled = true
+
+		// Insert into parent inodes.
+		if node.parent != nil {
+			var key = node.key
+			if key == nil {
+				key = node.inodes[0].Key()
+			}
+
+			node.parent.put(key, node.inodes[0].Key(), nil, node.pgid, 0)
+			node.key = node.inodes[0].Key()
+			common.Assert(len(node.key) > 0, "spill: zero-length node key")
+		}
+
+		// Update the statistics.
+		tx.stats.IncSpill(1)
+	}
+
+	// If the root node split and created a new root then we need to spill that
+	// as well. We'll clear out the children to make sure it doesn't try to respill.
+	if n.parent != nil && n.parent.pgid == 0 {
+		n.children = nil
+		return n.parent.spill()
+	}
+
+	return nil
+}
+
+// rebalance attempts to combine the node with sibling nodes if the node fill
+// size is below a threshold or if there are not enough keys.
+func (n *node) rebalance() {
+	if !n.unbalanced {
+		return
+	}
+	n.unbalanced = false
+
+	// Update statistics.
+	n.bucket.tx.stats.IncRebalance(1)
+
+	// Ignore if node is above threshold (25% when FillPercent is set to DefaultFillPercent) and has enough keys.
+	var threshold = int(float64(n.bucket.tx.db.pageSize)*n.bucket.FillPercent) / 2
+	if n.size() > threshold && len(n.inodes) > n.minKeys() {
+		return
+	}
+
+	// Root node has special handling.
+	if n.parent == nil {
+		// If root node is a branch and only has one node then collapse it.
+		if !n.isLeaf && len(n.inodes) == 1 {
+			// Move root's child up.
+			child := n.bucket.node(n.inodes[0].Pgid(), n)
+			n.isLeaf = child.isLeaf
+			n.inodes = child.inodes[:]
+			n.children = child.children
+
+			// Reparent all child nodes being moved.
+			for _, inode := range n.inodes {
+				if child, ok := n.bucket.nodes[inode.Pgid()]; ok {
+					child.parent = n
+				}
+			}
+
+			// Remove old child.
+			child.parent = nil
+			delete(n.bucket.nodes, child.pgid)
+			child.free()
+		}
+
+		return
+	}
+
+	// If node has no keys then just remove it.
+	if n.numChildren() == 0 {
+		n.parent.del(n.key)
+		n.parent.removeChild(n)
+		delete(n.bucket.nodes, n.pgid)
+		n.free()
+		n.parent.rebalance()
+		return
+	}
+
+	common.Assert(n.parent.numChildren() > 1, "parent must have at least 2 children")
+
+	// Merge with right sibling if idx == 0, otherwise left sibling.
+	var leftNode, rightNode *node
+	var useNextSibling = n.parent.childIndex(n) == 0
+	if useNextSibling {
+		leftNode = n
+		rightNode = n.nextSibling()
+	} else {
+		leftNode = n.prevSibling()
+		rightNode = n
+	}
+
+	// If both nodes are too small then merge them.
+	// Reparent all child nodes being moved.
+	for _, inode := range rightNode.inodes {
+		if child, ok := n.bucket.nodes[inode.Pgid()]; ok {
+			child.parent.removeChild(child)
+			child.parent = leftNode
+			child.parent.children = append(child.parent.children, child)
+		}
+	}
+
+	// Copy over inodes from right node to left node and remove right node.
+	leftNode.inodes = append(leftNode.inodes, rightNode.inodes...)
+	n.parent.del(rightNode.key)
+	n.parent.removeChild(rightNode)
+	delete(n.bucket.nodes, rightNode.pgid)
+	rightNode.free()
+
+	// Either this node or the sibling node was deleted from the parent so rebalance it.
+	n.parent.rebalance()
+}
+
+// removes a node from the list of in-memory children.
+// This does not affect the inodes.
+func (n *node) removeChild(target *node) {
+	for i, child := range n.children {
+		if child == target {
+			n.children = append(n.children[:i], n.children[i+1:]...)
+			return
+		}
+	}
+}
+
+// dereference causes the node to copy all its inode key/value references to heap memory.
+// This is required when the mmap is reallocated so inodes are not pointing to stale data.
+func (n *node) dereference() {
+	if n.key != nil {
+		key := make([]byte, len(n.key))
+		copy(key, n.key)
+		n.key = key
+		common.Assert(n.pgid == 0 || len(n.key) > 0, "dereference: zero-length node key on existing node")
+	}
+
+	for i := range n.inodes {
+		inode := &n.inodes[i]
+
+		key := make([]byte, len(inode.Key()))
+		copy(key, inode.Key())
+		inode.SetKey(key)
+		common.Assert(len(inode.Key()) > 0, "dereference: zero-length inode key")
+
+		value := make([]byte, len(inode.Value()))
+		copy(value, inode.Value())
+		inode.SetValue(value)
+	}
+
+	// Recursively dereference children.
+	for _, child := range n.children {
+		child.dereference()
+	}
+
+	// Update statistics.
+	n.bucket.tx.stats.IncNodeDeref(1)
+}
+
+// free adds the node's underlying page to the freelist.
+func (n *node) free() {
+	if n.pgid != 0 {
+		n.bucket.tx.db.freelist.Free(n.bucket.tx.meta.Txid(), n.bucket.tx.page(n.pgid))
+		n.pgid = 0
+	}
+}
+
+// dump writes the contents of the node to STDERR for debugging purposes.
+/*
+func (n *node) dump() {
+	// Write node header.
+	var typ = "branch"
+	if n.isLeaf {
+		typ = "leaf"
+	}
+	warnf("[NODE %d {type=%s count=%d}]", n.pgid, typ, len(n.inodes))
+
+	// Write out abbreviated version of each item.
