-----Original Message----- From: David Miller [mailto:davem@xxxxxxxxxxxxx] Sent: Tuesday, February 13, 2018 1:19 AM To: Atul Gupta <atul.gupta@xxxxxxxxxxx> Cc: davejwatson@xxxxxx; herbert@xxxxxxxxxxxxxxxxxxx; sd@xxxxxxxxxxxxxxx; linux-crypto@xxxxxxxxxxxxxxx; netdev@xxxxxxxxxxxxxxx; Ganesh GR <ganeshgr@xxxxxxxxxxx> Subject: Re: [Crypto v4 03/12] support for inline tls From: Atul Gupta <atul.gupta@xxxxxxxxxxx> Date: Mon, 12 Feb 2018 17:34:28 +0530 > +static int get_tls_prot(struct sock *sk) { > + struct tls_context *ctx = tls_get_ctx(sk); > + struct net_device *netdev; > + struct tls_device *dev; > + > + /* Device bound to specific IP */ > + if (inet_sk(sk)->inet_rcv_saddr) { > + netdev = find_netdev(sk); > + if (!netdev) > + goto out; > + > + /* Device supports Inline record processing */ > + if (!(netdev->features & NETIF_F_HW_TLS_INLINE)) > + goto out; > + > + mutex_lock(&device_mutex); > + list_for_each_entry(dev, &device_list, dev_list) { > + if (dev->netdev && dev->netdev(dev, netdev)) > + break; > + } > + mutex_unlock(&device_mutex); > + > + ctx->tx_conf = TLS_FULL_HW; > + if (dev->prot) > + dev->prot(dev, sk); What if the same IP address is configured on multiple interfaces? Thanks, I overlooked this point. The checks above were based on the premise that device chosen is indeed the one with Inline TLS enabled, net_device corresponding to specific IP address, feature configured for device from ethtool and net_device corresponds to Inline TLS driver registered with net tls. Case with same IP configured on multiple interface looks similar to INADDR_ANY below. The TLS_FULL_HW and modified hash routines handles devices with/without Inline TLS support. The first Inline TLS capable device updates sk_prot for TLS_FULL_HW. The tls_hw_hash listens on all interfaces and process device specific routine, the listen however succeeds for device on which connect is initiated and may not have the Inline TLS capability, such connection continues in TLS_BASE_TX or non-tls-offload mode. On the other hand, if Inline TLS capable device were to establish connection it updates the prot as required for offload mode to continue. > + } else { /* src address not known or INADDR_ANY */ > + mutex_lock(&device_mutex); > + list_for_each_entry(dev, &device_list, dev_list) { > + if (dev->feature && dev->feature(dev)) { > + ctx->tx_conf = TLS_FULL_HW; > + break; > + } > + } > + mutex_unlock(&device_mutex); > + update_sk_prot(sk, ctx); And I think this is even more of a stretch. Just because you find an inline TLS device on the global list doesn't mean traffic will necessarily flow through it once the connection is fully established and therefore be able to provide inline TLS offloading.
