qkc/serialize: port QuarkChain binary serialization (byte-compatible) (#12)

* qkc/serialize: port QuarkChain binary serialization (byte-compatible) Verbatim port of github.com/QuarkChain/goquarkchain/serialize, the reflection-based binary codec used (instead of RLP) for all slavechain headers, blocks and storage records. Pure stdlib, zero adaptations. Includes the original golden byte-vector tests. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * qkc/serialize: bound deserializeList allocation by remaining input deserializeList allocated reflect.MakeSlice(val.Type(), vlen, vlen) from the wire-supplied length before reading any element bytes, with no bound check. A malicious P2P packet could claim a huge element count (up to ~4B with bytesizeofslicelen:4) and force a massive allocation from a tiny packet — an OOM/DoS vector, since this codec decodes blocks/headers straight from peers. Reject a length larger than the remaining buffer (every list element consumes at least one byte, so such a length is always malformed and would fail below anyway). This does not change the decoding of any valid input; it is an intentional hardening divergence from the verbatim goquarkchain port, which has the same unbounded allocation. Addresses review feedback from @qzhodl. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * qkc/serialize: reject negative big.Int in fixed-size Uint128/Uint256 serializeFixSizeBigUint built the fixed-width bytes from big.Int.Bytes(), which returns the absolute value. A negative value therefore serialized identically to its positive counterpart (e.g. -1 as +1), silently changing it. Reject negatives, matching the variable-length serializeBigInt path which already does so. Non-negative encodings are unchanged. This is an intentional hardening divergence from the verbatim goquarkchain port, whose fixed-size path has the same silent sign loss while its variable-length path rejects negatives. Addresses review feedback from @qzhodl. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * qkc/serialize: reject big.Int that overflows the single-byte length prefix serializeBigInt writes the byte length as a single uint8, so a value needing more than 255 bytes (>2040 bits) had its length silently truncated and decoded to a different value — a non-roundtrippable encoding. Return an error when len(bytes) > 255. Values at or under the limit are unchanged and still round-trip. Addresses review feedback from @qzhodl. (Same single-byte length prefix exists in the verbatim goquarkchain port.) Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * qkc/serialize: reject trailing bytes in DeserializeFromBytes DeserializeFromBytes decoded from a fresh buffer but never checked that the buffer was fully consumed, so an input of valid data plus trailing garbage was silently accepted. Return an error when bytes remain after a successful decode. The lower-level Deserialize(bb, val) stays permissive so callers can stream multiple values from one buffer. All current callers pass an exact SerializeToBytes blob, so this only rejects malformed/over-long input. Addresses review feedback from @qzhodl. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * qkc/serialize: reject non-canonical nil-pointer presence markers The ser:"nil" presence marker is written as exactly 0 (absent) or 1 (present), but the deserializer treated any non-zero byte as present. That let distinct byte strings (e.g. 0x05 and 0x01) decode to the same value, a non-canonical encoding. Reject any marker other than 0 or 1. A ported table case used a non-canonical marker (0x03) to reach a truncated- buffer error; switch it to the canonical 0x01 so it still exercises that path, and add TestDeserializeNilMarkerCanonical to cover the rejection directly. Addresses review feedback from @qzhodl. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * qkc/serialize: reject uint encodings wider than the destination The variable-length reflect.Uint path reads a 1-byte length prefix, so it can deliver up to 255 bytes. Those bytes were folded into a uint64 accumulator with no width check, so an encoding longer than the uint width overflowed the accumulator and silently truncated the decoded value. Reject any encoding whose byte count exceeds the destination width (Type().Bits()/8). Fixed-width kinds always read exactly that many bytes, so this is a no-op for them. Addresses review feedback from @qzhodl. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * qkc/serialize: reject zero-byte list element types at registration The deserializeList bound (vlen <= remaining bytes) assumes every element consumes at least one byte, but the codec can encode zero-byte elements (struct{}, structs whose fields are all ignored/unexported, [0]T). Such a list serializes to just its length prefix, so a valid non-empty round-trip then failed the bound — the codec emitted bytes it refused to decode. Reject slices of zero-byte-encoding elements symmetrically at type registration (genTypeInfo via encodesZeroBytes), so serialize and deserialize agree and the allocation bound stays sound for the >=1-byte element types it protects. Byte slices and arrays are exempt (a byte is one byte; array length is fixed by the type, not read from input). QKC has no such list types, so this changes no valid encoding. The ported nil-slice-pointer case used struct{ uint } (embedded unexported field => zero bytes); give it an exported field so it stays a supported element while still exercising nil-pointer-to-slice. Addresses review feedback from @qzhodl. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * qkc/serialize: unmark visited types on return in encodesZeroBytes The visited set is meant to break reference cycles, but it was marked on the way down and never unmarked, so it tracked every type ever seen rather than the current recursion path. A struct with repeated fields of the same zero-byte type (e.g. struct{ A empty; B empty }) then hit a stale mark on the second field and was misclassified as non-zero, so a slice of it escaped the registration check — serialize emitted a bare length prefix that deserialize rejected, reintroducing the encode/decode asymmetry the check exists to prevent. defer delete(visited, typ) scopes the set to the current path: repeated sibling fields are evaluated independently, while a type still on the path is still treated as a cycle. Add TestZeroByteRepeatedFieldTypeRejected pinning the symmetric rejection. Addresses review feedback from @syntrust. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * qkc/serialize: reject nil value in SerializeWithTags SerializeWithTags called reflect.ValueOf(val).Type() unguarded, so an untyped nil panicked ("reflect: call of reflect.Value.Type on zero Value") instead of returning an error. DeserializeWithTags already rejects a nil target up front; add the symmetric check on the serialize side. A typed nil pointer still carries a type and continues to serialize as before. Addresses review feedback from @syntrust. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * qkc/serialize: bound length-prefix accumulation in getLen getLen accumulated the length prefix into a signed int with an unbounded left shift. Once the prefix reached the platform int width (4 bytes on 32-bit, 8 on 64-bit) a high-bit-set byte landed in the sign bit and silently produced a negative length. A negative length slipped past downstream len<=remaining checks and then panicked (reflect.MakeSlice: negative len / slice bounds out of range) on attacker-controlled input — confirmed with an 8-byte 0xFF prefix. Accumulate in uint64, reject a shift that would overflow it, and reject a value that does not fit a non-negative int. This changes no representable length (a wide prefix encoding a small value still decodes). Hardening divergence from goquarkchain. Addresses review feedback from @syntrust. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-23 05:06:43 +00:00 · 2026-06-11 20:27:26 +08:00 · 2026-06-11 20:27:26 +08:00 · 7a5b24894d
commit 7a5b24894d
parent 117e067f0f
16 changed files with 1902 additions and 0 deletions
--- a/qkc/serialize/bytebuffer.go
+++ b/qkc/serialize/bytebuffer.go
@ -0,0 +1,122 @@
+// Ported verbatim from github.com/QuarkChain/goquarkchain/serialize (byte-compatible).
+
+package serialize
+
+import (
+	"encoding/binary"
+	"fmt"
+	"math"
+)
+
+type ByteBuffer struct {
+	data     *[]byte
+	position int
+	size     int
+}
+
+func NewByteBuffer(bytes []byte) *ByteBuffer {
+	bb := ByteBuffer{&bytes, 0, len(bytes)}
+	return &bb
+}
+
+func (bb *ByteBuffer) GetOffset() int {
+	return bb.position
+}
+
+func (bb *ByteBuffer) getBytes(size int) ([]byte, error) {
+	if size > bb.size-bb.position {
+		return nil, fmt.Errorf("deser: buffer is shorter than expected")
+	}
+
+	bytes := (*bb.data)[bb.position : bb.position+size]
+	bb.position += size
+	return bytes, nil
+}
+
+func (bb *ByteBuffer) GetUInt8() (uint8, error) {
+	bytes, err := bb.getBytes(1)
+	if err != nil {
+		return 0, err
+	}
+
+	return uint8(bytes[0]), nil
+}
+
+func (bb *ByteBuffer) GetUInt16() (uint16, error) {
+	bytes, err := bb.getBytes(2)
+	if err != nil {
+		return 0, err
+	}
+
+	return binary.BigEndian.Uint16(bytes), nil
+}
+
+func (bb *ByteBuffer) GetUInt32() (uint32, error) {
+	bytes, err := bb.getBytes(4)
+	if err != nil {
+		return 0, err
+	}
+
+	return binary.BigEndian.Uint32(bytes), nil
+}
+
+func (bb *ByteBuffer) GetUInt64() (uint64, error) {
+	bytes, err := bb.getBytes(8)
+	if err != nil {
+		return 0, err
+	}
+
+	return binary.BigEndian.Uint64(bytes), nil
+}
+
+func (bb *ByteBuffer) getLen(byteSize int) (int, error) {
+	if byteSize < 1 {
+		return 0, fmt.Errorf("deser: bytesize in GetVarBytes should larger than 0")
+	}
+
+	b, err := bb.getBytes(byteSize)
+	if err != nil {
+		return 0, err
+	}
+
+	// Accumulate in uint64 and bound the result. Using a signed int with an
+	// unbounded left shift let a high length-prefix byte land in the sign bit
+	// once the prefix reached the platform int width (4 bytes on 32-bit, 8 on
+	// 64-bit), silently yielding a NEGATIVE length. A negative length slips past
+	// downstream "len <= remaining" checks and then panics (slice bounds /
+	// reflect.MakeSlice: negative len) on attacker-controlled input. Reject any
+	// prefix that overflows uint64 or does not fit a non-negative int. Hardening
+	// divergence from goquarkchain; it does not change any representable length.