+	for _, item := range n.inodes {
+		if n.isLeaf {
+			if item.flags&bucketLeafFlag != 0 {
+				bucket := (*bucket)(unsafe.Pointer(&item.value[0]))
+				warnf("+L %08x -> (bucket root=%d)", trunc(item.key, 4), bucket.root)
+			} else {
+				warnf("+L %08x -> %08x", trunc(item.key, 4), trunc(item.value, 4))
+			}
+		} else {
+			warnf("+B %08x -> pgid=%d", trunc(item.key, 4), item.pgid)
+		}
+	}
+	warn("")
+}
+*/
+
+func compareKeys(left, right []byte) int {
+	return bytes.Compare(left, right)
+}
+
+type nodes []*node
+
+func (s nodes) Len() int      { return len(s) }
+func (s nodes) Swap(i, j int) { s[i], s[j] = s[j], s[i] }
+func (s nodes) Less(i, j int) bool {
+	return bytes.Compare(s[i].inodes[0].Key(), s[j].inodes[0].Key()) == -1
+}

vendor/go.etcd.io/bbolt/tx.go

@@ -0,0 +1,858 @@
+package bbolt
+
+import (
+	"errors"
+	"fmt"
+	"io"
+	"os"
+	"runtime"
+	"sort"
+	"strings"
+	"sync/atomic"
+	"time"
+	"unsafe"
+
+	berrors "go.etcd.io/bbolt/errors"
+	"go.etcd.io/bbolt/internal/common"
+)
+
+// Tx represents a read-only or read/write transaction on the database.
+// Read-only transactions can be used for retrieving values for keys and creating cursors.
+// Read/write transactions can create and remove buckets and create and remove keys.
+//
+// IMPORTANT: You must commit or rollback transactions when you are done with
+// them. Pages can not be reclaimed by the writer until no more transactions
+// are using them. A long running read transaction can cause the database to
+// quickly grow.
+type Tx struct {
+	writable       bool
+	managed        bool
+	db             *DB
+	meta           *common.Meta
+	root           Bucket
+	pages          map[common.Pgid]*common.Page
+	stats          TxStats
+	commitHandlers []func()
+
+	// WriteFlag specifies the flag for write-related methods like WriteTo().
+	// Tx opens the database file with the specified flag to copy the data.
+	//
+	// By default, the flag is unset, which works well for mostly in-memory
+	// workloads. For databases that are much larger than available RAM,
+	// set the flag to syscall.O_DIRECT to avoid trashing the page cache.
+	WriteFlag int
+}
+
+// init initializes the transaction.
+func (tx *Tx) init(db *DB) {
+	tx.db = db
+	tx.pages = nil
+
+	// Copy the meta page since it can be changed by the writer.
+	tx.meta = &common.Meta{}
+	db.meta().Copy(tx.meta)
+
+	// Copy over the root bucket.
+	tx.root = newBucket(tx)
+	tx.root.InBucket = &common.InBucket{}
+	*tx.root.InBucket = *(tx.meta.RootBucket())
+
+	// Increment the transaction id and add a page cache for writable transactions.
+	if tx.writable {
+		tx.pages = make(map[common.Pgid]*common.Page)
+		tx.meta.IncTxid()
+	}
+}
+
+// ID returns the transaction id.
+func (tx *Tx) ID() int {
+	if tx == nil || tx.meta == nil {
+		return -1
+	}
+	return int(tx.meta.Txid())
+}
+
+// DB returns a reference to the database that created the transaction.
+func (tx *Tx) DB() *DB {
+	return tx.db
+}
+
+// Size returns current database size in bytes as seen by this transaction.
+func (tx *Tx) Size() int64 {
+	return int64(tx.meta.Pgid()) * int64(tx.db.pageSize)
+}
+
+// Writable returns whether the transaction can perform write operations.
+func (tx *Tx) Writable() bool {
+	return tx.writable
+}
+
+// Cursor creates a cursor associated with the root bucket.
+// All items in the cursor will return a nil value because all root bucket keys point to buckets.
+// The cursor is only valid as long as the transaction is open.
+// Do not use a cursor after the transaction is closed.
+func (tx *Tx) Cursor() *Cursor {
+	return tx.root.Cursor()
+}
+
+// Stats retrieves a copy of the current transaction statistics.
+func (tx *Tx) Stats() TxStats {
+	return tx.stats
+}
+
+// Inspect returns the structure of the database.
+func (tx *Tx) Inspect() BucketStructure {
+	return tx.root.Inspect()
+}
+
+// Bucket retrieves a bucket by name.
+// Returns nil if the bucket does not exist.
+// The bucket instance is only valid for the lifetime of the transaction.
+func (tx *Tx) Bucket(name []byte) *Bucket {
+	return tx.root.Bucket(name)
+}
+
+// CreateBucket creates a new bucket.
+// Returns an error if the bucket already exists, if the bucket name is blank, or if the bucket name is too long.
+// The bucket instance is only valid for the lifetime of the transaction.
+func (tx *Tx) CreateBucket(name []byte) (*Bucket, error) {
+	return tx.root.CreateBucket(name)
+}
+
+// CreateBucketIfNotExists creates a new bucket if it doesn't already exist.
+// Returns an error if the bucket name is blank, or if the bucket name is too long.
+// The bucket instance is only valid for the lifetime of the transaction.
+func (tx *Tx) CreateBucketIfNotExists(name []byte) (*Bucket, error) {
+	return tx.root.CreateBucketIfNotExists(name)
+}
+
+// DeleteBucket deletes a bucket.
+// Returns an error if the bucket cannot be found or if the key represents a non-bucket value.
+func (tx *Tx) DeleteBucket(name []byte) error {
+	return tx.root.DeleteBucket(name)
+}
+
+// MoveBucket moves a sub-bucket from the source bucket to the destination bucket.
+// Returns an error if
+//  1. the sub-bucket cannot be found in the source bucket;
+//  2. or the key already exists in the destination bucket;
+//  3. the key represents a non-bucket value.
+//
+// If src is nil, it means moving a top level bucket into the target bucket.
+// If dst is nil, it means converting the child bucket into a top level bucket.
+func (tx *Tx) MoveBucket(child []byte, src *Bucket, dst *Bucket) error {
+	if src == nil {
+		src = &tx.root
+	}
+	if dst == nil {
+		dst = &tx.root
+	}
+	return src.MoveBucket(child, dst)
+}
+
+// ForEach executes a function for each bucket in the root.
+// If the provided function returns an error then the iteration is stopped and
+// the error is returned to the caller.
+func (tx *Tx) ForEach(fn func(name []byte, b *Bucket) error) error {
+	return tx.root.ForEach(func(k, v []byte) error {
+		return fn(k, tx.root.Bucket(k))
+	})
+}
+
+// OnCommit adds a handler function to be executed after the transaction successfully commits.
+func (tx *Tx) OnCommit(fn func()) {
+	tx.commitHandlers = append(tx.commitHandlers, fn)
+}
+
+// Commit writes all changes to disk, updates the meta page and closes the transaction.
+// Returns an error if a disk write error occurs, or if Commit is
+// called on a read-only transaction.
