From patchwork Thu Aug 29 13:07:40 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 1155250 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=redhat.com Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 46K2w63hltz9sNf for ; Thu, 29 Aug 2019 23:07:50 +1000 (AEST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727433AbfH2NHm (ORCPT ); Thu, 29 Aug 2019 09:07:42 -0400 Received: from mx1.redhat.com ([209.132.183.28]:49055 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727066AbfH2NHm (ORCPT ); Thu, 29 Aug 2019 09:07:42 -0400 Received: from smtp.corp.redhat.com (int-mx08.intmail.prod.int.phx2.redhat.com [10.5.11.23]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id D6E0018012CA; Thu, 29 Aug 2019 13:07:41 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-120-255.rdu2.redhat.com [10.10.120.255]) by smtp.corp.redhat.com (Postfix) with ESMTP id D25B84116; Thu, 29 Aug 2019 13:07:40 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [PATCH net 1/7] rxrpc: Improve jumbo packet counting From: David Howells To: netdev@vger.kernel.org Cc: dhowells@redhat.com, linux-afs@lists.infradead.org, linux-kernel@vger.kernel.org Date: Thu, 29 Aug 2019 14:07:40 +0100 Message-ID: <156708406010.26102.14441247200172165798.stgit@warthog.procyon.org.uk> In-Reply-To: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> References: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.84 on 10.5.11.23 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.6.2 (mx1.redhat.com [10.5.110.63]); Thu, 29 Aug 2019 13:07:41 +0000 (UTC) Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Improve the information stored about jumbo packets so that we don't need to reparse them so much later. Signed-off-by: David Howells Reviewed-by: Jeffrey Altman --- net/rxrpc/ar-internal.h | 10 +++++++--- net/rxrpc/input.c | 23 ++++++++++++++--------- net/rxrpc/protocol.h | 9 +++++++++ 3 files changed, 30 insertions(+), 12 deletions(-) diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index 145335611af6..87cff6c218b6 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -185,11 +185,15 @@ struct rxrpc_host_header { * - max 48 bytes (struct sk_buff::cb) */ struct rxrpc_skb_priv { - union { - u8 nr_jumbo; /* Number of jumbo subpackets */ - }; + u8 nr_subpackets; /* Number of subpackets */ + u8 rx_flags; /* Received packet flags */ +#define RXRPC_SKB_INCL_LAST 0x01 /* - Includes last packet */ union { int remain; /* amount of space remaining for next write */ + + /* List of requested ACKs on subpackets */ + unsigned long rx_req_ack[(RXRPC_MAX_NR_JUMBO + BITS_PER_LONG - 1) / + BITS_PER_LONG]; }; struct rxrpc_host_header hdr; /* RxRPC packet header from this packet */ diff --git a/net/rxrpc/input.c b/net/rxrpc/input.c index dd47d465d1d3..ffcec5117954 100644 --- a/net/rxrpc/input.c +++ b/net/rxrpc/input.c @@ -347,7 +347,7 @@ static bool rxrpc_receiving_reply(struct rxrpc_call *call) } /* - * Scan a jumbo packet to validate its structure and to work out how many + * Scan a data packet to validate its structure and to work out how many * subpackets it contains. * * A jumbo packet is a collection of consecutive packets glued together with @@ -358,16 +358,21 @@ static bool rxrpc_receiving_reply(struct rxrpc_call *call) * the last are RXRPC_JUMBO_DATALEN in size. The last subpacket may be of any * size. */ -static bool rxrpc_validate_jumbo(struct sk_buff *skb) +static bool rxrpc_validate_data(struct sk_buff *skb) { struct rxrpc_skb_priv *sp = rxrpc_skb(skb); unsigned int offset = sizeof(struct rxrpc_wire_header); unsigned int len = skb->len; - int nr_jumbo = 1; u8 flags = sp->hdr.flags; - do { - nr_jumbo++; + for (;;) { + if (flags & RXRPC_REQUEST_ACK) + __set_bit(sp->nr_subpackets, sp->rx_req_ack); + sp->nr_subpackets++; + + if (!(flags & RXRPC_JUMBO_PACKET)) + break; + if (len - offset < RXRPC_JUMBO_SUBPKTLEN) goto protocol_error; if (flags & RXRPC_LAST_PACKET) @@ -376,9 +381,10 @@ static bool rxrpc_validate_jumbo(struct sk_buff *skb) if (skb_copy_bits(skb, offset, &flags, 1) < 0) goto protocol_error; offset += sizeof(struct rxrpc_jumbo_header); - } while (flags & RXRPC_JUMBO_PACKET); + } - sp->nr_jumbo = nr_jumbo; + if (flags & RXRPC_LAST_PACKET) + sp->rx_flags |= RXRPC_SKB_INCL_LAST; return true; protocol_error: @@ -1237,8 +1243,7 @@ int rxrpc_input_packet(struct sock *udp_sk, struct sk_buff *skb) if (sp->hdr.callNumber == 0 || sp->hdr.seq == 0) goto bad_message; - if (sp->hdr.flags & RXRPC_JUMBO_PACKET && - !rxrpc_validate_jumbo(skb)) + if (!rxrpc_validate_data(skb)) goto bad_message; break; diff --git a/net/rxrpc/protocol.h b/net/rxrpc/protocol.h index 99ce322d7caa..49bb972539aa 100644 --- a/net/rxrpc/protocol.h +++ b/net/rxrpc/protocol.h @@ -89,6 +89,15 @@ struct rxrpc_jumbo_header { #define RXRPC_JUMBO_DATALEN 1412 /* non-terminal jumbo packet data length */ #define RXRPC_JUMBO_SUBPKTLEN (RXRPC_JUMBO_DATALEN + sizeof(struct rxrpc_jumbo_header)) +/* + * The maximum number of subpackets that can possibly fit in a UDP packet is: + * + * ((max_IP - IP_hdr - UDP_hdr) / RXRPC_JUMBO_SUBPKTLEN) + 1 + * = ((65535 - 28 - 28) / 1416) + 1 + * = 46 non-terminal packets and 1 terminal packet. + */ +#define RXRPC_MAX_NR_JUMBO 47 + /*****************************************************************************/ /* * on-the-wire Rx ACK packet data payload From patchwork Thu Aug 29 13:07:47 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 1155251 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=redhat.com Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 46K2wD6GTmz9sMr for ; Thu, 29 Aug 2019 23:07:56 +1000 (AEST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728076AbfH2NHx (ORCPT ); Thu, 29 Aug 2019 09:07:53 -0400 Received: from mx1.redhat.com ([209.132.183.28]:12442 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727195AbfH2NHt (ORCPT ); Thu, 29 Aug 2019 09:07:49 -0400 Received: from smtp.corp.redhat.com (int-mx06.intmail.prod.int.phx2.redhat.com [10.5.11.16]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 367BD300BEB0; Thu, 29 Aug 2019 13:07:49 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-120-255.rdu2.redhat.com [10.10.120.255]) by smtp.corp.redhat.com (Postfix) with ESMTP id DFF3A5C1D6; Thu, 29 Aug 2019 13:07:47 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [PATCH net 2/7] rxrpc: Use info in skbuff instead of reparsing a jumbo packet From: David Howells To: netdev@vger.kernel.org Cc: dhowells@redhat.com, linux-afs@lists.infradead.org, linux-kernel@vger.kernel.org Date: Thu, 29 Aug 2019 14:07:47 +0100 Message-ID: <156708406708.26102.13192564522523294223.stgit@warthog.procyon.org.uk> In-Reply-To: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> References: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.16 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.45]); Thu, 29 Aug 2019 13:07:49 +0000 (UTC) Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Use the information now cached in the skbuff private data to avoid the need to reparse a jumbo packet. We can find all the subpackets by dead reckoning, so it's only necessary to note how many there are, whether the last one is flagged as LAST_PACKET and whether any have the REQUEST_ACK flag set. This is necessary as once recvmsg() can see the packet, it can start modifying it, such as doing in-place decryption. Fixes: 248f219cb8bc ("rxrpc: Rewrite the data and ack handling code") Signed-off-by: David Howells --- net/rxrpc/ar-internal.h | 3 - net/rxrpc/input.c | 231 +++++++++++++++++++++++------------------------ net/rxrpc/recvmsg.c | 41 +++++--- 3 files changed, 139 insertions(+), 136 deletions(-) diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index 87cff6c218b6..20d7907a5bc6 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -617,8 +617,7 @@ struct rxrpc_call { #define RXRPC_TX_ANNO_LAST 0x04 #define RXRPC_TX_ANNO_RESENT 0x08 -#define RXRPC_RX_ANNO_JUMBO 0x3f /* Jumbo subpacket number + 1 if not zero */ -#define RXRPC_RX_ANNO_JLAST 0x40 /* Set if last element of a jumbo packet */ +#define RXRPC_RX_ANNO_SUBPACKET 0x3f /* Subpacket number in jumbogram */ #define RXRPC_RX_ANNO_VERIFIED 0x80 /* Set if verified and decrypted */ rxrpc_seq_t tx_hard_ack; /* Dead slot in buffer; the first transmitted but * not hard-ACK'd packet follows this. diff --git a/net/rxrpc/input.c b/net/rxrpc/input.c index ffcec5117954..35b1a9368d80 100644 --- a/net/rxrpc/input.c +++ b/net/rxrpc/input.c @@ -405,10 +405,10 @@ static bool rxrpc_validate_data(struct sk_buff *skb) * (that information is encoded in the ACK packet). */ static void rxrpc_input_dup_data(struct rxrpc_call *call, rxrpc_seq_t seq, - u8 annotation, bool *_jumbo_bad) + bool is_jumbo, bool *_jumbo_bad) { /* Discard normal packets that are duplicates. */ - if (annotation == 0) + if (is_jumbo) return; /* Skip jumbo subpackets that are duplicates. When we've had three or @@ -428,19 +428,17 @@ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb) { struct rxrpc_skb_priv *sp = rxrpc_skb(skb); enum rxrpc_call_state state; - unsigned int offset = sizeof(struct rxrpc_wire_header); - unsigned int ix; + unsigned int j; rxrpc_serial_t serial = sp->hdr.serial, ack_serial = 0; - rxrpc_seq_t seq = sp->hdr.seq, hard_ack; - bool immediate_ack = false, jumbo_bad = false, queued; - u16 len; - u8 ack = 0, flags, annotation = 0; + rxrpc_seq_t seq0 = sp->hdr.seq, hard_ack; + bool immediate_ack = false, jumbo_bad = false; + u8 ack = 0; _enter("{%u,%u},{%u,%u}", - call->rx_hard_ack, call->rx_top, skb->len, seq); + call->rx_hard_ack, call->rx_top, skb->len, seq0); - _proto("Rx DATA %%%u { #%u f=%02x }", - sp->hdr.serial, seq, sp->hdr.flags); + _proto("Rx DATA %%%u { #%u f=%02x n=%u }", + sp->hdr.serial, seq0, sp->hdr.flags, sp->nr_subpackets); state = READ_ONCE(call->state); if (state >= RXRPC_CALL_COMPLETE) @@ -469,137 +467,136 @@ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb) !rxrpc_receiving_reply(call)) goto unlock; - call->ackr_prev_seq = seq; - + call->ackr_prev_seq = seq0; hard_ack = READ_ONCE(call->rx_hard_ack); - if (after(seq, hard_ack + call->rx_winsize)) { - ack = RXRPC_ACK_EXCEEDS_WINDOW; - ack_serial = serial; - goto ack; - } - flags = sp->hdr.flags; - if (flags & RXRPC_JUMBO_PACKET) { + if (sp->nr_subpackets > 1) { if (call->nr_jumbo_bad > 3) { ack = RXRPC_ACK_NOSPACE; ack_serial = serial; goto ack; } - annotation = 1; } -next_subpacket: - queued = false; - ix = seq & RXRPC_RXTX_BUFF_MASK; - len = skb->len; - if (flags & RXRPC_JUMBO_PACKET) - len = RXRPC_JUMBO_DATALEN; - - if (flags & RXRPC_LAST_PACKET) { - if (test_bit(RXRPC_CALL_RX_LAST, &call->flags) && - seq != call->rx_top) { - rxrpc_proto_abort("LSN", call, seq); - goto unlock; - } - } else { - if (test_bit(RXRPC_CALL_RX_LAST, &call->flags) && - after_eq(seq, call->rx_top)) { - rxrpc_proto_abort("LSA", call, seq); - goto unlock; + for (j = 0; j < sp->nr_subpackets; j++) { + rxrpc_serial_t serial = sp->hdr.serial + j; + rxrpc_seq_t seq = seq0 + j; + unsigned int ix = seq & RXRPC_RXTX_BUFF_MASK; + bool terminal = (j == sp->nr_subpackets - 1); + bool last = terminal && (sp->rx_flags & RXRPC_SKB_INCL_LAST); + u8 flags, annotation = j; + + _proto("Rx DATA+%u %%%u { #%x t=%u l=%u }", + j, serial, seq, terminal, last); + + if (last) { + if (test_bit(RXRPC_CALL_RX_LAST, &call->flags) && + seq != call->rx_top) { + rxrpc_proto_abort("LSN", call, seq); + goto unlock; + } + } else { + if (test_bit(RXRPC_CALL_RX_LAST, &call->flags) && + after_eq(seq, call->rx_top)) { + rxrpc_proto_abort("LSA", call, seq); + goto unlock; + } } - } - - trace_rxrpc_rx_data(call->debug_id, seq, serial, flags, annotation); - if (before_eq(seq, hard_ack)) { - ack = RXRPC_ACK_DUPLICATE; - ack_serial = serial; - goto skip; - } - if (flags & RXRPC_REQUEST_ACK && !ack) { - ack = RXRPC_ACK_REQUESTED; - ack_serial = serial; - } + flags = 0; + if (last) + flags |= RXRPC_LAST_PACKET; + if (!terminal) + flags |= RXRPC_JUMBO_PACKET; + if (test_bit(j, sp->rx_req_ack)) + flags |= RXRPC_REQUEST_ACK; + trace_rxrpc_rx_data(call->debug_id, seq, serial, flags, annotation); - if (call->rxtx_buffer[ix]) { - rxrpc_input_dup_data(call, seq, annotation, &jumbo_bad); - if (ack != RXRPC_ACK_DUPLICATE) { + if (before_eq(seq, hard_ack)) { ack = RXRPC_ACK_DUPLICATE; ack_serial = serial; + continue; } - immediate_ack = true; - goto skip; - } - - /* Queue the packet. We use a couple of memory barriers here as need - * to make sure that rx_top is perceived to be set after the buffer - * pointer and that the buffer pointer is set after the annotation and - * the skb data. - * - * Barriers against rxrpc_recvmsg_data() and rxrpc_rotate_rx_window() - * and also rxrpc_fill_out_ack(). - */ - rxrpc_get_skb(skb, rxrpc_skb_rx_got); - call->rxtx_annotations[ix] = annotation; - smp_wmb(); - call->rxtx_buffer[ix] = skb; - if (after(seq, call->rx_top)) { - smp_store_release(&call->rx_top, seq); - } else if (before(seq, call->rx_top)) { - /* Send an immediate ACK if we fill in a hole */ - if (!ack) { - ack = RXRPC_ACK_DELAY; - ack_serial = serial; - } - immediate_ack = true; - } - if (flags & RXRPC_LAST_PACKET) { - set_bit(RXRPC_CALL_RX_LAST, &call->flags); - trace_rxrpc_receive(call, rxrpc_receive_queue_last, serial, seq); - } else { - trace_rxrpc_receive(call, rxrpc_receive_queue, serial, seq); - } - queued = true; - if (after_eq(seq, call->rx_expect_next)) { - if (after(seq, call->rx_expect_next)) { - _net("OOS %u > %u", seq, call->rx_expect_next); - ack = RXRPC_ACK_OUT_OF_SEQUENCE; - ack_serial = serial; + if (call->rxtx_buffer[ix]) { + rxrpc_input_dup_data(call, seq, sp->nr_subpackets > 1, + &jumbo_bad); + if (ack != RXRPC_ACK_DUPLICATE) { + ack = RXRPC_ACK_DUPLICATE; + ack_serial = serial; + } + immediate_ack = true; + continue; } - call->rx_expect_next = seq + 1; - } -skip: - offset += len; - if (flags & RXRPC_JUMBO_PACKET) { - if (skb_copy_bits(skb, offset, &flags, 1) < 0) { - rxrpc_proto_abort("XJF", call, seq); - goto unlock; - } - offset += sizeof(struct rxrpc_jumbo_header); - seq++; - serial++; - annotation++; - if (flags & RXRPC_JUMBO_PACKET) - annotation |= RXRPC_RX_ANNO_JLAST; if (after(seq, hard_ack + call->rx_winsize)) { ack = RXRPC_ACK_EXCEEDS_WINDOW; ack_serial = serial; - if (!jumbo_bad) { - call->nr_jumbo_bad++; - jumbo_bad = true; + if (flags & RXRPC_JUMBO_PACKET) { + if (!jumbo_bad) { + call->nr_jumbo_bad++; + jumbo_bad = true; + } } + goto ack; } - _proto("Rx DATA Jumbo %%%u", serial); - goto next_subpacket; - } + if (flags & RXRPC_REQUEST_ACK && !ack) { + ack = RXRPC_ACK_REQUESTED; + ack_serial = serial; + } + + /* Queue the packet. We use a couple of memory barriers here as need + * to make sure that rx_top is perceived to be set after the buffer + * pointer and that the buffer pointer is set after the annotation and + * the skb data. + * + * Barriers against rxrpc_recvmsg_data() and rxrpc_rotate_rx_window() + * and also rxrpc_fill_out_ack(). + */ + rxrpc_get_skb(skb, rxrpc_skb_rx_got); + call->rxtx_annotations[ix] = annotation; + smp_wmb(); + call->rxtx_buffer[ix] = skb; + if (after(seq, call->rx_top)) { + smp_store_release(&call->rx_top, seq); + } else