[net-next,v2,11/39] rxrpc: Prepare to be able to send jumbo DATA packets

Message ID	20241204074710.990092-12-dhowells@redhat.com (mailing list archive)
State	New
Delegated to:	Netdev Maintainers
Headers	show Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.133.124]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 1BF601990C3 for <netdev@vger.kernel.org>; Wed, 4 Dec 2024 07:48:09 +0000 (UTC) From: David Howells <dhowells@redhat.com> To: netdev@vger.kernel.org Cc: David Howells <dhowells@redhat.com>, Marc Dionne <marc.dionne@auristor.com>, Yunsheng Lin <linyunsheng@huawei.com>, "David S. Miller" <davem@davemloft.net>, Eric Dumazet <edumazet@google.com>, Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>, linux-afs@lists.infradead.org, linux-kernel@vger.kernel.org Subject: [PATCH net-next v2 11/39] rxrpc: Prepare to be able to send jumbo DATA packets Date: Wed, 4 Dec 2024 07:46:39 +0000 Message-ID: <20241204074710.990092-12-dhowells@redhat.com> In-Reply-To: <20241204074710.990092-1-dhowells@redhat.com> References: <20241204074710.990092-1-dhowells@redhat.com> Precedence: bulk MIME-Version: 1.0 Content-Transfer-Encoding: 8bit
Series	rxrpc: Implement jumbo DATA transmission and RACK-TLP \| expand [net-next,v2,00/39] rxrpc: Implement jumbo DATA transmission and RACK-TLP [net-next,v2,01/39] ktime: Add us_to_ktime() [net-next,v2,02/39] rxrpc: Fix handling of received connection abort [net-next,v2,03/39] rxrpc: Use umin() and umax() rather than min_t()/max_t() where possible [net-next,v2,04/39] rxrpc: Clean up Tx header flags generation handling [net-next,v2,05/39] rxrpc: Don't set the MORE-PACKETS rxrpc wire header flag [net-next,v2,06/39] rxrpc: Show stats counter for received reason-0 ACKs [net-next,v2,07/39] rxrpc: Request an ACK on impending Tx stall [net-next,v2,08/39] rxrpc: Use a large kvec[] in rxrpc_local rather than every rxrpc_txbuf [net-next,v2,09/39] rxrpc: Implement path-MTU probing using padded PING ACKs (RFC8899) [net-next,v2,10/39] rxrpc: Separate the packet length from the data length in rxrpc_txbuf [net-next,v2,11/39] rxrpc: Prepare to be able to send jumbo DATA packets [net-next,v2,12/39] rxrpc: Add a tracepoint to show variables pertinent to jumbo packet size [net-next,v2,13/39] rxrpc: Fix CPU time starvation in I/O thread [net-next,v2,14/39] rxrpc: Fix injection of packet loss [net-next,v2,15/39] rxrpc: Only set DF=1 on initial DATA transmission [net-next,v2,16/39] rxrpc: Timestamp DATA packets before transmitting them [net-next,v2,17/39] rxrpc: Don't need barrier for ->tx_bottom and ->acks_hard_ack [net-next,v2,18/39] rxrpc: Implement progressive transmission queue struct [net-next,v2,19/39] rxrpc: call->acks_hard_ack is now the same call->tx_bottom, so remove it [net-next,v2,20/39] rxrpc: Replace call->acks_first_seq with tracking of the hard ACK point [net-next,v2,21/39] rxrpc: Display stats about jumbo packets transmitted and received [net-next,v2,22/39] rxrpc: Adjust names and types of congestion-related fields [net-next,v2,23/39] rxrpc: Use the new rxrpc_tx_queue struct to more efficiently process ACKs [net-next,v2,24/39] rxrpc: Store the DATA serial in the txqueue and use this in RTT calc [net-next,v2,25/39] rxrpc: Don't use received skbuff timestamps [net-next,v2,26/39] rxrpc: Generate