rxrpc: Don't use a ring buffer for call Tx queue

Change the way the Tx queueing works to make the following ends easier to
achieve:

 (1) The filling of packets, the encryption of packets and the transmission
     of packets can be handled in parallel by separate threads, rather than
     rxrpc_sendmsg() allocating, filling, encrypting and transmitting each
     packet before moving onto the next one.

 (2) Get rid of the fixed-size ring which sets a hard limit on the number
     of packets that can be retained in the ring.  This allows the number
     of packets to increase without having to allocate a very large ring or
     having variable-sized rings.

     [Note: the downside of this is that it's then less efficient to locate
     a packet for retransmission as we then have to step through a list and
     examine each buffer in the list.]

 (3) Allow the filler/encrypter to run ahead of the transmission window.

 (4) Make it easier to do zero copy UDP from the packet buffers.

 (5) Make it easier to do zero copy from userspace to the packet buffers -
     and thence to UDP (only if for unauthenticated connections).

To that end, the following changes are made:

 (1) Use the new rxrpc_txbuf struct instead of sk_buff for keeping packets
     to be transmitted in.  This allows them to be placed on multiple
     queues simultaneously.  An sk_buff isn't really necessary as it's
     never passed on to lower-level networking code.

 (2) Keep the transmissable packets in a linked list on the call struct
     rather than in a ring.  As a consequence, the annotation buffer isn't
     used either; rather a flag is set on the packet to indicate ackedness.

 (3) Use the RXRPC_CALL_TX_LAST flag to indicate that the last packet to be
     transmitted has been queued.  Add RXRPC_CALL_TX_ALL_ACKED to indicate
     that all packets up to and including the last got hard acked.

 (4) Wire headers are now stored in the txbuf rather than being concocted
     on the stack and they're stored immediately before the data, thereby
     allowing zerocopy of a single span.

 (5) Don't bother with instant-resend on transmission failure; rather,
     leave it for a timer or an ACK packet to trigger.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org

1 files changed, 37 insertions, 65 deletions

diff --git a/net/rxrpc/rxkad.c b/net/rxrpc/rxkad.c
index d87c99b36e01..8fc055587f0e 100644
--- a/net/rxrpc/rxkad.c
+++ b/net/rxrpc/rxkad.c
@@ -259,11 +259,10 @@ static void rxkad_free_call_crypto(struct rxrpc_call *call)
  * partially encrypt a packet (level 1 security)
  */
 static int rxkad_secure_packet_auth(const struct rxrpc_call *call,
-				    struct sk_buff *skb, u32 data_size,
+				    struct rxrpc_txbuf *txb,
 				    struct skcipher_request *req)
 {
-	struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
-	struct rxkad_level1_hdr hdr;
+	struct rxkad_level1_hdr *hdr = (void *)txb->data;
 	struct rxrpc_crypt iv;
 	struct scatterlist sg;
 	size_t pad;
@@ -271,22 +270,22 @@ static int rxkad_secure_packet_auth(const struct rxrpc_call *call,
 
 	_enter("");
 
-	check = sp->hdr.seq ^ call->call_id;
-	data_size |= (u32)check << 16;
-
-	hdr.data_size = htonl(data_size);
-	memcpy(skb->head, &hdr, sizeof(hdr));
+	check = txb->seq ^ ntohl(txb->wire.callNumber);
+	hdr->data_size = htonl((u32)check << 16 | txb->len);
 
-	pad = sizeof(struct rxkad_level1_hdr) + data_size;
+	txb->len += sizeof(struct rxkad_level1_hdr);
+	pad = txb->len;
 	pad = RXKAD_ALIGN - pad;
 	pad &= RXKAD_ALIGN - 1;
-	if (pad)
-		skb_put_zero(skb, pad);
+	if (pad) {
+		memset(txb->data + txb->offset, 0, pad);
+		txb->len += pad;
+	}
 
 	/* start the encryption afresh */
 	memset(&iv, 0, sizeof(iv));
 
-	sg_init_one(&sg, skb->head, 8);
+	sg_init_one(&sg, txb->data, 8);
 	skcipher_request_set_sync_tfm(req, call->conn->rxkad.cipher);
 	skcipher_request_set_callback(req, 0, NULL, NULL);
 	skcipher_request_set_crypt(req, &sg, &sg, 8, iv.x);
@@ -301,87 +300,60 @@ static int rxkad_secure_packet_auth(const struct rxrpc_call *call,
  * wholly encrypt a packet (level 2 security)
  */
 static int rxkad_secure_packet_encrypt(const struct rxrpc_call *call,
-				       struct sk_buff *skb,
-				       u32 data_size,
+				       struct rxrpc_txbuf *txb,
 				       struct skcipher_request *req)
 {
 	const struct rxrpc_key_token *token;
-	struct rxkad_level2_hdr rxkhdr;
-	struct rxrpc_skb_priv *sp;
+	struct rxkad_level2_hdr *rxkhdr = (void *)txb->data;
 	struct rxrpc_crypt iv;
-	struct scatterlist sg[16];
-	unsigned int len;
+	struct scatterlist sg;
 	size_t pad;
 	u16 check;
-	int err;
-
-	sp = rxrpc_skb(skb);
+	int ret;
 