if (before(seq, call->rx_top)) { + /* Send an immediate ACK if we fill in a hole */ + if (!ack) { + ack = RXRPC_ACK_DELAY; + ack_serial = serial; + } + immediate_ack = true; + } + + if (terminal) { + /* From this point on, we're not allowed to touch the + * packet any longer as its ref now belongs to the Rx + * ring. + */ + skb = NULL; + } - if (queued && flags & RXRPC_LAST_PACKET && !ack) { - ack = RXRPC_ACK_DELAY; - ack_serial = serial; + if (last) { + set_bit(RXRPC_CALL_RX_LAST, &call->flags); + if (!ack) { + ack = RXRPC_ACK_DELAY; + ack_serial = serial; + } + trace_rxrpc_receive(call, rxrpc_receive_queue_last, serial, seq); + } else { + trace_rxrpc_receive(call, rxrpc_receive_queue, serial, seq); + } + + if (after_eq(seq, call->rx_expect_next)) { + if (after(seq, call->rx_expect_next)) { + _net("OOS %u > %u", seq, call->rx_expect_next); + ack = RXRPC_ACK_OUT_OF_SEQUENCE; + ack_serial = serial; + } + call->rx_expect_next = seq + 1; + } } ack: @@ -612,7 +609,7 @@ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb) false, true, rxrpc_propose_ack_input_data); - if (sp->hdr.seq == READ_ONCE(call->rx_hard_ack) + 1) { + if (seq0 == READ_ONCE(call->rx_hard_ack) + 1) { trace_rxrpc_notify_socket(call->debug_id, serial); rxrpc_notify_socket(call); } diff --git a/net/rxrpc/recvmsg.c b/net/rxrpc/recvmsg.c index 9a7e1bc9791d..e49eacfaf4d6 100644 --- a/net/rxrpc/recvmsg.c +++ b/net/rxrpc/recvmsg.c @@ -177,7 +177,8 @@ static void rxrpc_rotate_rx_window(struct rxrpc_call *call) struct sk_buff *skb; rxrpc_serial_t serial; rxrpc_seq_t hard_ack, top; - u8 flags; + bool last = false; + u8 subpacket; int ix; _enter("%d", call->debug_id); @@ -191,10 +192,13 @@ static void rxrpc_rotate_rx_window(struct rxrpc_call *call) skb = call->rxtx_buffer[ix]; rxrpc_see_skb(skb, rxrpc_skb_rx_rotated); sp = rxrpc_skb(skb); - flags = sp->hdr.flags; - serial = sp->hdr.serial; - if (call->rxtx_annotations[ix] & RXRPC_RX_ANNO_JUMBO) - serial += (call->rxtx_annotations[ix] & RXRPC_RX_ANNO_JUMBO) - 1; + + subpacket = call->rxtx_annotations[ix] & RXRPC_RX_ANNO_SUBPACKET; + serial = sp->hdr.serial + subpacket; + + if (subpacket == sp->nr_subpackets - 1 && + sp->rx_flags & RXRPC_SKB_INCL_LAST) + last = true; call->rxtx_buffer[ix] = NULL; call->rxtx_annotations[ix] = 0; @@ -203,9 +207,8 @@ static void rxrpc_rotate_rx_window(struct rxrpc_call *call) rxrpc_free_skb(skb, rxrpc_skb_rx_freed); - _debug("%u,%u,%02x", hard_ack, top, flags); trace_rxrpc_receive(call, rxrpc_receive_rotate, serial, hard_ack); - if (flags & RXRPC_LAST_PACKET) { + if (last) { rxrpc_end_rx_phase(call, serial); } else { /* Check to see if there's an ACK that needs sending. */ @@ -233,18 +236,19 @@ static int rxrpc_verify_packet(struct rxrpc_call *call, struct sk_buff *skb, struct rxrpc_skb_priv *sp = rxrpc_skb(skb); rxrpc_seq_t seq = sp->hdr.seq; u16 cksum = sp->hdr.cksum; + u8 subpacket = annotation & RXRPC_RX_ANNO_SUBPACKET; _enter(""); /* For all but the head jumbo subpacket, the security checksum is in a * jumbo header immediately prior to the data. */ - if ((annotation & RXRPC_RX_ANNO_JUMBO) > 1) { + if (subpacket > 0) { __be16 tmp; if (skb_copy_bits(skb, offset - 2, &tmp, 2) < 0) BUG(); cksum = ntohs(tmp); - seq += (annotation & RXRPC_RX_ANNO_JUMBO) - 1; + seq += subpacket; } return call->conn->security->verify_packet(call, skb, offset, len, @@ -265,19 +269,18 @@ static int rxrpc_locate_data(struct rxrpc_call *call, struct sk_buff *skb, u8 *_annotation, unsigned int *_offset, unsigned int *_len) { + struct rxrpc_skb_priv *sp = rxrpc_skb(skb); unsigned int offset = sizeof(struct rxrpc_wire_header); unsigned int len; int ret; u8 annotation = *_annotation; + u8 subpacket = annotation & RXRPC_RX_ANNO_SUBPACKET; /* Locate the subpacket */ + offset += subpacket * RXRPC_JUMBO_SUBPKTLEN; len = skb->len - offset; - if ((annotation & RXRPC_RX_ANNO_JUMBO) > 0) { - offset += (((annotation & RXRPC_RX_ANNO_JUMBO) - 1) * - RXRPC_JUMBO_SUBPKTLEN); - len = (annotation & RXRPC_RX_ANNO_JLAST) ? - skb->len - offset : RXRPC_JUMBO_SUBPKTLEN; - } + if (subpacket < sp->nr_subpackets - 1) + len = RXRPC_JUMBO_DATALEN; if (!(annotation & RXRPC_RX_ANNO_VERIFIED)) { ret = rxrpc_verify_packet(call, skb, annotation, offset, len); @@ -303,6 +306,7 @@ static int rxrpc_recvmsg_data(struct socket *sock, struct rxrpc_call *call, { struct rxrpc_skb_priv *sp; struct sk_buff *skb; + rxrpc_serial_t serial; rxrpc_seq_t hard_ack, top, seq; size_t remain; bool last; @@ -339,9 +343,12 @@ static int rxrpc_recvmsg_data(struct socket *sock, struct rxrpc_call *call, rxrpc_see_skb(skb, rxrpc_skb_rx_seen); sp = rxrpc_skb(skb); - if (!(flags & MSG_PEEK)) + if (!(flags & MSG_PEEK)) { + serial = sp->hdr.serial; + serial += call->rxtx_annotations[ix] & RXRPC_RX_ANNO_SUBPACKET; trace_rxrpc_receive(call, rxrpc_receive_front, - sp->hdr.serial, seq); + serial, seq); + } if (msg) sock_recv_timestamp(msg, sock->sk, skb); From patchwork Thu Aug 29 13:07:54 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 1155252 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=redhat.com Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 46K2wL2qkvz9sML for ; Thu, 29 Aug 2019 23:08:02 +1000 (AEST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727308AbfH2NH6 (ORCPT ); Thu, 29 Aug 2019 09:07:58 -0400 Received: from mx1.redhat.com ([209.132.183.28]:42660 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1728094AbfH2NH4 (ORCPT ); Thu, 29 Aug 2019 09:07:56 -0400 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.phx2.redhat.com [10.5.11.15]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id F06D8106E293; Thu, 29 Aug 2019 13:07:55 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-120-255.rdu2.redhat.com [10.10.120.255]) by smtp.corp.redhat.com (Postfix) with ESMTP id 2258E5D713; Thu, 29 Aug 2019 13:07:54 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [PATCH net 3/7] rxrpc: Pass the input handler's data skb reference to the Rx ring From: David Howells To: netdev@vger.kernel.org Cc: dhowells@redhat.com, linux-afs@lists.infradead.org, linux-kernel@vger.kernel.org Date: Thu, 29 Aug 2019 14:07:54 +0100 Message-ID: <156708407439.26102.1440677467082820559.stgit@warthog.procyon.org.uk> In-Reply-To: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> References: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.15 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.6.2 (mx1.redhat.com [10.5.110.64]); Thu, 29 Aug 2019 13:07:56 +0000 (UTC) Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Pass the reference held on a DATA skb in the rxrpc input handler into the Rx ring rather than getting an additional ref for this and then dropping the original ref at the end. Signed-off-by: David Howells --- net/rxrpc/input.c | 20 +++++++++++++++----- 1 file changed, 15 insertions(+), 5 deletions(-) diff --git a/net/rxrpc/input.c b/net/rxrpc/input.c index 35b1a9368d80..140cede77655 100644 --- a/net/rxrpc/input.c +++ b/net/rxrpc/input.c @@ -422,7 +422,8 @@ static void rxrpc_input_dup_data(struct rxrpc_call *call, rxrpc_seq_t seq, } /* - * Process a DATA packet, adding the packet to the Rx ring. + * Process a DATA packet, adding the packet to the Rx ring. The caller's + * packet ref must be passed on or discarded. */ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb) { @@ -441,8 +442,10 @@ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb) sp->hdr.serial, seq0, sp->hdr.flags, sp->nr_subpackets); state = READ_ONCE(call->state); - if (state >= RXRPC_CALL_COMPLETE) + if (state >= RXRPC_CALL_COMPLETE) { + rxrpc_free_skb(skb, rxrpc_skb_rx_freed); return; + } if (call->state == RXRPC_CALL_SERVER_RECV_REQUEST) { unsigned long timo = READ_ONCE(call->next_req_timo); @@ -555,7 +558,8 @@ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb) * Barriers against rxrpc_recvmsg_data() and rxrpc_rotate_rx_window() * and also rxrpc_fill_out_ack(). */ - rxrpc_get_skb(skb, rxrpc_skb_rx_got); + if (!terminal) + rxrpc_get_skb(skb, rxrpc_skb_rx_got); call->rxtx_annotations[ix] = annotation; smp_wmb(); call->rxtx_buffer[ix] = skb; @@ -616,6 +620,7 @@ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb) unlock: spin_unlock(&call->input_lock); + rxrpc_free_skb(skb, rxrpc_skb_rx_freed); _leave(" [queued]"); } @@ -1024,7 +1029,7 @@ static void rxrpc_input_call_packet(struct rxrpc_call *call, switch (sp->hdr.type) { case RXRPC_PACKET_TYPE_DATA: rxrpc_input_data(call, skb); - break; + goto no_free; case RXRPC_PACKET_TYPE_ACK: rxrpc_input_ack(call, skb); @@ -1051,6 +1056,8 @@ static void rxrpc_input_call_packet(struct rxrpc_call *call, break; } + rxrpc_free_skb(skb, rxrpc_skb_rx_freed); +no_free: _leave(""); } @@ -1375,8 +1382,11 @@ int rxrpc_input_packet(struct sock *udp_sk, struct sk_buff *skb) mutex_unlock(&call->user_mutex); } + /* Process a call packet; this either discards or passes on the ref + * elsewhere. + */ rxrpc_input_call_packet(call, skb); - goto discard; + goto out; discard: rxrpc_free_skb(skb, rxrpc_skb_rx_freed); From patchwork Thu Aug 29 13:08:01 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 1155253 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=redhat.com Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 46K2wN4C1Jz9sML for ; Thu, 29 Aug 2019 23:08:04 +1000 (AEST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728127AbfH2NID (ORCPT ); Thu, 29 Aug 2019 09:08:03 -0400 Received: from mx1.redhat.com ([209.132.183.28]:40186 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727046AbfH2NIC (ORCPT ); Thu, 29 Aug 2019 09:08:02 -0400 Received: from smtp.corp.redhat.com (int-mx06.intmail.prod.int.phx2.redhat.com [10.5.11.16]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id C21BB7F746; Thu, 29 Aug 2019 13:08:02 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-120-255.rdu2.redhat.com [10.10.120.255]) by smtp.corp.redhat.com (Postfix) with ESMTP id E6A335C21E; Thu, 29 Aug 2019 13:08:01 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [PATCH net 4/7] rxrpc: Abstract out rxtx ring cleanup From: David Howells To: netdev@vger.kernel.org Cc: dhowells@redhat.com, linux-afs@lists.infradead.org, linux-kernel@vger.kernel.org Date: Thu, 29 Aug 2019 14:08:01 +0100 Message-ID: <156708408120.26102.7710356156597753765.stgit@warthog.procyon.org.uk> In-Reply-To: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> References: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.16 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.6.2 (mx1.redhat.com [10.5.110.71]); Thu, 29 Aug 2019 13:08:02 +0000 (UTC) Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Abstract out rxtx ring cleanup into its own function from its two callers. This makes it easier to apply the same changes to both. Signed-off-by: David Howells --- net/rxrpc/call_object.c | 33 +++++++++++++++++---------------- 1 file changed, 17 insertions(+), 16 deletions(-) diff --git a/net/rxrpc/call_object.c b/net/rxrpc/call_object.c index 217b12be9e08..c9ab2da957fe 100644 --- a/net/rxrpc/call_object.c +++ b/net/rxrpc/call_object.c @@ -421,6 +421,21 @@ void rxrpc_get_call(struct rxrpc_call *call, enum rxrpc_call_trace op) trace_rxrpc_call(call, op, n, here, NULL); } +/* + * Clean up the RxTx skb ring. + */ +static void rxrpc_cleanup_ring(struct rxrpc_call *call) +{ + int i; + + for (i = 0; i < RXRPC_RXTX_BUFF_SIZE; i++) { + rxrpc_free_skb(call->rxtx_buffer[i], + (call->tx_phase ? rxrpc_skb_tx_cleaned : + rxrpc_skb_rx_cleaned)); + call->rxtx_buffer[i] = NULL; + } +} + /* * Detach a call from its owning socket. */ @@ -429,7 +444,6 @@ void rxrpc_release_call(struct rxrpc_sock *rx, struct rxrpc_call *call) const void *here = __builtin_return_address(0); struct rxrpc_connection *conn = call->conn; bool put = false; - int i; _enter("{%d,%d}", call->debug_id, atomic_read(&call->usage)); @@ -479,13 +493,7 @@ void rxrpc_release_call(struct rxrpc_sock *rx, struct rxrpc_call *call) if (conn) rxrpc_disconnect_call(call); - for (i = 0; i < RXRPC_RXTX_BUFF_SIZE; i++) { - rxrpc_free_skb(call->rxtx_buffer[i], - (call->tx_phase ? rxrpc_skb_tx_cleaned : - rxrpc_skb_rx_cleaned)); - call->rxtx_buffer[i] = NULL; - } - + rxrpc_cleanup_ring(call); _leave(""); } @@ -568,8 +576,6 @@ static void rxrpc_rcu_destroy_call(struct rcu_head *rcu) */ void rxrpc_cleanup_call(struct rxrpc_call *call) { - int i; - _net("DESTROY CALL %d", call->debug_id); memset(&call->sock_node, 0xcd, sizeof(call->sock_node)); @@ -580,12 +586,7 @@ void rxrpc_cleanup_call(struct rxrpc_call *call) ASSERT(test_bit(RXRPC_CALL_RELEASED, &call->flags)); ASSERTCMP(call->conn, ==, NULL); - /* Clean up the Rx/Tx buffer */ - for (i = 0; i < RXRPC_RXTX_BUFF_SIZE; i++) - rxrpc_free_skb(call->rxtx_buffer[i], - (call->tx_phase ? rxrpc_skb_tx_cleaned : - rxrpc_skb_rx_cleaned)); - + rxrpc_cleanup_ring(call); rxrpc_free_skb(call->tx_pending, rxrpc_skb_tx_cleaned); call_rcu(&call->rcu, rxrpc_rcu_destroy_call); From patchwork Thu Aug 29 13:08:08 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 1155254 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=redhat.com Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 46K2wW4Lh2z9s5b for ; Thu, 29 Aug 2019 23:08:11 +1000 (AEST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728160AbfH2NIK (ORCPT ); Thu, 29 Aug 2019 09:08:10 -0400 Received: from mx1.redhat.com ([209.132.183.28]:18183 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727046AbfH2NIJ (ORCPT ); Thu, 29 Aug 2019 09:08:09 -0400 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.phx2.redhat.com [10.5.11.15]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id AD4403CA25; Thu, 29 Aug 2019 13:08:09 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-120-255.rdu2.redhat.com [10.10.120.255]) by smtp.corp.redhat.com (Postfix) with ESMTP id C6F5C5D772; Thu, 29 Aug 2019 13:08:08 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [PATCH net 5/7] rxrpc: Add a private skb flag to indicate transmission-phase skbs From: David Howells To: netdev@vger.kernel.org Cc: dhowells@redhat.com, linux-afs@lists.infradead.org, linux-kernel@vger.kernel.org Date: Thu, 29 Aug 2019 14:08:08 +0100 Message-ID: <156708408801.26102.11592449726244941243.stgit@warthog.procyon.org.uk> In-Reply-To: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> References: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.15 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.39]); Thu, 29 Aug 2019 13:08:09 +0000 (UTC) Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Add a flag in the private data on an skbuff to indicate that this is a transmission-phase buffer rather than a receive-phase buffer. Signed-off-by: David Howells --- net/rxrpc/ar-internal.h | 1 + net/rxrpc/sendmsg.c | 3 ++- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index 20d7907a5bc6..63d3a91ce5e9 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -188,6 +188,7 @@ struct rxrpc_skb_priv { u8 nr_subpackets; /* Number of subpackets */ u8 rx_flags; /* Received packet flags */ #define RXRPC_SKB_INCL_LAST 0x01 /* - Includes last packet */ +#define RXRPC_SKB_TX_BUFFER 0x02 /* - Is