Skip to content

Commit

Permalink
refactor: use modern Protobuf encoder (syncthing#9817)
Browse files Browse the repository at this point in the history
At a high level, this is what I've done and why:

- I'm moving the protobuf generation for the `protocol`, `discovery` and
`db` packages to the modern alternatives, and using `buf` to generate
because it's nice and simple.
- After trying various approaches on how to integrate the new types with
the existing code, I opted for splitting off our own data model types
from the on-the-wire generated types. This means we can have a
`FileInfo` type with nicer ergonomics and lots of methods, while the
protobuf generated type stays clean and close to the wire protocol. It
does mean copying between the two when required, which certainly adds a
small amount of inefficiency. If we want to walk this back in the future
and use the raw generated type throughout, that's possible, this however
makes the refactor smaller (!) as it doesn't change everything about the
type for everyone at the same time.
- I have simply removed in cold blood a significant number of old
database migrations. These depended on previous generations of generated
messages of various kinds and were annoying to support in the new
fashion. The oldest supported database version now is the one from
Syncthing 1.9.0 from Sep 7, 2020.
- I changed config structs to be regular manually defined structs.

For the sake of discussion, some things I tried that turned out not to
work...

### Embedding / wrapping

Embedding the protobuf generated structs in our existing types as a data
container and keeping our methods and stuff:

```
package protocol

type FileInfo struct {
  *generated.FileInfo
}
```

This generates a lot of problems because the internal shape of the
generated struct is quite different (different names, different types,
more pointers), because initializing it doesn't work like you'd expect
(i.e., you end up with an embedded nil pointer and a panic), and because
the types of child types don't get wrapped. That is, even if we also
have a similar wrapper around a `Vector`, that's not the type you get
when accessing `someFileInfo.Version`, you get the `*generated.Vector`
that doesn't have methods, etc.

### Aliasing

```
package protocol

type FileInfo = generated.FileInfo
```

Doesn't help because you can't attach methods to it, plus all the above.

### Generating the types into the target package like we do now and
attaching methods

This fails because of the different shape of the generated type (as in
the embedding case above) plus the generated struct already has a bunch
of methods that we can't necessarily override properly (like `String()`
and a bunch of getters).

### Methods to functions

I considered just moving all the methods we attach to functions in a
specific package, so that for example

```
package protocol

func (f FileInfo) Equal(other FileInfo) bool
```

would become

```
package fileinfos

func Equal(a, b *generated.FileInfo) bool
```

and this would mostly work, but becomes quite verbose and cumbersome,
and somewhat limits discoverability (you can't see what methods are
available on the type in auto completions, etc). In the end I did this
in some cases, like in the database layer where a lot of things like
`func (fv *FileVersion) IsEmpty() bool` becomes `func fvIsEmpty(fv
*generated.FileVersion)` because they were anyway just internal methods.

Fixes syncthing#8247
  • Loading branch information
calmh authored Dec 1, 2024
1 parent 2b8ee4c commit 77970d5
Show file tree
Hide file tree
Showing 203 changed files with 7,424 additions and 28,623 deletions.
47 changes: 30 additions & 17 deletions .golangci.yml
Original file line number Diff line number Diff line change
@@ -1,26 +1,39 @@
linters-settings:
maligned:
suggest-new: true

linters:
enable-all: true
disable:
- goimports
- cyclop
- depguard
- lll
- gochecknoinits
- exhaustive
- exhaustruct
- funlen
- gci
- gochecknoglobals
- gochecknoinits
- gocognit
- goconst
- gocyclo
- godox
- gofmt
- goimports
- gomoddirectives
- inamedparam
- interfacebloat
- ireturn
- lll
- maintidx
- nestif
- nonamedreturns
- paralleltest
- protogetter
- scopelint
- gocyclo
- funlen
- tagalign
- tagliatelle
- testpackage
- varnamelen
- wsl
- gocognit
- godox

service:
golangci-lint-version: 1.21.x
prepare:
- rm -f go.sum # 1.12 -> 1.13 issues with QUIC-go
- GO111MODULE=on go mod vendor
- go run build.go assets
issues:
exclude-dirs:
- internal/gen
- cmd/dev
- repos
12 changes: 12 additions & 0 deletions buf.gen.yaml
Original file line number Diff line number Diff line change
@@ -0,0 +1,12 @@
version: v2
managed:
enabled: true
override:
- file_option: go_package_prefix
value: github.com/syncthing/syncthing/internal/gen
plugins:
- remote: buf.build/protocolbuffers/go:v1.35.1
out: .
opt: module=github.com/syncthing/syncthing
inputs:
- directory: proto
10 changes: 10 additions & 0 deletions buf.yaml
Original file line number Diff line number Diff line change
@@ -0,0 +1,10 @@
version: v2
modules:
- path: proto
name: github.com/syncthing/syncthing
lint:
use:
- STANDARD
breaking:
use:
- WIRE_JSON
27 changes: 3 additions & 24 deletions build.go
Original file line number Diff line number Diff line change
Expand Up @@ -925,22 +925,9 @@ func updateDependencies() {
}