+func (tx *Tx) Commit() (err error) {
+	txId := tx.ID()
+	lg := tx.db.Logger()
+	if lg != discardLogger {
+		lg.Debugf("Committing transaction %d", txId)
+		defer func() {
+			if err != nil {
+				lg.Errorf("Committing transaction failed: %v", err)
+			} else {
+				lg.Debugf("Committing transaction %d successfully", txId)
+			}
+		}()
+	}
+
+	common.Assert(!tx.managed, "managed tx commit not allowed")
+	if tx.db == nil {
+		return berrors.ErrTxClosed
+	} else if !tx.writable {
+		return berrors.ErrTxNotWritable
+	}
+
+	// TODO(benbjohnson): Use vectorized I/O to write out dirty pages.
+
+	// Rebalance nodes which have had deletions.
+	var startTime = time.Now()
+	tx.root.rebalance()
+	if tx.stats.GetRebalance() > 0 {
+		tx.stats.IncRebalanceTime(time.Since(startTime))
+	}
+
+	opgid := tx.meta.Pgid()
+
+	// spill data onto dirty pages.
+	startTime = time.Now()
+	if err = tx.root.spill(); err != nil {
+		lg.Errorf("spilling data onto dirty pages failed: %v", err)
+		tx.rollback()
+		return err
+	}
+	tx.stats.IncSpillTime(time.Since(startTime))
+
+	// Free the old root bucket.
+	tx.meta.RootBucket().SetRootPage(tx.root.RootPage())
+
+	// Free the old freelist because commit writes out a fresh freelist.
+	if tx.meta.Freelist() != common.PgidNoFreelist {
+		tx.db.freelist.Free(tx.meta.Txid(), tx.db.page(tx.meta.Freelist()))
+	}
+
+	if !tx.db.NoFreelistSync {
+		err = tx.commitFreelist()
+		if err != nil {
+			lg.Errorf("committing freelist failed: %v", err)
+			return err
+		}
+	} else {
+		tx.meta.SetFreelist(common.PgidNoFreelist)
+	}
+
+	// If the high water mark has moved up then attempt to grow the database.
+	if tx.meta.Pgid() > opgid {
+		_ = errors.New("")
+		// gofail: var lackOfDiskSpace string
+		// tx.rollback()
+		// return errors.New(lackOfDiskSpace)
+		if err = tx.db.grow(int(tx.meta.Pgid()+1) * tx.db.pageSize); err != nil {
+			lg.Errorf("growing db size failed, pgid: %d, pagesize: %d, error: %v", tx.meta.Pgid(), tx.db.pageSize, err)
+			tx.rollback()
+			return err
+		}
+	}
+
+	// Write dirty pages to disk.
+	startTime = time.Now()
+	if err = tx.write(); err != nil {
+		lg.Errorf("writing data failed: %v", err)
+		tx.rollback()
+		return err
+	}
+
+	// If strict mode is enabled then perform a consistency check.
+	if tx.db.StrictMode {
+		ch := tx.Check()
+		var errs []string
+		for {
+			chkErr, ok := <-ch
+			if !ok {
+				break
+			}
+			errs = append(errs, chkErr.Error())
+		}
+		if len(errs) > 0 {
+			panic("check fail: " + strings.Join(errs, "\n"))
+		}
+	}
+
+	// Write meta to disk.
+	if err = tx.writeMeta(); err != nil {
+		lg.Errorf("writeMeta failed: %v", err)
+		tx.rollback()
+		return err
+	}
+	tx.stats.IncWriteTime(time.Since(startTime))
+
+	// Finalize the transaction.
+	tx.close()
+
+	// Execute commit handlers now that the locks have been removed.
+	for _, fn := range tx.commitHandlers {
+		fn()
+	}
+
+	return nil
+}
+
+func (tx *Tx) commitFreelist() error {
+	// Allocate new pages for the new free list. This will overestimate
+	// the size of the freelist but not underestimate the size (which would be bad).
+	p, err := tx.allocate((tx.db.freelist.EstimatedWritePageSize() / tx.db.pageSize) + 1)
+	if err != nil {
+		tx.rollback()
+		return err
+	}
+
+	tx.db.freelist.Write(p)
+	tx.meta.SetFreelist(p.Id())
+
+	return nil
+}
+
+// Rollback closes the transaction and ignores all previous updates. Read-only
+// transactions must be rolled back and not committed.
+func (tx *Tx) Rollback() error {
+	common.Assert(!tx.managed, "managed tx rollback not allowed")
+	if tx.db == nil {
+		return berrors.ErrTxClosed
+	}
+	tx.nonPhysicalRollback()
+	return nil
+}
+
+// nonPhysicalRollback is called when user calls Rollback directly, in this case we do not need to reload the free pages from disk.
+func (tx *Tx) nonPhysicalRollback() {
+	if tx.db == nil {
+		return
+	}
+	if tx.writable {
+		tx.db.freelist.Rollback(tx.meta.Txid())
+	}
+	tx.close()
+}
+
+// rollback needs to reload the free pages from disk in case some system error happens like fsync error.
+func (tx *Tx) rollback() {
+	if tx.db == nil {
+		return
+	}
+	if tx.writable {
+		tx.db.freelist.Rollback(tx.meta.Txid())
+		// When mmap fails, the `data`, `dataref` and `datasz` may be reset to
+		// zero values, and there is no way to reload free page IDs in this case.
+		if tx.db.data != nil {
+			if !tx.db.hasSyncedFreelist() {
+				// Reconstruct free page list by scanning the DB to get the whole free page list.
+				// Note: scanning the whole db is heavy if your db size is large in NoSyncFreeList mode.
+				tx.db.freelist.NoSyncReload(tx.db.freepages())
+			} else {
+				// Read free page list from freelist page.
+				tx.db.freelist.Reload(tx.db.page(tx.db.meta().Freelist()))
+			}
+		}
+	}
+	tx.close()
+}
+
+func (tx *Tx) close() {
+	if tx.db == nil {
+		return
+	}
+	if tx.writable {
+		// Grab freelist stats.
+		var freelistFreeN = tx.db.freelist.FreeCount()
+		var freelistPendingN = tx.db.freelist.PendingCount()
+		var freelistAlloc = tx.db.freelist.EstimatedWritePageSize()
+
+		// Remove transaction ref & writer lock.
+		tx.db.rwtx = nil
+		tx.db.rwlock.Unlock()
+
+		// Merge statistics.