+	var size uint64 = 0
+	for i := 0; i < byteSize; i++ {
+		if size > math.MaxUint64>>8 {
+			return 0, fmt.Errorf("deser: length prefix of %d bytes overflows", byteSize)
+		}
+		size = (size << 8) | uint64(b[i])
+	}
+	if size > uint64(math.MaxInt) {
+		return 0, fmt.Errorf("deser: length prefix %d exceeds maximum %d", size, uint64(math.MaxInt))
+	}
+
+	return int(size), nil
+}
+
+func (bb *ByteBuffer) GetVarBytes(byteSizeOfSliceLen int) ([]byte, error) {
+	size, err := bb.getLen(byteSizeOfSliceLen)
+	if err != nil {
+		return nil, err
+	}
+
+	bs, err := bb.getBytes(size)
+	if err != nil {
+		return nil, err
+	}
+
+	bytes := make([]byte, size, size)
+	copy(bytes, bs)
+	return bytes, nil
+}
+
+func (bb *ByteBuffer) Remaining() int {
+	return bb.size - bb.position
+}
--- a/qkc/serialize/bytebuffer_getlen_test.go
+++ b/qkc/serialize/bytebuffer_getlen_test.go
@ -0,0 +1,47 @@
+package serialize
+
+import "testing"
+
+// TestGetLenRejectsOverflow pins that getLen never returns a negative or wrapped
+// length. A length prefix at/above the platform int width with the high bit set
+// previously overflowed the signed accumulator into a negative value, which slid
+// past downstream "len <= remaining" checks and panicked (reflect.MakeSlice:
+// negative len) on attacker-controlled input.
+func TestGetLenRejectsOverflow(t *testing.T) {
+	// 8-byte prefix, high bit set: > MaxInt on every platform -> error.
+	bb := NewByteBuffer([]byte{0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF})
+	if n, err := bb.getLen(8); err == nil {
+		t.Fatalf("expected an error for an over-large 8-byte length prefix, got %d", n)
+	}
+
+	// A wide (10-byte) prefix encoding a small value still decodes: leading
+	// zeros must not be mistaken for overflow.
+	small := NewByteBuffer([]byte{0, 0, 0, 0, 0, 0, 0, 0, 0x12, 0x34})
+	if n, err := small.getLen(10); err != nil || n != 0x1234 {
+		t.Fatalf("small value with wide prefix: n=%d err=%v, want 4660", n, err)
+	}
+
+	// getLen must never return a negative length for any input/width.
+	for _, w := range []int{1, 2, 4, 8} {
+		b := make([]byte, w)
+		for i := range b {
+			b[i] = 0xFF
+		}
+		if n, err := NewByteBuffer(b).getLen(w); err == nil && n < 0 {
+			t.Fatalf("getLen(%d) returned negative length %d", w, n)
+		}
+	}
+}
+
+// TestDeserializeWideLenPrefixNoPanic ensures an attacker-controlled over-wide
+// slice length prefix yields an error, not a panic.
+func TestDeserializeWideLenPrefixNoPanic(t *testing.T) {
+	type wide struct {
+		V []uint32 `bytesizeofslicelen:"8"`
+	}
+	var out wide
+	err := Deserialize(NewByteBuffer([]byte{0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF}), &out)
+	if err == nil {
+		t.Fatal("expected an error for an over-large length prefix")
+	}
+}
--- a/qkc/serialize/deserializer.go
+++ b/qkc/serialize/deserializer.go
@ -0,0 +1,307 @@
+// Ported verbatim from github.com/QuarkChain/goquarkchain/serialize (byte-compatible).
+
+package serialize
+
+import (
+	"errors"
+	"fmt"
+	"math/big"
+	"reflect"
+)
+
+var (
+	errNoPointer          = errors.New("deser: interface given to Deserialize must be a pointer")
+	errDeserializeIntoNil = errors.New("deser: pointer given to Deserialize must not be nil")
+)
+
+func Deserialize(bb *ByteBuffer, val interface{}) error {
+	return DeserializeWithTags(bb, val, Tags{ByteSizeOfSliceLen: 1})
+}
+
+func DeserializeWithTags(bb *ByteBuffer, val interface{}, ts Tags) error {
+	if val == nil {
+		return errDeserializeIntoNil
+	}
+
+	rval := reflect.ValueOf(val)
+	rtyp := rval.Type()
+	if rtyp.Kind() != reflect.Ptr {
+		return errNoPointer
+	}
+	if rval.IsNil() {
+		return errDeserializeIntoNil
+	}
+
+	info, err := cachedTypeInfo(rtyp.Elem())
+	if err != nil {
+		return err
+	}
+
+	err = info.deserializer(bb, rval.Elem(), ts)
+	return err
+}
+
+func DeserializeFromBytes(b []byte, val interface{}) error {
+	bb := NewByteBuffer(b)
+	if err := Deserialize(bb, val); err != nil {
+		return err
+	}
+	// Reject inputs that decode successfully but leave bytes unconsumed: valid
+	// data followed by trailing garbage must not be silently accepted. (The
+	// lower-level Deserialize stays permissive so callers can stream multiple
+	// values from one buffer.)
+	if bb.Remaining() != 0 {
+		return fmt.Errorf("deser: %d trailing byte(s) after decoding", bb.Remaining())
+	}
+	return nil
+}
+
+func makeDeserializer(typ reflect.Type) (deserializer, error) {
+	kind := typ.Kind()
+	switch {
+	//check Ptr first and add optional byte output if ts is nilok,
+	//then get serializer for typ.Elem() which is not a ptr
+	case kind == reflect.Ptr:
+		return deserializePtr, nil
+	case kind != reflect.Ptr && reflect.PtrTo(typ).Implements(serializableInterface):
+		return deserializeSerializableInterface, nil
+	case typ.AssignableTo(bigInt):
+		return deserializeBigIntNoPtr, nil
+	case isUint(kind):
+		return deserializeUint, nil
+	case kind == reflect.Bool:
+		return deserializeBool, nil
+	case kind == reflect.String:
+		return deserializeString, nil
+	case kind == reflect.Slice && isByte(typ.Elem()):
+		return deserializeByteSlice, nil
+	case kind == reflect.Array && isByte(typ.Elem()):
+		return deserializeByteArray, nil
+	case kind == reflect.Slice || kind == reflect.Array:
+		return deserializeList, nil
+	case kind == reflect.Struct:
+		return deserializeStruct, nil
+	default:
+		return nil, fmt.Errorf("type %v is not serializable", typ)
+	}
+}
+
+func deserializeSerializableInterface(bb *ByteBuffer, val reflect.Value, ts Tags) error {
+	return val.Addr().Interface().(Serializable).Deserialize(bb)
+}
+
+func deserializeUint(bb *ByteBuffer, val reflect.Value, ts Tags) error {
+	kind := val.Type().Kind()
+	var bytes []byte
+	var err error
+	switch {
+	case kind > reflect.Uint && kind <= reflect.Uintptr:
+		bytes, err = bb.getBytes(val.Type().Bits() / 8)
+		break
+	case kind == reflect.Uint:
+		bytes, err = bb.GetVarBytes(1)
+		break
+	default:
+		err = fmt.Errorf("deser: invalid Uint type: %s", val.Type().Name())
+		break
+	}
+
+	if err == nil {
+		// Reject encodings wider than the destination. The variable-length
+		// reflect.Uint path (GetVarBytes, 1-byte length prefix) can otherwise
+		// deliver more bytes than the uint width, overflowing the uint64
+		// accumulator below and silently truncating the value. Fixed-width kinds
+		// always read exactly Bits()/8 bytes, so this never triggers for them.
+		if len(bytes) > val.Type().Bits()/8 {
+			return fmt.Errorf("deser: uint encoding too long: %d bytes exceeds the %d-bit destination width", len(bytes), val.Type().Bits())
+		}
+		var ui uint64 = 0
+		for i := 0; i < len(bytes); i++ {
+			ui = ui<<8 | uint64(bytes[i])
+		}
+		val.SetUint(ui)
+	}
+
+	return err
+}
+
+func deserializeFixSizeBigUint(bb *ByteBuffer, val *big.Int, size int) error {
+	bytes, err := bb.getBytes(size)
+	if err == nil {
+		val.SetBytes(bytes)
+	}
+
+	return err
+}
+
+func deserializeBigIntNoPtr(bb *ByteBuffer, val reflect.Value, ts Tags) error {
+	return deserializeBigInt(bb, val.Addr())
+}
+
+func deserializeBigInt(bb *ByteBuffer, val reflect.Value) error {
+	bytes, err := bb.GetVarBytes(1)
+	if err != nil {
+		return err
+	}
+
+	i := val.Interface().(*big.Int)
+	if i == nil {
+		i = new(big.Int)
+		val.Set(reflect.ValueOf(i))
+	}
+
+	i.SetBytes(bytes)
+	return nil
+}
+
+func deserializeBool(bb *ByteBuffer, val reflect.Value, ts Tags) error {
+	b, err := bb.getBytes(1)
+	if err == nil {
+		switch b[0] {
+		case 0x00:
+			val.SetBool(false)
+		case 0x01:
+			val.SetBool(true)
+		default:
+			err = fmt.Errorf("deser: invalid boolean value: %d", b[0])
+		}
+	}
+
+	return err
+}
+
+// FixedSizeBytes
+func deserializeByteArray(bb *ByteBuffer, val reflect.Value, ts Tags) error {
+	if val.Kind() != reflect.Array {
+		return fmt.Errorf("deser: invalid byte array type: %s", val.Kind())
+	}
+	if val.Type().Elem().Kind() != reflect.Uint8 {
+		return fmt.Errorf("deser: invalid byte array type: [%d]%s", val.Len(), val.Kind())
+	}
+
+	bytes, err := bb.getBytes(val.Len())
+	if err == nil {
+		reflect.Copy(val, reflect.ValueOf(bytes))
+	}
+
+	return err
+}
+
+// deserializePrependedSizeBytes
+func deserializeByteSlice(bb *ByteBuffer, val reflect.Value, ts Tags) error {
+	bytes, err := bb.GetVarBytes(ts.ByteSizeOfSliceLen)
+	if err == nil {
+		val.SetBytes(bytes)
+	}
+
+	return err
+}
+
+func deserializeList(bb *ByteBuffer, val reflect.Value, ts Tags) error {
+	typeinfo, err := cachedTypeInfo(val.Type().Elem())
+	if err != nil {
+		return err
+	}
+
+	var vlen int = 0
+	if val.Kind() == reflect.Slice {
+		vlen, err = bb.getLen(ts.ByteSizeOfSliceLen)
+		if err != nil {
+			return err
+		}
+
+		// Bound the element count by the bytes left in the buffer before
+		// allocating. Every element of a registrable slice consumes at least one
+		// byte — zero-byte element types are rejected at registration (see
+		// encodesZeroBytes in genTypeInfo) — so a vlen larger than Remaining() can
+		// only come from a malformed/malicious input (it would fail below when
+		// reading elements anyway). Without this guard a tiny packet claiming a
+		// huge count (up to ~4B with bytesizeofslicelen:4) forces a massive
+		// MakeSlice allocation — an OOM/DoS vector, since this codec decodes
+		// blocks/headers straight from peers. This is an intentional hardening
+		// divergence from goquarkchain; it does not change the decoding of any
+		// valid input.
