Socket ether linux step2

[devnexen]

No description provided.

[arnaud-lb]

This should use a class, e.g. TcpPacket

[arnaud-lb]

This belongs to the IP layer, and should be in a separate object:

final readonly class SocketEthernetInfo // or EthernetPacket ?
{
    public Socket $socket;
    public int    $protocol;
    public string $macsrc;
    public string $macdst;

    public Ipv4Packet|Ipv6Packet|null $payload;

    public int $headersize;
    public string $rawpacket;
}

final readonly class Ipv4Packet
{
    public string $srcaddr;
    public string $dscaddr;
    // ... (other fields)

    public ?object $payload;

    public int $headersize;
    public string $rawpacket;
}

final readonly class TcpPacket
{
    public string $srcport;
    public string $dstport;
    // ...

    public int $headersize;
    public string $rawpacket;
}

[arnaud-lb]

NB: We should probably namespace those

[arnaud-lb]

Suggested change

	payload = ((unsigned char *)e + sizeof(struct ethhdr));
	payload = ((unsigned char *)e + ETH_HLEN);

[arnaud-lb]

Suggested change

	payload = ((unsigned char *)e + sizeof(struct ethhdr));
	payload = ((unsigned char *)e + ETH_HLEN);

[arnaud-lb]

We should not trust ip->ihl

[arnaud-lb]

We could expose the payload here, like we do for ETH_P_IP

[arnaud-lb]

This can be a SocketEthernetInfo

[arnaud-lb]

I suggested this, but I think that rawpayload is not the right name after all. The payload field represents the contents of the eth packet, but the rawpayload field represents the eth packet itself. rawpacket would be a better name than rawpayload.

[bukka]

This has a bit of class design so it should be probably at least posted to internals (it's not like adding a constant or a simple funciton). I agree that there should be some namespacing as well possibly. Definitely not the current names that will easily result in BC break for appliations.