transmit buffer */ union { int remain; /* amount of space remaining for next write */ diff --git a/net/rxrpc/sendmsg.c b/net/rxrpc/sendmsg.c index bae14438f869..472dc3b7d91f 100644 --- a/net/rxrpc/sendmsg.c +++ b/net/rxrpc/sendmsg.c @@ -336,6 +336,8 @@ static int rxrpc_send_data(struct rxrpc_sock *rx, if (!skb) goto maybe_error; + sp = rxrpc_skb(skb); + sp->rx_flags |= RXRPC_SKB_TX_BUFFER; rxrpc_new_skb(skb, rxrpc_skb_tx_new); _debug("ALLOC SEND %p", skb); @@ -346,7 +348,6 @@ static int rxrpc_send_data(struct rxrpc_sock *rx, skb_reserve(skb, call->conn->security_size); skb->len += call->conn->security_size; - sp = rxrpc_skb(skb); sp->remain = chunk; if (sp->remain > skb_tailroom(skb)) sp->remain = skb_tailroom(skb); From patchwork Thu Aug 29 13:08:14 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 1155255 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=redhat.com Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 46K2wj2Lv5z9sNF for ; Thu, 29 Aug 2019 23:08:21 +1000 (AEST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728179AbfH2NIS (ORCPT ); Thu, 29 Aug 2019 09:08:18 -0400 Received: from mx1.redhat.com ([209.132.183.28]:47422 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727046AbfH2NIR (ORCPT ); Thu, 29 Aug 2019 09:08:17 -0400 Received: from smtp.corp.redhat.com (int-mx06.intmail.prod.int.phx2.redhat.com [10.5.11.16]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id D1B6B18C893F; Thu, 29 Aug 2019 13:08:16 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-120-255.rdu2.redhat.com [10.10.120.255]) by smtp.corp.redhat.com (Postfix) with ESMTP id A86FD5C21E; Thu, 29 Aug 2019 13:08:15 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [PATCH net 6/7] rxrpc: Use the tx-phase skb flag to simplify tracing From: David Howells To: netdev@vger.kernel.org Cc: dhowells@redhat.com, linux-afs@lists.infradead.org, linux-kernel@vger.kernel.org Date: Thu, 29 Aug 2019 14:08:14 +0100 Message-ID: <156708409493.26102.10580279561538010561.stgit@warthog.procyon.org.uk> In-Reply-To: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> References: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.16 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.6.2 (mx1.redhat.com [10.5.110.70]); Thu, 29 Aug 2019 13:08:16 +0000 (UTC) Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Use the previously-added transmit-phase skbuff private flag to simplify the socket buffer tracing a bit. Which phase the skbuff comes from can now be divined from the skb rather than having to be guessed from the call state. We can also reduce the number of rxrpc_skb_trace values by eliminating the difference between Tx and Rx in the symbols. Signed-off-by: David Howells --- include/trace/events/rxrpc.h | 51 ++++++++++++++++++------------------------ net/rxrpc/ar-internal.h | 1 + net/rxrpc/call_event.c | 8 +++---- net/rxrpc/call_object.c | 6 ++--- net/rxrpc/conn_event.c | 6 ++--- net/rxrpc/input.c | 22 +++++++++--------- net/rxrpc/local_event.c | 4 ++- net/rxrpc/output.c | 6 ++--- net/rxrpc/peer_event.c | 10 ++++---- net/rxrpc/recvmsg.c | 6 ++--- net/rxrpc/sendmsg.c | 10 ++++---- net/rxrpc/skbuff.c | 15 +++++++----- 12 files changed, 69 insertions(+), 76 deletions(-) diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index fa06b528c73c..e2356c51883b 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -23,20 +23,15 @@ #define __RXRPC_DECLARE_TRACE_ENUMS_ONCE_ONLY enum rxrpc_skb_trace { - rxrpc_skb_rx_cleaned, - rxrpc_skb_rx_freed, - rxrpc_skb_rx_got, - rxrpc_skb_rx_lost, - rxrpc_skb_rx_purged, - rxrpc_skb_rx_received, - rxrpc_skb_rx_rotated, - rxrpc_skb_rx_seen, - rxrpc_skb_tx_cleaned, - rxrpc_skb_tx_freed, - rxrpc_skb_tx_got, - rxrpc_skb_tx_new, - rxrpc_skb_tx_rotated, - rxrpc_skb_tx_seen, + rxrpc_skb_cleaned, + rxrpc_skb_freed, + rxrpc_skb_got, + rxrpc_skb_lost, + rxrpc_skb_new, + rxrpc_skb_purged, + rxrpc_skb_received, + rxrpc_skb_rotated, + rxrpc_skb_seen, }; enum rxrpc_local_trace { @@ -228,20 +223,15 @@ enum rxrpc_tx_point { * Declare tracing information enums and their string mappings for display. */ #define rxrpc_skb_traces \ - EM(rxrpc_skb_rx_cleaned, "Rx CLN") \ - EM(rxrpc_skb_rx_freed, "Rx FRE") \ - EM(rxrpc_skb_rx_got, "Rx GOT") \ - EM(rxrpc_skb_rx_lost, "Rx *L*") \ - EM(rxrpc_skb_rx_purged, "Rx PUR") \ - EM(rxrpc_skb_rx_received, "Rx RCV") \ - EM(rxrpc_skb_rx_rotated, "Rx ROT") \ - EM(rxrpc_skb_rx_seen, "Rx SEE") \ - EM(rxrpc_skb_tx_cleaned, "Tx CLN") \ - EM(rxrpc_skb_tx_freed, "Tx FRE") \ - EM(rxrpc_skb_tx_got, "Tx GOT") \ - EM(rxrpc_skb_tx_new, "Tx NEW") \ - EM(rxrpc_skb_tx_rotated, "Tx ROT") \ - E_(rxrpc_skb_tx_seen, "Tx SEE") + EM(rxrpc_skb_cleaned, "CLN") \ + EM(rxrpc_skb_freed, "FRE") \ + EM(rxrpc_skb_got, "GOT") \ + EM(rxrpc_skb_lost, "*L*") \ + EM(rxrpc_skb_new, "NEW") \ + EM(rxrpc_skb_purged, "PUR") \ + EM(rxrpc_skb_received, "RCV") \ + EM(rxrpc_skb_rotated, "ROT") \ + E_(rxrpc_skb_seen, "SEE") #define rxrpc_local_traces \ EM(rxrpc_local_got, "GOT") \ @@ -650,6 +640,7 @@ TRACE_EVENT(rxrpc_skb, TP_STRUCT__entry( __field(struct sk_buff *, skb ) __field(enum rxrpc_skb_trace, op ) + __field(u8, flags ) __field(int, usage ) __field(int, mod_count ) __field(const void *, where ) @@ -657,14 +648,16 @@ TRACE_EVENT(rxrpc_skb, TP_fast_assign( __entry->skb = skb; + __entry->flags = rxrpc_skb(skb)->rx_flags; __entry->op = op; __entry->usage = usage; __entry->mod_count = mod_count; __entry->where = where; ), - TP_printk("s=%p %s u=%d m=%d p=%pSR", + TP_printk("s=%p %cx %s u=%d m=%d p=%pSR", __entry->skb, + __entry->flags & RXRPC_SKB_TX_BUFFER ? 'T' : 'R', __print_symbolic(__entry->op, rxrpc_skb_traces), __entry->usage, __entry->mod_count, diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index 63d3a91ce5e9..2d5294f3e62f 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -185,6 +185,7 @@ struct rxrpc_host_header { * - max 48 bytes (struct sk_buff::cb) */ struct rxrpc_skb_priv { + atomic_t nr_ring_pins; /* Number of rxtx ring pins */ u8 nr_subpackets; /* Number of subpackets */ u8 rx_flags; /* Received packet flags */ #define RXRPC_SKB_INCL_LAST 0x01 /* - Includes last packet */ diff --git a/net/rxrpc/call_event.c b/net/rxrpc/call_event.c index c767679bfa5d..cedbbb3a7c2e 100644 --- a/net/rxrpc/call_event.c +++ b/net/rxrpc/call_event.c @@ -199,7 +199,7 @@ static void rxrpc_resend(struct rxrpc_call *call, unsigned long now_j) continue; skb = call->rxtx_buffer[ix]; - rxrpc_see_skb(skb, rxrpc_skb_tx_seen); + rxrpc_see_skb(skb, rxrpc_skb_seen); if (anno_type == RXRPC_TX_ANNO_UNACK) { if (ktime_after(skb->tstamp, max_age)) { @@ -255,18 +255,18 @@ static void rxrpc_resend(struct rxrpc_call *call, unsigned long now_j) continue; skb = call->rxtx_buffer[ix]; - rxrpc_get_skb(skb, rxrpc_skb_tx_got); + rxrpc_get_skb(skb, rxrpc_skb_got); spin_unlock_bh(&call->lock); if (rxrpc_send_data_packet(call, skb, true) < 0) { - rxrpc_free_skb(skb, rxrpc_skb_tx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); return; } if (rxrpc_is_client_call(call)) rxrpc_expose_client_call(call); - rxrpc_free_skb(skb, rxrpc_skb_tx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); spin_lock_bh(&call->lock); /* We need to clear the retransmit state, but there are two diff --git a/net/rxrpc/call_object.c b/net/rxrpc/call_object.c index c9ab2da957fe..014548c259ce 100644 --- a/net/rxrpc/call_object.c +++ b/net/rxrpc/call_object.c @@ -429,9 +429,7 @@ static void rxrpc_cleanup_ring(struct rxrpc_call *call) int i; for (i = 0; i < RXRPC_RXTX_BUFF_SIZE; i++) { - rxrpc_free_skb(call->rxtx_buffer[i], - (call->tx_phase ? rxrpc_skb_tx_cleaned : - rxrpc_skb_rx_cleaned)); + rxrpc_free_skb(call->rxtx_buffer[i], rxrpc_skb_cleaned); call->rxtx_buffer[i] = NULL; } } @@ -587,7 +585,7 @@ void rxrpc_cleanup_call(struct rxrpc_call *call) ASSERTCMP(call->conn, ==, NULL); rxrpc_cleanup_ring(call); - rxrpc_free_skb(call->tx_pending, rxrpc_skb_tx_cleaned); + rxrpc_free_skb(call->tx_pending, rxrpc_skb_cleaned); call_rcu(&call->rcu, rxrpc_rcu_destroy_call); } diff --git a/net/rxrpc/conn_event.c b/net/rxrpc/conn_event.c index df6624c140be..a1ceef4f5cd0 100644 --- a/net/rxrpc/conn_event.c +++ b/net/rxrpc/conn_event.c @@ -472,7 +472,7 @@ void rxrpc_process_connection(struct work_struct *work) /* go through the conn-level event packets, releasing the ref on this * connection that each one has when we've finished with it */ while ((skb = skb_dequeue(&conn->rx_queue))) { - rxrpc_see_skb(skb, rxrpc_skb_rx_seen); + rxrpc_see_skb(skb, rxrpc_skb_seen); ret = rxrpc_process_event(conn, skb, &abort_code); switch (ret) { case -EPROTO: @@ -484,7 +484,7 @@ void rxrpc_process_connection(struct work_struct *work) goto requeue_and_leave; case -ECONNABORTED: default: - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); break; } } @@ -501,6 +501,6 @@ void rxrpc_process_connection(struct work_struct *work) protocol_error: if (rxrpc_abort_connection(conn, ret, abort_code) < 0) goto requeue_and_leave; - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); goto out; } diff --git a/net/rxrpc/input.c b/net/rxrpc/input.c index 140cede77655..31090bdf1fae 100644 --- a/net/rxrpc/input.c +++ b/net/rxrpc/input.c @@ -233,7 +233,7 @@ static bool rxrpc_rotate_tx_window(struct rxrpc_call *call, rxrpc_seq_t to, ix = call->tx_hard_ack & RXRPC_RXTX_BUFF_MASK; skb = call->rxtx_buffer[ix]; annotation = call->rxtx_annotations[ix]; - rxrpc_see_skb(skb, rxrpc_skb_tx_rotated); + rxrpc_see_skb(skb, rxrpc_skb_rotated); call->rxtx_buffer[ix] = NULL; call->rxtx_annotations[ix] = 0; skb->next = list; @@ -258,7 +258,7 @@ static bool rxrpc_rotate_tx_window(struct rxrpc_call *call, rxrpc_seq_t to, skb = list; list = skb->next; skb_mark_not_on_list(skb); - rxrpc_free_skb(skb, rxrpc_skb_tx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); } return rot_last; @@ -443,7 +443,7 @@ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb) state = READ_ONCE(call->state); if (state >= RXRPC_CALL_COMPLETE) { - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); return; } @@ -559,7 +559,7 @@ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb) * and also rxrpc_fill_out_ack(). */ if (!terminal) - rxrpc_get_skb(skb, rxrpc_skb_rx_got); + rxrpc_get_skb(skb, rxrpc_skb_got); call->rxtx_annotations[ix] = annotation; smp_wmb(); call->rxtx_buffer[ix] = skb; @@ -620,7 +620,7 @@ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb) unlock: spin_unlock(&call->input_lock); - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); _leave(" [queued]"); } @@ -1056,7 +1056,7 @@ static void rxrpc_input_call_packet(struct rxrpc_call *call, break; } - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); no_free: _leave(""); } @@ -1119,7 +1119,7 @@ static void rxrpc_post_packet_to_local(struct rxrpc_local *local, skb_queue_tail(&local->event_queue, skb); rxrpc_queue_local(local); } else { - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); } } @@ -1134,7 +1134,7 @@ static void rxrpc_reject_packet(struct rxrpc_local *local, struct sk_buff *skb) skb_queue_tail(&local->reject_queue, skb); rxrpc_queue_local(local); } else { - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); } } @@ -1198,7 +1198,7 @@ int rxrpc_input_packet(struct sock *udp_sk, struct sk_buff *skb) if (skb->tstamp == 0) skb->tstamp = ktime_get_real(); - rxrpc_new_skb(skb, rxrpc_skb_rx_received); + rxrpc_new_skb(skb, rxrpc_skb_received); skb_pull(skb, sizeof(struct udphdr)); @@ -1215,7 +1215,7 @@ int rxrpc_input_packet(struct sock *udp_sk, struct sk_buff *skb) static int lose; if ((lose++ & 7) == 7) { trace_rxrpc_rx_lose(sp); - rxrpc_free_skb(skb, rxrpc_skb_rx_lost); + rxrpc_free_skb(skb, rxrpc_skb_lost); return 0; } } @@ -1389,7 +1389,7 @@ int rxrpc_input_packet(struct sock *udp_sk, struct sk_buff *skb) goto out; discard: - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); out: trace_rxrpc_rx_done(0, 0); return 0; diff --git a/net/rxrpc/local_event.c b/net/rxrpc/local_event.c index e93a78f7c05e..3ce6d628cd75 100644 --- a/net/rxrpc/local_event.c +++ b/net/rxrpc/local_event.c @@ -90,7 +90,7 @@ void rxrpc_process_local_events(struct rxrpc_local *local) if (skb) { struct rxrpc_skb_priv *sp = rxrpc_skb(skb); - rxrpc_see_skb(skb, rxrpc_skb_rx_seen); + rxrpc_see_skb(skb, rxrpc_skb_seen); _debug("{%d},{%u}", local->debug_id, sp->hdr.type); switch (sp->hdr.type) { @@ -108,7 +108,7 @@ void rxrpc_process_local_events(struct rxrpc_local *local) break; } - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); } _leave(""); diff --git a/net/rxrpc/output.c b/net/rxrpc/output.c index 369e516c4bdf..935bb60fff56 100644 --- a/net/rxrpc/output.c +++ b/net/rxrpc/output.c @@ -565,7 +565,7 @@ void rxrpc_reject_packets(struct rxrpc_local *local) memset(&whdr, 0, sizeof(whdr)); while ((skb = skb_dequeue(&local->reject_queue))) { - rxrpc_see_skb(skb, rxrpc_skb_rx_seen); + rxrpc_see_skb(skb, rxrpc_skb_seen); sp = rxrpc_skb(skb); switch (skb->mark) { @@ -581,7 +581,7 @@ void rxrpc_reject_packets(struct rxrpc_local *local) ioc = 2; break; default: - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); continue; } @@ -606,7 +606,7 @@ void rxrpc_reject_packets(struct rxrpc_local *local) rxrpc_tx_point_reject); } - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); } _leave(""); diff --git a/net/rxrpc/peer_event.c b/net/rxrpc/peer_event.c index 7666ec72d37e..c97ebdc043e4 100644 --- a/net/rxrpc/peer_event.c +++ b/net/rxrpc/peer_event.c @@ -163,11 +163,11 @@ void rxrpc_error_report(struct sock *sk) _leave("UDP socket errqueue empty"); return; } - rxrpc_new_skb(skb, rxrpc_skb_rx_received); + rxrpc_new_skb(skb, rxrpc_skb_received); serr = SKB_EXT_ERR(skb); if (!skb->len && serr->ee.ee_origin == SO_EE_ORIGIN_TIMESTAMPING) { _leave("UDP empty message"); - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); return; } @@ -177,7 +177,7 @@ void rxrpc_error_report(struct sock *sk) peer = NULL; if (!peer) { rcu_read_unlock(); - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); _leave(" [no peer]"); return; } @@ -189,7 +189,7 @@ void rxrpc_error_report(struct sock *sk) serr->ee.ee_code == ICMP_FRAG_NEEDED)) { rxrpc_adjust_mtu(peer, serr); rcu_read_unlock(); - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); rxrpc_put_peer(peer); _leave(" [MTU update]"); return; @@ -197,7 +197,7 @@ void rxrpc_error_report(struct sock *sk) rxrpc_store_error(peer, serr); rcu_read_unlock(); - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); rxrpc_put_peer(peer); _leave(""); diff --git a/net/rxrpc/recvmsg.c b/net/rxrpc/recvmsg.c index e49eacfaf4d6..3b0becb12041 100644 --- a/net/rxrpc/recvmsg.c +++ b/net/rxrpc/recvmsg.c @@ -190,7 +190,7 @@ static void rxrpc_rotate_rx_window(struct rxrpc_call *call) hard_ack++; ix = hard_ack & RXRPC_RXTX_BUFF_MASK; skb = call->rxtx_buffer[ix]; - rxrpc_see_skb(skb, rxrpc_skb_rx_rotated); + rxrpc_see_skb(skb, rxrpc_skb_rotated); sp = rxrpc_skb(skb); subpacket = call->rxtx_annotations[ix] & RXRPC_RX_ANNO_SUBPACKET; @@ -205,7 +205,7 @@ static void rxrpc_rotate_rx_window(struct rxrpc_call *call) /* Barrier against rxrpc_input_data(). */ smp_store_release(&call->rx_hard_ack, hard_ack); - rxrpc_free_skb(skb, rxrpc_skb_rx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); trace_rxrpc_receive(call, rxrpc_receive_rotate, serial, hard_ack); if (last) { @@ -340,7 +340,7 @@ static int rxrpc_recvmsg_data(struct socket *sock, struct rxrpc_call *call, break; } smp_rmb(); - rxrpc_see_skb(skb, rxrpc_skb_rx_seen); + rxrpc_see_skb(skb, rxrpc_skb_seen); sp = rxrpc_skb(skb); if (!