+		if vlen > bb.Remaining() {
+			return fmt.Errorf("deser: list length %d exceeds remaining buffer %d", vlen, bb.Remaining())
+		}
+
+		newv := reflect.MakeSlice(val.Type(), vlen, vlen)
+		reflect.Copy(newv, val)
+		val.Set(newv)
+	} else if val.Kind() == reflect.Array {
+		vlen = val.Len()
+	}
+
+	for i := 0; i < vlen; i++ {
+		if err := typeinfo.deserializer(bb, val.Index(i), Tags{ByteSizeOfSliceLen: 1}); err != nil {
+			return err
+		}
+	}
+
+	return nil
+}
+
+func deserializeStruct(bb *ByteBuffer, val reflect.Value, ts Tags) error {
+	fields, err := structFields(val.Type())
+	if err != nil {
+		return err
+	}
+
+	for _, f := range fields {
+		err := f.info.deserializer(bb, val.Field(f.index), f.tags)
+		if err != nil {
+			return fmt.Errorf("%s for %v%s", err.Error(), val.Type(), "."+val.Type().Field(f.index).Name)
+		}
+	}
+
+	return nil
+}
+
+func deserializeString(bb *ByteBuffer, val reflect.Value, ts Tags) error {
+	b, err := bb.GetVarBytes(4)
+	if err != nil {
+		return err
+	}
+
+	val.SetString(string(b))
+	return nil
+}
+
+func deserializePtr(bb *ByteBuffer, val reflect.Value, ts Tags) error {
+	typ := val.Type()
+	typeinfo, err := cachedTypeInfo(typ.Elem())
+	if err != nil {
+		return err
+	}
+
+	if ts.NilOK {
+		b, err := bb.GetUInt8()
+		if err != nil {
+			return err
+		}
+
+		if b == 0 {
+			// set the pointer to nil.
+			val.Set(reflect.Zero(typ))
+			return nil
+		}
+		// The presence marker is written as exactly 0 or 1 (see serializePtr), so
+		// any other value is a non-canonical encoding and must be rejected —
+		// otherwise e.g. 0x05 would decode identically to 0x01.
+		if b != 1 {
+			return fmt.Errorf("deser: invalid nil-pointer presence marker %d (must be 0 or 1)", b)
+		}
+	}
+
+	newval := val
+	if val.IsNil() {
+		newval = reflect.New(typ.Elem())
+	}
+
+	err = typeinfo.deserializer(bb, newval.Elem(), ts)
+	if err == nil {
+		val.Set(newval)
+	}
+
+	return err
+}
--- a/qkc/serialize/deserializer_bound_test.go
+++ b/qkc/serialize/deserializer_bound_test.go
@ -0,0 +1,35 @@
+package serialize
+
+import "testing"
+
+// TestDeserializeListBound guards the OOM/DoS vector where a tiny packet claims a
+// huge list length: deserializeList must reject a length larger than the
+// remaining buffer instead of pre-allocating it.
+func TestDeserializeListBound(t *testing.T) {
+	type bigList struct {
+		V []uint32 `bytesizeofslicelen:"4"`
+	}
+	// 4-byte length prefix = 0xFFFFFFFF (~4 billion), followed by no element bytes.
+	input := unhex("FFFFFFFF")
+	var out bigList
+	if err := Deserialize(NewByteBuffer(input), &out); err == nil {
+		t.Fatal("expected an error for an over-large list length, got nil (would over-allocate ~4B elements)")
+	}
+}
+
+// TestDeserializeListValidStillDecodes confirms the bound does not reject valid
+// input where length <= remaining bytes.
+func TestDeserializeListValidStillDecodes(t *testing.T) {
+	type list struct {
+		V []uint32 `bytesizeofslicelen:"4"`
+	}
+	// len = 2 (4 bytes), then two uint32s (8 bytes): 0x00000007, 0x00000009.
+	input := unhex("000000020000000700000009")
+	var out list
+	if err := Deserialize(NewByteBuffer(input), &out); err != nil {
+		t.Fatalf("valid list rejected: %v", err)
+	}
+	if len(out.V) != 2 || out.V[0] != 7 || out.V[1] != 9 {
+		t.Fatalf("bad decode: %v", out.V)
+	}
+}
--- a/qkc/serialize/deserializer_nilmarker_test.go
+++ b/qkc/serialize/deserializer_nilmarker_test.go
@ -0,0 +1,36 @@
+package serialize
+
+import "testing"
+
+type withNilPtr struct {
+	P *uint32 `ser:"nil"`
+}
+
+// TestDeserializeNilMarkerCanonical checks that the ser:"nil" presence marker
+// only accepts 0 or 1. The serializer writes exactly those two values, so any
+// other byte is a non-canonical encoding and must be rejected.
+func TestDeserializeNilMarkerCanonical(t *testing.T) {
+	// marker 0x00 -> nil pointer.
+	var nilCase withNilPtr
+	if err := Deserialize(NewByteBuffer([]byte{0x00}), &nilCase); err != nil {
+		t.Fatalf("marker 0 should decode to nil: %v", err)
+	}
+	if nilCase.P != nil {
+		t.Fatalf("marker 0 should leave pointer nil, got %v", *nilCase.P)
+	}
+
+	// marker 0x01 followed by a 4-byte uint32 -> present.
+	var present withNilPtr
+	if err := Deserialize(NewByteBuffer([]byte{0x01, 0x00, 0x00, 0x00, 0x07}), &present); err != nil {
+		t.Fatalf("marker 1 should decode the value: %v", err)
+	}
+	if present.P == nil || *present.P != 7 {
+		t.Fatalf("marker 1 should decode value 7, got %v", present.P)
+	}
+
+	// marker 0x05 must be rejected (non-canonical "present" byte).
+	var bad withNilPtr
+	if err := Deserialize(NewByteBuffer([]byte{0x05, 0x00, 0x00, 0x00, 0x07}), &bad); err == nil {
+		t.Fatal("expected an error for a non-0/1 presence marker")
+	}
+}
--- a/qkc/serialize/deserializer_test.go
+++ b/qkc/serialize/deserializer_test.go
@ -0,0 +1,301 @@
+// Ported verbatim from github.com/QuarkChain/goquarkchain/serialize (byte-compatible).
+
+package serialize
+
+import (
+	"encoding/hex"
+	"fmt"
+	"math/big"
+	"reflect"
+	"strings"
+	"testing"
+)
+
+type testDataForDeserialize struct {
+	input string
+	ptr   interface{}
+	value interface{}
+	error string
+}
+
+type simplestruct struct {
+	A uint
+	B string
+}
+
+var (
+	veryBigInt = big.NewInt(0).Add(
+		big.NewInt(0).Lsh(big.NewInt(0xFFFFFFFFFFFFFF), 16),
+		big.NewInt(0xFFFF),
+	)
+)
+
+type hasIgnoredField struct {
+	A uint
+	B uint `ser:"-"`
+	C uint32
+}
+
+var deserdata = []testDataForDeserialize{
+	// booleans
+	{input: "01", ptr: new(bool), value: true},
+	{input: "00", ptr: new(bool), value: false},
+	{input: "02", ptr: new(bool), error: "deser: invalid boolean value: 2"},
+
+	// integers
+
+	{input: "00", ptr: new(uint8), value: uint8(0)},
+	{input: "05", ptr: new(uint8), value: uint8(5)},
+	{input: "0000", ptr: new(uint16), value: uint16(0)},
+	{input: "0005", ptr: new(uint16), value: uint16(5)},
+	{input: "05", ptr: new(uint16), error: "deser: buffer is shorter than expected"},
+	{input: "00000000", ptr: new(uint32), value: uint32(0)},
+	{input: "00000505", ptr: new(uint32), value: uint32(0x0505)},
+	{input: "05050505", ptr: new(uint32), value: uint32(0x05050505)},
+	{input: "0000000000000000", ptr: new(uint64), value: uint64(0)},
+	{input: "0000000000000505", ptr: new(uint64), value: uint64(0x0505)},
+	{input: "0505050505050505", ptr: new(uint64), value: uint64(0x0505050505050505)},
+	{input: "0100", ptr: new(uint), value: uint(0)},
+	{input: "020505", ptr: new(uint), value: uint(0x0505)},
+	{input: "080505050505050505", ptr: new(uint), value: uint(0x0505050505050505)},
+
+	{input: "00000000000000000000000000000001", ptr: new(Uint128), value: *newUint128(1)},
+	{input: "00000000000000000000000000000080", ptr: new(Uint128), value: *newUint128(128)},
+	{input: "0000000000000000FFFFFFFFFFFFFFFF", ptr: new(Uint128), value: *newUint128(0xFFFFFFFFFFFFFFFF)},
+	{input: "05", ptr: new(Uint128), error: "deser: buffer is shorter than expected"},
+	{input: "0000000000000000000000000000000000000000000000000000000000000001", ptr: new(Uint256), value: *newUint256(1)},
+	{input: "0000000000000000000000000000000000000000000000000000000000000080", ptr: new(Uint256), value: *newUint256(128)},
+	{input: "000000000000000000000000000000000000000000000000FFFFFFFFFFFFFFFF", ptr: new(Uint256), value: *newUint256(0xFFFFFFFFFFFFFFFF)},
+	{input: "05", ptr: new(Uint256), error: "deser: buffer is shorter than expected"},
+
+	//uint slices
+	{input: "00", ptr: new([]uint), value: []uint{}},
+	{input: "080102030405060708", ptr: new([]uint8), value: []uint8{1, 2, 3, 4, 5, 6, 7, 8}},
+	{input: "080000000100000002000000030000000400000005000000060000000700000008", ptr: new([]uint32), value: []uint32{1, 2, 3, 4, 5, 6, 7, 8}},
+	{input: "050102", ptr: new([]uint8), error: "deser: buffer is shorter than expected"},
+
+	// arrays
+	{input: "0102030405", ptr: new([5]uint8), value: [5]uint8{1, 2, 3, 4, 5}},
+	{input: "0000000100000002000000030000000400000005", ptr: new([5]uint32), value: [5]uint32{1, 2, 3, 4, 5}},
