Before v1.1.18, the following code would not print any message, i.e. dlp.DecodeLayers returns a nil error:
import (
"fmt"
"github.com/google/gopacket"
"github.com/google/gopacket/layers"
)
func main() {
dlp := gopacket.NewDecodingLayerParser(layer.LayerTypeEthernet, &layers.Ethernet{}, &layers.IPv4{}, &layers.TCP{}, &layersUDP{})
decoded := []gopacket.LayerType{}
empty := []byte{}
err := dlp.DecodeLayers(empty, &decoded)
if err != nil {
fmt.Printf("Error from dlp parser: %v\n", err)
}
}In v1.1.18 this behavior changed and an error is now returned from the first layer, i.e. layers.Ethernet in this case:
Error from dlp parser: Ethernet packet too small
This seems to be due to 04f6565 (PR #660), namely the following change:
@@ -153,23 +304,15 @@ func (l *DecodingLayerParser) DecodeLayers(data []byte, decoded *[]LayerType) (e
if !l.IgnorePanic {
defer panicToError(&err)
}
- typ := l.first
- *decoded = (*decoded)[:0] // Truncated decoded layers.
- for len(data) > 0 {
- decoder, ok := l.decoders[typ]
- if !ok {
- if l.IgnoreUnsupported {
- return nil
- }
- return UnsupportedLayerType(typ)
- } else if err = decoder.DecodeFromBytes(data, l.df); err != nil {
- return err
+ typ, err := l.decodeFunc(data, decoded)
+ if typ != LayerTypeZero {
+ // no decoder
+ if l.IgnoreUnsupported {
+ return nil
}
- *decoded = append(*decoded, typ)
- typ = decoder.NextLayerType()
- data = decoder.LayerPayload()
+ return UnsupportedLayerType(typ)
}
- return nil
+ return err
}We have code that depends on this behavior and strictly speaking this is an API breakage, so my suggestion would be to keep returning nil for empty data. The other option of course is to check len(data) != 0 before calling (*DecodingLayerParser).DecodeLayers in our but this would be rather cumbersome.
/cc @notti @yerden
Before v1.1.18, the following code would not print any message, i.e.
dlp.DecodeLayersreturns anilerror:In v1.1.18 this behavior changed and an error is now returned from the first layer, i.e.
layers.Ethernetin this case:This seems to be due to 04f6565 (PR #660), namely the following change:
We have code that depends on this behavior and strictly speaking this is an API breakage, so my suggestion would be to keep returning
nilfor emptydata. The other option of course is to checklen(data) != 0before calling(*DecodingLayerParser).DecodeLayersin our but this would be rather cumbersome./cc @notti @yerden