 	_enter("");
 
-	check = sp->hdr.seq ^ call->call_id;
+	check = txb->seq ^ ntohl(txb->wire.callNumber);
 
-	rxkhdr.data_size = htonl(data_size | (u32)check << 16);
-	rxkhdr.checksum = 0;
-	memcpy(skb->head, &rxkhdr, sizeof(rxkhdr));
+	rxkhdr->data_size = htonl(txb->len | (u32)check << 16);
+	rxkhdr->checksum = 0;
 
-	pad = sizeof(struct rxkad_level2_hdr) + data_size;
+	txb->len += sizeof(struct rxkad_level2_hdr);
+	pad = txb->len;
 	pad = RXKAD_ALIGN - pad;
 	pad &= RXKAD_ALIGN - 1;
-	if (pad)
-		skb_put_zero(skb, pad);
+	if (pad) {
+		memset(txb->data + txb->offset, 0, pad);
+		txb->len += pad;
+	}
 
 	/* encrypt from the session key */
 	token = call->conn->params.key->payload.data[0];
 	memcpy(&iv, token->kad->session_key, sizeof(iv));
 
-	sg_init_one(&sg[0], skb->head, sizeof(rxkhdr));
+	sg_init_one(&sg, txb->data, txb->len);
 	skcipher_request_set_sync_tfm(req, call->conn->rxkad.cipher);
 	skcipher_request_set_callback(req, 0, NULL, NULL);
-	skcipher_request_set_crypt(req, &sg[0], &sg[0], sizeof(rxkhdr), iv.x);
-	crypto_skcipher_encrypt(req);
-
-	/* we want to encrypt the skbuff in-place */
-	err = -EMSGSIZE;
-	if (skb_shinfo(skb)->nr_frags > 16)
-		goto out;
-
-	len = round_up(data_size, RXKAD_ALIGN);
-
-	sg_init_table(sg, ARRAY_SIZE(sg));
-	err = skb_to_sgvec(skb, sg, 8, len);
-	if (unlikely(err < 0))
-		goto out;
-	skcipher_request_set_crypt(req, sg, sg, len, iv.x);
-	crypto_skcipher_encrypt(req);
-
-	_leave(" = 0");
-	err = 0;
-
-out:
+	skcipher_request_set_crypt(req, &sg, &sg, txb->len, iv.x);
+	ret = crypto_skcipher_encrypt(req);
 	skcipher_request_zero(req);
-	return err;
+	return ret;
 }
 
 /*
  * checksum an RxRPC packet header
  */
-static int rxkad_secure_packet(struct rxrpc_call *call,
-			       struct sk_buff *skb,
-			       size_t data_size)
+static int rxkad_secure_packet(struct rxrpc_call *call, struct rxrpc_txbuf *txb)
 {
-	struct rxrpc_skb_priv *sp;
 	struct skcipher_request	*req;
 	struct rxrpc_crypt iv;
 	struct scatterlist sg;
 	u32 x, y;
 	int ret;
 
-	sp = rxrpc_skb(skb);
-
-	_enter("{%d{%x}},{#%u},%zu,",
+	_enter("{%d{%x}},{#%u},%u,",
 	       call->debug_id, key_serial(call->conn->params.key),
-	       sp->hdr.seq, data_size);
+	       txb->seq, txb->len);
 
 	if (!call->conn->rxkad.cipher)
 		return 0;
@@ -398,9 +370,9 @@ static int rxkad_secure_packet(struct rxrpc_call *call,
 	memcpy(&iv, call->conn->rxkad.csum_iv.x, sizeof(iv));
 
 	/* calculate the security checksum */
-	x = (call->cid & RXRPC_CHANNELMASK) << (32 - RXRPC_CIDSHIFT);
-	x |= sp->hdr.seq & 0x3fffffff;
-	call->crypto_buf[0] = htonl(call->call_id);
+	x = (ntohl(txb->wire.cid) & RXRPC_CHANNELMASK) << (32 - RXRPC_CIDSHIFT);
+	x |= txb->seq & 0x3fffffff;
+	call->crypto_buf[0] = txb->wire.callNumber;
 	call->crypto_buf[1] = htonl(x);
 
 	sg_init_one(&sg, call->crypto_buf, 8);
@@ -414,17 +386,17 @@ static int rxkad_secure_packet(struct rxrpc_call *call,
 	y = (y >> 16) & 0xffff;
 	if (y == 0)
 		y = 1; /* zero checksums are not permitted */
-	sp->hdr.cksum = y;
+	txb->wire.cksum = htons(y);
 
 	switch (call->conn->params.security_level) {
 	case RXRPC_SECURITY_PLAIN:
 		ret = 0;
 		break;
 	case RXRPC_SECURITY_AUTH:
-		ret = rxkad_secure_packet_auth(call, skb, data_size, req);
+		ret = rxkad_secure_packet_auth(call, txb, req);
 		break;
 	case RXRPC_SECURITY_ENCRYPT:
-		ret = rxkad_secure_packet_encrypt(call, skb, data_size, req);
+		ret = rxkad_secure_packet_encrypt(call, txb, req);
 		break;
 	default:
 		ret = -EPERM;