+	{input: "", ptr: new([0]uint), value: [0]uint{}}, // zero sized arrays
+	{input: "0102", ptr: new([5]uint8), error: "deser: buffer is shorter than expected"},
+
+	// byte slices
+	{input: "0101", ptr: new([]byte), value: []byte{1}},
+	{input: "00", ptr: new([]byte), value: []byte{}},
+	{input: "0D6162636465666768696A6B6C6D", ptr: new([]byte), value: []byte("abcdefghijklm")},
+	{input: "0D0102", ptr: new([]byte), error: "deser: buffer is shorter than expected"},
+
+	// byte slices, strings
+	{input: "00", ptr: new([]byte), value: []byte{}},
+	{input: "017E", ptr: new([]byte), value: []byte{0x7E}},
+	{input: "0180", ptr: new([]byte), value: []byte{0x80}},
+	{input: "03010203", ptr: new([]byte), value: []byte{1, 2, 3}},
+	{input: "04010203", ptr: new([]byte), error: "deser: buffer is shorter than expected"},
+
+	//SerializableList interface
+	{input: "00000000", ptr: new(LargeBytes), value: LargeBytes{[]byte{}}},
+	{input: "000000017E", ptr: new(LargeBytes), value: LargeBytes{[]byte{0x7E}}},
+	{input: "0000000180", ptr: new(LargeBytes), value: LargeBytes{[]byte{0x80}}},
+	{input: "00000003010203", ptr: new(LargeBytes), value: LargeBytes{[]byte{1, 2, 3}}},
+	{input: "03010203", ptr: new(LargeBytes), error: "deser: buffer is shorter than expected for serialize.LargeBytes.Value"},
+
+	// byte arrays
+	{input: "00", ptr: new([0]byte), value: [0]byte{}},
+	{input: "02", ptr: new([1]byte), value: [1]byte{2}},
+	{input: "80", ptr: new([1]byte), value: [1]byte{128}},
+	{input: "0102030405", ptr: new([5]byte), value: [5]byte{1, 2, 3, 4, 5}},
+
+	// strings
+	{input: "0000000100", ptr: new(string), value: "\000"},
+	{input: "0000000D6162636465666768696A6B6C6D", ptr: new(string), value: "abcdefghijklm"},
+	{input: "0D6162636465666768696A6B6C6D", ptr: new(string), error: "deser: buffer is shorter than expected"},
+
+	// big ints
+	{input: "0101", ptr: new(*big.Int), value: big.NewInt(1)},
+	{input: "09FFFFFFFFFFFFFFFFFF", ptr: new(*big.Int), value: veryBigInt},
+	{input: "0110", ptr: new(big.Int), value: *big.NewInt(16)}, // non-pointer also works
+	{input: "0210", ptr: new(big.Int), error: "deser: buffer is shorter than expected"},
+
+	// structs
+	{input: "0301020300", ptr: new(structForTest), value: newStructForTest(&[]byte{1, 2, 3}, nil)},
+	{input: "030102030103040506", ptr: new(structForTest), value: newStructForTest(&[]byte{1, 2, 3}, &[]byte{4, 5, 6})},
+	// To present (canonical marker 01) but its slice buffer is truncated. (The
+	// presence marker must be exactly 0 or 1; non-canonical markers like 03 are
+	// rejected — covered by TestDeserializeNilMarkerCanonical.)
+	{input: "0301020301040506", ptr: new(structForTest), error: "deser: buffer is shorter than expected for serialize.structForTest.To"},
+
+	// structs
+	{
+		input: "010500000003343434",
+		ptr:   new(simplestruct),
+		value: simplestruct{5, "444"},
+	},
+
+	// struct tag "-"
+	{
+		input: "010100000002",
+		ptr:   new(hasIgnoredField),
+		value: hasIgnoredField{A: 1, C: 2},
+	},
+
+	// pointers
+	{input: "0100", ptr: new(*[]byte), value: &[]byte{0}},
+	{input: "0100", ptr: new(*uint), value: uintp(0)},
+	{input: "0107", ptr: new(*uint), value: uintp(7)},
+	{input: "0180", ptr: new(*uint), value: uintp(0x80)},
+	{input: "010109", ptr: new(*[]uint), value: &[]uint{9}},
+	{input: "010403030303", ptr: new(*[][]byte), value: &[][]byte{{3, 3, 3, 3}}},
+
+	// do not support interface{}, need to know the real type
+	{input: "02", ptr: new(int), error: "type int is not serializable"},
+	{input: "00", ptr: new(interface{}), error: "type interface {} is not serializable"},
+}
+
+func uintp(i uint) *uint { return &i }
+
+func runTests(t *testing.T, deserialize func([]byte, interface{}) error) {
+	for i, test := range deserdata {
+		input, err := hex.DecodeString(test.input)
+		if err != nil {
+			t.Errorf("test %d: invalid hex input %q", i, test.input)
+			continue
+		}
+		err = deserialize(input, test.ptr)
+		if err != nil && test.error == "" {
+			t.Errorf("test %d: unexpected Deserialize error: %v\ndecoding into %T\ninput %q",
+				i, err, test.ptr, test.input)
+			continue
+		}
+		if test.error != "" && fmt.Sprint(err) != test.error {
+			t.Errorf("test %d: Deserialize error mismatch\ngot  %v\nwant %v\ndecoding into %T\ninput %q",
+				i, err, test.error, test.ptr, test.input)
+			continue
+		}
+		deref := reflect.ValueOf(test.ptr).Elem().Interface()
+		if err == nil && !reflect.DeepEqual(deref, test.value) {
+			t.Errorf("test %d: value mismatch\ngot  %#v\nwant %#v\ndecoding into %T\ninput %q",
+				i, deref, test.value, test.ptr, test.input)
+		}
+	}
+}
+
+func TestDeserialize(t *testing.T) {
+	runTests(t, func(input []byte, into interface{}) error {
+		return Deserialize(NewByteBuffer(input), into)
+	})
+}
+
+func ExampleDeserialize() {
+	input, _ := hex.DecodeString("010a0000001400000006666F6F626172")
+
+	type example struct {
+		A       uint
+		B       uint32
+		private uint // private fields are Ignored
+		String  string
+	}
+
+	var s example
+	err := Deserialize(NewByteBuffer(input), &s)
+	if err != nil {
+		fmt.Printf("Error: %v\n", err)
+	} else {
+		fmt.Printf("Deserialized value: %#v\n", s)
+	}
+	// Output:
+	// Deserialized value: serialize.example{A:0xa, B:0x14, private:0x0, String:"foobar"}
+}
+
+func ExampleDeserialize_structTagNilAndIgnore() {
+	// In this example, we'll use the "nil" struct tag to change
+	// how a pointer-typed field is deserialized. The input contains an RLP
+	// list of one element, an empty string.
+	input := []byte{0x00, 0x01, 0x03, 0x04, 0x05, 0x06}
+
+	s := new(structForTest)
+	Deserialize(NewByteBuffer(input), &s)
+	fmt.Printf("From = %v\n", *s.From)
+	fmt.Printf("To = %v\n", *s.To)
+
+	// Output:
+	// From = []
+	// To = [4 5 6]
+}
+
+func ExampleDeserialize_structTagNil() {
+	// In this example, we'll use the "nil" struct tag to change
+	// how a pointer-typed field is deserialized. The input contains an RLP
+	// list of one element, an empty string.
+	input := []byte{0x00}
+
+	// This type uses the normal rules.
+	// The empty input string is deserialized as a pointer to an empty Go string.
+	var normalRules struct {
+		String *[]byte
+	}
+	Deserialize(NewByteBuffer(input), &normalRules)
+	fmt.Printf("normal: String = %v\n", *normalRules.String)
+
+	// This type uses the struct tag.
+	// The empty input string is deserialized as a nil pointer.
+	var withEmptyOK struct {
+		String *[]byte `ser:"nil"`
+	}
+	Deserialize(NewByteBuffer(input), &withEmptyOK)
+	fmt.Printf("with nil tag: String = %v\n", withEmptyOK.String)
+
+	// Output:
+	// normal: String = []
+	// with nil tag: String = <nil>
+}
+
+func BenchmarkDeserialize(b *testing.B) {
+	enc := encodeTestSlice(90000)
+	b.SetBytes(int64(len(enc)))
+	b.ReportAllocs()
+	b.ResetTimer()
+
+	for i := 0; i < b.N; i++ {
+		var s []uint
+		bb := NewByteBuffer(enc)
+		if err := Deserialize(bb, &s); err != nil {
+			b.Fatalf("Deserialize error: %v", err)
+		}
+	}
+}
+
+func BenchmarkDeserializeIntSliceReuse(b *testing.B) {
+	enc := encodeTestSlice(100000)
+	b.SetBytes(int64(len(enc)))
+	b.ReportAllocs()
+	b.ResetTimer()
+
+	var s []uint
+	for i := 0; i < b.N; i++ {
+		bb := NewByteBuffer(enc)
+		if err := Deserialize(bb, &s); err != nil {
+			b.Fatalf("Deserialize error: %v", err)
+		}
+	}
+}
+
+func encodeTestSlice(n uint) []byte {
+	s := make([]uint, n)
+	for i := uint(0); i < n; i++ {
+		s[i] = i
+	}
+	b, err := SerializeToBytes(s)
+	if err != nil {
+		panic(fmt.Sprintf("encode error: %v", err))
+	}
+	return b
+}
+
+func unhex(str string) []byte {
+	b, err := hex.DecodeString(strings.Replace(str, " ", "", -1))
+	if err != nil {
+		panic(fmt.Sprintf("invalid hex string: %q", str))
+	}
+	return b
+}
--- a/qkc/serialize/deserializer_trailing_test.go
+++ b/qkc/serialize/deserializer_trailing_test.go
@ -0,0 +1,19 @@
+package serialize
+
+import "testing"
+
+// TestDeserializeFromBytesRejectsTrailing checks that valid data followed by
+// trailing bytes is rejected rather than silently accepted.
