From 40570375356c874b1578e05c1dcc3ff7c1322dbe Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 5 Apr 2022 16:35:38 -0700 Subject: tcp: add accessors to read/set tp->snd_cwnd We had various bugs over the years with code breaking the assumption that tp->snd_cwnd is greater than zero. Lately, syzbot reported the WARN_ON_ONCE(!tp->prior_cwnd) added in commit 8b8a321ff72c ("tcp: fix zero cwnd in tcp_cwnd_reduction") can trigger, and without a repro we would have to spend considerable time finding the bug. Instead of complaining too late, we want to catch where and when tp->snd_cwnd is set to an illegal value. Signed-off-by: Eric Dumazet Suggested-by: Yuchung Cheng Cc: Neal Cardwell Acked-by: Yuchung Cheng Link: https://lore.kernel.org/r/20220405233538.947344-1-eric.dumazet@gmail.com Signed-off-by: Jakub Kicinski --- include/trace/events/tcp.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/trace') diff --git a/include/trace/events/tcp.h b/include/trace/events/tcp.h index 521059d8dc0a..edcd6369de10 100644 --- a/include/trace/events/tcp.h +++ b/include/trace/events/tcp.h @@ -279,7 +279,7 @@ TRACE_EVENT(tcp_probe, __entry->data_len = skb->len - __tcp_hdrlen(th); __entry->snd_nxt = tp->snd_nxt; __entry->snd_una = tp->snd_una; - __entry->snd_cwnd = tp->snd_cwnd; + __entry->snd_cwnd = tcp_snd_cwnd(tp); __entry->snd_wnd = tp->snd_wnd; __entry->rcv_wnd = tp->rcv_wnd; __entry->ssthresh = tcp_current_ssthresh(sk); -- cgit v1.3-8-gc7d7 From 15fcdf6ae116d1e8da4ff76c6fd82514f5ea501b Mon Sep 17 00:00:00 2001 From: Ping Gan Date: Wed, 6 Apr 2022 09:09:56 +0800 Subject: tcp: Add tracepoint for tcp_set_ca_state The congestion status of a tcp flow may be updated since there is congestion between tcp sender and receiver. It makes sense to add tracepoint for congestion status set function to summate cc status duration and evaluate the performance of network and congestion algorithm. the backgound of this patch is below. Link: https://github.com/iovisor/bcc/pull/3899 Signed-off-by: Ping Gan Reviewed-by: Eric Dumazet Link: https://lore.kernel.org/r/20220406010956.19656-1-jacky_gam_2001@163.com Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 12 +++--------- include/trace/events/tcp.h | 45 +++++++++++++++++++++++++++++++++++++++++++++ net/ipv4/tcp_cong.c | 12 ++++++++++++ 3 files changed, 60 insertions(+), 9 deletions(-) (limited to 'include/trace') diff --git a/include/net/tcp.h b/include/net/tcp.h index 54aa2f02bb63..6d50a662bf89 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1139,15 +1139,6 @@ static inline bool tcp_ca_needs_ecn(const struct sock *sk) return icsk->icsk_ca_ops->flags & TCP_CONG_NEEDS_ECN; } -static inline void tcp_set_ca_state(struct sock *sk, const u8 ca_state) -{ - struct inet_connection_sock *icsk = inet_csk(sk); - - if (icsk->icsk_ca_ops->set_state) - icsk->icsk_ca_ops->set_state(sk, ca_state); - icsk->icsk_ca_state = ca_state; -} - static inline void tcp_ca_event(struct sock *sk, const enum tcp_ca_event event) { const struct inet_connection_sock *icsk = inet_csk(sk); @@ -1156,6 +1147,9 @@ static inline void tcp_ca_event(struct sock *sk, const enum tcp_ca_event event) icsk->icsk_ca_ops->cwnd_event(sk, event); } +/* From tcp_cong.c */ +void tcp_set_ca_state(struct sock *sk, const u8 ca_state); + /* From tcp_rate.c */ void tcp_rate_skb_sent(struct sock *sk, struct sk_buff *skb); void tcp_rate_skb_delivered(struct sock *sk, struct sk_buff *skb, diff --git a/include/trace/events/tcp.h b/include/trace/events/tcp.h index edcd6369de10..901b440238d5 100644 --- a/include/trace/events/tcp.h +++ b/include/trace/events/tcp.h @@ -371,6 +371,51 @@ DEFINE_EVENT(tcp_event_skb, tcp_bad_csum, TP_ARGS(skb) ); +TRACE_EVENT(tcp_cong_state_set, + + TP_PROTO(struct sock *sk, const u8 ca_state), + + TP_ARGS(sk, ca_state), + + TP_STRUCT__entry( + __field(const void *, skaddr) + __field(__u16, sport) + __field(__u16, dport) + __array(__u8, saddr, 4) + __array(__u8, daddr, 4) + __array(__u8, saddr_v6, 16) + __array(__u8, daddr_v6, 16) + __field(__u8, cong_state) + ), + + TP_fast_assign( + struct inet_sock *inet = inet_sk(sk); + __be32 *p32; + + __entry->skaddr = sk; + + __entry->sport = ntohs(inet->inet_sport); + __entry->dport = ntohs(inet->inet_dport); + + p32 = (__be32 *) __entry->saddr; + *p32 = inet->inet_saddr; + + p32 = (__be32 *) __entry->daddr; + *p32 = inet->inet_daddr; + + TP_STORE_ADDRS(__entry, inet->inet_saddr, inet->inet_daddr, + sk->sk_v6_rcv_saddr, sk->sk_v6_daddr); + + __entry->cong_state = ca_state; + ), + + TP_printk("sport=%hu dport=%hu saddr=%pI4 daddr=%pI4 saddrv6=%pI6c daddrv6=%pI6c cong_state=%u", + __entry->sport, __entry->dport, + __entry->saddr, __entry->daddr, + __entry->saddr_v6, __entry->daddr_v6, + __entry->cong_state) +); + #endif /* _TRACE_TCP_H */ /* This part must be outside protection */ diff --git a/net/ipv4/tcp_cong.c b/net/ipv4/tcp_cong.c index d854bcfb9906..d3cae40749e8 100644 --- a/net/ipv4/tcp_cong.c +++ b/net/ipv4/tcp_cong.c @@ -16,6 +16,7 @@ #include #include #include +#include static DEFINE_SPINLOCK(tcp_cong_list_lock); static LIST_HEAD(tcp_cong_list); @@ -33,6 +34,17 @@ struct tcp_congestion_ops *tcp_ca_find(const char *name) return NULL; } +void tcp_set_ca_state(struct sock *sk, const u8 ca_state) +{ + struct inet_connection_sock *icsk = inet_csk(sk); + + trace_tcp_cong_state_set(sk, ca_state); + + if (icsk->icsk_ca_ops->set_state) + icsk->icsk_ca_ops->set_state(sk, ca_state); + icsk->icsk_ca_state = ca_state; +} + /* Must be called with rcu lock held */ static struct tcp_congestion_ops *tcp_ca_find_autoload(struct net *net, const char *name) -- cgit v1.3-8-gc7d7 From 9f8ed577c28813410614b418bad42285840c1a00 Mon Sep 17 00:00:00 2001 From: Menglong Dong Date: Thu, 7 Apr 2022 14:20:50 +0800 Subject: net: skb: rename SKB_DROP_REASON_PTYPE_ABSENT As David Ahern suggested, the reasons for skb drops should be more general and not be code based. Therefore, rename SKB_DROP_REASON_PTYPE_ABSENT to SKB_DROP_REASON_UNHANDLED_PROTO, which is used for the cases of no L3 protocol handler, no L4 protocol handler, version extensions, etc. From previous discussion, now we have the aim to make these reasons more abstract and users based, avoiding code based. Signed-off-by: Menglong Dong Reviewed-by: David Ahern Signed-off-by: David S. Miller --- include/linux/skbuff.h | 8 +++----- include/trace/events/skb.h | 2 +- net/core/dev.c | 8 +++----- 3 files changed, 7 insertions(+), 11 deletions(-) (limited to 'include/trace') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 2394441fa3dd..173bc35a10a3 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -408,11 +408,9 @@ enum skb_drop_reason { */ SKB_DROP_REASON_XDP, /* dropped by XDP in input path */ SKB_DROP_REASON_TC_INGRESS, /* dropped in TC ingress HOOK */ - SKB_DROP_REASON_PTYPE_ABSENT, /* not packet_type found to handle - * the skb. For an etner packet, - * this means that L3 protocol is - * not supported - */ + SKB_DROP_REASON_UNHANDLED_PROTO, /* protocol not implemented + * or not supported + */ SKB_DROP_REASON_SKB_CSUM, /* sk_buff checksum computation * error */ diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h index e1670e1e4934..85abd7cbd221 100644 --- a/include/trace/events/skb.h +++ b/include/trace/events/skb.h @@ -50,7 +50,7 @@ EM(SKB_DROP_REASON_CPU_BACKLOG, CPU_BACKLOG) \ EM(SKB_DROP_REASON_XDP, XDP) \ EM(SKB_DROP_REASON_TC_INGRESS, TC_INGRESS) \ - EM(SKB_DROP_REASON_PTYPE_ABSENT, PTYPE_ABSENT) \ + EM(SKB_DROP_REASON_UNHANDLED_PROTO, UNHANDLED_PROTO) \ EM(SKB_DROP_REASON_SKB_CSUM, SKB_CSUM) \ EM(SKB_DROP_REASON_SKB_GSO_SEG, SKB_GSO_SEG) \ EM(SKB_DROP_REASON_SKB_UCOPY_FAULT, SKB_UCOPY_FAULT) \ diff --git a/net/core/dev.c b/net/core/dev.c index e027410e861b..ba853e878007 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5375,13 +5375,11 @@ check_vlan_id: *ppt_prev = pt_prev; } else { drop: - if (!deliver_exact) { + if (!deliver_exact) dev_core_stats_rx_dropped_inc(skb->dev); - kfree_skb_reason(skb, SKB_DROP_REASON_PTYPE_ABSENT); - } else { + else dev_core_stats_rx_nohandler_inc(skb->dev); - kfree_skb(skb); - } + kfree_skb_reason(skb, SKB_DROP_REASON_UNHANDLED_PROTO); /* Jamal, now you will not able to escape explaining * me how you were going to use this. :-) */ -- cgit v1.3-8-gc7d7 From b384c95a861eebf47e88695cf6a29f34e0b10b0f Mon Sep 17 00:00:00 2001 From: Menglong Dong Date: Thu, 7 Apr 2022 14:20:52 +0800 Subject: net: icmp: add skb drop reasons to icmp protocol Replace kfree_skb() used in icmp_rcv() and icmpv6_rcv() with kfree_skb_reason(). In order to get the reasons of the skb drops after icmp message handle, we change the return type of 'handler()' in 'struct icmp_control' from 'bool' to 'enum skb_drop_reason'. This may change its original intention, as 'false' means failure, but 'SKB_NOT_DROPPED_YET' means success now. Therefore, all 'handler' and the call of them need to be handled. Following 'handler' functions are involved: icmp_unreach() icmp_redirect() icmp_echo() icmp_timestamp() icmp_discard() And following new drop reasons are added: SKB_DROP_REASON_ICMP_CSUM SKB_DROP_REASON_INVALID_PROTO The reason 'INVALID_PROTO' is introduced for the case that the packet doesn't follow rfc 1122 and is dropped. This is not a common case, and I believe we can locate the problem from the data in the packet. For now, this 'INVALID_PROTO' is used for the icmp broadcasts with wrong types. Maybe there should be a document file for these reasons. For example, list all the case that causes the 'UNHANDLED_PROTO' and 'INVALID_PROTO' drop reason. Therefore, users can locate their problems according to the document. Reviewed-by: Hao Peng Reviewed-by: Jiang Biao Signed-off-by: Menglong