+		tx.db.statlock.Lock()
+		tx.db.stats.FreePageN = freelistFreeN
+		tx.db.stats.PendingPageN = freelistPendingN
+		tx.db.stats.FreeAlloc = (freelistFreeN + freelistPendingN) * tx.db.pageSize
+		tx.db.stats.FreelistInuse = freelistAlloc
+		tx.db.stats.TxStats.add(&tx.stats)
+		tx.db.statlock.Unlock()
+	} else {
+		tx.db.removeTx(tx)
+	}
+
+	// Clear all references.
+	tx.db = nil
+	tx.meta = nil
+	tx.root = Bucket{tx: tx}
+	tx.pages = nil
+}
+
+// Copy writes the entire database to a writer.
+// This function exists for backwards compatibility.
+//
+// Deprecated: Use WriteTo() instead.
+func (tx *Tx) Copy(w io.Writer) error {
+	_, err := tx.WriteTo(w)
+	return err
+}
+
+// WriteTo writes the entire database to a writer.
+// If err == nil then exactly tx.Size() bytes will be written into the writer.
+func (tx *Tx) WriteTo(w io.Writer) (n int64, err error) {
+	// Attempt to open reader with WriteFlag
+	f, err := tx.db.openFile(tx.db.path, os.O_RDONLY|tx.WriteFlag, 0)
+	if err != nil {
+		return 0, err
+	}
+	defer func() {
+		if cerr := f.Close(); err == nil {
+			err = cerr
+		}
+	}()
+
+	// Generate a meta page. We use the same page data for both meta pages.
+	buf := make([]byte, tx.db.pageSize)
+	page := (*common.Page)(unsafe.Pointer(&buf[0]))
+	page.SetFlags(common.MetaPageFlag)
+	*page.Meta() = *tx.meta
+
+	// Write meta 0.
+	page.SetId(0)
+	page.Meta().SetChecksum(page.Meta().Sum64())
+	nn, err := w.Write(buf)
+	n += int64(nn)
+	if err != nil {
+		return n, fmt.Errorf("meta 0 copy: %s", err)
+	}
+
+	// Write meta 1 with a lower transaction id.
+	page.SetId(1)
+	page.Meta().DecTxid()
+	page.Meta().SetChecksum(page.Meta().Sum64())
+	nn, err = w.Write(buf)
+	n += int64(nn)
+	if err != nil {
+		return n, fmt.Errorf("meta 1 copy: %s", err)
+	}
+
+	// Move past the meta pages in the file.
+	if _, err := f.Seek(int64(tx.db.pageSize*2), io.SeekStart); err != nil {
+		return n, fmt.Errorf("seek: %s", err)
+	}
+
+	// Copy data pages.
+	wn, err := io.CopyN(w, f, tx.Size()-int64(tx.db.pageSize*2))
+	n += wn
+	if err != nil {
+		return n, err
+	}
+
+	return n, nil
+}
+
+// CopyFile copies the entire database to file at the given path.
+// A reader transaction is maintained during the copy so it is safe to continue
+// using the database while a copy is in progress.
+func (tx *Tx) CopyFile(path string, mode os.FileMode) error {
+	f, err := tx.db.openFile(path, os.O_RDWR|os.O_CREATE|os.O_TRUNC, mode)
+	if err != nil {
+		return err
+	}
+
+	_, err = tx.WriteTo(f)
+	if err != nil {
+		_ = f.Close()
+		return err
+	}
+	return f.Close()
+}
+
+// allocate returns a contiguous block of memory starting at a given page.
+func (tx *Tx) allocate(count int) (*common.Page, error) {
+	lg := tx.db.Logger()
+	p, err := tx.db.allocate(tx.meta.Txid(), count)
+	if err != nil {
+		lg.Errorf("allocating failed, txid: %d, count: %d, error: %v", tx.meta.Txid(), count, err)
+		return nil, err
+	}
+
+	// Save to our page cache.
+	tx.pages[p.Id()] = p
+
+	// Update statistics.
+	tx.stats.IncPageCount(int64(count))
+	tx.stats.IncPageAlloc(int64(count * tx.db.pageSize))
+
+	return p, nil
+}
+
+// write writes any dirty pages to disk.
+func (tx *Tx) write() error {
+	// Sort pages by id.
+	lg := tx.db.Logger()
+	pages := make(common.Pages, 0, len(tx.pages))
+	for _, p := range tx.pages {
+		pages = append(pages, p)
+	}
+	// Clear out page cache early.
+	tx.pages = make(map[common.Pgid]*common.Page)
+	sort.Sort(pages)
+
+	// Write pages to disk in order.
+	for _, p := range pages {
+		rem := (uint64(p.Overflow()) + 1) * uint64(tx.db.pageSize)
+		offset := int64(p.Id()) * int64(tx.db.pageSize)
+		var written uintptr
+
+		// Write out page in "max allocation" sized chunks.
+		for {
+			sz := rem
+			if sz > maxAllocSize-1 {
+				sz = maxAllocSize - 1
+			}
+			buf := common.UnsafeByteSlice(unsafe.Pointer(p), written, 0, int(sz))
+
+			if _, err := tx.db.ops.writeAt(buf, offset); err != nil {
+				lg.Errorf("writeAt failed, offset: %d: %w", offset, err)
+				return err
+			}
+
+			// Update statistics.
+			tx.stats.IncWrite(1)
+
+			// Exit inner for loop if we've written all the chunks.
+			rem -= sz
+			if rem == 0 {
+				break
+			}
+
+			// Otherwise move offset forward and move pointer to next chunk.
+			offset += int64(sz)
+			written += uintptr(sz)
+		}
+	}
+
+	// Ignore file sync if flag is set on DB.
+	if !tx.db.NoSync || common.IgnoreNoSync {
+		// gofail: var beforeSyncDataPages struct{}
+		if err := fdatasync(tx.db); err != nil {
+			lg.Errorf("[GOOS: %s, GOARCH: %s] fdatasync failed: %w", runtime.GOOS, runtime.GOARCH, err)
+			return err
+		}
+	}
+
+	// Put small pages back to page pool.
+	for _, p := range pages {
+		// Ignore page sizes over 1 page.
+		// These are allocated using make() instead of the page pool.
+		if int(p.Overflow()) != 0 {
+			continue
+		}
+
+		buf := common.UnsafeByteSlice(unsafe.Pointer(p), 0, 0, tx.db.pageSize)
+
+		// See https://go.googlesource.com/go/+/f03c9202c43e0abb130669852082117ca50aa9b1
+		for i := range buf {
+			buf[i] = 0
+		}
+		tx.db.pagePool.Put(buf) //nolint:staticcheck
+	}
+
+	return nil
+}
+
+// writeMeta writes the meta to the disk.