rtt_min [net-next,v2,27/39] rxrpc: Adjust the rxrpc_rtt_rx tracepoint [net-next,v2,28/39] rxrpc: Display userStatus in rxrpc_rx_ack trace [net-next,v2,29/39] rxrpc: Fix the calculation and use of RTO [net-next,v2,30/39] rxrpc: Fix initial resend timeout [net-next,v2,31/39] rxrpc: Send jumbo DATA packets [net-next,v2,32/39] rxrpc: Don't allocate a txbuf for an ACK transmission [net-next,v2,33/39] rxrpc: Use irq-disabling spinlocks between app and I/O thread [net-next,v2,34/39] rxrpc: Tidy up the ACK parsing a bit [net-next,v2,35/39] rxrpc: Add a reason indicator to the tx_data tracepoint [net-next,v2,36/39] rxrpc: Add a reason indicator to the tx_ack tracepoint [net-next,v2,37/39] rxrpc: Manage RTT per-call rather than per-peer [net-next,v2,38/39] rxrpc: Fix request for an ACK when cwnd is minimum [net-next,v2,39/39] rxrpc: Implement RACK/TLP to deal with transmission stalls [RFC8985]

Context	Check	Description
netdev/series_format	fail	Series longer than 15 patches
netdev/tree_selection	success	Clearly marked for net-next
netdev/ynl	success	Generated files up to date; no warnings/errors; no diff in generated;
netdev/fixes_present	success	Fixes tag not required for -next series
netdev/header_inline	success	No static functions without inline keyword in header files
netdev/build_32bit	success	Errors and warnings before: 3 this patch: 3
netdev/build_tools	success	No tools touched, skip
netdev/cc_maintainers	warning	1 maintainers not CCed: horms@kernel.org
netdev/build_clang	fail	Errors and warnings before: 3 this patch: 5
netdev/verify_signedoff	success	Signed-off-by tag matches author and committer
netdev/deprecated_api	success	None detected
netdev/check_selftest	success	No net selftest shell script
netdev/verify_fixes	success	No Fixes tag
netdev/build_allmodconfig_warn	success	Errors and warnings before: 307 this patch: 307
netdev/checkpatch	warning	WARNING: line length of 100 exceeds 80 columns WARNING: line length of 81 exceeds 80 columns WARNING: line length of 82 exceeds 80 columns WARNING: line length of 89 exceeds 80 columns WARNING: line length of 90 exceeds 80 columns WARNING: line length of 94 exceeds 80 columns WARNING: line length of 97 exceeds 80 columns
netdev/build_clang_rust	success	No Rust files in patch. Skipping build
netdev/kdoc	success	Errors and warnings before: 0 this patch: 0
netdev/source_inline	success	Was 0 now: 0

diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index a5c0bc917641..4386b2e6cca5 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -832,6 +832,7 @@ struct rxrpc_txbuf { __be16 cksum; /* Checksum to go in header */ unsigned short ack_rwind; /* ACK receive window */ u8 /*enum rxrpc_propose_ack_trace*/ ack_why; /* If ack, why */ + bool jumboable; /* Can be non-terminal jumbo subpacket */ u8 nr_kvec; /* Amount of kvec[] used */ struct kvec kvec[3]; }; @@ -862,6 +863,21 @@ static inline rxrpc_serial_t rxrpc_get_next_serial(struct rxrpc_connection *conn return serial; } +/* + * Allocate the next serial n numbers on a connection. 