func proto() {
pv := protobufVersion()
repo := "https://github.com/gogo/protobuf.git"
path := filepath.Join("repos", "protobuf")

runPrint(goCmd, "install", fmt.Sprintf("github.com/gogo/protobuf/protoc-gen-gogofast@%v", pv))
os.MkdirAll("repos", 0o755)

if _, err := os.Stat(path); err != nil {
runPrint("git", "clone", repo, path)
} else {
runPrintInDir(path, "git", "fetch")
}
runPrintInDir(path, "git", "checkout", pv)

runPrint(goCmd, "generate", "github.com/syncthing/syncthing/cmd/stdiscosrv")
runPrint(goCmd, "generate", "proto/generate.go")
// buf needs to be installed
// https://buf.build/docs/installation/
runPrint("buf", "generate")
}

func testmocks() {
Expand Down Expand Up @@ -1483,14 +1470,6 @@ func (t target) BinaryName() string {
return t.binaryName
}

func protobufVersion() string {
bs, err := runError(goCmd, "list", "-f", "{{.Version}}", "-m", "github.com/gogo/protobuf")
if err != nil {
log.Fatal("Getting protobuf version:", err)
}
return string(bs)
}

func currentAndLatestVersions(n int) ([]string, error) {
bs, err := runError("git", "tag", "--sort", "taggerdate")
if err != nil {
Expand Down
20 changes: 12 additions & 8 deletions cmd/dev/stdisco/main.go
Original file line number Diff line number Diff line change
Expand Up @@ -15,6 +15,9 @@ import (
"strings"
"time"

"google.golang.org/protobuf/proto"

"github.com/syncthing/syncthing/internal/gen/discoproto"
_ "github.com/syncthing/syncthing/lib/automaxprocs"
"github.com/syncthing/syncthing/lib/beacon"
"github.com/syncthing/syncthing/lib/discover"
Expand Down Expand Up @@ -75,32 +78,33 @@ func recv(bc beacon.Interface) {
continue
}

var ann discover.Announce
ann.Unmarshal(data[4:])
var ann discoproto.Announce
proto.Unmarshal(data[4:], &ann)

if ann.ID == myID {
id, _ := protocol.DeviceIDFromBytes(ann.Id)
if id == myID {
// This is one of our own fake packets, don't print it.
continue
}

// Print announcement details for the first packet from a given
// device ID and source address, or if -all was given.
key := ann.ID.String() + src.String()
key := id.String() + src.String()
if all || !seen[key] {
log.Printf("Announcement from %v\n", src)
log.Printf(" %v at %s\n", ann.ID, strings.Join(ann.Addresses, ", "))
log.Printf(" %v at %s\n", id, strings.Join(ann.Addresses, ", "))
seen[key] = true
}
}
}

// sends fake discovery announcements once every second
func send(bc beacon.Interface) {
ann := discover.Announce{
ID: myID,
ann := &discoproto.Announce{
Id: myID[:],
Addresses: []string{"tcp://fake.example.com:12345"},
}
bs, _ := ann.Marshal()
bs, _ := proto.Marshal(ann)

for {
bc.Send(bs)
Expand Down
1 change: 1 addition & 0 deletions cmd/infra/strelaypoolsrv/main.go
Original file line number Diff line number Diff line change
Expand Up @@ -23,6 +23,7 @@ import (
lru "github.com/hashicorp/golang-lru/v2"
"github.com/prometheus/client_golang/prometheus"
"github.com/prometheus/client_golang/prometheus/promhttp"

"github.com/syncthing/syncthing/cmd/infra/strelaypoolsrv/auto"
"github.com/syncthing/syncthing/lib/assets"
_ "github.com/syncthing/syncthing/lib/automaxprocs"
Expand Down
4 changes: 1 addition & 3 deletions cmd/infra/ursrv/serve/serve.go
Original file line number Diff line number Diff line change
Expand Up @@ -17,17 +17,15 @@ import (
"log/slog"
"net"
"net/http"
_ "net/http/pprof"
"os"
"regexp"
"strings"
"time"

_ "net/http/pprof"

"github.com/prometheus/client_golang/prometheus"
"github.com/prometheus/client_golang/prometheus/promhttp"
"github.com/puzpuzpuz/xsync/v3"

"github.com/syncthing/syncthing/lib/build"
"github.com/syncthing/syncthing/lib/geoip"
"github.com/syncthing/syncthing/lib/s3"
Expand Down
28 changes: 16 additions & 12 deletions cmd/stdiscosrv/amqp.go
Original file line number Diff line number Diff line change
Expand Up @@ -13,16 +13,20 @@ import (
"log"

amqp "github.com/rabbitmq/amqp091-go"
"github.com/syncthing/syncthing/lib/protocol"
"github.com/thejerf/suture/v4"
"google.golang.org/protobuf/proto"