+func TestDeserializeFromBytesRejectsTrailing(t *testing.T) {
+	// uint8 consumes exactly one byte; the second byte is trailing garbage.
+	var v uint8
+	if err := DeserializeFromBytes([]byte{0x05, 0xff}, &v); err == nil {
+		t.Fatal("expected an error for trailing bytes after a complete decode")
+	}
+
+	// An exact input still decodes.
+	var ok uint8
+	if err := DeserializeFromBytes([]byte{0x05}, &ok); err != nil || ok != 5 {
+		t.Fatalf("exact input should decode cleanly: err=%v v=%d", err, ok)
+	}
+}
--- a/qkc/serialize/deserializer_uintwidth_test.go
+++ b/qkc/serialize/deserializer_uintwidth_test.go
@ -0,0 +1,38 @@
+package serialize
+
+import (
+	"math/bits"
+	"testing"
+)
+
+// TestDeserializeUintWidthBound checks that the variable-length reflect.Uint
+// path rejects an encoding wider than the platform uint: such an input would
+// overflow the uint64 accumulator and silently truncate the decoded value.
+func TestDeserializeUintWidthBound(t *testing.T) {
+	width := bits.UintSize / 8 // bytes in a platform uint (4 on 32-bit, 8 on 64-bit)
+
+	// One byte past the platform width: must be rejected.
+	over := make([]byte, 1+width+1)
+	over[0] = byte(width + 1) // 1-byte length prefix
+	for i := 1; i < len(over); i++ {
+		over[i] = 0xFF
+	}
+	var u uint
+	if err := Deserialize(NewByteBuffer(over), &u); err == nil {
+		t.Fatalf("expected an error for a %d-byte uint encoding (platform width %d bytes)", width+1, width)
+	}
+
+	// Exactly the platform width (max uint) still decodes.
+	in := make([]byte, 1+width)
+	in[0] = byte(width)
+	for i := 1; i < len(in); i++ {
+		in[i] = 0xFF
+	}
+	var ok uint
+	if err := Deserialize(NewByteBuffer(in), &ok); err != nil {
+		t.Fatalf("%d-byte uint rejected: %v", width, err)
+	}
+	if ok != ^uint(0) {
+		t.Fatalf("expected max uint, got %d", ok)
+	}
+}
--- a/qkc/serialize/serializer.go
+++ b/qkc/serialize/serializer.go
@ -0,0 +1,321 @@
+// Ported verbatim from github.com/QuarkChain/goquarkchain/serialize (byte-compatible).
+
+package serialize
+
+import (
+	"encoding/binary"
+	"errors"
+	"fmt"
+	"math/big"
+	"reflect"
+)
+
+// errSerializeNil is returned when Serialize is given an untyped nil, which
+// carries no type to encode. Mirrors DeserializeWithTags rejecting a nil target
+// (rather than panicking in reflect.ValueOf(nil).Type()).
+var errSerializeNil = errors.New("ser: nil value given to Serialize")
+
+func Serialize(w *[]byte, val interface{}) error {
+	return SerializeWithTags(w, val, Tags{ByteSizeOfSliceLen: 1})
+}
+
+func SerializeWithTags(w *[]byte, val interface{}, ts Tags) error {
+	if val == nil {
+		return errSerializeNil
+	}
+
+	rval := reflect.ValueOf(val)
+	ti, err := cachedTypeInfo(rval.Type())
+	if err != nil {
+		return err
+	}
+
+	return ti.serializer(rval, w, ts)
+}
+
+// SerializeToBytes returns the serialize result of val.
+func SerializeToBytes(val interface{}) ([]byte, error) {
+	w := make([]byte, 0, 512)
+	if err := Serialize(&w, val); err != nil {
+		return nil, err
+	}
+
+	return w, nil
+}
+
+func makeSerializer(typ reflect.Type) (serializer, error) {
+	kind := typ.Kind()
+	switch {
+	//check Ptr first and add optional byte output if ts is nilok,
+	//then get serializer for typ.Elem() which is not a ptr
+	case kind == reflect.Ptr:
+		return serializePtr, nil
+	case kind != reflect.Ptr && reflect.PtrTo(typ).Implements(serializableInterface):
+		return serializeSerializableInterface, nil
+	case typ.AssignableTo(bigInt):
+		return serializeBigIntNoPtr, nil
+	case isUint(kind):
+		return serializeUint, nil
+	case kind == reflect.Bool:
+		return serializeBool, nil
+	case kind == reflect.String:
+		return serializeString, nil
+	case kind == reflect.Slice && isByte(typ.Elem()):
+		return serializeByteSlice, nil
+	case kind == reflect.Array && isByte(typ.Elem()):
+		return serializeByteArray, nil
+	case kind == reflect.Slice || kind == reflect.Array:
+		return serializeList, nil
+	case kind == reflect.Struct:
+		return serializeStruct, nil
+	default:
+		return nil, fmt.Errorf("type %v is not serializable", typ)
+	}
+}
+
+func serializeSerializableInterface(val reflect.Value, w *[]byte, tags Tags) error {
+	if !val.CanAddr() {
+		return fmt.Errorf("ser: unaddressable value of type %v, Serialize is pointer method", val.Type())
+	}
+
+	return val.Addr().Interface().(Serializable).Serialize(w)
+}
+
+func prefillByteArray(size int, barray []byte) ([]byte, error) {
+	len := len(barray)
+	if len > size {
+		return nil, errors.New("barray len is larger then expected size")
+	}
+	if len == size {
+		return barray, nil
+	}
+
+	bytes := make([]byte, size, size)
+	var startIndex = size - len
+	copy(bytes[startIndex:], barray)
+	return bytes, nil
+}
+
+func serializeFixSizeBigUint(val *big.Int, size int, w *[]byte) error {
+	if val == nil {
+		bytes := make([]byte, size, size)
+		*w = append(*w, bytes...)
+		return nil
+	}
+	// big.Int.Bytes() returns the absolute value, so without this check a
+	// negative value would serialize identically to its positive counterpart
+	// (e.g. -1 as +1), silently changing it. Reject it, matching the
+	// variable-length serializeBigInt path. This is an intentional hardening
+	// divergence from the verbatim goquarkchain port, whose fixed-size path has
+	// the same silent sign loss; it does not change any valid (non-negative)
+	// encoding.
+	if val.Sign() < 0 {
+		return fmt.Errorf("ser: cannot serialize negative big.Int")
+	}
+	bytes, err := prefillByteArray(size, val.Bytes())
+	if err == nil {
+		*w = append(*w, bytes...)
+	}
+
+	return err
+}
+
+func serializeBigIntNoPtr(val reflect.Value, w *[]byte, ts Tags) error {
+	i := val.Interface().(big.Int)
+	return serializeBigInt(&i, w)
+}
+
+func serializeBigInt(i *big.Int, w *[]byte) error {
+	var bytes []byte
+	if cmp := i.Cmp(big.NewInt(0)); cmp == -1 {
+		return fmt.Errorf("ser: cannot serialize negative *big.Int")
+	} else if cmp > 0 {
+		bytes = i.Bytes()
+	}
+
+	// The length is written as a single byte, so a value needing more than 255
+	// bytes (>2040 bits) would have its length truncated by uint8() and decode
+	// to a different value — a non-roundtrippable encoding. Reject it. (Real
+	// QKC quantities are far below this; this guards the general codec.)
+	if len(bytes) > 255 {
+		return fmt.Errorf("ser: big.Int too large to serialize: %d bytes exceeds the single-byte length prefix (max 255)", len(bytes))
+	}
+
+	*w = append(*w, uint8(len(bytes)))
+	*w = append(*w, bytes...)
+	return nil
+}
+
+func serializeBool(val reflect.Value, w *[]byte, ts Tags) error {
+	if val.Bool() {
+		*w = append(*w, 0x01)
+	} else {
+		*w = append(*w, 0x00)
+	}
+
+	return nil
+}
+
+func serializeByteArray(val reflect.Value, w *[]byte, ts Tags) error {
+	if !val.CanAddr() {
+		// Slice requires the value to be addressable.
+		// Make it addressable by copying.
+		copy := reflect.New(val.Type()).Elem()
+		copy.Set(val)
+		val = copy
+	}
+
+	size := val.Len()
+	slice := val.Slice(0, size).Bytes()
+
+	*w = append(*w, slice...)
+	return nil
+}
+
+func writeListLen(w *[]byte, len int, byteSizeOfSliceLen int) error {
+	sizeBytes := make([]byte, byteSizeOfSliceLen)
+	for i := byteSizeOfSliceLen - 1; i >= 0 && len != 0; i-- {
+		sizeBytes[i] = byte(len)
+		len = len >> 8
+	}
+	if len > 0 {
+		return errors.New("barray len is larger then expected size")
+	}
+
+	*w = append(*w, sizeBytes...)
+	return nil
+}
+
+// serializePrependedSizeBytes
+func serializeByteSlice(val reflect.Value, w *[]byte, ts Tags) error {
+	err := writeListLen(w, val.Len(), ts.ByteSizeOfSliceLen)
+	if err != nil {
+		return err
+	}
+
+	bytes := val.Bytes()
+	*w = append(*w, bytes...)