0 must be skipped. + */ +static inline rxrpc_serial_t rxrpc_get_next_serials(struct rxrpc_connection *conn, + unsigned int n) +{ + rxrpc_serial_t serial; + + serial = conn->tx_serial; + if (serial + n <= n) + serial = 1; + conn->tx_serial = serial + n; + return serial; +} + /* * af_rxrpc.c */ @@ -1176,7 +1192,7 @@ int rxrpc_send_abort_packet(struct rxrpc_call *); void rxrpc_send_conn_abort(struct rxrpc_connection *conn); void rxrpc_reject_packet(struct rxrpc_local *local, struct sk_buff *skb); void rxrpc_send_keepalive(struct rxrpc_peer *); -void rxrpc_transmit_one(struct rxrpc_call *call, struct rxrpc_txbuf *txb); +void rxrpc_transmit_data(struct rxrpc_call *call, struct rxrpc_txbuf *txb, int n); /* * peer_event.c diff --git a/net/rxrpc/call_event.c b/net/rxrpc/call_event.c index 1d889b6f0366..3379adfaaf65 100644 --- a/net/rxrpc/call_event.c +++ b/net/rxrpc/call_event.c @@ -124,7 +124,7 @@ void rxrpc_resend(struct rxrpc_call *call, struct sk_buff *ack_skb) ktime_sub(resend_at, now)); txb->flags |= RXRPC_TXBUF_RESENT; - rxrpc_transmit_one(call, txb); + rxrpc_transmit_data(call, txb, 1); did_send = true; now = ktime_get_real(); @@ -164,7 +164,7 @@ void rxrpc_resend(struct rxrpc_call *call, struct sk_buff *ack_skb) unacked = true; txb->flags |= RXRPC_TXBUF_RESENT; - rxrpc_transmit_one(call, txb); + rxrpc_transmit_data(call, txb, 1); did_send = true; rxrpc_inc_stat(call->rxnet, stat_tx_data_retrans); now = ktime_get_real(); @@ -231,15 +231,12 @@ static void rxrpc_close_tx_phase(struct rxrpc_call *call) } } -static bool rxrpc_tx_window_has_space(struct rxrpc_call *call) +static unsigned int rxrpc_tx_window_space(struct rxrpc_call *call) { - unsigned int winsize = umin(call->tx_winsize, call->cong_cwnd + call->cong_extra); - rxrpc_seq_t window = call->acks_hard_ack, wtop = window + winsize; - rxrpc_seq_t tx_top = call->tx_top; - int space; + int winsize = umin(call->tx_winsize, call->cong_cwnd + call->cong_extra); + int in_flight = call->tx_top - call->acks_hard_ack; - space = wtop - tx_top; - return space > 0; + return max(winsize - in_flight, 0); } /* @@ -247,7 +244,7 @@ static bool rxrpc_tx_window_has_space(struct rxrpc_call *call) */ static void rxrpc_decant_prepared_tx(struct rxrpc_call *call) { - struct rxrpc_txbuf *txb; + int space = rxrpc_tx_window_space(call); if (!test_bit(RXRPC_CALL_EXPOSED, &call->flags)) { if (list_empty(&call->tx_sendmsg)) @@ -255,22 +252,33 @@ static void rxrpc_decant_prepared_tx(struct rxrpc_call *call) rxrpc_expose_client_call(call); } - while ((txb = list_first_entry_or_null(&call->tx_sendmsg, - struct rxrpc_txbuf, call_link))) { + while (space > 0) { + struct rxrpc_txbuf *head = NULL, *txb; + int count = 0, limit = min(space, 1); + + if (list_empty(&call->tx_sendmsg)) + break; + spin_lock(&call->tx_lock); - list_del(&txb->call_link); + do { + txb = list_first_entry(&call->tx_sendmsg, + struct rxrpc_txbuf, call_link); + if (!head) + head = txb; + list_move_tail(&txb->call_link, &call->tx_buffer); + count++; + if (!txb->jumboable) + break; + } while (count < limit && !list_empty(&call->tx_sendmsg)); + spin_unlock(&call->tx_lock); call->tx_top = txb->seq; - list_add_tail(&txb->call_link, &call->tx_buffer); - if (txb->flags & RXRPC_LAST_PACKET) rxrpc_close_tx_phase(call); - rxrpc_transmit_one(call, txb); - - if (!rxrpc_tx_window_has_space(call)) - break; + space -= count; + rxrpc_transmit_data(call, head, count); } } @@ -285,7 +293,7 @@ static void rxrpc_transmit_some_data(struct rxrpc_call *call) case RXRPC_CALL_SERVER_SEND_REPLY: case