+func (tx *Tx) writeMeta() error {
+	// gofail: var beforeWriteMetaError string
+	// return errors.New(beforeWriteMetaError)
+
+	// Create a temporary buffer for the meta page.
+	lg := tx.db.Logger()
+	buf := make([]byte, tx.db.pageSize)
+	p := tx.db.pageInBuffer(buf, 0)
+	tx.meta.Write(p)
+
+	// Write the meta page to file.
+	if _, err := tx.db.ops.writeAt(buf, int64(p.Id())*int64(tx.db.pageSize)); err != nil {
+		lg.Errorf("writeAt failed, pgid: %d, pageSize: %d, error: %v", p.Id(), tx.db.pageSize, err)
+		return err
+	}
+	if !tx.db.NoSync || common.IgnoreNoSync {
+		// gofail: var beforeSyncMetaPage struct{}
+		if err := fdatasync(tx.db); err != nil {
+			lg.Errorf("[GOOS: %s, GOARCH: %s] fdatasync failed: %w", runtime.GOOS, runtime.GOARCH, err)
+			return err
+		}
+	}
+
+	// Update statistics.
+	tx.stats.IncWrite(1)
+
+	return nil
+}
+
+// page returns a reference to the page with a given id.
+// If page has been written to then a temporary buffered page is returned.
+func (tx *Tx) page(id common.Pgid) *common.Page {
+	// Check the dirty pages first.
+	if tx.pages != nil {
+		if p, ok := tx.pages[id]; ok {
+			p.FastCheck(id)
+			return p
+		}
+	}
+
+	// Otherwise return directly from the mmap.
+	p := tx.db.page(id)
+	p.FastCheck(id)
+	return p
+}
+
+// forEachPage iterates over every page within a given page and executes a function.
+func (tx *Tx) forEachPage(pgidnum common.Pgid, fn func(*common.Page, int, []common.Pgid)) {
+	stack := make([]common.Pgid, 10)
+	stack[0] = pgidnum
+	tx.forEachPageInternal(stack[:1], fn)
+}
+
+func (tx *Tx) forEachPageInternal(pgidstack []common.Pgid, fn func(*common.Page, int, []common.Pgid)) {
+	p := tx.page(pgidstack[len(pgidstack)-1])
+
+	// Execute function.
+	fn(p, len(pgidstack)-1, pgidstack)
+
+	// Recursively loop over children.
+	if p.IsBranchPage() {
+		for i := 0; i < int(p.Count()); i++ {
+			elem := p.BranchPageElement(uint16(i))
+			tx.forEachPageInternal(append(pgidstack, elem.Pgid()), fn)
+		}
+	}
+}
+
+// Page returns page information for a given page number.
+// This is only safe for concurrent use when used by a writable transaction.
+func (tx *Tx) Page(id int) (*common.PageInfo, error) {
+	if tx.db == nil {
+		return nil, berrors.ErrTxClosed
+	} else if common.Pgid(id) >= tx.meta.Pgid() {
+		return nil, nil
+	}
+
+	if tx.db.freelist == nil {
+		return nil, berrors.ErrFreePagesNotLoaded
+	}
+
+	// Build the page info.
+	p := tx.db.page(common.Pgid(id))
+	info := &common.PageInfo{
+		ID:            id,
+		Count:         int(p.Count()),
+		OverflowCount: int(p.Overflow()),
+	}
+
+	// Determine the type (or if it's free).
+	if tx.db.freelist.Freed(common.Pgid(id)) {
+		info.Type = "free"
+	} else {
+		info.Type = p.Typ()
+	}
+
+	return info, nil
+}
+
+// TxStats represents statistics about the actions performed by the transaction.
+type TxStats struct {
+	// Page statistics.
+	//
+	// DEPRECATED: Use GetPageCount() or IncPageCount()
+	PageCount int64 // number of page allocations
+	// DEPRECATED: Use GetPageAlloc() or IncPageAlloc()
+	PageAlloc int64 // total bytes allocated
+
+	// Cursor statistics.
+	//
+	// DEPRECATED: Use GetCursorCount() or IncCursorCount()
+	CursorCount int64 // number of cursors created
+
+	// Node statistics
+	//
+	// DEPRECATED: Use GetNodeCount() or IncNodeCount()
+	NodeCount int64 // number of node allocations
+	// DEPRECATED: Use GetNodeDeref() or IncNodeDeref()
+	NodeDeref int64 // number of node dereferences
+
+	// Rebalance statistics.
+	//
+	// DEPRECATED: Use GetRebalance() or IncRebalance()
+	Rebalance int64 // number of node rebalances
+	// DEPRECATED: Use GetRebalanceTime() or IncRebalanceTime()
+	RebalanceTime time.Duration // total time spent rebalancing
+
+	// Split/Spill statistics.
+	//
+	// DEPRECATED: Use GetSplit() or IncSplit()
+	Split int64 // number of nodes split
+	// DEPRECATED: Use GetSpill() or IncSpill()
+	Spill int64 // number of nodes spilled
+	// DEPRECATED: Use GetSpillTime() or IncSpillTime()
+	SpillTime time.Duration // total time spent spilling
+
+	// Write statistics.
+	//
+	// DEPRECATED: Use GetWrite() or IncWrite()
+	Write int64 // number of writes performed
+	// DEPRECATED: Use GetWriteTime() or IncWriteTime()
+	WriteTime time.Duration // total time spent writing to disk
+}
+
+func (s *TxStats) add(other *TxStats) {
+	s.IncPageCount(other.GetPageCount())
+	s.IncPageAlloc(other.GetPageAlloc())
+	s.IncCursorCount(other.GetCursorCount())
+	s.IncNodeCount(other.GetNodeCount())
+	s.IncNodeDeref(other.GetNodeDeref())
+	s.IncRebalance(other.GetRebalance())
+	s.IncRebalanceTime(other.GetRebalanceTime())
+	s.IncSplit(other.GetSplit())
+	s.IncSpill(other.GetSpill())
+	s.IncSpillTime(other.GetSpillTime())
+	s.IncWrite(other.GetWrite())
+	s.IncWriteTime(other.GetWriteTime())
+}
+
+// Sub calculates and returns the difference between two sets of transaction stats.
+// This is useful when obtaining stats at two different points and time and
+// you need the performance counters that occurred within that time span.