"github.com/syncthing/syncthing/internal/gen/discosrv"
"github.com/syncthing/syncthing/internal/protoutil"
"github.com/syncthing/syncthing/lib/protocol"
)

type amqpReplicator struct {
suture.Service
broker string
sender *amqpSender
receiver *amqpReceiver
outbox chan ReplicationRecord
outbox chan *discosrv.ReplicationRecord
}

func newAMQPReplicator(broker, clientID string, db database) *amqpReplicator {
Expand All @@ -31,7 +35,7 @@ func newAMQPReplicator(broker, clientID string, db database) *amqpReplicator {
sender := &amqpSender{
broker: broker,
clientID: clientID,
outbox: make(chan ReplicationRecord, replicationOutboxSize),
outbox: make(chan *discosrv.ReplicationRecord, replicationOutboxSize),
}
svc.Add(sender)

Expand All @@ -47,18 +51,18 @@ func newAMQPReplicator(broker, clientID string, db database) *amqpReplicator {
broker: broker,
sender: sender,
receiver: receiver,
outbox: make(chan ReplicationRecord, replicationOutboxSize),
outbox: make(chan *discosrv.ReplicationRecord, replicationOutboxSize),
}
}

func (s *amqpReplicator) send(key *protocol.DeviceID, ps []DatabaseAddress, seen int64) {
func (s *amqpReplicator) send(key *protocol.DeviceID, ps []*discosrv.DatabaseAddress, seen int64) {
s.sender.send(key, ps, seen)
}

type amqpSender struct {
broker string
clientID string
outbox chan ReplicationRecord
outbox chan *discosrv.ReplicationRecord
}

func (s *amqpSender) Serve(ctx context.Context) error {
Expand All @@ -73,12 +77,12 @@ func (s *amqpSender) Serve(ctx context.Context) error {
for {
select {
case rec := <-s.outbox:
size := rec.Size()
size := proto.Size(rec)
if len(buf) < size {
buf = make([]byte, size)
}

n, err := rec.MarshalTo(buf)
n, err := protoutil.MarshalTo(buf, rec)
if err != nil {
replicationSendsTotal.WithLabelValues("error").Inc()
return fmt.Errorf("replication marshal: %w", err)
Expand Down Expand Up @@ -111,8 +115,8 @@ func (s *amqpSender) String() string {
return fmt.Sprintf("amqpSender(%q)", s.broker)
}

func (s *amqpSender) send(key *protocol.DeviceID, ps []DatabaseAddress, seen int64) {
item := ReplicationRecord{
func (s *amqpSender) send(key *protocol.DeviceID, ps []*discosrv.DatabaseAddress, seen int64) {
item := &discosrv.ReplicationRecord{
Key: key[:],
Addresses: ps,
Seen: seen,
Expand Down Expand Up @@ -158,8 +162,8 @@ func (s *amqpReceiver) Serve(ctx context.Context) error {
continue
}

var rec ReplicationRecord
if err := rec.Unmarshal(msg.Body); err != nil {
var rec discosrv.ReplicationRecord
if err := proto.Unmarshal(msg.Body, &rec); err != nil {
replicationRecvsTotal.WithLabelValues("error").Inc()
return fmt.Errorf("replication unmarshal: %w", err)
}
Expand Down
7 changes: 4 additions & 3 deletions cmd/stdiscosrv/apisrv.go
Original file line number Diff line number Diff line change
Expand Up @@ -28,6 +28,7 @@ import (
"sync"
"time"

"github.com/syncthing/syncthing/internal/gen/discosrv"
"github.com/syncthing/syncthing/lib/protocol"
"github.com/syncthing/syncthing/lib/stringutil"
)
Expand All @@ -52,7 +53,7 @@ type apiSrv struct {
}

type replicator interface {
send(key *protocol.DeviceID, addrs []DatabaseAddress, seen int64)
send(key *protocol.DeviceID, addrs []*discosrv.DatabaseAddress, seen int64)
}

type requestID int64
Expand Down Expand Up @@ -312,7 +313,7 @@ func (s *apiSrv) handleAnnounce(deviceID protocol.DeviceID, addresses []string)
slices.Sort(addresses)
addresses = slices.Compact(addresses)

dbAddrs := make([]DatabaseAddress, len(addresses))
dbAddrs := make([]*discosrv.DatabaseAddress, len(addresses))
for i := range addresses {
dbAddrs[i].Address = addresses[i]
dbAddrs[i].Expires = expire
Expand Down Expand Up @@ -511,7 +512,7 @@ func (lrw *loggingResponseWriter) WriteHeader(code int) {
lrw.ResponseWriter.WriteHeader(code)
}

func addressStrs(dbAddrs []DatabaseAddress) []string {
func addressStrs(dbAddrs []*discosrv.DatabaseAddress) []string {
res := make([]string, len(dbAddrs))
for i, a := range dbAddrs {
res[i] = a.Address
Expand Down
Loading

0 comments on commit 77970d5

Please sign in to comment.