+	return nil
+}
+
+// PrependedSizeListSerializer
+func serializeList(val reflect.Value, w *[]byte, ts Tags) error {
+	typeinfo, err := cachedTypeInfo(val.Type().Elem())
+	if err != nil {
+		return err
+	}
+
+	if val.Kind() == reflect.Slice {
+		err = writeListLen(w, val.Len(), ts.ByteSizeOfSliceLen)
+		if err != nil {
+			return err
+		}
+	}
+
+	vlen := val.Len()
+	for i := 0; i < vlen; i++ {
+		if err := typeinfo.serializer(val.Index(i), w, Tags{ByteSizeOfSliceLen: 1}); err != nil {
+			return err
+		}
+	}
+
+	return nil
+}
+
+func serializeStruct(val reflect.Value, w *[]byte, ts Tags) error {
+	fields, err := structFields(val.Type())
+	if err != nil {
+		return err
+	}
+
+	for _, f := range fields {
+		if err := f.info.serializer(val.Field(f.index), w, f.tags); err != nil {
+			return err
+		}
+	}
+
+	return nil
+}
+
+func SerializeStructWithout(val reflect.Value, w *[]byte, excludeList map[string]bool) error {
+	fields, err := structFields(val.Type())
+	if err != nil {
+		return err
+	}
+
+	for _, f := range fields {
+		if excludeList != nil {
+			if _, ok := excludeList[f.name]; ok {
+				continue
+			}
+		}
+
+		if err := f.info.serializer(val.Field(f.index), w, f.tags); err != nil {
+			return err
+		}
+	}
+
+	return nil
+}
+
+func serializeString(val reflect.Value, w *[]byte, ts Tags) error {
+	s := val.String()
+
+	sizeBytes := make([]byte, 4, 4)
+	binary.BigEndian.PutUint32(sizeBytes, uint32(val.Len()))
+
+	*w = append(*w, sizeBytes...)
+	*w = append(*w, s...)
+	return nil
+}
+
+func serializePtr(val reflect.Value, w *[]byte, ts Tags) error {
+	typ := val.Type()
+	typeinfo, err := cachedTypeInfo(typ.Elem())
+	if err != nil {
+		return err
+	}
+	switch {
+	case val.IsNil() && ts.NilOK:
+		*w = append(*w, 0)
+		return nil
+	case val.IsNil() && typ.Implements(serializableInterface):
+		zero := reflect.New(typ.Elem())
+		return typeinfo.serializer(zero.Elem(), w, ts)
+	case val.IsNil():
+		zero := reflect.Zero(typ.Elem())
+		return typeinfo.serializer(zero, w, ts)
+	default:
+		if ts.NilOK {
+			*w = append(*w, 1)
+		}
+		return typeinfo.serializer(val.Elem(), w, ts)
+	}
+}
+
+func serializeUint(val reflect.Value, w *[]byte, ts Tags) error {
+	kind := val.Type().Kind()
+	value := val.Uint()
+	switch kind {
+	case reflect.Uint8:
+		*w = append(*w, byte(value))
+		return nil
+	case reflect.Uint16:
+		*w = append(*w, byte(value>>8), byte(value))
+		break
+	case reflect.Uint32:
+		*w = append(*w, byte(value>>24), byte(value>>16), byte(value>>8), byte(value))
+		return nil
+	case reflect.Uint64:
+		*w = append(*w, byte(value>>56), byte(value>>48), byte(value>>40), byte(value>>32),
+			byte(value>>24), byte(value>>16), byte(value>>8), byte(value))
+		return nil
+	case reflect.Uint:
+		//As Uint would be 32/64 bit, so
+		var bi big.Int
+		return serializeBigInt(bi.SetUint64(val.Uint()), w)
+	default:
+		return fmt.Errorf("ser: invalid Uint type: %s", val.Type().Name())
+	}
+	return nil
+}
--- a/qkc/serialize/serializer_bound_test.go
+++ b/qkc/serialize/serializer_bound_test.go
@ -0,0 +1,36 @@
+package serialize
+
+import (
+	"math/big"
+	"testing"
+)
+
+// TestSerializeBigIntLengthBound guards the variable-length big.Int encoding:
+// its length prefix is a single byte, so a value needing more than 255 bytes
+// would truncate the prefix and decode to a different value. It must error
+// instead, and values at/under the limit must still round-trip.
+func TestSerializeBigIntLengthBound(t *testing.T) {
+	// 2^2048 needs 257 bytes (> 255): must be rejected.
+	tooBig := new(big.Int).Lsh(big.NewInt(1), 2048)
+	var w []byte
+	if err := Serialize(&w, tooBig); err == nil {
+		t.Fatal("expected an error serializing a big.Int needing >255 bytes (non-roundtrippable length prefix)")
+	}
+
+	// 2^2032 needs exactly 255 bytes: the maximum that fits, must round-trip.
+	maxFit := new(big.Int).Lsh(big.NewInt(1), 8*254)
+	if got := len(maxFit.Bytes()); got != 255 {
+		t.Fatalf("test setup: expected 255 bytes, got %d", got)
+	}
+	w = nil
+	if err := Serialize(&w, maxFit); err != nil {
+		t.Fatalf("255-byte big.Int rejected: %v", err)
+	}
+	var back big.Int
+	if err := Deserialize(NewByteBuffer(w), &back); err != nil {
+		t.Fatalf("deserialize: %v", err)
+	}
+	if back.Cmp(maxFit) != 0 {
+		t.Fatalf("round-trip mismatch")
+	}
+}
--- a/qkc/serialize/serializer_nil_test.go
+++ b/qkc/serialize/serializer_nil_test.go
@ -0,0 +1,25 @@
+package serialize
+
+import "testing"
+
+// TestSerializeNilValueRejected checks that an untyped nil is rejected with an
+// error rather than panicking, symmetric with DeserializeWithTags rejecting a
+// nil target. A typed nil pointer is still serializable.
+func TestSerializeNilValueRejected(t *testing.T) {
+	var w []byte
+	if err := Serialize(&w, nil); err == nil {
+		t.Fatal("expected an error serializing untyped nil")
+	}
+	if _, err := SerializeToBytes(nil); err == nil {
+		t.Fatal("expected an error from SerializeToBytes(nil)")
+	}
+
+	// A typed nil pointer carries a type, so it still serializes (as its
+	// element's zero value).
+	type foo struct{ X uint32 }
+	var p *foo
+	w = nil
+	if err := Serialize(&w, p); err != nil {
+		t.Fatalf("typed nil pointer should serialize: %v", err)
+	}
+}
--- a/qkc/serialize/serializer_test.go
+++ b/qkc/serialize/serializer_test.go
@ -0,0 +1,213 @@
+// Ported verbatim from github.com/QuarkChain/goquarkchain/serialize (byte-compatible).
+
+package serialize
+
+import (
+	"bytes"
+	"errors"
+	"fmt"
+	"math/big"
+	"testing"
+)
+
+type serializableStruct struct {
+	val uint64
+	err error
+}
+
+func (e *serializableStruct) Serialize(w *[]byte) error {
+	if e.err != nil {
+		return e.err
+	} else {
+		*w = append(*w, new(big.Int).SetUint64(e.val).Bytes()...)
+	}
+	return nil
+}
+
+func (e *serializableStruct) Deserialize(bb *ByteBuffer) error {
+	if e == nil {
+
+	} else if bb.Remaining() == 4 {
+		e = nil
+	} else {
+		e.val = 1
+	}
+
+	return nil
+}
+
+type structForTest struct {
+	From         *[]byte
+	To           *[]byte `json:"to"                 ser:"nil"`
+	IgnoredField int     `json:"ignore"             ser:"-"`
+	privateField int     //private field will be Ignored
+}
+
+func newStructForTest(from, to *[]byte) structForTest {
+	s := structForTest{from, to, 0, 0}
+	return s
+}
+
+type testDataForSerialize struct {
+	val           interface{}
+	output, error string
+}
+
+func newUint128(val uint64) *Uint128 {
+	ui := new(Uint128)
+	ui.Value = new(big.Int).SetUint64(val)
+	return ui
+}
+
+func newUint256(val uint64) *Uint256 {
+	ui := new(Uint256)
+	ui.Value = new(big.Int).SetUint64(val)
+	return ui
+}
+
+type LargeBytes struct {
+	Value []byte `bytesizeofslicelen:"4"`
+}
+
+var serdata = []testDataForSerialize{
+	// booleans
+	{val: true, output: "01"},
+	{val: false, output: "00"},
+
+	// integers
+	{val: uint8(0), output: "00"},
+	{val: uint8(128), output: "80"},
+	{val: uint16(0), output: "0000"},
+	{val: uint16(128), output: "0080"},
+	{val: uint32(0), output: "00000000"},
+	{val: uint32(128), output: "00000080"},
+	{val: uint32(1024), output: "00000400"},
+	{val: uint32(0xFFFFFFFF), output: "FFFFFFFF"},
+	{val: uint64(0), output: "0000000000000000"},
+	{val: uint64(128), output: "0000000000000080"},
+	{val: uint64(0xFFFFFFFF), output: "00000000FFFFFFFF"},
+	{val: uint64(0xFFFFFFFFFFFFFFFF), output: "FFFFFFFFFFFFFFFF"},
+	{val: uint(0), output: "00"},
+	{val: uint(128), output: "0180"},
+	{val: uint(1024), output: "020400"},
+	{val: uint(0xFFFFFFFF), output: "04FFFFFFFF"},
+
+	{val: newUint128(0), output: "00000000000000000000000000000000"},
+	{val: newUint128(128), output: "00000000000000000000000000000080"},
+	{val: newUint128(0xFFFFFFFFFFFFFFFF), output: "0000000000000000FFFFFFFFFFFFFFFF"},
+	{val: newUint256(0), output: "0000000000000000000000000000000000000000000000000000000000000000"},
+	{val: newUint256(128), output: "0000000000000000000000000000000000000000000000000000000000000080"},
+	{val: newUint256(0xFFFFFFFFFFFFFFFF), output: "000000000000000000000000000000000000000000000000FFFFFFFFFFFFFFFF"},
+
+	// big integers (should match uint for small values)
+	{val: big.NewInt(0), output: "00"},
+	{val: big.NewInt(1), output: "0101"},
+	{val: big.NewInt(128), output: "0180"},
+	{val: big.NewInt(256), output: "020100"},