+func (s *TxStats) Sub(other *TxStats) TxStats {
+	var diff TxStats
+	diff.PageCount = s.GetPageCount() - other.GetPageCount()
+	diff.PageAlloc = s.GetPageAlloc() - other.GetPageAlloc()
+	diff.CursorCount = s.GetCursorCount() - other.GetCursorCount()
+	diff.NodeCount = s.GetNodeCount() - other.GetNodeCount()
+	diff.NodeDeref = s.GetNodeDeref() - other.GetNodeDeref()
+	diff.Rebalance = s.GetRebalance() - other.GetRebalance()
+	diff.RebalanceTime = s.GetRebalanceTime() - other.GetRebalanceTime()
+	diff.Split = s.GetSplit() - other.GetSplit()
+	diff.Spill = s.GetSpill() - other.GetSpill()
+	diff.SpillTime = s.GetSpillTime() - other.GetSpillTime()
+	diff.Write = s.GetWrite() - other.GetWrite()
+	diff.WriteTime = s.GetWriteTime() - other.GetWriteTime()
+	return diff
+}
+
+// GetPageCount returns PageCount atomically.
+func (s *TxStats) GetPageCount() int64 {
+	return atomic.LoadInt64(&s.PageCount)
+}
+
+// IncPageCount increases PageCount atomically and returns the new value.
+func (s *TxStats) IncPageCount(delta int64) int64 {
+	return atomic.AddInt64(&s.PageCount, delta)
+}
+
+// GetPageAlloc returns PageAlloc atomically.
+func (s *TxStats) GetPageAlloc() int64 {
+	return atomic.LoadInt64(&s.PageAlloc)
+}
+
+// IncPageAlloc increases PageAlloc atomically and returns the new value.
+func (s *TxStats) IncPageAlloc(delta int64) int64 {
+	return atomic.AddInt64(&s.PageAlloc, delta)
+}
+
+// GetCursorCount returns CursorCount atomically.
+func (s *TxStats) GetCursorCount() int64 {
+	return atomic.LoadInt64(&s.CursorCount)
+}
+
+// IncCursorCount increases CursorCount atomically and return the new value.
+func (s *TxStats) IncCursorCount(delta int64) int64 {
+	return atomic.AddInt64(&s.CursorCount, delta)
+}
+
+// GetNodeCount returns NodeCount atomically.
+func (s *TxStats) GetNodeCount() int64 {
+	return atomic.LoadInt64(&s.NodeCount)
+}
+
+// IncNodeCount increases NodeCount atomically and returns the new value.
+func (s *TxStats) IncNodeCount(delta int64) int64 {
+	return atomic.AddInt64(&s.NodeCount, delta)
+}
+
+// GetNodeDeref returns NodeDeref atomically.
+func (s *TxStats) GetNodeDeref() int64 {
+	return atomic.LoadInt64(&s.NodeDeref)
+}
+
+// IncNodeDeref increases NodeDeref atomically and returns the new value.
+func (s *TxStats) IncNodeDeref(delta int64) int64 {
+	return atomic.AddInt64(&s.NodeDeref, delta)
+}
+
+// GetRebalance returns Rebalance atomically.
+func (s *TxStats) GetRebalance() int64 {
+	return atomic.LoadInt64(&s.Rebalance)
+}
+
+// IncRebalance increases Rebalance atomically and returns the new value.
+func (s *TxStats) IncRebalance(delta int64) int64 {
+	return atomic.AddInt64(&s.Rebalance, delta)
+}
+
+// GetRebalanceTime returns RebalanceTime atomically.
+func (s *TxStats) GetRebalanceTime() time.Duration {
+	return atomicLoadDuration(&s.RebalanceTime)
+}
+
+// IncRebalanceTime increases RebalanceTime atomically and returns the new value.
+func (s *TxStats) IncRebalanceTime(delta time.Duration) time.Duration {
+	return atomicAddDuration(&s.RebalanceTime, delta)
+}
+
+// GetSplit returns Split atomically.
+func (s *TxStats) GetSplit() int64 {
+	return atomic.LoadInt64(&s.Split)
+}
+
+// IncSplit increases Split atomically and returns the new value.
+func (s *TxStats) IncSplit(delta int64) int64 {
+	return atomic.AddInt64(&s.Split, delta)
+}
+
+// GetSpill returns Spill atomically.
+func (s *TxStats) GetSpill() int64 {
+	return atomic.LoadInt64(&s.Spill)
+}
+
+// IncSpill increases Spill atomically and returns the new value.
+func (s *TxStats) IncSpill(delta int64) int64 {
+	return atomic.AddInt64(&s.Spill, delta)
+}
+
+// GetSpillTime returns SpillTime atomically.
+func (s *TxStats) GetSpillTime() time.Duration {
+	return atomicLoadDuration(&s.SpillTime)
+}
+
+// IncSpillTime increases SpillTime atomically and returns the new value.
+func (s *TxStats) IncSpillTime(delta time.Duration) time.Duration {
+	return atomicAddDuration(&s.SpillTime, delta)
+}
+
+// GetWrite returns Write atomically.
+func (s *TxStats) GetWrite() int64 {
+	return atomic.LoadInt64(&s.Write)
+}
+
+// IncWrite increases Write atomically and returns the new value.
+func (s *TxStats) IncWrite(delta int64) int64 {
+	return atomic.AddInt64(&s.Write, delta)
+}
+
+// GetWriteTime returns WriteTime atomically.
+func (s *TxStats) GetWriteTime() time.Duration {
+	return atomicLoadDuration(&s.WriteTime)
+}
+
+// IncWriteTime increases WriteTime atomically and returns the new value.
+func (s *TxStats) IncWriteTime(delta time.Duration) time.Duration {
+	return atomicAddDuration(&s.WriteTime, delta)
+}
+
+func atomicAddDuration(ptr *time.Duration, du time.Duration) time.Duration {
+	return time.Duration(atomic.AddInt64((*int64)(unsafe.Pointer(ptr)), int64(du)))
+}
+
+func atomicLoadDuration(ptr *time.Duration) time.Duration {
+	return time.Duration(atomic.LoadInt64((*int64)(unsafe.Pointer(ptr))))
+}

vendor/go.etcd.io/bbolt/tx_check.go

@@ -0,0 +1,290 @@
+package bbolt
+
+import (
+	"encoding/hex"
+	"fmt"
+
+	"go.etcd.io/bbolt/internal/common"
+)
+
+// Check performs several consistency checks on the database for this transaction.