Dong Reviewed-by: David Ahern Signed-off-by: David S. Miller --- include/linux/skbuff.h | 5 ++++ include/net/ping.h | 2 +- include/trace/events/skb.h | 2 ++ net/ipv4/icmp.c | 75 +++++++++++++++++++++++++++------------------- net/ipv4/ping.c | 14 +++++---- net/ipv6/icmp.c | 24 +++++++++------ 6 files changed, 75 insertions(+), 47 deletions(-) (limited to 'include/trace') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 173bc35a10a3..9b81ba497665 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -442,6 +442,11 @@ enum skb_drop_reason { SKB_DROP_REASON_TAP_TXFILTER, /* dropped by tx filter implemented * at tun/tap, e.g., check_filter() */ + SKB_DROP_REASON_ICMP_CSUM, /* ICMP checksum error */ + SKB_DROP_REASON_INVALID_PROTO, /* the packet doesn't follow RFC + * 2211, such as a broadcasts + * ICMP_TIMESTAMP + */ SKB_DROP_REASON_MAX, }; diff --git a/include/net/ping.h b/include/net/ping.h index 2fe78874318c..b68fbfdb606f 100644 --- a/include/net/ping.h +++ b/include/net/ping.h @@ -76,7 +76,7 @@ int ping_recvmsg(struct sock *sk, struct msghdr *msg, size_t len, int noblock, int ping_common_sendmsg(int family, struct msghdr *msg, size_t len, void *user_icmph, size_t icmph_len); int ping_queue_rcv_skb(struct sock *sk, struct sk_buff *skb); -bool ping_rcv(struct sk_buff *skb); +enum skb_drop_reason ping_rcv(struct sk_buff *skb); #ifdef CONFIG_PROC_FS void *ping_seq_start(struct seq_file *seq, loff_t *pos, sa_family_t family); diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h index 85abd7cbd221..42647114fffe 100644 --- a/include/trace/events/skb.h +++ b/include/trace/events/skb.h @@ -61,6 +61,8 @@ EM(SKB_DROP_REASON_HDR_TRUNC, HDR_TRUNC) \ EM(SKB_DROP_REASON_TAP_FILTER, TAP_FILTER) \ EM(SKB_DROP_REASON_TAP_TXFILTER, TAP_TXFILTER) \ + EM(SKB_DROP_REASON_ICMP_CSUM, ICMP_CSUM) \ + EM(SKB_DROP_REASON_INVALID_PROTO, INVALID_PROTO) \ EMe(SKB_DROP_REASON_MAX, MAX) #undef EM diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index 72a375c7f417..236debd9fded 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -186,7 +186,7 @@ EXPORT_SYMBOL(icmp_err_convert); */ struct icmp_control { - bool (*handler)(struct sk_buff *skb); + enum skb_drop_reason (*handler)(struct sk_buff *skb); short error; /* This ICMP is classed as an error message */ }; @@ -839,8 +839,9 @@ static bool icmp_tag_validation(int proto) * ICMP_PARAMETERPROB. */ -static bool icmp_unreach(struct sk_buff *skb) +static enum skb_drop_reason icmp_unreach(struct sk_buff *skb) { + enum skb_drop_reason reason = SKB_NOT_DROPPED_YET; const struct iphdr *iph; struct icmphdr *icmph; struct net *net; @@ -860,8 +861,10 @@ static bool icmp_unreach(struct sk_buff *skb) icmph = icmp_hdr(skb); iph = (const struct iphdr *)skb->data; - if (iph->ihl < 5) /* Mangled header, drop. */ + if (iph->ihl < 5) { /* Mangled header, drop. */ + reason = SKB_DROP_REASON_IP_INHDR; goto out_err; + } switch (icmph->type) { case ICMP_DEST_UNREACH: @@ -941,10 +944,10 @@ static bool icmp_unreach(struct sk_buff *skb) icmp_socket_deliver(skb, info); out: - return true; + return reason; out_err: __ICMP_INC_STATS(net, ICMP_MIB_INERRORS); - return false; + return reason ?: SKB_DROP_REASON_NOT_SPECIFIED; } @@ -952,20 +955,20 @@ out_err: * Handle ICMP_REDIRECT. */ -static bool icmp_redirect(struct sk_buff *skb) +static enum skb_drop_reason icmp_redirect(struct sk_buff *skb) { if (skb->len < sizeof(struct iphdr)) { __ICMP_INC_STATS(dev_net(skb->dev), ICMP_MIB_INERRORS); - return false; + return SKB_DROP_REASON_PKT_TOO_SMALL; } if (!pskb_may_pull(skb, sizeof(struct iphdr))) { /* there aught to be a stat */ - return false; + return SKB_DROP_REASON_NOMEM; } icmp_socket_deliver(skb, ntohl(icmp_hdr(skb)->un.gateway)); - return true; + return SKB_NOT_DROPPED_YET; } /* @@ -982,7 +985,7 @@ static bool icmp_redirect(struct sk_buff *skb) * See also WRT handling of options once they are done and working. */ -static bool icmp_echo(struct sk_buff *skb) +static enum skb_drop_reason icmp_echo(struct sk_buff *skb) { struct icmp_bxm icmp_param; struct net *net; @@ -990,7 +993,7 @@ static bool icmp_echo(struct sk_buff *skb) net = dev_net(skb_dst(skb)->dev); /* should there be an ICMP stat for ignored echos? */ if (net->ipv4.sysctl_icmp_echo_ignore_all) - return true; + return SKB_NOT_DROPPED_YET; icmp_param.data.icmph = *icmp_hdr(skb); icmp_param.skb = skb; @@ -1001,10 +1004,10 @@ static bool icmp_echo(struct sk_buff *skb) if (icmp_param.data.icmph.type == ICMP_ECHO) icmp_param.data.icmph.type = ICMP_ECHOREPLY; else if (!icmp_build_probe(skb, &icmp_param.data.icmph)) - return true; + return SKB_NOT_DROPPED_YET; icmp_reply(&icmp_param, skb); - return true; + return SKB_NOT_DROPPED_YET; } /* Helper for icmp_echo and icmpv6_echo_reply. @@ -1122,7 +1125,7 @@ EXPORT_SYMBOL_GPL(icmp_build_probe); * MUST be accurate to a few minutes. * MUST be updated at least at 15Hz. */ -static bool icmp_timestamp(struct sk_buff *skb) +static enum skb_drop_reason icmp_timestamp(struct sk_buff *skb) { struct icmp_bxm icmp_param; /* @@ -1147,17 +1150,17 @@ static bool icmp_timestamp(struct sk_buff *skb) icmp_param.data_len = 0; icmp_param.head_len = sizeof(struct icmphdr) + 12; icmp_reply(&icmp_param, skb); - return true; + return SKB_NOT_DROPPED_YET; out_err: __ICMP_INC_STATS(dev_net(skb_dst(skb)->dev), ICMP_MIB_INERRORS); - return false; + return SKB_DROP_REASON_PKT_TOO_SMALL; } -static bool icmp_discard(struct sk_buff *skb) +static enum skb_drop_reason icmp_discard(struct sk_buff *skb) { /* pretend it was a success */ - return true; + return SKB_NOT_DROPPED_YET; } /* @@ -1165,18 +1168,20 @@ static bool icmp_discard(struct sk_buff *skb) */ int icmp_rcv(struct sk_buff *skb) { - struct icmphdr *icmph; + enum skb_drop_reason reason = SKB_DROP_REASON_NOT_SPECIFIED; struct rtable *rt = skb_rtable(skb); struct net *net = dev_net(rt->dst.dev); - bool success; + struct icmphdr *icmph; if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) { struct sec_path *sp = skb_sec_path(skb); int nh; if (!(sp && sp->xvec[sp->len - 1]->props.flags & - XFRM_STATE_ICMP)) + XFRM_STATE_ICMP)) { + reason = SKB_DROP_REASON_XFRM_POLICY; goto drop; + } if (!pskb_may_pull(skb, sizeof(*icmph) + sizeof(struct iphdr))) goto drop; @@ -1184,8 +1189,11 @@ int icmp_rcv(struct sk_buff *skb) nh = skb_network_offset(skb); skb_set_network_header(skb, sizeof(*icmph)); - if (!xfrm4_policy_check_reverse(NULL, XFRM_POLICY_IN, skb)) + if (!xfrm4_policy_check_reverse(NULL, XFRM_POLICY_IN, + skb)) { + reason = SKB_DROP_REASON_XFRM_POLICY; goto drop; + } skb_set_network_header(skb, nh); } @@ -1207,13 +1215,13 @@ int icmp_rcv(struct sk_buff *skb) /* We can't use icmp_pointers[].handler() because it is an array of * size NR_ICMP_TYPES + 1 (19 elements) and PROBE has code 42. */ - success = icmp_echo(skb); - goto success_check; + reason = icmp_echo(skb); + goto reason_check; } if (icmph->type == ICMP_EXT_ECHOREPLY) { - success = ping_rcv(skb); - goto success_check; + reason = ping_rcv(skb); + goto reason_check; } /* @@ -1222,8 +1230,10 @@ int icmp_rcv(struct sk_buff *skb) * RFC 1122: 3.2.2 Unknown ICMP messages types MUST be silently * discarded. */ - if (icmph->type > NR_ICMP_TYPES) + if (icmph->type > NR_ICMP_TYPES) { + reason = SKB_DROP_REASON_UNHANDLED_PROTO; goto error; + } /* * Parse the ICMP message @@ -1239,27 +1249,30 @@ int icmp_rcv(struct sk_buff *skb) if ((icmph->type == ICMP_ECHO || icmph->type == ICMP_TIMESTAMP) && net->ipv4.sysctl_icmp_echo_ignore_broadcasts) { + reason = SKB_DROP_REASON_INVALID_PROTO; goto error; } if (icmph->type != ICMP_ECHO && icmph->type != ICMP_TIMESTAMP && icmph->type != ICMP_ADDRESS && icmph->type != ICMP_ADDRESSREPLY) { + reason = SKB_DROP_REASON_INVALID_PROTO; goto error; } } - success = icmp_pointers[icmph->type].handler(skb); -success_check: - if (success) { + reason = icmp_pointers[icmph->type].handler(skb); +reason_check: + if (!reason) { consume_skb(skb); return NET_RX_SUCCESS; } drop: - kfree_skb(skb); + kfree_skb_reason(skb, reason); return NET_RX_DROP; csum_error: + reason = SKB_DROP_REASON_ICMP_CSUM; __ICMP_INC_STATS(net, ICMP_MIB_CSUMERRORS); error: __ICMP_INC_STATS(net, ICMP_MIB_INERRORS); diff --git a/net/ipv4/ping.c b/net/ipv4/ping.c index 9878c7696acf..2897fcf71211 100644 --- a/net/ipv4/ping.c +++ b/net/ipv4/ping.c @@ -961,12 +961,12 @@ EXPORT_SYMBOL_GPL(ping_queue_rcv_skb); * All we need to do is get the socket. */ -bool ping_rcv(struct sk_buff *skb) +enum skb_drop_reason ping_rcv(struct sk_buff *skb) { + enum skb_drop_reason reason = SKB_DROP_REASON_NO_SOCKET; struct sock *sk; struct net *net = dev_net(skb->dev); struct icmphdr *icmph = icmp_hdr(skb); - bool rc = false; /* We assume the packet has already been checked by icmp_rcv */ @@ -981,15 +981,17 @@ bool ping_rcv(struct sk_buff *skb) struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC); pr_debug("rcv on socket %p\n", sk); - if (skb2 && !ping_queue_rcv_skb(sk, skb2)) - rc = true; + if (skb2) + reason = __ping_queue_rcv_skb(sk, skb2); + else + reason = SKB_DROP_REASON_NOMEM; sock_put(sk); } - if (!rc) + if (reason) pr_debug("no socket, dropping\n"); - return rc; + return reason; } EXPORT_SYMBOL_GPL(ping_rcv); diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c index e6b978ea0e87..01c8003c9fc9 100644 --- a/net/ipv6/icmp.c +++ b/net/ipv6/icmp.c @@ -864,21 +864,23 @@ out: static int icmpv6_rcv(struct sk_buff *skb) { + enum skb_drop_reason reason = SKB_DROP_REASON_NOT_SPECIFIED; struct net *net = dev_net(skb->dev); struct net_device *dev = icmp6_dev(skb); struct inet6_dev *idev = __in6_dev_get(dev); const struct in6_addr *saddr, *daddr; struct icmp6hdr *hdr; u8 type; - bool success = false; if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb)) { struct sec_path *sp = skb_sec_path(skb); int nh; if (!