+	{val: big.NewInt(1024), output: "020400"},
+	{val: big.NewInt(0xFFFFFF), output: "03FFFFFF"},
+	{val: big.NewInt(0xFFFFFFFFFFFFFF), output: "07FFFFFFFFFFFFFF"},
+	{
+		val:    big.NewInt(0).SetBytes(unhex("102030405060708090A0B0C0D0E0F2")),
+		output: "0F102030405060708090A0B0C0D0E0F2",
+	},
+	{
+		val:    big.NewInt(0).SetBytes(unhex("0100020003000400050006000700080009000A000B000C000D000E01")),
+		output: "1C0100020003000400050006000700080009000A000B000C000D000E01",
+	},
+
+	// non-pointer big.Int
+	{val: *big.NewInt(0), output: "00"},
+	{val: *big.NewInt(0xFFFFFF), output: "03FFFFFF"},
+
+	// negative ints are not supported
+	{val: big.NewInt(-1), error: "ser: cannot serialize negative *big.Int"},
+
+	// byte slices, strings
+	{val: []byte{}, output: "00"},
+	{val: []byte{0x7E}, output: "017E"},
+	{val: []byte{0x80}, output: "0180"},
+	{val: []byte{1, 2, 3}, output: "03010203"},
+
+	{val: &LargeBytes{}, output: "00000000"},
+	{val: &LargeBytes{[]byte{0x7E}}, output: "000000017E"},
+	{val: &LargeBytes{[]byte{0x80}}, output: "0000000180"},
+	{val: &LargeBytes{[]byte{1, 2, 3}}, output: "00000003010203"},
+
+	{val: "", output: "00000000"},
+	{val: "\x7E", output: "000000017E"},
+	{val: "\x80", output: "0000000180"},
+	{val: "dog", output: "00000003646F67"},
+
+	// slices
+	{val: []uint8{}, output: "00"},
+	{val: []uint8{1, 2, 3}, output: "03010203"},
+	{val: []uint32{}, output: "00"},
+	{val: []uint32{1, 2, 3}, output: "03000000010000000200000003"},
+
+	//Array
+	{val: [3]uint8{1, 2, 3}, output: "010203"},
+	{val: [3]uint32{1, 2, 3}, output: "000000010000000200000003"},
+
+	// structs
+	{val: newStructForTest(&[]byte{1, 2, 3}, nil), output: "0301020300"},
+	{val: newStructForTest(&[]byte{1, 2, 3}, &[]byte{4, 5, 6}), output: "030102030103040506"},
+	{val: newStructForTest(nil, &[]byte{4, 5, 6}), output: "000103040506"},
+
+	// nil
+	// as nilOk default value is false, serialize will use default
+	// value to serialize instead of nil
+	{val: (*uint)(nil), output: "00"},
+	{val: (*string)(nil), output: "00000000"},
+	{val: (*[]byte)(nil), output: "00"},
+	{val: (*[10]byte)(nil), output: "00000000000000000000"},
+	{val: (*big.Int)(nil), output: "00"},
+	{val: (*[]string)(nil), output: "00"},
+	{val: (*[10]string)(nil), output: "00000000000000000000000000000000000000000000000000000000000000000000000000000000"},
+	// Exported field so the element encodes to >= 1 byte: a slice whose element
+	// encodes to zero bytes (e.g. the embedded-unexported struct{ uint }) is now
+	// rejected at registration — see TestZeroByteListElementRejected.
+	{val: (*[]struct{ X uint })(nil), output: "00"},
+
+	// interfaces
+	// Serializer
+	{val: (*serializableStruct)(nil), output: ""},
+	{val: &serializableStruct{val: 0xFFFF}, output: "FFFF"},
+	{val: &serializableStruct{1, errors.New("test error")}, error: "test error"},
+
+	// int is not support
+	{val: int(0), error: "type int is not serializable"},
+	{val: (*interface{})(nil), error: "type interface {} is not serializable"},
+}
+
+func runEncTests(t *testing.T, f func(val interface{}) ([]byte, error)) {
+	for i, test := range serdata {
+		output, err := f(test.val)
+		if err != nil && test.error == "" {
+			t.Errorf("test %d: unexpected error: %v\nvalue %#v\ntype %T",
+				i, err, test.val, test.val)
+			continue
+		}
+		if test.error != "" && fmt.Sprint(err) != test.error {
+			t.Errorf("test %d: error mismatch\ngot   %v\nwant  %v\nvalue %#v\ntype  %T",
+				i, err, test.error, test.val, test.val)
+			continue
+		}
+		if err == nil && !bytes.Equal(output, unhex(test.output)) {
+			t.Errorf("test %d: output mismatch:\ngot   %X\nwant  %s\nvalue %#v\ntype  %T",
+				i, output, test.output, test.val, test.val)
+		}
+	}
+}
+
+func TestSerialize(t *testing.T) {
+	runEncTests(t, func(val interface{}) ([]byte, error) {
+		b := make([]byte, 0, 1)
+		err := Serialize(&b, val)
+		return b, err
+	})
+}
+
+func TestSerializeToBytes(t *testing.T) {
+	runEncTests(t, SerializeToBytes)
+}
--- a/qkc/serialize/typecache.go
+++ b/qkc/serialize/typecache.go
@ -0,0 +1,224 @@
+// Ported verbatim from github.com/QuarkChain/goquarkchain/serialize (byte-compatible).
+// Modified from go-ethereum under GNU Lesser General Public License
+
+package serialize
+
+import (
+	"fmt"
+	"reflect"
+	"strconv"
+	"strings"
+	"sync"
+)
+
+var (
+	typeCacheMutex   sync.RWMutex
+	typeCache        = make(map[reflect.Type]*typeinfo)
+	fieldsCacheMutex sync.RWMutex
+	fieldsCache      = make(map[reflect.Type][]field)
+)
+
+type typeinfo struct {
+	serializer
+	deserializer
+}
+
+// represents struct Tags
+type Tags struct {
+	// ser:"nil" controls whether empty input results in a nil pointer.
+	NilOK bool
+	// ser:"-" ignores fields.
+	Ignored bool
+	// bytesizeofslicelen: number
+	ByteSizeOfSliceLen int
+}
+
+type deserializer func(*ByteBuffer, reflect.Value, Tags) error
+
+type serializer func(reflect.Value, *[]byte, Tags) error
+
+func cachedTypeInfo(typ reflect.Type) (*typeinfo, error) {
+	typeCacheMutex.RLock()
+	info := typeCache[typ]
+	typeCacheMutex.RUnlock()
+	if info != nil {
+		return info, nil
+	}
+
+	typeCacheMutex.Lock()
+	defer typeCacheMutex.Unlock()
+
+	info, err := genTypeInfo(typ)
+	if err != nil {
+		return nil, err
+	}
+
+	typeCache[typ] = info
+	return typeCache[typ], err
+}
+
+type field struct {
+	index int
+	info  *typeinfo
+	tags  Tags
+	name  string
+}
+
+func structFields(typ reflect.Type) (fields []field, err error) {
+	fieldsCacheMutex.RLock()
+	flds := fieldsCache[typ]
+	fieldsCacheMutex.RUnlock()
+	if flds != nil {
+		return flds, nil
+	}
+	for i := 0; i < typ.NumField(); i++ {
+		if f := typ.Field(i); f.PkgPath == "" { // exported
+			tags, err := parseStructTag(typ, i)
+			if err != nil {
+				return nil, err
+			}
+			if tags.Ignored {
+				continue
+			}
+			info, err := cachedTypeInfo(f.Type)
+			if err != nil {
+				return nil, err
+			}
+			fields = append(fields, field{i, info, tags, f.Name})
+		}
+	}
+	fieldsCacheMutex.Lock()
+	defer fieldsCacheMutex.Unlock()
+	fieldsCache[typ] = fields
+
+	return fields, nil
+}
+
+func parseStructTag(typ reflect.Type, fi int) (Tags, error) {
+	f := typ.Field(fi)
+	var ts Tags
+	ts.ByteSizeOfSliceLen = 1
+	for _, t := range strings.Split(f.Tag.Get("ser"), ",") {
+		switch t = strings.TrimSpace(t); t {
+		case "":
+		case "-":
+			ts.Ignored = true
+		case "nil": // nil equal to optional in PyQuackChain
+			ts.NilOK = true
+		default:
+			return ts, fmt.Errorf("ser: unknown struct tag %q on %v.%s", t, typ, f.Name)
+		}
+	}
+	// bytesizeofslicelen use to specify the number of bytes used to save a slice len
+	// only slice is useful
+	if f.Type.Kind() == reflect.Slice {
+		for _, t := range strings.Split(f.Tag.Get("bytesizeofslicelen"), ",") {
+			t = strings.TrimSpace(t)
+			if t != "" {
+				num, err := strconv.Atoi(t)
+				if err != nil {
+					return ts, err
+				}
+
+				ts.ByteSizeOfSliceLen = num
+			}
+		}
+	}
+
+	return ts, nil
+}
+
+// encodesZeroBytes reports whether a value of typ, serialized as a list element
+// (default tags, i.e. no nil marker), can encode to zero bytes — e.g. struct{},
+// a struct whose fields are all ignored/unexported, a zero-length array, or a
+// pointer/array/struct recursively composed of those. genTypeInfo rejects slices
+// of such elements: they carry no per-element byte cost, which both makes the
+// bytes-remaining bound in deserializeList reject a valid round-trip and leaves
+// that bound unable to limit allocation.
+//
+// It mirrors makeSerializer's dispatch and must stay lock-free (no cachedTypeInfo
+// / structFields), as it runs inside genTypeInfo while typeCacheMutex is held.
+// The visited set breaks reference cycles, conservatively treating an in-progress
+// type as non-zero. Serializable types are assumed non-zero, since their custom
+// encoding can't be inspected statically (QKC's all write >= 1 byte).
+func encodesZeroBytes(typ reflect.Type, visited map[reflect.Type]bool) bool {
+	if visited[typ] {
+		return false
+	}
+	// Mark typ only for the current recursion path and unmark on return, so the
+	// set tracks ancestors, not every type ever seen. Without the delete, the
+	// same zero-byte type used as repeated sibling fields (e.g.