+// An error is returned if any inconsistency is found.
+//
+// It can be safely run concurrently on a writable transaction. However, this
+// incurs a high cost for large databases and databases with a lot of subbuckets
+// because of caching. This overhead can be removed if running on a read-only
+// transaction, however, it is not safe to execute other writer transactions at
+// the same time.
+//
+// It also allows users to provide a customized `KVStringer` implementation,
+// so that bolt can generate human-readable diagnostic messages.
+func (tx *Tx) Check(options ...CheckOption) <-chan error {
+	chkConfig := checkConfig{
+		kvStringer: HexKVStringer(),
+	}
+	for _, op := range options {
+		op(&chkConfig)
+	}
+
+	ch := make(chan error)
+	go func() {
+		// Close the channel to signal completion.
+		defer close(ch)
+		tx.check(chkConfig, ch)
+	}()
+	return ch
+}
+
+func (tx *Tx) check(cfg checkConfig, ch chan error) {
+	// Force loading free list if opened in ReadOnly mode.
+	tx.db.loadFreelist()
+
+	// Check if any pages are double freed.
+	freed := make(map[common.Pgid]bool)
+	all := make([]common.Pgid, tx.db.freelist.Count())
+	tx.db.freelist.Copyall(all)
+	for _, id := range all {
+		if freed[id] {
+			ch <- fmt.Errorf("page %d: already freed", id)
+		}
+		freed[id] = true
+	}
+
+	// Track every reachable page.
+	reachable := make(map[common.Pgid]*common.Page)
+	reachable[0] = tx.page(0) // meta0
+	reachable[1] = tx.page(1) // meta1
+	if tx.meta.Freelist() != common.PgidNoFreelist {
+		for i := uint32(0); i <= tx.page(tx.meta.Freelist()).Overflow(); i++ {
+			reachable[tx.meta.Freelist()+common.Pgid(i)] = tx.page(tx.meta.Freelist())
+		}
+	}
+
+	if cfg.pageId == 0 {
+		// Check the whole db file, starting from the root bucket and
+		// recursively check all child buckets.
+		tx.recursivelyCheckBucket(&tx.root, reachable, freed, cfg.kvStringer, ch)
+
+		// Ensure all pages below high water mark are either reachable or freed.
+		for i := common.Pgid(0); i < tx.meta.Pgid(); i++ {
+			_, isReachable := reachable[i]
+			if !isReachable && !freed[i] {
+				ch <- fmt.Errorf("page %d: unreachable unfreed", int(i))
+			}
+		}
+	} else {
+		// Check the db file starting from a specified pageId.
+		if cfg.pageId < 2 || cfg.pageId >= uint64(tx.meta.Pgid()) {
+			ch <- fmt.Errorf("page ID (%d) out of range [%d, %d)", cfg.pageId, 2, tx.meta.Pgid())
+			return
+		}
+
+		tx.recursivelyCheckPage(common.Pgid(cfg.pageId), reachable, freed, cfg.kvStringer, ch)
+	}
+}
+
+func (tx *Tx) recursivelyCheckPage(pageId common.Pgid, reachable map[common.Pgid]*common.Page, freed map[common.Pgid]bool,
+	kvStringer KVStringer, ch chan error) {
+	tx.checkInvariantProperties(pageId, reachable, freed, kvStringer, ch)
+	tx.recursivelyCheckBucketInPage(pageId, reachable, freed, kvStringer, ch)
+}
+
+func (tx *Tx) recursivelyCheckBucketInPage(pageId common.Pgid, reachable map[common.Pgid]*common.Page, freed map[common.Pgid]bool,
+	kvStringer KVStringer, ch chan error) {
+	p := tx.page(pageId)
+
+	switch {
+	case p.IsBranchPage():
+		for i := range p.BranchPageElements() {
+			elem := p.BranchPageElement(uint16(i))
+			tx.recursivelyCheckBucketInPage(elem.Pgid(), reachable, freed, kvStringer, ch)
+		}
+	case p.IsLeafPage():
+		for i := range p.LeafPageElements() {
+			elem := p.LeafPageElement(uint16(i))
+			if elem.IsBucketEntry() {
+				inBkt := common.NewInBucket(pageId, 0)
+				tmpBucket := Bucket{
+					InBucket:    &inBkt,
+					rootNode:    &node{isLeaf: p.IsLeafPage()},
+					FillPercent: DefaultFillPercent,
+					tx:          tx,
+				}
+				if child := tmpBucket.Bucket(elem.Key()); child != nil {
+					tx.recursivelyCheckBucket(child, reachable, freed, kvStringer, ch)
+				}
+			}
+		}
+	default:
+		ch <- fmt.Errorf("unexpected page type (flags: %x) for pgId:%d", p.Flags(), pageId)
+	}
+}
+
+func (tx *Tx) recursivelyCheckBucket(b *Bucket, reachable map[common.Pgid]*common.Page, freed map[common.Pgid]bool,
+	kvStringer KVStringer, ch chan error) {
+	// Ignore inline buckets.
+	if b.RootPage() == 0 {
+		return
+	}
+
+	tx.checkInvariantProperties(b.RootPage(), reachable, freed, kvStringer, ch)
+
+	// Check each bucket within this bucket.
+	_ = b.ForEachBucket(func(k []byte) error {
+		if child := b.Bucket(k); child != nil {
+			tx.recursivelyCheckBucket(child, reachable, freed, kvStringer, ch)
+		}
+		return nil
+	})
+}
+
+func (tx *Tx) checkInvariantProperties(pageId common.Pgid, reachable map[common.Pgid]*common.Page, freed map[common.Pgid]bool,
+	kvStringer KVStringer, ch chan error) {
+	tx.forEachPage(pageId, func(p *common.Page, _ int, stack []common.Pgid) {
+		verifyPageReachable(p, tx.meta.Pgid(), stack, reachable, freed, ch)
+	})
+
+	tx.recursivelyCheckPageKeyOrder(pageId, kvStringer.KeyToString, ch)
+}
+
+func verifyPageReachable(p *common.Page, hwm common.Pgid, stack []common.Pgid, reachable map[common.Pgid]*common.Page, freed map[common.Pgid]bool, ch chan error) {
+	if p.Id() > hwm {
+		ch <- fmt.Errorf("page %d: out of bounds: %d (stack: %v)", int(p.Id()), int(hwm), stack)
+	}
+
+	// Ensure each page is only referenced once.