RXRPC_CALL_CLIENT_SEND_REQUEST: - if (!rxrpc_tx_window_has_space(call)) + if (!rxrpc_tx_window_space(call)) return; if (list_empty(&call->tx_sendmsg)) { rxrpc_inc_stat(call->rxnet, stat_tx_data_underflow); diff --git a/net/rxrpc/input.c b/net/rxrpc/input.c index fd08d813ef29..8398fa10ee8d 100644 --- a/net/rxrpc/input.c +++ b/net/rxrpc/input.c @@ -693,9 +693,12 @@ static void rxrpc_input_ack_trailer(struct rxrpc_call *call, struct sk_buff *skb { struct rxrpc_skb_priv *sp = rxrpc_skb(skb); struct rxrpc_peer *peer = call->peer; - unsigned int max_data; + unsigned int max_data, capacity; bool wake = false; - u32 rwind = ntohl(trailer->rwind); + u32 max_mtu = ntohl(trailer->maxMTU); + //u32 if_mtu = ntohl(trailer->ifMTU); + u32 rwind = ntohl(trailer->rwind); + u32 jumbo_max = ntohl(trailer->jumbo_max); if (rwind > RXRPC_TX_MAX_WINDOW) rwind = RXRPC_TX_MAX_WINDOW; @@ -706,24 +709,29 @@ static void rxrpc_input_ack_trailer(struct rxrpc_call *call, struct sk_buff *skb call->tx_winsize = rwind; } - if (trailer->jumbo_max == 0) { - /* The peer says it supports pmtu discovery */ - peer->ackr_adv_pmtud = true; - } else { - peer->ackr_adv_pmtud = false; - } - - max_data = ntohl(trailer->maxMTU); - peer->ackr_max_data = max_data; + max_mtu = clamp(max_mtu, 500, 65535); + peer->ackr_max_data = max_mtu; - if (max_data < peer->max_data) { - trace_rxrpc_pmtud_reduce(peer, sp->hdr.serial, max_data, + if (max_mtu < peer->max_data) { + trace_rxrpc_pmtud_reduce(peer, sp->hdr.serial, max_mtu, rxrpc_pmtud_reduce_ack); write_seqcount_begin(&peer->mtu_lock); - peer->max_data = max_data; + peer->max_data = max_mtu; write_seqcount_end(&peer->mtu_lock); } + max_data = umin(max_mtu, peer->max_data); + capacity = max_data; + capacity += sizeof(struct rxrpc_jumbo_header); /* First subpacket has main hdr, not jumbo */ + capacity /= sizeof(struct rxrpc_jumbo_header) + RXRPC_JUMBO_DATALEN; + + if (jumbo_max == 0) { + /* The peer says it supports pmtu discovery */ + peer->ackr_adv_pmtud = true; + } else { + peer->ackr_adv_pmtud = false; + } + if (wake) wake_up(&call->waitq); } diff --git a/net/rxrpc/insecure.c b/net/rxrpc/insecure.c index d665f486be5f..e068f9b79d02 100644 --- a/net/rxrpc/insecure.c +++ b/net/rxrpc/insecure.c @@ -25,6 +25,8 @@ static struct rxrpc_txbuf *none_alloc_txbuf(struct rxrpc_call *call, size_t rema static int none_secure_packet(struct rxrpc_call *call, struct rxrpc_txbuf *txb) { txb->pkt_len = txb->len; + if (txb->len == RXRPC_JUMBO_DATALEN) + txb->jumboable = true; return 0; } diff --git a/net/rxrpc/output.c b/net/rxrpc/output.c index df9af4ad4260..aededdd474d7 100644 --- a/net/rxrpc/output.c +++ b/net/rxrpc/output.c @@ -377,9 +377,10 @@ int rxrpc_send_abort_packet(struct rxrpc_call *call) */ static size_t rxrpc_prepare_data_subpacket(struct rxrpc_call *call, struct rxrpc_txbuf *txb, rxrpc_serial_t serial, - int subpkt) + int subpkt, int nr_subpkts) { struct rxrpc_wire_header *whdr = txb->kvec[0].iov_base; + struct rxrpc_jumbo_header *jumbo = (void *)(whdr + 1) - sizeof(*jumbo); enum rxrpc_req_ack_trace why; struct rxrpc_connection *conn = call->conn; struct kvec *kv = &call->local->kvec[subpkt]; @@ -399,6 +400,11 @@ static size_t rxrpc_prepare_data_subpacket(struct rxrpc_call *call, struct rxrpc flags = txb->flags & RXRPC_TXBUF_WIRE_FLAGS; last = txb->flags & RXRPC_LAST_PACKET; + if (subpkt < nr_subpkts - 1) { + len = RXRPC_JUMBO_DATALEN; + goto dont_set_request_ack; + } + more = (!list_is_last(&txb->call_link, &call->tx_buffer) || !list_empty(&call->tx_sendmsg)); @@ -436,13 +442,25 @@ static size_t rxrpc_prepare_data_subpacket(struct rxrpc_call *call, struct rxrpc } dont_set_request_ack: - whdr->flags = flags; - whdr->serial = htonl(txb->serial); - whdr->cksum = txb->cksum; - whdr->serviceId = htons(conn->service_id); - kv->iov_base = whdr; - len += sizeof(*whdr); - // TODO: Convert into a jumbo header for tail subpackets + /* The jumbo header overlays the wire header in the txbuf. */ + if (subpkt < nr_subpkts - 1) + flags |= RXRPC_JUMBO_PACKET; + else + flags &= ~RXRPC_JUMBO_PACKET; + if (subpkt == 0) { + whdr->flags = flags; + whdr->serial = htonl(txb->serial); + whdr->cksum = txb->cksum; + whdr->serviceId = htons(conn->service_id); + kv->iov_base = whdr; + len += sizeof(*whdr); + } else { + jumbo->flags = flags; + jumbo->pad = 0; + jumbo->cksum = txb->cksum; + kv->iov_base = jumbo; + len += sizeof(*jumbo); + } trace_rxrpc_tx_data(call, txb->seq, txb->serial, flags, false); kv->iov_len = len; @@ -450,18 +468,22 @@ static size_t rxrpc_prepare_data_subpacket(struct rxrpc_call *call, struct rxrpc } /* - * Prepare a packet for transmission. + * Prepare a (jumbo) packet for transmission. */ -static size_t rxrpc_prepare_data_packet(struct rxrpc_call *call, struct rxrpc_txbuf *txb) +static size_t rxrpc_prepare_data_packet(struct rxrpc_call *call, struct rxrpc_txbuf *head, int n) { + struct rxrpc_txbuf *txb = head; rxrpc_serial_t serial; size_t len = 0; /* Each transmission of a Tx packet needs a new serial number */ - serial = rxrpc_get_next_serial(call->conn); + serial = rxrpc_get_next_serials(call->conn, n); - len += rxrpc_prepare_data_subpacket(call, txb, serial, 0); - // TODO: Loop around adding tail subpackets + for (int i = 0; i < n; i++) { + len += rxrpc_prepare_data_subpacket(call, txb, serial, i, n); + serial++; + txb = list_next_entry(txb, call_link); + } return len; } @@ -469,16 +491,24 @@ static size_t rxrpc_prepare_data_packet(struct rxrpc_call *call, struct rxrpc_tx /* * Set timeouts after transmitting a packet. */ -static void rxrpc_tstamp_data_packets(struct rxrpc_call *call, struct rxrpc_txbuf *txb) +static void rxrpc_tstamp_data_packets(struct rxrpc_call *call, struct rxrpc_txbuf *txb, int n) { + rxrpc_serial_t serial; ktime_t now = ktime_get_real(); bool ack_requested = txb->flags & RXRPC_REQUEST_ACK; + int i; call->tx_last_sent = now; - txb->last_sent = now; + + for (i = 0; i < n; i++) { + txb->last_sent = now; + ack_requested |= txb->flags & RXRPC_REQUEST_ACK; + serial = txb->serial; + txb = list_next_entry(txb, call_link); + } if (ack_requested) { - rxrpc_begin_rtt_probe(call, txb->serial, now, rxrpc_rtt_tx_data); + rxrpc_begin_rtt_probe(call, serial, now, rxrpc_rtt_tx_data); call->peer->rtt_last_req = now; if (call->peer->rtt_count > 1) { @@ -502,7 +532,7 @@ static void rxrpc_tstamp_data_packets(struct rxrpc_call *call, struct rxrpc_txbu /* * send a packet through the transport endpoint */ -static int rxrpc_send_data_packet(struct rxrpc_call *call, struct rxrpc_txbuf *txb) +static int rxrpc_send_data_packet(struct rxrpc_call *call, struct rxrpc_txbuf *txb, int n) { struct rxrpc_connection *conn = call->conn; enum rxrpc_tx_point frag; @@ -512,7 +542,7 @@ static int rxrpc_send_data_packet(struct rxrpc_call *call, struct rxrpc_txbuf *t _enter("%x,{%d}", txb->seq, txb->pkt_len); - len = rxrpc_prepare_data_packet(call, txb); + len = rxrpc_prepare_data_packet(call, txb, n); if (IS_ENABLED(CONFIG_AF_RXRPC_INJECT_LOSS)) { static int lose; @@ -524,7 +554,7 @@ static int rxrpc_send_data_packet(struct rxrpc_call *call, struct rxrpc_txbuf *t } } - iov_iter_kvec(&msg.msg_iter, WRITE, call->local->kvec, 1, len); + iov_iter_kvec(&msg.msg_iter, WRITE, call->local->kvec, n, len); msg.msg_name = &call->peer->srx.transport; msg.msg_namelen = call->peer->srx.transport_len; @@ -537,7 +567,7 @@ static int rxrpc_send_data_packet(struct rxrpc_call *call, struct rxrpc_txbuf *t * yet. */ if (txb->seq == call->tx_transmitted + 1) - call->tx_transmitted = txb->seq; + call->tx_transmitted = txb->seq + n - 1; /* send the packet with the don't fragment bit set if we currently * think it's small enough */ @@ -568,7 +598,7 @@ static int rxrpc_send_data_packet(struct rxrpc_call *call, struct rxrpc_txbuf *t } rxrpc_tx_backoff(call, ret); - if (ret == -EMSGSIZE && frag == rxrpc_tx_point_call_data_frag) { + if (ret == -EMSGSIZE && frag == rxrpc_tx_point_call_data_nofrag) { rxrpc_local_dont_fragment(conn->local, false); frag = rxrpc_tx_point_call_data_frag; goto retry; @@ -576,7 +606,7 @@ static int rxrpc_send_data_packet(struct rxrpc_call *call, struct rxrpc_txbuf *t done: if (ret >= 0) { - rxrpc_tstamp_data_packets(call, txb); + rxrpc_tstamp_data_packets(call, txb, n); } else { /* Cancel the call if the initial transmission fails, * particularly if that's due to network routing issues that @@ -776,13 +806,13 @@ static inline void rxrpc_instant_resend(struct rxrpc_call *call, } /* - * Transmit one packet. + * Transmit a packet, possibly gluing several subpackets together. */ -void rxrpc_transmit_one(struct rxrpc_call *call, struct rxrpc_txbuf *txb) +void rxrpc_transmit_data(struct rxrpc_call *call, struct rxrpc_txbuf *txb, int n) { int ret; - ret = rxrpc_send_data_packet(call, txb); + ret = rxrpc_send_data_packet(call, txb, n); if (ret < 0) { switch (ret) { case -ENETUNREACH: diff --git a/net/rxrpc/rxkad.c b/net/rxrpc/rxkad.c index 755897fab626..62b09d23ec08 100644 --- a/net/rxrpc/rxkad.c +++ b/net/rxrpc/rxkad.c @@ -392,15 +392,28 @@ static int rxkad_secure_packet(struct rxrpc_call *call, struct rxrpc_txbuf *txb) break; case RXRPC_SECURITY_AUTH: ret = rxkad_secure_packet_auth(call, txb, req); + if (txb->alloc_size == RXRPC_JUMBO_DATALEN) + txb->jumboable = true; break; case RXRPC_SECURITY_ENCRYPT: ret = rxkad_secure_packet_encrypt(call, txb, req); + if (txb->alloc_size == RXRPC_JUMBO_DATALEN) + txb->jumboable = true; break; default: ret = -EPERM; break; } + /* Clear excess space in the packet */ + if (txb->pkt_len < txb->alloc_size) { + struct rxrpc_wire_header *whdr = txb->kvec[0].iov_base; + size_t gap = txb->alloc_size - txb->pkt_len; + void *p = whdr + 1; + + memset(p + txb->pkt_len, 0, gap); + } + skcipher_request_free(req); _leave(" = %d [set %x]", ret, y); return ret;

[net-next,v2,11/39] rxrpc: Prepare to be able to send jumbo DATA packets

Checks

Commit Message

Patch