(flags & MSG_PEEK)) { diff --git a/net/rxrpc/sendmsg.c b/net/rxrpc/sendmsg.c index 472dc3b7d91f..6a1547b270fe 100644 --- a/net/rxrpc/sendmsg.c +++ b/net/rxrpc/sendmsg.c @@ -176,7 +176,7 @@ static int rxrpc_queue_packet(struct rxrpc_sock *rx, struct rxrpc_call *call, skb->tstamp = ktime_get_real(); ix = seq & RXRPC_RXTX_BUFF_MASK; - rxrpc_get_skb(skb, rxrpc_skb_tx_got); + rxrpc_get_skb(skb, rxrpc_skb_got); call->rxtx_annotations[ix] = annotation; smp_wmb(); call->rxtx_buffer[ix] = skb; @@ -248,7 +248,7 @@ static int rxrpc_queue_packet(struct rxrpc_sock *rx, struct rxrpc_call *call, } out: - rxrpc_free_skb(skb, rxrpc_skb_tx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); _leave(" = %d", ret); return ret; } @@ -289,7 +289,7 @@ static int rxrpc_send_data(struct rxrpc_sock *rx, skb = call->tx_pending; call->tx_pending = NULL; - rxrpc_see_skb(skb, rxrpc_skb_tx_seen); + rxrpc_see_skb(skb, rxrpc_skb_seen); copied = 0; do { @@ -338,7 +338,7 @@ static int rxrpc_send_data(struct rxrpc_sock *rx, sp = rxrpc_skb(skb); sp->rx_flags |= RXRPC_SKB_TX_BUFFER; - rxrpc_new_skb(skb, rxrpc_skb_tx_new); + rxrpc_new_skb(skb, rxrpc_skb_new); _debug("ALLOC SEND %p", skb); @@ -440,7 +440,7 @@ static int rxrpc_send_data(struct rxrpc_sock *rx, return ret; call_terminated: - rxrpc_free_skb(skb, rxrpc_skb_tx_freed); + rxrpc_free_skb(skb, rxrpc_skb_freed); _leave(" = %d", call->error); return call->error; diff --git a/net/rxrpc/skbuff.c b/net/rxrpc/skbuff.c index 9ad5045b7c2f..8e6f45f84b9b 100644 --- a/net/rxrpc/skbuff.c +++ b/net/rxrpc/skbuff.c @@ -14,7 +14,8 @@ #include #include "ar-internal.h" -#define select_skb_count(op) (op >= rxrpc_skb_tx_cleaned ? &rxrpc_n_tx_skbs : &rxrpc_n_rx_skbs) +#define is_tx_skb(skb) (rxrpc_skb(skb)->rx_flags & RXRPC_SKB_TX_BUFFER) +#define select_skb_count(skb) (is_tx_skb(skb) ? &rxrpc_n_tx_skbs : &rxrpc_n_rx_skbs) /* * Note the allocation or reception of a socket buffer. @@ -22,7 +23,7 @@ void rxrpc_new_skb(struct sk_buff *skb, enum rxrpc_skb_trace op) { const void *here = __builtin_return_address(0); - int n = atomic_inc_return(select_skb_count(op)); + int n = atomic_inc_return(select_skb_count(skb)); trace_rxrpc_skb(skb, op, refcount_read(&skb->users), n, here); } @@ -33,7 +34,7 @@ void rxrpc_see_skb(struct sk_buff *skb, enum rxrpc_skb_trace op) { const void *here = __builtin_return_address(0); if (skb) { - int n = atomic_read(select_skb_count(op)); + int n = atomic_read(select_skb_count(skb)); trace_rxrpc_skb(skb, op, refcount_read(&skb->users), n, here); } } @@ -44,7 +45,7 @@ void rxrpc_see_skb(struct sk_buff *skb, enum rxrpc_skb_trace op) void rxrpc_get_skb(struct sk_buff *skb, enum rxrpc_skb_trace op) { const void *here = __builtin_return_address(0); - int n = atomic_inc_return(select_skb_count(op)); + int n = atomic_inc_return(select_skb_count(skb)); trace_rxrpc_skb(skb, op, refcount_read(&skb->users), n, here); skb_get(skb); } @@ -58,7 +59,7 @@ void rxrpc_free_skb(struct sk_buff *skb, enum rxrpc_skb_trace op) if (skb) { int n; CHECK_SLAB_OKAY(&skb->users); - n = atomic_dec_return(select_skb_count(op)); + n = atomic_dec_return(select_skb_count(skb)); trace_rxrpc_skb(skb, op, refcount_read(&skb->users), n, here); kfree_skb(skb); } @@ -72,8 +73,8 @@ void rxrpc_purge_queue(struct sk_buff_head *list) const void *here = __builtin_return_address(0); struct sk_buff *skb; while ((skb = skb_dequeue((list))) != NULL) { - int n = atomic_dec_return(select_skb_count(rxrpc_skb_rx_purged)); - trace_rxrpc_skb(skb, rxrpc_skb_rx_purged, + int n = atomic_dec_return(select_skb_count(skb)); + trace_rxrpc_skb(skb, rxrpc_skb_purged, refcount_read(&skb->users), n, here); kfree_skb(skb); } From patchwork Thu Aug 29 13:08:22 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 1155256 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=fail (p=none dis=none) header.from=redhat.com Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 46K2wr0Yfwz9sML for ; Thu, 29 Aug 2019 23:08:28 +1000 (AEST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728216AbfH2NIZ (ORCPT ); Thu, 29 Aug 2019 09:08:25 -0400 Received: from mx1.redhat.com ([209.132.183.28]:19898 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727285AbfH2NIY (ORCPT ); Thu, 29 Aug 2019 09:08:24 -0400 Received: from smtp.corp.redhat.com (int-mx06.intmail.prod.int.phx2.redhat.com [10.5.11.16]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id D63173078A23; Thu, 29 Aug 2019 13:08:23 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-120-255.rdu2.redhat.com [10.10.120.255]) by smtp.corp.redhat.com (Postfix) with ESMTP id CEFE45C1D6; Thu, 29 Aug 2019 13:08:22 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [PATCH net 7/7] rxrpc: Use skb_unshare() rather than skb_cow_data() From: David Howells To: netdev@vger.kernel.org Cc: dhowells@redhat.com, linux-afs@lists.infradead.org, linux-kernel@vger.kernel.org Date: Thu, 29 Aug 2019 14:08:22 +0100 Message-ID: <156708410208.26102.7491617612573289871.stgit@warthog.procyon.org.uk> In-Reply-To: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> References: <156708405310.26102.7954021163316252673.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.16 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.48]); Thu, 29 Aug 2019 13:08:23 +0000 (UTC) Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org The in-place decryption routines in AF_RXRPC's rxkad security module currently call skb_cow_data() to make sure the data isn't shared and that the skb can be written over. This has a problem, however, as the softirq handler may be still holding a ref or the Rx ring may be holding multiple refs when skb_cow_data() is called in rxkad_verify_packet() - and so skb_shared() returns true and __pskb_pull_tail() dislikes that. If this occurs, something like the following report will be generated. kernel BUG at net/core/skbuff.c:1463! ... RIP: 0010:pskb_expand_head+0x253/0x2b0 ... Call Trace: __pskb_pull_tail+0x49/0x460 skb_cow_data+0x6f/0x300 rxkad_verify_packet+0x18b/0xb10 [rxrpc] rxrpc_recvmsg_data.isra.11+0x4a8/0xa10 [rxrpc] rxrpc_kernel_recv_data+0x126/0x240 [rxrpc] afs_extract_data+0x51/0x2d0 [kafs] afs_deliver_fs_fetch_data+0x188/0x400 [kafs] afs_deliver_to_call+0xac/0x430 [kafs] afs_wait_for_call_to_complete+0x22f/0x3d0 [kafs] afs_make_call+0x282/0x3f0 [kafs] afs_fs_fetch_data+0x164/0x300 [kafs] afs_fetch_data+0x54/0x130 [kafs] afs_readpages+0x20d/0x340 [kafs] read_pages+0x66/0x180 __do_page_cache_readahead+0x188/0x1a0 ondemand_readahead+0x17d/0x2e0 generic_file_read_iter+0x740/0xc10 __vfs_read+0x145/0x1a0 vfs_read+0x8c/0x140 ksys_read+0x4a/0xb0 do_syscall_64+0x43/0xf0 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Fix this by using skb_unshare() instead in the input path for DATA packets that have a security index != 0. Non-DATA packets don't need in-place encryption and neither do unencrypted DATA packets. Fixes: 248f219cb8bc ("rxrpc: Rewrite the data and ack handling code") Reported-by: Julian Wollrath Signed-off-by: David Howells --- include/trace/events/rxrpc.h | 12 ++++++++---- net/rxrpc/ar-internal.h | 1 + net/rxrpc/input.c | 18 ++++++++++++++++++ net/rxrpc/rxkad.c | 32 +++++++++----------------------- net/rxrpc/skbuff.c | 25 ++++++++++++++++++++----- 5 files changed, 56 insertions(+), 32 deletions(-) diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index e2356c51883b..a13a62db3565 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -32,6 +32,8 @@ enum rxrpc_skb_trace { rxrpc_skb_received, rxrpc_skb_rotated, rxrpc_skb_seen, + rxrpc_skb_unshared, + rxrpc_skb_unshared_nomem, }; enum rxrpc_local_trace { @@ -231,7 +233,9 @@ enum rxrpc_tx_point { EM(rxrpc_skb_purged, "PUR") \ EM(rxrpc_skb_received, "RCV") \ EM(rxrpc_skb_rotated, "ROT") \ - E_(rxrpc_skb_seen, "SEE") + EM(rxrpc_skb_seen, "SEE") \ + EM(rxrpc_skb_unshared, "UNS") \ + E_(rxrpc_skb_unshared_nomem, "US0") #define rxrpc_local_traces \ EM(rxrpc_local_got, "GOT") \ @@ -633,9 +637,9 @@ TRACE_EVENT(rxrpc_call, TRACE_EVENT(rxrpc_skb, TP_PROTO(struct sk_buff *skb, enum rxrpc_skb_trace op, - int usage, int mod_count, const void *where), + int usage, int mod_count, u8 flags, const void *where), - TP_ARGS(skb, op, usage, mod_count, where), + TP_ARGS(skb, op, usage, mod_count, flags, where), TP_STRUCT__entry( __field(struct sk_buff *, skb ) @@ -648,7 +652,7 @@ TRACE_EVENT(rxrpc_skb, TP_fast_assign( __entry->skb = skb; - __entry->flags = rxrpc_skb(skb)->rx_flags; + __entry->flags = flags; __entry->op = op; __entry->usage = usage; __entry->mod_count = mod_count; diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index 2d5294f3e62f..852e58781fda 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -1110,6 +1110,7 @@ void rxrpc_kernel_data_consumed(struct rxrpc_call *, struct sk_buff *); void rxrpc_packet_destructor(struct sk_buff *); void rxrpc_new_skb(struct sk_buff *, enum rxrpc_skb_trace); void rxrpc_see_skb(struct sk_buff *, enum rxrpc_skb_trace); +void rxrpc_eaten_skb(struct sk_buff *, enum rxrpc_skb_trace); void rxrpc_get_skb(struct sk_buff *, enum rxrpc_skb_trace); void rxrpc_free_skb(struct sk_buff *, enum rxrpc_skb_trace); void rxrpc_purge_queue(struct sk_buff_head *); diff --git a/net/rxrpc/input.c b/net/rxrpc/input.c index 31090bdf1fae..d122c53c8697 100644 --- a/net/rxrpc/input.c +++ b/net/rxrpc/input.c @@ -1249,6 +1249,24 @@ int rxrpc_input_packet(struct sock *udp_sk, struct sk_buff *skb) goto bad_message; if (!rxrpc_validate_data(skb)) goto bad_message; + + /* Unshare the packet so that it can be modified for in-place + * decryption. + */ + if (sp->hdr.securityIndex != 0) { + struct sk_buff *nskb = skb_unshare(skb, GFP_ATOMIC); + if (!nskb) { + rxrpc_eaten_skb(skb, rxrpc_skb_unshared_nomem); + goto out; + } + + if (nskb != skb) { + rxrpc_eaten_skb(skb, rxrpc_skb_received); + rxrpc_new_skb(skb, rxrpc_skb_unshared); + skb = nskb; + sp = rxrpc_skb(skb); + } + } break; case RXRPC_PACKET_TYPE_CHALLENGE: diff --git a/net/rxrpc/rxkad.c b/net/rxrpc/rxkad.c index ae8cd8926456..c60c520fde7c 100644 --- a/net/rxrpc/rxkad.c +++ b/net/rxrpc/rxkad.c @@ -187,10 +187,8 @@ static int rxkad_secure_packet_encrypt(const struct rxrpc_call *call, struct rxrpc_skb_priv *sp; struct rxrpc_crypt iv; struct scatterlist sg[16]; - struct sk_buff *trailer; unsigned int len; u16 check; - int nsg; int err; sp = rxrpc_skb(skb); @@ -214,15 +212,14 @@ static int rxkad_secure_packet_encrypt(const struct rxrpc_call *call, crypto_skcipher_encrypt(req); /* we want to encrypt the skbuff in-place */ - nsg = skb_cow_data(skb, 0, &trailer); - err = -ENOMEM; - if (nsg < 0 || nsg > 16) + err = -EMSGSIZE; + if (skb_shinfo(skb)->nr_frags > 16) goto out; len = data_size + call->conn->size_align - 1; len &= ~(call->conn->size_align - 1); - sg_init_table(sg, nsg); + sg_init_table(sg, ARRAY_SIZE(sg)); err = skb_to_sgvec(skb, sg, 0, len); if (unlikely(err < 0)) goto out; @@ -319,11 +316,10 @@ static int rxkad_verify_packet_1(struct rxrpc_call *call, struct sk_buff *skb, struct rxkad_level1_hdr sechdr; struct rxrpc_crypt iv; struct scatterlist sg[16]; - struct sk_buff *trailer; bool aborted; u32 data_size, buf; u16 check; - int nsg, ret; + int ret; _enter(""); @@ -336,11 +332,7 @@ static int rxkad_verify_packet_1(struct rxrpc_call *call, struct sk_buff *skb, /* Decrypt the skbuff in-place. TODO: We really want to decrypt * directly into the target buffer. */ - nsg = skb_cow_data(skb, 0, &trailer); - if (nsg < 0 || nsg > 16) - goto nomem; - - sg_init_table(sg, nsg); + sg_init_table(sg, ARRAY_SIZE(sg)); ret = skb_to_sgvec(skb, sg, offset, 8); if (unlikely(ret < 0)) return ret; @@ -388,10 +380,6 @@ static int rxkad_verify_packet_1(struct rxrpc_call *call, struct sk_buff *skb, if (aborted) rxrpc_send_abort_packet(call); return -EPROTO; - -nomem: - _leave(" = -ENOMEM"); - return -ENOMEM; } /* @@ -406,7 +394,6 @@ static int rxkad_verify_packet_2(struct rxrpc_call *call, struct sk_buff *skb, struct rxkad_level2_hdr sechdr; struct rxrpc_crypt iv; struct scatterlist _sg[4], *sg; - struct sk_buff *trailer; bool aborted; u32 data_size, buf; u16 check; @@ -423,12 +410,11 @@ static int rxkad_verify_packet_2(struct rxrpc_call *call, struct sk_buff *skb, /* Decrypt the skbuff in-place. TODO: We really want to decrypt * directly into the target buffer. */ - nsg = skb_cow_data(skb, 0, &trailer); - if (nsg < 0) - goto nomem; - sg = _sg; - if (unlikely(nsg > 4)) { + nsg = skb_shinfo(skb)->nr_frags; + if (nsg <= 4) { + nsg = 4; + } else { sg = kmalloc_array(nsg, sizeof(*sg), GFP_NOIO); if (!sg) goto nomem; diff --git a/net/rxrpc/skbuff.c b/net/rxrpc/skbuff.c index 8e6f45f84b9b..0348d2bf6f7d 100644 --- a/net/rxrpc/skbuff.c +++ b/net/rxrpc/skbuff.c @@ -24,7 +24,8 @@ void rxrpc_new_skb(struct sk_buff *skb, enum rxrpc_skb_trace op) { const void *here = __builtin_return_address(0); int n = atomic_inc_return(select_skb_count(skb)); - trace_rxrpc_skb(skb, op, refcount_read(&skb->users), n, here); + trace_rxrpc_skb(skb, op, refcount_read(&skb->users), n, + rxrpc_skb(skb)->rx_flags, here); } /* @@ -35,7 +36,8 @@ void rxrpc_see_skb(struct sk_buff *skb, enum rxrpc_skb_trace op) const void *here = __builtin_return_address(0); if (skb) { int n = atomic_read(select_skb_count(skb)); - trace_rxrpc_skb(skb, op, refcount_read(&skb->users), n, here); + trace_rxrpc_skb(skb, op, refcount_read(&skb->users), n, + rxrpc_skb(skb)->rx_flags, here); } } @@ -46,10 +48,21 @@ void rxrpc_get_skb(struct sk_buff *skb, enum rxrpc_skb_trace op) { const void *here = __builtin_return_address(0); int n = atomic_inc_return(select_skb_count(skb)); - trace_rxrpc_skb(skb, op, refcount_read(&skb->users), n, here); + trace_rxrpc_skb(skb, op, refcount_read(&skb->users), n, + rxrpc_skb(skb)->rx_flags, here); skb_get(skb); } +/* + * Note the dropping of a ref on a socket buffer by the core. + */ +void rxrpc_eaten_skb(struct sk_buff *skb, enum rxrpc_skb_trace op) +{ + const void *here = __builtin_return_address(0); + int n = atomic_inc_return(&rxrpc_n_rx_skbs); + trace_rxrpc_skb(skb, op, 0, n, 0, here); +} + /* * Note the destruction of a socket buffer. */ @@ -60,7 +73,8 @@ void rxrpc_free_skb(struct sk_buff *skb, enum rxrpc_skb_trace op) int n; CHECK_SLAB_OKAY(&skb->users); n = atomic_dec_return(select_skb_count(skb)); - trace_rxrpc_skb(skb, op, refcount_read(&skb->users), n, here); + trace_rxrpc_skb(skb, op, refcount_read(&skb->users), n, + rxrpc_skb(skb)->rx_flags, here); kfree_skb(skb); } } @@ -75,7 +89,8 @@ void rxrpc_purge_queue(struct sk_buff_head *list) while ((skb = skb_dequeue((list))) != NULL) { int n = atomic_dec_return(select_skb_count(skb)); trace_rxrpc_skb(skb, rxrpc_skb_purged, - refcount_read(&skb->users), n, here); + refcount_read(&skb->users), n, + rxrpc_skb(skb)->rx_flags, here); kfree_skb(skb); } }