+	for i := common.Pgid(0); i <= common.Pgid(p.Overflow()); i++ {
+		var id = p.Id() + i
+		if _, ok := reachable[id]; ok {
+			ch <- fmt.Errorf("page %d: multiple references (stack: %v)", int(id), stack)
+		}
+		reachable[id] = p
+	}
+
+	// We should only encounter un-freed leaf and branch pages.
+	if freed[p.Id()] {
+		ch <- fmt.Errorf("page %d: reachable freed", int(p.Id()))
+	} else if !p.IsBranchPage() && !p.IsLeafPage() {
+		ch <- fmt.Errorf("page %d: invalid type: %s (stack: %v)", int(p.Id()), p.Typ(), stack)
+	}
+}
+
+// recursivelyCheckPageKeyOrder verifies database consistency with respect to b-tree
+// key order constraints:
+//   - keys on pages must be sorted
+//   - keys on children pages are between 2 consecutive keys on the parent's branch page).
+func (tx *Tx) recursivelyCheckPageKeyOrder(pgId common.Pgid, keyToString func([]byte) string, ch chan error) {
+	tx.recursivelyCheckPageKeyOrderInternal(pgId, nil, nil, nil, keyToString, ch)
+}
+
+// recursivelyCheckPageKeyOrderInternal verifies that all keys in the subtree rooted at `pgid` are:
+//   - >=`minKeyClosed` (can be nil)
+//   - <`maxKeyOpen` (can be nil)
+//   - Are in right ordering relationship to their parents.
+//     `pagesStack` is expected to contain IDs of pages from the tree root to `pgid` for the clean debugging message.
+func (tx *Tx) recursivelyCheckPageKeyOrderInternal(
+	pgId common.Pgid, minKeyClosed, maxKeyOpen []byte, pagesStack []common.Pgid,
+	keyToString func([]byte) string, ch chan error) (maxKeyInSubtree []byte) {
+
+	p := tx.page(pgId)
+	pagesStack = append(pagesStack, pgId)
+	switch {
+	case p.IsBranchPage():
+		// For branch page we navigate ranges of all subpages.
+		runningMin := minKeyClosed
+		for i := range p.BranchPageElements() {
+			elem := p.BranchPageElement(uint16(i))
+			verifyKeyOrder(elem.Pgid(), "branch", i, elem.Key(), runningMin, maxKeyOpen, ch, keyToString, pagesStack)
+
+			maxKey := maxKeyOpen
+			if i < len(p.BranchPageElements())-1 {
+				maxKey = p.BranchPageElement(uint16(i + 1)).Key()
+			}
+			maxKeyInSubtree = tx.recursivelyCheckPageKeyOrderInternal(elem.Pgid(), elem.Key(), maxKey, pagesStack, keyToString, ch)
+			runningMin = maxKeyInSubtree
+		}
+		return maxKeyInSubtree
+	case p.IsLeafPage():
+		runningMin := minKeyClosed
+		for i := range p.LeafPageElements() {
+			elem := p.LeafPageElement(uint16(i))
+			verifyKeyOrder(pgId, "leaf", i, elem.Key(), runningMin, maxKeyOpen, ch, keyToString, pagesStack)
+			runningMin = elem.Key()
+		}
+		if p.Count() > 0 {
+			return p.LeafPageElement(p.Count() - 1).Key()
+		}
+	default:
+		ch <- fmt.Errorf("unexpected page type (flags: %x) for pgId:%d", p.Flags(), pgId)
+	}
+	return maxKeyInSubtree
+}
+
+/***
+ * verifyKeyOrder checks whether an entry with given #index on pgId (pageType: "branch|leaf") that has given "key",
+ * is within range determined by (previousKey..maxKeyOpen) and reports found violations to the channel (ch).
+ */
+func verifyKeyOrder(pgId common.Pgid, pageType string, index int, key []byte, previousKey []byte, maxKeyOpen []byte, ch chan error, keyToString func([]byte) string, pagesStack []common.Pgid) {
+	if index == 0 && previousKey != nil && compareKeys(previousKey, key) > 0 {
+		ch <- fmt.Errorf("the first key[%d]=(hex)%s on %s page(%d) needs to be >= the key in the ancestor (%s). Stack: %v",
+			index, keyToString(key), pageType, pgId, keyToString(previousKey), pagesStack)
+	}
+	if index > 0 {
+		cmpRet := compareKeys(previousKey, key)
+		if cmpRet > 0 {
+			ch <- fmt.Errorf("key[%d]=(hex)%s on %s page(%d) needs to be > (found <) than previous element (hex)%s. Stack: %v",
+				index, keyToString(key), pageType, pgId, keyToString(previousKey), pagesStack)
+		}
+		if cmpRet == 0 {
+			ch <- fmt.Errorf("key[%d]=(hex)%s on %s page(%d) needs to be > (found =) than previous element (hex)%s. Stack: %v",
+				index, keyToString(key), pageType, pgId, keyToString(previousKey), pagesStack)
+		}
+	}
+	if maxKeyOpen != nil && compareKeys(key, maxKeyOpen) >= 0 {
+		ch <- fmt.Errorf("key[%d]=(hex)%s on %s page(%d) needs to be < than key of the next element in ancestor (hex)%s. Pages stack: %v",
+			index, keyToString(key), pageType, pgId, keyToString(previousKey), pagesStack)
+	}
+}
+
+// ===========================================================================================
+
+type checkConfig struct {
+	kvStringer KVStringer
+	pageId     uint64
+}
+
+type CheckOption func(options *checkConfig)
+
+func WithKVStringer(kvStringer KVStringer) CheckOption {
+	return func(c *checkConfig) {
+		c.kvStringer = kvStringer
+	}
+}
+
+// WithPageId sets a page ID from which the check command starts to check
+func WithPageId(pageId uint64) CheckOption {
+	return func(c *checkConfig) {
+		c.pageId = pageId
+	}
+}
+
+// KVStringer allows to prepare human-readable diagnostic messages.
+type KVStringer interface {
+	KeyToString([]byte) string
+	ValueToString([]byte) string
+}
+
+// HexKVStringer serializes both key & value to hex representation.
+func HexKVStringer() KVStringer {
+	return hexKvStringer{}
+}
+
+type hexKvStringer struct{}
+
+func (_ hexKvStringer) KeyToString(key []byte) string {
+	return hex.EncodeToString(key)
+}
+
+func (_ hexKvStringer) ValueToString(value []byte) string {
+	return hex.EncodeToString(value)
+}