+	// struct{ A empty; B empty }) would hit a stale mark on its second occurrence
+	// and be misclassified as non-zero, breaking serialize/deserialize symmetry.
+	// A type still on the path is a genuine cycle, kept conservatively as non-zero.
+	visited[typ] = true
+	defer delete(visited, typ)
+
+	switch {
+	case typ.Kind() == reflect.Ptr:
+		// A list element pointer carries no nil marker, so it costs exactly what
+		// its element costs.
+		return encodesZeroBytes(typ.Elem(), visited)
+	case reflect.PtrTo(typ).Implements(serializableInterface):
+		return false
+	case typ.AssignableTo(bigInt):
+		return false
+	case isUint(typ.Kind()):
+		return false
+	case typ.Kind() == reflect.Bool:
+		return false
+	case typ.Kind() == reflect.String:
+		return false
+	case typ.Kind() == reflect.Slice:
+		// A slice always writes a length prefix (>= 1 byte).
+		return false
+	case typ.Kind() == reflect.Array:
+		if typ.Len() == 0 {
+			return true
+		}
+		if isByte(typ.Elem()) {
+			return false
+		}
+		return encodesZeroBytes(typ.Elem(), visited)
+	case typ.Kind() == reflect.Struct:
+		for i := 0; i < typ.NumField(); i++ {
+			f := typ.Field(i)
+			if f.PkgPath != "" { // unexported fields are skipped by the codec
+				continue
+			}
+			tags, err := parseStructTag(typ, i)
+			if err != nil || tags.Ignored {
+				continue
+			}
+			if tags.NilOK {
+				return false // a nil marker byte is always written
+			}
+			if !encodesZeroBytes(f.Type, visited) {
+				return false
+			}
+		}
+		return true
+	default:
+		return false
+	}
+}
+
+func genTypeInfo(typ reflect.Type) (info *typeinfo, err error) {
+	info = new(typeinfo)
+	if info.serializer, err = makeSerializer(typ); err != nil {
+		return nil, err
+	}
+	if info.deserializer, err = makeDeserializer(typ); err != nil {
+		return nil, err
+	}
+	// A slice whose element encodes to zero bytes is unsupported: such a list
+	// serializes to just its length prefix, so deserializeList's bytes-remaining
+	// bound would reject the valid round-trip (and could not bound allocation
+	// anyway). Reject it symmetrically here so serialize and deserialize agree.
+	// Byte slices are exempt (a byte is one byte); arrays are exempt (their length
+	// is fixed by the type, not read from input, so there is nothing to bound).
+	if typ.Kind() == reflect.Slice && !isByte(typ.Elem()) && encodesZeroBytes(typ.Elem(), make(map[reflect.Type]bool)) {
+		return nil, fmt.Errorf("ser: list element type %v encodes to zero bytes, which is unsupported", typ.Elem())
+	}
+	return info, nil
+}
--- a/qkc/serialize/typecache_zerobyte_test.go
+++ b/qkc/serialize/typecache_zerobyte_test.go
@ -0,0 +1,86 @@
+package serialize
+
+import "testing"
+
+// allIgnoredFields encodes to zero bytes: A is ignored, b is unexported.
+type allIgnoredFields struct {
+	A int `ser:"-"`
+	b int
+}
+
+// TestZeroByteListElementRejected pins the contract that a slice whose element
+// encodes to zero bytes is unsupported and is rejected SYMMETRICALLY — both
+// serialize and deserialize must fail. (Previously serialize emitted just the
+// length prefix while deserialize refused it, an encode/decode asymmetry.)
+func TestZeroByteListElementRejected(t *testing.T) {
+	cases := []struct {
+		name string
+		ser  func() error
+		des  func() error
+	}{
+		{
+			"struct{}",
+			func() error { _, e := SerializeToBytes([]struct{}{{}, {}}); return e },
+			func() error { var v []struct{}; return DeserializeFromBytes([]byte{0x02}, &v) },
+		},
+		{
+			"all ignored/unexported fields",
+			func() error { _, e := SerializeToBytes([]allIgnoredFields{{}, {}}); return e },
+			func() error { var v []allIgnoredFields; return DeserializeFromBytes([]byte{0x02}, &v) },
+		},
+		{
+			"zero-length array element",
+			func() error { _, e := SerializeToBytes([][0]uint32{{}, {}}); return e },
+			func() error { var v [][0]uint32; return DeserializeFromBytes([]byte{0x02}, &v) },
+		},
+	}
+	for _, c := range cases {
+		if err := c.ser(); err == nil {
+			t.Errorf("%s: SerializeToBytes should be rejected, got nil", c.name)
+		}
+		if err := c.des(); err == nil {
+			t.Errorf("%s: DeserializeFromBytes should be rejected, got nil", c.name)
+		}
+	}
+	_ = allIgnoredFields{}.b // silence unused-field linters
+}
+
+type emptyElem struct{}
+
+// repeatedEmpties has two fields of the SAME zero-byte type, so it still encodes
+// to zero bytes. It exercises the visited-map reset in encodesZeroBytes: a stale
+// mark would misclassify the second field as non-zero, letting []repeatedEmpties
+// serialize to a bare length prefix that deserialize then rejects (an asymmetry).
+type repeatedEmpties struct {
+	A emptyElem
+	B emptyElem
+}
+
+// TestZeroByteRepeatedFieldTypeRejected pins that a slice whose element repeats a
+// zero-byte type is rejected SYMMETRICALLY (both serialize and deserialize fail).
+func TestZeroByteRepeatedFieldTypeRejected(t *testing.T) {
+	if _, err := SerializeToBytes([]repeatedEmpties{{}, {}}); err == nil {
+		t.Error("SerializeToBytes([]repeatedEmpties) should be rejected, got nil")
+	}
+	var v []repeatedEmpties
+	if err := DeserializeFromBytes([]byte{0x02}, &v); err == nil {
+		t.Error("DeserializeFromBytes into []repeatedEmpties should be rejected, got nil")
+	}
+}
+
+// TestNonZeroByteListStillRoundTrips guards against over-rejection: a slice
+// whose element consumes >= 1 byte must still encode and decode normally.
+func TestNonZeroByteListStillRoundTrips(t *testing.T) {
+	in := []uint32{7, 9}
+	b, err := SerializeToBytes(in)
+	if err != nil {
+		t.Fatalf("serialize: %v", err)
+	}
+	var out []uint32
+	if err := DeserializeFromBytes(b, &out); err != nil {
+		t.Fatalf("deserialize: %v", err)
+	}
+	if len(out) != 2 || out[0] != 7 || out[1] != 9 {
+		t.Fatalf("round-trip mismatch: %v", out)
+	}
+}
--- a/qkc/serialize/utils.go
+++ b/qkc/serialize/utils.go
@ -0,0 +1,58 @@
+// Ported verbatim from github.com/QuarkChain/goquarkchain/serialize (byte-compatible).
+
+package serialize
+
+import (
+	"math/big"
+	"reflect"
+)
+
+var (
+	serializableInterface = reflect.TypeOf(new(Serializable)).Elem()
+	bigInt                = reflect.TypeOf(big.Int{})
+	typUint128            = reflect.TypeOf(Uint128{})
+	typUint256            = reflect.TypeOf(Uint256{})
+	big0                  = big.NewInt(0)
+)
+
+type BigUint struct {
+	Value *big.Int
+}
+
+type Uint128 BigUint
+type Uint256 BigUint
+
+func (ui *Uint128) Serialize(w *[]byte) error {
+	return serializeFixSizeBigUint(ui.Value, 16, w)
+}
+
+func (ui *Uint128) Deserialize(bb *ByteBuffer) error {
+	if ui.Value == nil {
+		ui.Value = new(big.Int)
+	}
+	return deserializeFixSizeBigUint(bb, ui.Value, 16)
+}
+
+func (ui *Uint256) Serialize(w *[]byte) error {
+	return serializeFixSizeBigUint(ui.Value, 32, w)
+}
+
+func (ui *Uint256) Deserialize(bb *ByteBuffer) error {
+	if ui.Value == nil {
+		ui.Value = new(big.Int)
+	}
+	return deserializeFixSizeBigUint(bb, ui.Value, 32)
+}
+
+type Serializable interface {
+	Serialize(w *[]byte) error
+	Deserialize(bb *ByteBuffer) error
+}
+
+func isUint(k reflect.Kind) bool {
+	return k >= reflect.Uint && k <= reflect.Uintptr
+}
+
+func isByte(typ reflect.Type) bool {
+	return typ.Kind() == reflect.Uint8 && !typ.Implements(serializableInterface)
+}
--- a/qkc/serialize/utils_negative_test.go
+++ b/qkc/serialize/utils_negative_test.go
@ -0,0 +1,34 @@
+package serialize
+
+import (
+	"math/big"
+	"testing"
+)
+
+// TestFixedSizeUintRejectsNegative guards against silent sign loss: because
+// big.Int.Bytes() returns the absolute value, a negative Uint128/Uint256 would
+// otherwise serialize identically to its positive counterpart. It must error
+// instead.
+func TestFixedSizeUintRejectsNegative(t *testing.T) {
+	cases := []struct {
+		name string
+		ser  func() error
+	}{
+		{"Uint128", func() error { var w []byte; return (&Uint128{Value: big.NewInt(-1)}).Serialize(&w) }},
+		{"Uint256", func() error { var w []byte; return (&Uint256{Value: big.NewInt(-1)}).Serialize(&w) }},
+	}
+	for _, tc := range cases {
+		if err := tc.ser(); err == nil {
+			t.Fatalf("%s: expected an error serializing a negative value, got nil (sign silently dropped)", tc.name)
+		}
+	}
+
+	// Non-negative values still serialize to their fixed width.
+	var w []byte
+	if err := (&Uint256{Value: big.NewInt(1)}).Serialize(&w); err != nil {
+		t.Fatalf("positive Uint256 rejected: %v", err)
+	}
+	if len(w) != 32 {
+		t.Fatalf("Uint256 should serialize to 32 bytes, got %d", len(w))
+	}
+}