(sp && sp->xvec[sp->len - 1]->props.flags & - XFRM_STATE_ICMP)) + XFRM_STATE_ICMP)) { + reason = SKB_DROP_REASON_XFRM_POLICY; goto drop_no_count; + } if (!pskb_may_pull(skb, sizeof(*hdr) + sizeof(struct ipv6hdr))) goto drop_no_count; @@ -886,8 +888,11 @@ static int icmpv6_rcv(struct sk_buff *skb) nh = skb_network_offset(skb); skb_set_network_header(skb, sizeof(*hdr)); - if (!xfrm6_policy_check_reverse(NULL, XFRM_POLICY_IN, skb)) + if (!xfrm6_policy_check_reverse(NULL, XFRM_POLICY_IN, + skb)) { + reason = SKB_DROP_REASON_XFRM_POLICY; goto drop_no_count; + } skb_set_network_header(skb, nh); } @@ -924,11 +929,11 @@ static int icmpv6_rcv(struct sk_buff *skb) break; case ICMPV6_ECHO_REPLY: - success = ping_rcv(skb); + reason = ping_rcv(skb); break; case ICMPV6_EXT_ECHO_REPLY: - success = ping_rcv(skb); + reason = ping_rcv(skb); break; case ICMPV6_PKT_TOOBIG: @@ -994,19 +999,20 @@ static int icmpv6_rcv(struct sk_buff *skb) /* until the v6 path can be better sorted assume failure and * preserve the status quo behaviour for the rest of the paths to here */ - if (success) - consume_skb(skb); + if (reason) + kfree_skb_reason(skb, reason); else - kfree_skb(skb); + consume_skb(skb); return 0; csum_error: + reason = SKB_DROP_REASON_ICMP_CSUM; __ICMP6_INC_STATS(dev_net(dev), idev, ICMP6_MIB_CSUMERRORS); discard_it: __ICMP6_INC_STATS(dev_net(dev), idev, ICMP6_MIB_INERRORS); drop_no_count: - kfree_skb(skb); + kfree_skb_reason(skb, reason); return 0; } -- cgit v1.3-8-gc7d7 From c4eb664191b4a5ff6856478f903924176697719e Mon Sep 17 00:00:00 2001 From: Menglong Dong Date: Wed, 13 Apr 2022 16:15:53 +0800 Subject: net: ipv4: add skb drop reasons to ip_error() Eventually, I find out the handler function for inputting route lookup fail: ip_error(). The drop reasons we used in ip_error() are almost corresponding to IPSTATS_MIB_*, and following new reasons are introduced: SKB_DROP_REASON_IP_INADDRERRORS SKB_DROP_REASON_IP_INNOROUTES Isn't the name SKB_DROP_REASON_IP_HOSTUNREACH and SKB_DROP_REASON_IP_NETUNREACH more accurate? To make them corresponding to IPSTATS_MIB_*, we keep their name still. Signed-off-by: Menglong Dong Reviewed-by: Jiang Biao Reviewed-by: Hao Peng Signed-off-by: David S. Miller --- include/linux/skbuff.h | 6 ++++++ include/trace/events/skb.h | 2 ++ net/ipv4/route.c | 6 +++++- 3 files changed, 13 insertions(+), 1 deletion(-) (limited to 'include/trace') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 0cbd6ada957c..886e83ac4b70 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -447,6 +447,12 @@ enum skb_drop_reason { * 2211, such as a broadcasts * ICMP_TIMESTAMP */ + SKB_DROP_REASON_IP_INADDRERRORS, /* host unreachable, corresponding + * to IPSTATS_MIB_INADDRERRORS + */ + SKB_DROP_REASON_IP_INNOROUTES, /* network unreachable, corresponding + * to IPSTATS_MIB_INADDRERRORS + */ SKB_DROP_REASON_MAX, }; diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h index 42647114fffe..0acac7e5a019 100644 --- a/include/trace/events/skb.h +++ b/include/trace/events/skb.h @@ -63,6 +63,8 @@ EM(SKB_DROP_REASON_TAP_TXFILTER, TAP_TXFILTER) \ EM(SKB_DROP_REASON_ICMP_CSUM, ICMP_CSUM) \ EM(SKB_DROP_REASON_INVALID_PROTO, INVALID_PROTO) \ + EM(SKB_DROP_REASON_IP_INADDRERRORS, IP_INADDRERRORS) \ + EM(SKB_DROP_REASON_IP_INNOROUTES, IP_INNOROUTES) \ EMe(SKB_DROP_REASON_MAX, MAX) #undef EM diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 80f96170876c..e839d424b861 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -945,6 +945,7 @@ static int ip_error(struct sk_buff *skb) struct inet_peer *peer; unsigned long now; struct net *net; + SKB_DR(reason); bool send; int code; @@ -964,10 +965,12 @@ static int ip_error(struct sk_buff *skb) if (!IN_DEV_FORWARD(in_dev)) { switch (rt->dst.error) { case EHOSTUNREACH: + SKB_DR_SET(reason, IP_INADDRERRORS); __IP_INC_STATS(net, IPSTATS_MIB_INADDRERRORS); break; case ENETUNREACH: + SKB_DR_SET(reason, IP_INNOROUTES); __IP_INC_STATS(net, IPSTATS_MIB_INNOROUTES); break; } @@ -983,6 +986,7 @@ static int ip_error(struct sk_buff *skb) break; case ENETUNREACH: code = ICMP_NET_UNREACH; + SKB_DR_SET(reason, IP_INNOROUTES); __IP_INC_STATS(net, IPSTATS_MIB_INNOROUTES); break; case EACCES: @@ -1009,7 +1013,7 @@ static int ip_error(struct sk_buff *skb) if (send) icmp_send(skb, ICMP_DEST_UNREACH, code, 0); -out: kfree_skb(skb); +out: kfree_skb_reason(skb, reason); return 0; } -- cgit v1.3-8-gc7d7 From 2edc1a383fda8d2f580216292dfd9daeae691e47 Mon Sep 17 00:00:00 2001 From: Menglong Dong Date: Wed, 13 Apr 2022 16:15:55 +0800 Subject: net: ip: add skb drop reasons to ip forwarding Replace kfree_skb() which is used in ip6_forward() and ip_forward() with kfree_skb_reason(). The new drop reason 'SKB_DROP_REASON_PKT_TOO_BIG' is introduced for the case that the length of the packet exceeds MTU and can't fragment. Signed-off-by: Menglong Dong Reviewed-by: Jiang Biao Reviewed-by: Hao Peng Signed-off-by: David S. Miller --- include/linux/skbuff.h | 3 +++ include/trace/events/skb.h | 1 + net/ipv4/ip_forward.c | 13 ++++++++++--- net/ipv6/ip6_output.c | 9 ++++++--- 4 files changed, 20 insertions(+), 6 deletions(-) (limited to 'include/trace') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 886e83ac4b70..0ef11df1bc67 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -453,6 +453,9 @@ enum skb_drop_reason { SKB_DROP_REASON_IP_INNOROUTES, /* network unreachable, corresponding * to IPSTATS_MIB_INADDRERRORS */ + SKB_DROP_REASON_PKT_TOO_BIG, /* packet size is too big (maybe exceed + * the MTU) + */ SKB_DROP_REASON_MAX, }; diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h index 0acac7e5a019..2da72a9a5764 100644 --- a/include/trace/events/skb.h +++ b/include/trace/events/skb.h @@ -65,6 +65,7 @@ EM(SKB_DROP_REASON_INVALID_PROTO, INVALID_PROTO) \ EM(SKB_DROP_REASON_IP_INADDRERRORS, IP_INADDRERRORS) \ EM(SKB_DROP_REASON_IP_INNOROUTES, IP_INNOROUTES) \ + EM(SKB_DROP_REASON_PKT_TOO_BIG, PKT_TOO_BIG) \ EMe(SKB_DROP_REASON_MAX, MAX) #undef EM diff --git a/net/ipv4/ip_forward.c b/net/ipv4/ip_forward.c index 92ba3350274b..e3aa436a1bdf 100644 --- a/net/ipv4/ip_forward.c +++ b/net/ipv4/ip_forward.c @@ -90,6 +90,7 @@ int ip_forward(struct sk_buff *skb) struct rtable *rt; /* Route we use */ struct ip_options *opt = &(IPCB(skb)->opt); struct net *net; + SKB_DR(reason); /* that should never happen */ if (skb->pkt_type != PACKET_HOST) @@ -101,8 +102,10 @@ int ip_forward(struct sk_buff *skb) if (skb_warn_if_lro(skb)) goto drop; - if (!xfrm4_policy_check(NULL, XFRM_POLICY_FWD, skb)) + if (!xfrm4_policy_check(NULL, XFRM_POLICY_FWD, skb)) { + SKB_DR_SET(reason, XFRM_POLICY); goto drop; + } if (IPCB(skb)->opt.router_alert && ip_call_ra_chain(skb)) return NET_RX_SUCCESS; @@ -118,8 +121,10 @@ int ip_forward(struct sk_buff *skb) if (ip_hdr(skb)->ttl <= 1) goto too_many_hops; - if (!xfrm4_route_forward(skb)) + if (!xfrm4_route_forward(skb)) { + SKB_DR_SET(reason, XFRM_POLICY); goto drop; + } rt = skb_rtable(skb); @@ -132,6 +137,7 @@ int ip_forward(struct sk_buff *skb) IP_INC_STATS(net, IPSTATS_MIB_FRAGFAILS); icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, htonl(mtu)); + SKB_DR_SET(reason, PKT_TOO_BIG); goto drop; } @@ -169,7 +175,8 @@ too_many_hops: /* Tell the sender its packet died... */ __IP_INC_STATS(net, IPSTATS_MIB_INHDRERRORS); icmp_send(skb, ICMP_TIME_EXCEEDED, ICMP_EXC_TTL, 0); + SKB_DR_SET(reason, IP_INHDR); drop: - kfree_skb(skb); + kfree_skb_reason(skb, reason); return NET_RX_DROP; } diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index e23f058166af..3e729cee6486 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -469,6 +469,7 @@ int ip6_forward(struct sk_buff *skb) struct inet6_skb_parm *opt = IP6CB(skb); struct net *net = dev_net(dst->dev); struct inet6_dev *idev; + SKB_DR(reason); u32 mtu; idev = __in6_dev_get_safely(dev_get_by_index_rcu(net, IP6CB(skb)->iif)); @@ -518,7 +519,7 @@ int ip6_forward(struct sk_buff *skb) icmpv6_send(skb, ICMPV6_TIME_EXCEED, ICMPV6_EXC_HOPLIMIT, 0); __IP6_INC_STATS(net, idev, IPSTATS_MIB_INHDRERRORS); - kfree_skb(skb); + kfree_skb_reason(skb, SKB_DROP_REASON_IP_INHDR); return -ETIMEDOUT; } @@ -537,6 +538,7 @@ int ip6_forward(struct sk_buff *skb) if (!xfrm6_route_forward(skb)) { __IP6_INC_STATS(net, idev, IPSTATS_MIB_INDISCARDS); + SKB_DR_SET(reason, XFRM_POLICY); goto drop; } dst = skb_dst(skb); @@ -596,7 +598,7 @@ int ip6_forward(struct sk_buff *skb) __IP6_INC_STATS(net, idev, IPSTATS_MIB_INTOOBIGERRORS); __IP6_INC_STATS(net, ip6_dst_idev(dst), IPSTATS_MIB_FRAGFAILS); - kfree_skb(skb); + kfree_skb_reason(skb, SKB_DROP_REASON_PKT_TOO_BIG); return -EMSGSIZE; } @@ -618,8 +620,9 @@ int ip6_forward(struct sk_buff *skb) error: __IP6_INC_STATS(net, idev, IPSTATS_MIB_INADDRERRORS); + SKB_DR_SET(reason, IP_INADDRERRORS); drop: - kfree_skb(skb); + kfree_skb_reason(skb, reason); return -EINVAL; } -- cgit v1.3-8-gc7d7 From da40b613f89c43c58986e6f30560ad6573a4d569 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 15 Apr 2022 17:10:41 -0700 Subject: tcp: add drop reason support to tcp_validate_incoming() Creates four new drop reasons for the following cases: 1) packet being rejected by RFC 7323 PAWS check 2) packet being rejected by SEQUENCE check 3) Invalid RST packet 4) Invalid SYN packet Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/skbuff.h | 6 ++++++ include/trace/events/skb.h | 5 +++++ net/ipv4/tcp_input.c | 7 ++++++- 3 files changed, 17 insertions(+), 1 deletion(-) (limited to 'include/trace') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 0ef11df1bc67..a903da1fa0ed 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -381,6 +381,12 @@ enum skb_drop_reason { * the ofo queue, corresponding to * LINUX_MIB_TCPOFOMERGE */ + SKB_DROP_REASON_TCP_RFC7323_PAWS, /* PAWS check, corresponding to + * LINUX_MIB_PAWSESTABREJECTED + */ + SKB_DROP_REASON_TCP_INVALID_SEQUENCE, /* Not acceptable SEQ field */ + SKB_DROP_REASON_TCP_RESET, /* Invalid RST packet */ + SKB_DROP_REASON_TCP_INVALID_SYN, /* Incoming packet has unexpected SYN flag */ SKB_DROP_REASON_IP_OUTNOROUTES, /* route lookup failed */ SKB_DROP_REASON_BPF_CGROUP_EGRESS, /* dropped by * BPF_PROG_TYPE_CGROUP_SKB diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h index 2da72a9a5764..820dacd14bad 100644 --- a/include/trace/events/skb.h +++ b/include/trace/events/skb.h @@ -37,6 +37,11 @@ EM(SKB_DROP_REASON_TCP_OLD_DATA, TCP_OLD_DATA) \ EM(SKB_DROP_REASON_TCP_OVERWINDOW, TCP_OVERWINDOW) \ EM(SKB_DROP_REASON_TCP_OFOMERGE, TCP_OFOMERGE) \ + EM(SKB_DROP_REASON_TCP_RFC7323_PAWS, TCP_RFC7323_PAWS) \ + EM(SKB_DROP_REASON_TCP_INVALID_SEQUENCE, \ + TCP_INVALID_SEQUENCE) \ + EM(SKB_DROP_REASON_TCP_RESET, TCP_RESET) \ + EM(SKB_DROP_REASON_TCP_INVALID_SYN, TCP_INVALID_SYN) \ EM(SKB_DROP_REASON_IP_OUTNOROUTES, IP_OUTNOROUTES) \ EM(SKB_DROP_REASON_BPF_CGROUP_EGRESS, \ BPF_CGROUP_EGRESS) \ diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index b2d5fbef6ce3..9a1cb3f48c3f 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -5667,6 +5667,7 @@ static bool tcp_validate_incoming(struct sock *sk, struct sk_buff *skb, const struct tcphdr *th, int syn_inerr) { struct tcp_sock *tp = tcp_sk(sk); + SKB_DR(reason); /* RFC1323: H1. Apply PAWS check first. */ if (tcp_fast_parse_options(sock_net(sk), skb, th, tp) && @@ -5678,6 +5679,7 @@ static bool tcp_validate_incoming(struct sock *sk, struct sk_buff *skb, LINUX_MIB_TCPACKSKIPPEDPAWS, &tp->last_oow_ack_time)) tcp_send_dupack(sk, skb); + SKB_DR_SET(reason, TCP_RFC7323_PAWS); goto discard; } /* Reset is accepted even if it did not pass PAWS. */ @@ -5701,6 +5703,7 @@ static bool tcp_validate_incoming(struct sock *sk, struct sk_buff *skb, } else if (tcp_reset_check(sk, skb)) { goto reset; } + SKB_DR_SET(reason, TCP_INVALID_SEQUENCE); goto discard; } @@ -5743,6 +5746,7 @@ static bool tcp_validate_incoming(struct sock *sk, struct sk_buff *skb, sk->sk_state == TCP_ESTABLISHED) tcp_fastopen_active_disable(sk); tcp_send_challenge_ack(sk); + SKB_DR_SET(reason, TCP_RESET); goto discard; } @@ -5757,6 +5761,7 @@ syn_challenge: TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS); NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSYNCHALLENGE); tcp_send_challenge_ack(sk); + SKB_DR_SET(reason, TCP_INVALID_SYN); goto discard; } @@ -5765,7 +5770,7 @@ syn_challenge: return true; discard: - tcp_drop(sk, skb); + tcp_drop_reason(sk, skb, reason); return false; reset: -- cgit v1.3-8-gc7d7 From 669da7a71890b2b2a31a7e9571c0fdf1123e26ef Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 15 Apr 2022 17:10:43 -0700 Subject: tcp: add drop reasons to tcp_rcv_state_process() Add basic support for drop reasons in tcp_rcv_state_process() Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/skbuff.h | 3 +++ include/trace/events/skb.h | 3 +++ net/ipv4/tcp_input.c | 24 +++++++++++++++++------- 3 files changed, 23 insertions(+), 7 deletions(-) (limited to 'include/trace') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index a903da1fa0ed..6f1410b5ff13 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -387,6 +387,9 @@ enum skb_drop_reason { SKB_DROP_REASON_TCP_INVALID_SEQUENCE, /* Not acceptable SEQ field */ SKB_DROP_REASON_TCP_RESET, /* Invalid RST packet */ SKB_DROP_REASON_TCP_INVALID_SYN, /* Incoming packet has unexpected SYN flag */ + SKB_DROP_REASON_TCP_CLOSE, /* TCP socket in CLOSE state */ + SKB_DROP_REASON_TCP_FASTOPEN, /* dropped by FASTOPEN request socket */ + SKB_DROP_REASON_TCP_OLD_ACK, /* TCP ACK is old, but in window */ SKB_DROP_REASON_IP_OUTNOROUTES, /* route lookup failed */ SKB_DROP_REASON_BPF_CGROUP_EGRESS, /* dropped by * BPF_PROG_TYPE_CGROUP_SKB diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h index 820dacd14bad..fbe21ad038bc 100644 --- a/include/trace/events/skb.h +++ b/include/trace/events/skb.h @@ -42,6 +42,9 @@ TCP_INVALID_SEQUENCE) \ EM(SKB_DROP_REASON_TCP_RESET, TCP_RESET) \ EM(SKB_DROP_REASON_TCP_INVALID_SYN, TCP_INVALID_SYN) \ + EM(SKB_DROP_REASON_TCP_CLOSE, TCP_CLOSE) \ + EM(SKB_DROP_REASON_TCP_FASTOPEN, TCP_FASTOPEN) \ + EM(SKB_DROP_REASON_TCP_OLD_ACK, TCP_OLD_ACK) \ EM(SKB_DROP_REASON_IP_OUTNOROUTES, IP_OUTNOROUTES) \ EM(SKB_DROP_REASON_BPF_CGROUP_EGRESS, \ BPF_CGROUP_EGRESS) \ diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index f95a8368981d..85fae79c894d 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -6413,21 +6413,26 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb) struct request_sock *req; int queued = 0; bool acceptable; + SKB_DR(reason); switch (sk->sk_state) { case TCP_CLOSE: + SKB_DR_SET(reason, TCP_CLOSE); goto discard; case TCP_LISTEN: if (th->ack) return 1; - if (th->rst) + if (th->rst) { + SKB_DR_SET(reason, TCP_RESET); goto discard; - + } if (th->syn) { - if (th->fin) + if (th->fin) { + SKB_DR_SET(reason, TCP_FLAGS); goto discard; + } /* It is possible that we process SYN packets from backlog, * so we need to make sure to disable BH and RCU right there. */ @@ -6442,6 +6447,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb) consume_skb(skb); return 0; } + SKB_DR_SET(reason, TCP_FLAGS); goto discard; case TCP_SYN_SENT: @@ -6468,13 +6474,16 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb) WARN_ON_ONCE(sk->sk_state != TCP_SYN_RECV && sk->sk_state != TCP_FIN_WAIT1); - if (!tcp_check_req(sk, skb, req, true, &req_stolen)) + if (!tcp_check_req(sk, skb, req, true, &req_stolen)) { + SKB_DR_SET(reason, TCP_FASTOPEN); goto discard; + } } - if (!th->ack && !th->rst && !th->syn) + if (!th->ack && !th->rst && !th->syn) { + SKB_DR_SET(reason, TCP_FLAGS); goto discard; - + } if (!tcp_validate_incoming(sk, skb, th, 0)) return 0; @@ -6487,6 +6496,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb) if (sk->sk_state == TCP_SYN_RECV) return 1; /* send one RST */ tcp_send_challenge_ack(sk); + SKB_DR_SET(reason, TCP_OLD_ACK); goto discard; } switch (sk->sk_state) { @@ -6647,7 +6657,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb) if (!queued) { discard: - tcp_drop(sk, skb); + tcp_drop_reason(sk, skb, reason); } return 0; -- cgit v1.3-8-gc7d7 From 4b506af9c5b8de0da34097d50d9448dfb33d70c3 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 15 Apr 2022 17:10:44 -0700 Subject: tcp: add two drop reasons for tcp_ack() Add TCP_TOO_OLD_ACK and TCP_ACK_UNSENT_DATA drop reasons so that tcp_rcv_established() can report them. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/skbuff.h | 2 ++ include/trace/events/skb.h | 3 +++ net/ipv4/tcp_input.c | 7 ++++--- 3 files changed, 9 insertions(+), 3 deletions(-) (limited to 'include/trace') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 6f1410b5ff13..9ff5557b1909 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -390,6 +390,8 @@ enum skb_drop_reason { SKB_DROP_REASON_TCP_CLOSE, /* TCP socket in CLOSE state */ SKB_DROP_REASON_TCP_FASTOPEN, /* dropped by FASTOPEN request socket */ SKB_DROP_REASON_TCP_OLD_ACK, /* TCP ACK is old, but in window */ + SKB_DROP_REASON_TCP_TOO_OLD_ACK, /* TCP ACK is too old */ + SKB_DROP_REASON_TCP_ACK_UNSENT_DATA, /* TCP ACK for data we haven't sent yet */ SKB_DROP_REASON_IP_OUTNOROUTES, /* route lookup failed */ SKB_DROP_REASON_BPF_CGROUP_EGRESS, /* dropped by * BPF_PROG_TYPE_CGROUP_SKB diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h index fbe21ad038bc..eab0b09223f3 100644 --- a/include/trace/events/skb.h +++ b/include/trace/events/skb.h @@ -45,6 +45,9 @@ EM(SKB_DROP_REASON_TCP_CLOSE, TCP_CLOSE) \ EM(SKB_DROP_REASON_TCP_FASTOPEN, TCP_FASTOPEN) \ EM(SKB_DROP_REASON_TCP_OLD_ACK, TCP_OLD_ACK) \ + EM(SKB_DROP_REASON_TCP_TOO_OLD_ACK, TCP_TOO_OLD_ACK) \ + EM(SKB_DROP_REASON_TCP_ACK_UNSENT_DATA, \ + TCP_ACK_UNSENT_DATA) \ EM(SKB_DROP_REASON_IP_OUTNOROUTES, IP_OUTNOROUTES) \ EM(SKB_DROP_REASON_BPF_CGROUP_EGRESS, \ BPF_CGROUP_EGRESS) \ diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 85fae79c894d..8a68785b0405 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -3766,7 +3766,7 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag) if (before(ack, prior_snd_una - tp->max_window)) { if (!(flag & FLAG_NO_CHALLENGE_ACK)) tcp_send_challenge_ack(sk); - return -1; + return -SKB_DROP_REASON_TCP_TOO_OLD_ACK; } goto old_ack; } @@ -3775,7 +3775,7 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag) * this segment (RFC793 Section 3.9). */ if (after(ack, tp->snd_nxt)) - return -1; + return -SKB_DROP_REASON_TCP_ACK_UNSENT_DATA; if (after(ack, prior_snd_una)) { flag |= FLAG_SND_UNA_ADVANCED; @@ -5962,7 +5962,8 @@ slow_path: return; step5: - if (tcp_ack(sk, skb, FLAG_SLOWPATH | FLAG_UPDATE_TS_RECENT) < 0) + reason = tcp_ack(sk, skb, FLAG_SLOWPATH | FLAG_UPDATE_TS_RECENT); + if (reason < 0) goto discard; tcp_rcv_rtt_measure_ts(sk, skb); -- cgit v1.3-8-gc7d7 From e7c89ae4078eab24af71ba26b91642e819a4bd7f Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 15 Apr 2022 17:10:45 -0700 Subject: tcp: add drop reason support to tcp_prune_ofo_queue() Add one reason for packets dropped from OFO queue because of memory pressure. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/skbuff.h | 1 + include/trace/events/skb.h | 2 ++ net/ipv4/tcp_input.c | 3 ++- 3 files changed, 5 insertions(+), 1 deletion(-) (limited to 'include/trace') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 9ff5557b1909..ad15ad208b56 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -392,6 +392,7 @@ enum skb_drop_reason { SKB_DROP_REASON_TCP_OLD_ACK, /* TCP ACK is old, but in window */ SKB_DROP_REASON_TCP_TOO_OLD_ACK, /* TCP ACK is too old */ SKB_DROP_REASON_TCP_ACK_UNSENT_DATA, /* TCP ACK for data we haven't sent yet */ + SKB_DROP_REASON_TCP_OFO_QUEUE_PRUNE, /* pruned from TCP OFO queue */ SKB_DROP_REASON_IP_OUTNOROUTES, /* route lookup failed */ SKB_DROP_REASON_BPF_CGROUP_EGRESS, /* dropped by * BPF_PROG_TYPE_CGROUP_SKB diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h index eab0b09223f3..73d7a6e594cb 100644 --- a/include/trace/events/skb.h +++ b/include/trace/events/skb.h @@ -48,6 +48,8 @@ EM(SKB_DROP_REASON_TCP_TOO_OLD_ACK, TCP_TOO_OLD_ACK) \ EM(SKB_DROP_REASON_TCP_ACK_UNSENT_DATA, \ TCP_ACK_UNSENT_DATA) \ + EM(SKB_DROP_REASON_TCP_OFO_QUEUE_PRUNE, \ + TCP_OFO_QUEUE_PRUNE) \ EM(SKB_DROP_REASON_IP_OUTNOROUTES, IP_OUTNOROUTES) \ EM(SKB_DROP_REASON_BPF_CGROUP_EGRESS, \ BPF_CGROUP_EGRESS) \ diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 8a68785b0405..a1077adeb1b6 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -5334,7 +5334,8 @@ static bool tcp_prune_ofo_queue(struct sock *sk) prev = rb_prev(node); rb_erase(node, &tp->out_of_order_queue); goal -= rb_to_skb(node)->truesize; - tcp_drop(sk, rb_to_skb(node)); + tcp_drop_reason(sk, rb_to_skb(node), + SKB_DROP_REASON_TCP_OFO_QUEUE_PRUNE); if (!prev || goal <= 0) { sk_mem_reclaim(sk); if (atomic_read(&sk->sk_rmem_alloc) <= sk->sk_rcvbuf && -- cgit v1.3-8-gc7d7 From 8fbf195798b56e1e87f62d01be636a6425c304c2 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 15 Apr 2022 17:10:48 -0700 Subject: tcp: add drop reason support to tcp_ofo_queue() packets in OFO queue might be redundant, and dropped. tcp_drop() is no longer needed. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/linux/skbuff.h | 1 + include/trace/events/skb.h | 1 + net/ipv4/tcp_input.c | 9 ++------- 3 files changed, 4 insertions(+), 7 deletions(-) (limited to 'include/trace') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index ad15ad208b56..84d78df60453 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -393,6 +393,7 @@ enum skb_drop_reason { SKB_DROP_REASON_TCP_TOO_OLD_ACK, /* TCP ACK is too old */ SKB_DROP_REASON_TCP_ACK_UNSENT_DATA, /* TCP ACK for data we haven't sent yet */ SKB_DROP_REASON_TCP_OFO_QUEUE_PRUNE, /* pruned from TCP OFO queue */ + SKB_DROP_REASON_TCP_OFO_DROP, /* data already in receive queue */ SKB_DROP_REASON_IP_OUTNOROUTES, /* route lookup failed */ SKB_DROP_REASON_BPF_CGROUP_EGRESS, /* dropped by * BPF_PROG_TYPE_CGROUP_SKB diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h index 73d7a6e594cb..a477bf907498 100644 --- a/include/trace/events/skb.h +++ b/include/trace/events/skb.h @@ -37,6 +37,7 @@ EM(SKB_DROP_REASON_TCP_OLD_DATA, TCP_OLD_DATA) \ EM(SKB_DROP_REASON_TCP_OVERWINDOW, TCP_OVERWINDOW) \ EM(SKB_DROP_REASON_TCP_OFOMERGE, TCP_OFOMERGE) \ + EM(SKB_DROP_REASON_TCP_OFO_DROP, TCP_OFO_DROP) \ EM(SKB_DROP_REASON_TCP_RFC7323_PAWS, TCP_RFC7323_PAWS) \ EM(SKB_DROP_REASON_TCP_INVALID_SEQUENCE, \ TCP_INVALID_SEQUENCE) \ diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 339cc3d40745..cf2dc19bb8c7 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -4674,7 +4674,7 @@ static bool tcp_ooo_try_coalesce(struct sock *sk, { bool res = tcp_try_coalesce(sk, to, from, fragstolen); - /* In case tcp_drop() is called later, update to->gso_segs */ + /* In case tcp_drop_reason() is called later, update to->gso_segs */ if (res) { u32 gso_segs = max_t(u16, 1, skb_shinfo(to)->gso_segs) + max_t(u16, 1, skb_shinfo(from)->gso_segs); @@ -4691,11 +4691,6 @@ static void tcp_drop_reason(struct sock *sk, struct sk_buff *skb, kfree_skb_reason(skb, reason); } -static void tcp_drop(struct sock *sk, struct sk_buff *skb) -{ - tcp_drop_reason(sk, skb, SKB_DROP_REASON_NOT_SPECIFIED); -} - /* This one checks to see if we can put data from the * out_of_order queue into the receive_queue. */ @@ -4723,7 +4718,7 @@ static void tcp_ofo_queue(struct sock *sk) rb_erase(&skb->rbnode, &tp->out_of_order_queue); if (unlikely(!after(TCP_SKB_CB(skb)->end_seq, tp->rcv_nxt))) { - tcp_drop(sk, skb); + tcp_drop_reason(sk, skb, SKB_DROP_REASON_TCP_OFO_DROP); continue; } -- cgit v1.3-8-gc7d7 From c6af2aa9ffc9763826607bc2664ef3ea4475ed18 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Tue, 29 Mar 2022 17:27:33 +0200 Subject: swiotlb: make the swiotlb_init interface more useful Pass a boolean flag to indicate if swiotlb needs to be enabled based on the addressing needs, and replace the verbose argument with a set of flags, including one to force enable bounce buffering. Note that this patch removes the possibility to force xen-swiotlb use with the swiotlb=force parameter on the command line on x86 (arm and arm64 never supported that), but this interface will be restored shortly. Signed-off-by: Christoph Hellwig Reviewed-by: Konrad Rzeszutek Wilk Tested-by: Boris Ostrovsky --- arch/arm/mm/init.c | 6 +----- arch/arm64/mm/init.c | 6 +----- arch/ia64/mm/init.c | 4 +--- arch/mips/cavium-octeon/dma-octeon.c | 2 +- arch/mips/loongson64/dma.c | 2 +- arch/mips/sibyte/common/dma.c | 2 +- arch/powerpc/mm/mem.c | 3 ++- arch/powerpc/platforms/pseries/setup.c | 3 --- arch/riscv/mm/init.c | 8 +------- arch/s390/mm/init.c | 3 +-- arch/x86/kernel/pci-dma.c | 15 +++++++-------- drivers/xen/swiotlb-xen.c | 4 ++-- include/linux/swiotlb.h | 15 +++++++-------- include/trace/events/swiotlb.h | 29 ++++++++++------------------ kernel/dma/swiotlb.c | 35 ++++++++++++++++++---------------- 15 files changed, 55 insertions(+), 82 deletions(-) (limited to 'include/trace') diff --git a/arch/arm/mm/init.c b/arch/arm/mm/init.c index fe249ea91908..ce64bdb55a16 100644 --- a/arch/arm/mm/init.c +++ b/arch/arm/mm/init.c @@ -271,11 +271,7 @@ static void __init free_highpages(void) void __init mem_init(void) { #ifdef CONFIG_ARM_LPAE - if (swiotlb_force == SWIOTLB_FORCE || - max_pfn > arm_dma_pfn_limit) - swiotlb_init(1); - else - swiotlb_force = SWIOTLB_NO_FORCE; + swiotlb_init(max_pfn > arm_dma_pfn_limit, SWIOTLB_VERBOSE); #endif set_max_mapnr(pfn_to_page(max_pfn) - mem_map); diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c index 1e7b1550e2fc..bd4095b7fb40 100644 --- a/arch/arm64/mm/init.c +++ b/arch/arm64/mm/init.c @@ -398,11 +398,7 @@ void __init bootmem_init(void) */ void __init mem_init(void) { - if (swiotlb_force == SWIOTLB_FORCE || - max_pfn > PFN_DOWN(arm64_dma_phys_limit)) - swiotlb_init(1); - else if (!xen_swiotlb_detect()) - swiotlb_force = SWIOTLB_NO_FORCE; + swiotlb_init(max_pfn > PFN_DOWN(arm64_dma_phys_limit), SWIOTLB_VERBOSE); /* this will put all unused low memory onto the freelists */ memblock_free_all(); diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c index 5d165607bf35..3c3e15b22608 100644 --- a/arch/ia64/mm/init.c +++ b/arch/ia64/mm/init.c @@ -437,9 +437,7 @@ mem_init (void) if (iommu_detected) break; #endif -#ifdef CONFIG_SWIOTLB - swiotlb_init(1); -#endif + swiotlb_init(true, SWIOTLB_VERBOSE); } while (0); #ifdef CONFIG_FLATMEM diff --git a/arch/mips/cavium-octeon/dma-octeon.c b/arch/mips/cavium-octeon/dma-octeon.c index fb7547e21726..9fbba6a8fa4c 100644 --- a/arch/mips/cavium-octeon/dma-octeon.c +++ b/arch/mips/cavium-octeon/dma-octeon.c @@ -235,5 +235,5 @@ void __init plat_swiotlb_setup(void) #endif swiotlb_adjust_size(swiotlbsize); - swiotlb_init(1); + swiotlb_init(true, SWIOTLB_VERBOSE); } diff --git a/arch/mips/loongson64/dma.c b/arch/mips/loongson64/dma.c index 364f2f27c872..8220a1bc0db6 100644 --- a/arch/mips/loongson64/dma.c +++ b/arch/mips/loongson64/dma.c @@ -24,5 +24,5 @@ phys_addr_t dma_to_phys(struct device *dev, dma_addr_t daddr) void __init plat_swiotlb_setup(void) { - swiotlb_init(1); + swiotlb_init(true, SWIOTLB_VERBOSE); } diff --git a/arch/mips/sibyte/common/dma.c b/arch/mips/sibyte/common/dma.c index eb47a94f3583..c5c2c782aff6 100644 --- a/arch/mips/sibyte/common/dma.c +++ b/arch/mips/sibyte/common/dma.c @@ -10,5 +10,5 @@ void __init plat_swiotlb_setup(void) { - swiotlb_init(1); + swiotlb_init(true, SWIOTLB_VERBOSE); } diff --git a/arch/powerpc/mm/mem.c b/arch/powerpc/mm/mem.c index 4d221d033804..74ca516c3e7e 100644 --- a/arch/powerpc/mm/mem.c +++ b/arch/powerpc/mm/mem.c @@ -17,6 +17,7 @@ #include #include +#include #include #include #include @@ -251,7 +252,7 @@ void __init mem_init(void) if (is_secure_guest()) svm_swiotlb_init(); else - swiotlb_init(0); + swiotlb_init(ppc_swiotlb_enable, 0); #endif high_memory = (void *) __va(max_low_pfn * PAGE_SIZE); diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c index 955ff8aa1644..0f74b2284773 100644 --- a/arch/powerpc/platforms/pseries/setup.c +++ b/arch/powerpc/platforms/pseries/setup.c @@ -849,9 +849,6 @@ static void __init pSeries_setup_arch(void) } ppc_md.pcibios_root_bridge_prepare = pseries_root_bridge_prepare; - - if (swiotlb_force == SWIOTLB_FORCE) - ppc_swiotlb_enable = 1; } static void pseries_panic(char *str) diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c index 9535bea8688c..181ffd322eaf 100644 --- a/arch/riscv/mm/init.c +++ b/arch/riscv/mm/init.c @@ -120,13 +120,7 @@ void __init mem_init(void) BUG_ON(!mem_map); #endif /* CONFIG_FLATMEM */ -#ifdef CONFIG_SWIOTLB - if (swiotlb_force == SWIOTLB_FORCE || - max_pfn > PFN_DOWN(dma32_phys_limit)) - swiotlb_init(1); - else - swiotlb_force = SWIOTLB_NO_FORCE; -#endif + swiotlb_init(max_pfn > PFN_DOWN(dma32_phys_limit), SWIOTLB_VERBOSE); memblock_free_all(); print_vm_layout(); diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c index 86ffd0d51fd5..6fb6bf64326f 100644 --- a/arch/s390/mm/init.c +++ b/arch/s390/mm/init.c @@ -185,8 +185,7 @@ static void pv_init(void) return; /* make sure bounce buffers are shared */ - swiotlb_force = SWIOTLB_FORCE; - swiotlb_init(1); + swiotlb_init(true, SWIOTLB_FORCE | SWIOTLB_VERBOSE); swiotlb_update_mem_attributes(); } diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c index 04140e20ef1a..a705a199bf8a 100644 --- a/arch/x86/kernel/pci-dma.c +++ b/arch/x86/kernel/pci-dma.c @@ -39,6 +39,7 @@ int iommu_detected __read_mostly = 0; #ifdef CONFIG_SWIOTLB bool x86_swiotlb_enable; +static unsigned int x86_swiotlb_flags; static void __init pci_swiotlb_detect(void) { @@ -58,16 +59,16 @@ static void __init pci_swiotlb_detect(void) * bounce buffers as the hypervisor can't access arbitrary VM memory * that is not explicitly shared with it. */ - if (cc_platform_has(CC_ATTR_GUEST_MEM_ENCRYPT)) - swiotlb_force = SWIOTLB_FORCE; - - if (swiotlb_force == SWIOTLB_FORCE) + if (cc_platform_has(CC_ATTR_GUEST_MEM_ENCRYPT)) { x86_swiotlb_enable = true; + x86_swiotlb_flags |= SWIOTLB_FORCE; + } } #else static inline void __init pci_swiotlb_detect(void) { } +#define x86_swiotlb_flags 0 #endif /* CONFIG_SWIOTLB */ #ifdef CONFIG_SWIOTLB_XEN @@ -75,8 +76,7 @@ static bool xen_swiotlb; static void __init pci_xen_swiotlb_init(void) { - if (!xen_initial_domain() && !x86_swiotlb_enable && - swiotlb_force != SWIOTLB_FORCE) + if (!xen_initial_domain() && !x86_swiotlb_enable) return; x86_swiotlb_enable = true; xen_swiotlb = true; @@ -120,8 +120,7 @@ void __init pci_iommu_alloc(void) gart_iommu_hole_init(); amd_iommu_detect(); detect_intel_iommu(); - if (x86_swiotlb_enable) - swiotlb_init(0); + swiotlb_init(x86_swiotlb_enable, x86_swiotlb_flags); } /* diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c index 485cd06ed39e..c2da3eb4826e 100644 --- a/drivers/xen/swiotlb-xen.c +++ b/drivers/xen/swiotlb-xen.c @@ -251,7 +251,7 @@ retry: panic("%s (rc:%d)", xen_swiotlb_error(XEN_SWIOTLB_EFIXUP), rc); } - if (swiotlb_init_with_tbl(start, nslabs, true)) + if (swiotlb_init_with_tbl(start, nslabs, SWIOTLB_VERBOSE)) panic("Cannot allocate SWIOTLB buffer"); } #endif /* CONFIG_X86 */ @@ -376,7 +376,7 @@ static dma_addr_t xen_swiotlb_map_page(struct device *dev, struct page *page, /* * Oh well, have to allocate and map a bounce buffer. */ - trace_swiotlb_bounced(dev, dev_addr, size, swiotlb_force); + trace_swiotlb_bounced(dev, dev_addr, size); map = swiotlb_tbl_map_single(dev, phys, size, size, 0, dir, attrs); if (map == (phys_addr_t)DMA_MAPPING_ERROR) diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h index b48b26bfa0ed..ae0407173e84 100644 --- a/include/linux/swiotlb.h +++ b/include/linux/swiotlb.h @@ -13,11 +13,8 @@ struct device; struct page; struct scatterlist; -enum swiotlb_force { - SWIOTLB_NORMAL, /* Default - depending on HW DMA mask etc. */ - SWIOTLB_FORCE, /* swiotlb=force */ - SWIOTLB_NO_FORCE, /* swiotlb=noforce */ -}; +#define SWIOTLB_VERBOSE (1 << 0) /* verbose initialization */ +#define SWIOTLB_FORCE (1 << 1) /* force bounce buffering */ /* * Maximum allowable number of contiguous slabs to map, @@ -36,8 +33,7 @@ enum swiotlb_force { /* default to 64MB */ #define IO_TLB_DEFAULT_SIZE (64UL<<20) -extern void swiotlb_init(int verbose); -int swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose); +int swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, unsigned int flags); unsigned long swiotlb_size_or_default(void); extern int swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs); int swiotlb_init_late(size_t size); @@ -126,13 +122,16 @@ static inline bool is_swiotlb_force_bounce(struct device *dev) return mem && mem->force_bounce; } +void swiotlb_init(bool addressing_limited, unsigned int flags); void __init swiotlb_exit(void); unsigned int swiotlb_max_segment(void); size_t swiotlb_max_mapping_size(struct device *dev); bool is_swiotlb_active(struct device *dev); void __init swiotlb_adjust_size(unsigned long size); #else -#define swiotlb_force SWIOTLB_NO_FORCE +static inline void swiotlb_init(bool addressing_limited, unsigned int flags) +{ +} static inline bool is_swiotlb_buffer(struct device *dev, phys_addr_t paddr) { return false; diff --git a/include/trace/events/swiotlb.h b/include/trace/events/swiotlb.h index 705be43b71ab..da05c9ebd224 100644 --- a/include/trace/events/swiotlb.h +++ b/include/trace/events/swiotlb.h @@ -8,20 +8,15 @@ #include TRACE_EVENT(swiotlb_bounced, - - TP_PROTO(struct device *dev, - dma_addr_t dev_addr, - size_t size, - enum swiotlb_force swiotlb_force), - - TP_ARGS(dev, dev_addr, size, swiotlb_force), + TP_PROTO(struct device *dev, dma_addr_t dev_addr, size_t size), + TP_ARGS(dev, dev_addr, size), TP_STRUCT__entry( - __string( dev_name, dev_name(dev) ) - __field( u64, dma_mask ) - __field( dma_addr_t, dev_addr ) - __field( size_t, size ) - __field( enum swiotlb_force, swiotlb_force ) + __string(dev_name, dev_name(dev)) + __field(u64, dma_mask) + __field(dma_addr_t, dev_addr) + __field(size_t, size) + __field(bool, force) ), TP_fast_assign( @@ -29,19 +24,15 @@ TRACE_EVENT(swiotlb_bounced, __entry->dma_mask = (dev->dma_mask ? *dev->dma_mask : 0); __entry->dev_addr = dev_addr; __entry->size = size; - __entry->swiotlb_force = swiotlb_force; + __entry->force = is_swiotlb_force_bounce(dev); ), - TP_printk("dev_name: %s dma_mask=%llx dev_addr=%llx " - "size=%zu %s", + TP_printk("dev_name: %s dma_mask=%llx dev_addr=%llx size=%zu %s", __get_str(dev_name), __entry->dma_mask, (unsigned long long)__entry->dev_addr, __entry->size, - __print_symbolic(__entry->swiotlb_force, - { SWIOTLB_NORMAL, "NORMAL" }, - { SWIOTLB_FORCE, "FORCE" }, - { SWIOTLB_NO_FORCE, "NO_FORCE" })) + __entry->force ? "FORCE" : "NORMAL") ); #endif /* _TRACE_SWIOTLB_H */ diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c index 9a4fe6e48a07..86e877a96b82 100644 --- a/kernel/dma/swiotlb.c +++ b/kernel/dma/swiotlb.c @@ -62,7 +62,8 @@ #define INVALID_PHYS_ADDR (~(phys_addr_t)0) -enum swiotlb_force swiotlb_force; +static bool swiotlb_force_bounce; +static bool swiotlb_force_disable; struct io_tlb_mem io_tlb_default_mem; @@ -81,9 +82,9 @@ setup_io_tlb_npages(char *str) if (*str == ',') ++str; if (!strcmp(str, "force")) - swiotlb_force = SWIOTLB_FORCE; + swiotlb_force_bounce = true; else if (!strcmp(str, "noforce")) - swiotlb_force = SWIOTLB_NO_FORCE; + swiotlb_force_disable = true; return 0; } @@ -202,7 +203,7 @@ static void swiotlb_init_io_tlb_mem(struct io_tlb_mem *mem, phys_addr_t start, mem->index = 0; mem->late_alloc = late_alloc; - if (swiotlb_force == SWIOTLB_FORCE) + if (swiotlb_force_bounce) mem->force_bounce = true; spin_lock_init(&mem->lock); @@ -224,12 +225,13 @@ static void swiotlb_init_io_tlb_mem(struct io_tlb_mem *mem, phys_addr_t start, return; } -int __init swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose) +int __init swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, + unsigned int flags) { struct io_tlb_mem *mem = &io_tlb_default_mem; size_t alloc_size; - if (swiotlb_force == SWIOTLB_NO_FORCE) + if (swiotlb_force_disable) return 0; /* protect against double initialization */ @@ -243,8 +245,9 @@ int __init swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose) __func__, alloc_size, PAGE_SIZE); swiotlb_init_io_tlb_mem(mem, __pa(tlb), nslabs, false); + mem->force_bounce = flags & SWIOTLB_FORCE; - if (verbose) + if (flags & SWIOTLB_VERBOSE) swiotlb_print_info(); return 0; } @@ -253,20 +256,21 @@ int __init swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose) * Statically reserve bounce buffer space and initialize bounce buffer data * structures for the software IO TLB used to implement the DMA API. */ -void __init -swiotlb_init(int verbose) +void __init swiotlb_init(bool addressing_limit, unsigned int flags) { size_t bytes = PAGE_ALIGN(default_nslabs << IO_TLB_SHIFT); void *tlb; - if (swiotlb_force == SWIOTLB_NO_FORCE) + if (!addressing_limit && !swiotlb_force_bounce) + return; + if (swiotlb_force_disable) return; /* Get IO TLB memory from the low pages */ tlb = memblock_alloc_low(bytes, PAGE_SIZE); if (!tlb) goto fail; - if (swiotlb_init_with_tbl(tlb, default_nslabs, verbose)) + if (swiotlb_init_with_tbl(tlb, default_nslabs, flags)) goto fail_free_mem; return; @@ -289,7 +293,7 @@ int swiotlb_init_late(size_t size) unsigned int order; int rc = 0; - if (swiotlb_force == SWIOTLB_NO_FORCE) + if (swiotlb_force_disable) return 0; /* @@ -328,7 +332,7 @@ swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs) struct io_tlb_mem *mem = &io_tlb_default_mem; unsigned long bytes = nslabs << IO_TLB_SHIFT; - if (swiotlb_force == SWIOTLB_NO_FORCE) + if (swiotlb_force_disable) return 0; /* protect against double initialization */ @@ -353,7 +357,7 @@ void __init swiotlb_exit(void) unsigned long tbl_vaddr; size_t tbl_size, slots_size; - if (swiotlb_force == SWIOTLB_FORCE) + if (swiotlb_force_bounce) return; if (!mem->nslabs) @@ -704,8 +708,7 @@ dma_addr_t swiotlb_map(struct device *dev, phys_addr_t paddr, size_t size, phys_addr_t swiotlb_addr; dma_addr_t dma_addr; - trace_swiotlb_bounced(dev, phys_to_dma(dev, paddr), size, - swiotlb_force); + trace_swiotlb_bounced(dev, phys_to_dma(dev, paddr), size); swiotlb_addr = swiotlb_tbl_map_single(dev, paddr, size, size, 0, dir, attrs); -- cgit v1.3-8-gc7d7 From d9fdd02d4265c4c4fd5264b32b46034239192b7e Mon Sep 17 00:00:00 2001 From: Geliang Tang Date: Fri, 22 Apr 2022 14:55:42 -0700 Subject: mptcp: dump infinite_map field in mptcp_dump_mpext In trace event class mptcp_dump_mpext, dump the newly added infinite_map field of struct mptcp_dump_mpext too. Signed-off-by: Geliang Tang Signed-off-by: Mat Martineau Signed-off-by: David S. Miller --- include/trace/events/mptcp.h | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/mptcp.h b/include/trace/events/mptcp.h index f8e28e686c65..563e48617374 100644 --- a/include/trace/events/mptcp.h +++ b/include/trace/events/mptcp.h @@ -84,6 +84,7 @@ DECLARE_EVENT_CLASS(mptcp_dump_mpext, __field(u8, reset_transient) __field(u8, reset_reason) __field(u8, csum_reqd) + __field(u8, infinite_map) ), TP_fast_assign( @@ -102,9 +103,10 @@ DECLARE_EVENT_CLASS(mptcp_dump_mpext, __entry->reset_transient = mpext->reset_transient; __entry->reset_reason = mpext->reset_reason; __entry->csum_reqd = mpext->csum_reqd; + __entry->infinite_map = mpext->infinite_map; ), - TP_printk("data_ack=%llu data_seq=%llu subflow_seq=%u data_len=%u csum=%x use_map=%u dsn64=%u data_fin=%u use_ack=%u ack64=%u mpc_map=%u frozen=%u reset_transient=%u reset_reason=%u csum_reqd=%u", + TP_printk("data_ack=%llu data_seq=%llu subflow_seq=%u data_len=%u csum=%x use_map=%u dsn64=%u data_fin=%u use_ack=%u ack64=%u mpc_map=%u frozen=%u reset_transient=%u reset_reason=%u csum_reqd=%u infinite_map=%u", __entry->data_ack, __entry->data_seq, __entry->subflow_seq, __entry->data_len, __entry->csum, __entry->use_map, @@ -112,7 +114,7 @@ DECLARE_EVENT_CLASS(mptcp_dump_mpext, __entry->use_ack, __entry->ack64, __entry->mpc_map, __entry->frozen, __entry->reset_transient, __entry->reset_reason, - __entry->csum_reqd) + __entry->csum_reqd, __entry->infinite_map) ); DEFINE_EVENT(mptcp_dump_mpext, mptcp_sendmsg_frag, -- cgit v1.3-8-gc7d7 From c277f1411d7bc2831d48bde5c18d3b7f87c61646 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Mon, 21 Mar 2022 15:13:06 -0700 Subject: f2fs: introduce data read/write showing path info This was used in Android for a long time. Let's upstream it. Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/file.c | 58 ++++++++++++++++++++++++---- include/trace/events/f2fs.h | 94 +++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 145 insertions(+), 7 deletions(-) (limited to 'include/trace') diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 5b89af0f27f0..f08e6208e183 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -4329,17 +4329,39 @@ out: static ssize_t f2fs_file_read_iter(struct kiocb *iocb, struct iov_iter *to) { struct inode *inode = file_inode(iocb->ki_filp); + const loff_t pos = iocb->ki_pos; ssize_t ret; if (!f2fs_is_compress_backend_ready(inode)) return -EOPNOTSUPP; - if (f2fs_should_use_dio(inode, iocb, to)) - return f2fs_dio_read_iter(iocb, to); + if (trace_f2fs_dataread_start_enabled()) { + char *p = f2fs_kmalloc(F2FS_I_SB(inode), PATH_MAX, GFP_KERNEL); + char *path; + + if (!p) + goto skip_read_trace; + + path = dentry_path_raw(file_dentry(iocb->ki_filp), p, PATH_MAX); + if (IS_ERR(path)) { + kfree(p); + goto skip_read_trace; + } - ret = filemap_read(iocb, to, 0); - if (ret > 0) - f2fs_update_iostat(F2FS_I_SB(inode), APP_BUFFERED_READ_IO, ret); + trace_f2fs_dataread_start(inode, pos, iov_iter_count(to), + current->pid, path, current->comm); + kfree(p); + } +skip_read_trace: + if (f2fs_should_use_dio(inode, iocb, to)) { + ret = f2fs_dio_read_iter(iocb, to); + } else { + ret = filemap_read(iocb, to, 0); + if (ret > 0) + f2fs_update_iostat(F2FS_I_SB(inode), APP_BUFFERED_READ_IO, ret); + } + if (trace_f2fs_dataread_end_enabled()) + trace_f2fs_dataread_end(inode, pos, ret); return ret; } @@ -4631,14 +4653,36 @@ static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) /* Possibly preallocate the blocks for the write. */ target_size = iocb->ki_pos + iov_iter_count(from); preallocated = f2fs_preallocate_blocks(iocb, from, dio); - if (preallocated < 0) + if (preallocated < 0) { ret = preallocated; - else + } else { + if (trace_f2fs_datawrite_start_enabled()) { + char *p = f2fs_kmalloc(F2FS_I_SB(inode), + PATH_MAX, GFP_KERNEL); + char *path; + + if (!p) + goto skip_write_trace; + path = dentry_path_raw(file_dentry(iocb->ki_filp), + p, PATH_MAX); + if (IS_ERR(path)) { + kfree(p); + goto skip_write_trace; + } + trace_f2fs_datawrite_start(inode, orig_pos, orig_count, + current->pid, path, current->comm); + kfree(p); + } +skip_write_trace: /* Do the actual write. */ ret = dio ? f2fs_dio_write_iter(iocb, from, &may_need_sync): f2fs_buffered_write_iter(iocb, from); + if (trace_f2fs_datawrite_end_enabled()) + trace_f2fs_datawrite_end(inode, orig_pos, ret); + } + /* Don't leave any preallocated blocks around past i_size. */ if (preallocated && i_size_read(inode) < target_size) { f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h index 1779e133cea0..4d1ad64d4cab 100644 --- a/include/trace/events/f2fs.h +++ b/include/trace/events/f2fs.h @@ -2067,6 +2067,100 @@ TRACE_EVENT(f2fs_fiemap, __entry->ret) ); +DECLARE_EVENT_CLASS(f2fs__rw_start, + + TP_PROTO(struct inode *inode, loff_t offset, int bytes, + pid_t pid, char *pathname, char *command), + + TP_ARGS(inode, offset, bytes, pid, pathname, command), + + TP_STRUCT__entry( + __string(pathbuf, pathname) + __field(loff_t, offset) + __field(int, bytes) + __field(loff_t, i_size) + __string(cmdline, command) + __field(pid_t, pid) + __field(ino_t, ino) + ), + + TP_fast_assign( + /* + * Replace the spaces in filenames and cmdlines + * because this screws up the tooling that parses + * the traces. + */ + __assign_str(pathbuf, pathname); + (void)strreplace(__get_str(pathbuf), ' ', '_'); + __entry->offset = offset; + __entry->bytes = bytes; + __entry->i_size = i_size_read(inode); + __assign_str(cmdline, command); + (void)strreplace(__get_str(cmdline), ' ', '_'); + __entry->pid = pid; + __entry->ino = inode->i_ino; + ), + + TP_printk("entry_name %s, offset %llu, bytes %d, cmdline %s," + " pid %d, i_size %llu, ino %lu", + __get_str(pathbuf), __entry->offset, __entry->bytes, + __get_str(cmdline), __entry->pid, __entry->i_size, + (unsigned long) __entry->ino) +); + +DECLARE_EVENT_CLASS(f2fs__rw_end, + + TP_PROTO(struct inode *inode, loff_t offset, int bytes), + + TP_ARGS(inode, offset, bytes), + + TP_STRUCT__entry( + __field(ino_t, ino) + __field(loff_t, offset) + __field(int, bytes) + ), + + TP_fast_assign( + __entry->ino = inode->i_ino; + __entry->offset = offset; + __entry->bytes = bytes; + ), + + TP_printk("ino %lu, offset %llu, bytes %d", + (unsigned long) __entry->ino, + __entry->offset, __entry->bytes) +); + +DEFINE_EVENT(f2fs__rw_start, f2fs_dataread_start, + + TP_PROTO(struct inode *inode, loff_t offset, int bytes, + pid_t pid, char *pathname, char *command), + + TP_ARGS(inode, offset, bytes, pid, pathname, command) +); + +DEFINE_EVENT(f2fs__rw_end, f2fs_dataread_end, + + TP_PROTO(struct inode *inode, loff_t offset, int bytes), + + TP_ARGS(inode, offset, bytes) +); + +DEFINE_EVENT(f2fs__rw_start, f2fs_datawrite_start, + + TP_PROTO(struct inode *inode, loff_t offset, int bytes, + pid_t pid, char *pathname, char *command), + + TP_ARGS(inode, offset, bytes, pid, pathname, command) +); + +DEFINE_EVENT(f2fs__rw_end, f2fs_datawrite_end, + + TP_PROTO(struct inode *inode, loff_t offset, int bytes), + + TP_ARGS(inode, offset, bytes) +); + #endif /* _TRACE_F2FS_H */ /* This part must be outside protection */ -- cgit v1.3-8-gc7d7 From c3d438eeb5413111889edef10cb5fcc2c0fb8bc9 Mon Sep 17 00:00:00 2001 From: Lukasz Luba Date: Wed, 27 Apr 2022 09:08:06 +0100 Subject: arch_topology: Trace the update thermal pressure Add trace event to capture the moment of the call for updating the thermal pressure value. It's helpful to investigate how often those events occur in a system dealing with throttling. This trace event is needed since the old 'cdev_update' might not be used by some drivers. The old 'cdev_update' trace event only provides a cooling state value: [0, n]. That state value then needs additional tools to translate it: state -> freq -> capacity -> thermal pressure. This new trace event just stores proper thermal pressure value in the trace buffer, no need for additional logic. This is helpful for cooperation when someone can simply sends to the list the trace buffer output from the platform (no need from additional information from other subsystems). There are also platforms which due to some design reasons don't use cooling devices and thus don't trigger old 'cdev_update' trace event. They are also important and measuring latency for the thermal signal raising/decaying characteristics is in scope. This new trace event would cover them as well. We already have a trace point 'pelt_thermal_tp' which after a change to trace event can be paired with this new 'thermal_pressure_update' and derive more insight what is going on in the system under thermal pressure (and why). Signed-off-by: Lukasz Luba Acked-by: Sudeep Holla Link: https://lore.kernel.org/r/20220427080806.1906-1-lukasz.luba@arm.com Signed-off-by: Greg Kroah-Hartman --- drivers/base/arch_topology.c | 5 +++++ include/trace/events/thermal_pressure.h | 29 +++++++++++++++++++++++++++++ 2 files changed, 34 insertions(+) create mode 100644 include/trace/events/thermal_pressure.h (limited to 'include/trace') diff --git a/drivers/base/arch_topology.c b/drivers/base/arch_topology.c index f73b836047cf..579c851a2bd7 100644 --- a/drivers/base/arch_topology.c +++ b/drivers/base/arch_topology.c @@ -19,6 +19,9 @@ #include #include +#define CREATE_TRACE_POINTS +#include + static DEFINE_PER_CPU(struct scale_freq_data __rcu *, sft_data); static struct cpumask scale_freq_counters_mask; static bool scale_freq_invariant; @@ -195,6 +198,8 @@ void topology_update_thermal_pressure(const struct cpumask *cpus, th_pressure = max_capacity - capacity; + trace_thermal_pressure_update(cpu, th_pressure); + for_each_cpu(cpu, cpus) WRITE_ONCE(per_cpu(thermal_pressure, cpu), th_pressure); } diff --git a/include/trace/events/thermal_pressure.h b/include/trace/events/thermal_pressure.h new file mode 100644 index 000000000000..b68680201360 --- /dev/null +++ b/include/trace/events/thermal_pressure.h @@ -0,0 +1,29 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM thermal_pressure + +#if !defined(_TRACE_THERMAL_PRESSURE_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_THERMAL_PRESSURE_H + +#include + +TRACE_EVENT(thermal_pressure_update, + TP_PROTO(int cpu, unsigned long thermal_pressure), + TP_ARGS(cpu, thermal_pressure), + + TP_STRUCT__entry( + __field(unsigned long, thermal_pressure) + __field(int, cpu) + ), + + TP_fast_assign( + __entry->thermal_pressure = thermal_pressure; + __entry->cpu = cpu; + ), + + TP_printk("cpu=%d thermal_pressure=%lu", __entry->cpu, __entry->thermal_pressure) +); +#endif /* _TRACE_THERMAL_PRESSURE_H */ + +/* This part must be outside protection */ +#include -- cgit v1.3-8-gc7d7 From 3db1de0e582c358dd013f3703cd55b5fe4076436 Mon Sep 17 00:00:00 2001 From: Daeho Jeong Date: Thu, 28 Apr 2022 11:18:09 -0700 Subject: f2fs: change the current atomic write way Current atomic write has three major issues like below. - keeps the updates in non-reclaimable memory space and they are even hard to be migrated, which is not good for contiguous memory allocation. - disk spaces used for atomic files cannot be garbage collected, so this makes it difficult for the filesystem to be defragmented. - If atomic write operations hit the threshold of either memory usage or garbage collection failure count, All the atomic write operations will fail immediately. To resolve the issues, I will keep a COW inode internally for all the updates to be flushed from memory, when we need to flush them out in a situation like high memory pressure. These COW inodes will be tagged as orphan inodes to be reclaimed in case of sudden power-cut or system failure during atomic writes. Signed-off-by: Daeho Jeong Signed-off-by: Jaegeuk Kim --- fs/f2fs/data.c | 180 +++++++++++++-------- fs/f2fs/debug.c | 12 +- fs/f2fs/f2fs.h | 33 +--- fs/f2fs/file.c | 49 +++--- fs/f2fs/gc.c | 27 +--- fs/f2fs/inode.c | 3 +- fs/f2fs/namei.c | 28 ++-- fs/f2fs/node.c | 4 - fs/f2fs/node.h | 1 - fs/f2fs/segment.c | 380 +++++++++++++++----------------------------- fs/f2fs/segment.h | 4 +- fs/f2fs/super.c | 6 +- include/trace/events/f2fs.h | 22 --- 13 files changed, 303 insertions(+), 446 deletions(-) (limited to 'include/trace') diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 9a1a526f2092..8763a4690aaf 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -69,8 +69,7 @@ static bool __is_cp_guaranteed(struct page *page) if (f2fs_is_compressed_page(page)) return false; - if ((S_ISREG(inode->i_mode) && - (f2fs_is_atomic_file(inode) || IS_NOQUOTA(inode))) || + if ((S_ISREG(inode->i_mode) && IS_NOQUOTA(inode)) || page_private_gcing(page)) return true; return false; @@ -2563,7 +2562,12 @@ int f2fs_do_write_data_page(struct f2fs_io_info *fio) bool ipu_force = false; int err = 0; - set_new_dnode(&dn, inode, NULL, NULL, 0); + /* Use COW inode to make dnode_of_data for atomic write */ + if (f2fs_is_atomic_file(inode)) + set_new_dnode(&dn, F2FS_I(inode)->cow_inode, NULL, NULL, 0); + else + set_new_dnode(&dn, inode, NULL, NULL, 0); + if (need_inplace_update(fio) && f2fs_lookup_extent_cache(inode, page->index, &ei)) { fio->old_blkaddr = ei.blk + page->index - ei.fofs; @@ -2600,6 +2604,7 @@ got_it: err = -EFSCORRUPTED; goto out_writepage; } + /* * If current allocation needs SSR, * it had better in-place writes for updated data. @@ -3313,6 +3318,100 @@ unlock_out: return err; } +static int __find_data_block(struct inode *inode, pgoff_t index, + block_t *blk_addr) +{ + struct dnode_of_data dn; + struct page *ipage; + struct extent_info ei = {0, }; + int err = 0; + + ipage = f2fs_get_node_page(F2FS_I_SB(inode), inode->i_ino); + if (IS_ERR(ipage)) + return PTR_ERR(ipage); + + set_new_dnode(&dn, inode, ipage, ipage, 0); + + if (f2fs_lookup_extent_cache(inode, index, &ei)) { + dn.data_blkaddr = ei.blk + index - ei.fofs; + } else { + /* hole case */ + err = f2fs_get_dnode_of_data(&dn, index, LOOKUP_NODE); + if (err) { + dn.data_blkaddr = NULL_ADDR; + err = 0; + } + } + *blk_addr = dn.data_blkaddr; + f2fs_put_dnode(&dn); + return err; +} + +static int __reserve_data_block(struct inode *inode, pgoff_t index, + block_t *blk_addr, bool *node_changed) +{ + struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + struct dnode_of_data dn; + struct page *ipage; + int err = 0; + + f2fs_do_map_lock(sbi, F2FS_GET_BLOCK_PRE_AIO, true); + + ipage = f2fs_get_node_page(sbi, inode->i_ino); + if (IS_ERR(ipage)) { + err = PTR_ERR(ipage); + goto unlock_out; + } + set_new_dnode(&dn, inode, ipage, ipage, 0); + + err = f2fs_get_block(&dn, index); + + *blk_addr = dn.data_blkaddr; + *node_changed = dn.node_changed; + f2fs_put_dnode(&dn); + +unlock_out: + f2fs_do_map_lock(sbi, F2FS_GET_BLOCK_PRE_AIO, false); + return err; +} + +static int prepare_atomic_write_begin(struct f2fs_sb_info *sbi, + struct page *page, loff_t pos, unsigned int len, + block_t *blk_addr, bool *node_changed) +{ + struct inode *inode = page->mapping->host; + struct inode *cow_inode = F2FS_I(inode)->cow_inode; + pgoff_t index = page->index; + int err = 0; + block_t ori_blk_addr; + + /* If pos is beyond the end of file, reserve a new block in COW inode */ + if ((pos & PAGE_MASK) >= i_size_read(inode)) + return __reserve_data_block(cow_inode, index, blk_addr, + node_changed); + + /* Look for the block in COW inode first */ + err = __find_data_block(cow_inode, index, blk_addr); + if (err) + return err; + else if (*blk_addr != NULL_ADDR) + return 0; + + /* Look for the block in the original inode */ + err = __find_data_block(inode, index, &ori_blk_addr); + if (err) + return err; + + /* Finally, we should reserve a new block in COW inode for the update */ + err = __reserve_data_block(cow_inode, index, blk_addr, node_changed); + if (err) + return err; + + if (ori_blk_addr != NULL_ADDR) + *blk_addr = ori_blk_addr; + return 0; +} + static int f2fs_write_begin(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned flags, struct page **pagep, void **fsdata) @@ -3321,7 +3420,7 @@ static int f2fs_write_begin(struct file *file, struct address_space *mapping, struct f2fs_sb_info *sbi = F2FS_I_SB(inode); struct page *page = NULL; pgoff_t index = ((unsigned long long) pos) >> PAGE_SHIFT; - bool need_balance = false, drop_atomic = false; + bool need_balance = false; block_t blkaddr = NULL_ADDR; int err = 0; @@ -3332,14 +3431,6 @@ static int f2fs_write_begin(struct file *file, struct address_space *mapping, goto fail; } - if ((f2fs_is_atomic_file(inode) && - !f2fs_available_free_memory(sbi, INMEM_PAGES)) || - is_inode_flag_set(inode, FI_ATOMIC_REVOKE_REQUEST)) { - err = -ENOMEM; - drop_atomic = true; - goto fail; - } - /* * We should check this at this moment to avoid deadlock on inode page * and #0 page. The locking rule for inline_data conversion should be: @@ -3387,7 +3478,11 @@ repeat: *pagep = page; - err = prepare_write_begin(sbi, page, pos, len, + if (f2fs_is_atomic_file(inode)) + err = prepare_atomic_write_begin(sbi, page, pos, len, + &blkaddr, &need_balance); + else + err = prepare_write_begin(sbi, page, pos, len, &blkaddr, &need_balance); if (err) goto fail; @@ -3443,8 +3538,6 @@ repeat: fail: f2fs_put_page(page, 1); f2fs_write_failed(inode, pos + len); - if (drop_atomic) - f2fs_drop_inmem_pages_all(sbi, false); return err; } @@ -3488,8 +3581,12 @@ static int f2fs_write_end(struct file *file, set_page_dirty(page); if (pos + copied > i_size_read(inode) && - !f2fs_verity_in_progress(inode)) + !f2fs_verity_in_progress(inode)) { f2fs_i_size_write(inode, pos + copied); + if (f2fs_is_atomic_file(inode)) + f2fs_i_size_write(F2FS_I(inode)->cow_inode, + pos + copied); + } unlock_out: f2fs_put_page(page, 1); f2fs_update_time(F2FS_I_SB(inode), REQ_TIME); @@ -3522,9 +3619,6 @@ void f2fs_invalidate_folio(struct folio *folio, size_t offset, size_t length) inode->i_ino == F2FS_COMPRESS_INO(sbi)) clear_page_private_data(&folio->page); - if (page_private_atomic(&folio->page)) - return f2fs_drop_inmem_page(inode, &folio->page); - folio_detach_private(folio); } @@ -3534,10 +3628,6 @@ int f2fs_release_page(struct page *page, gfp_t wait) if (PageDirty(page)) return 0; - /* This is atomic written page, keep Private */ - if (page_private_atomic(page)) - return 0; - if (test_opt(F2FS_P_SB(page), COMPRESS_CACHE)) { struct inode *inode = page->mapping->host; @@ -3563,18 +3653,6 @@ static bool f2fs_dirty_data_folio(struct address_space *mapping, folio_mark_uptodate(folio); BUG_ON(folio_test_swapcache(folio)); - if (f2fs_is_atomic_file(inode) && !f2fs_is_commit_atomic_write(inode)) { - if (!page_private_atomic(&folio->page)) { - f2fs_register_inmem_page(inode, &folio->page); - return true; - } - /* - * Previously, this page has been registered, we just - * return here. - */ - return false; - } - if (!folio_test_dirty(folio)) { filemap_dirty_folio(mapping, folio); f2fs_update_dirty_folio(inode, folio); @@ -3654,42 +3732,14 @@ out: int f2fs_migrate_page(struct address_space *mapping, struct page *newpage, struct page *page, enum migrate_mode mode) { - int rc, extra_count; - struct f2fs_inode_info *fi = F2FS_I(mapping->host); - bool atomic_written = page_private_atomic(page); + int rc, extra_count = 0; BUG_ON(PageWriteback(page)); - /* migrating an atomic written page is safe with the inmem_lock hold */ - if (atomic_written) { - if (mode != MIGRATE_SYNC) - return -EBUSY; - if (!mutex_trylock(&fi->inmem_lock)) - return -EAGAIN; - } - - /* one extra reference was held for atomic_write page */ - extra_count = atomic_written ? 1 : 0; rc = migrate_page_move_mapping(mapping, newpage, page, extra_count); - if (rc != MIGRATEPAGE_SUCCESS) { - if (atomic_written) - mutex_unlock(&fi->inmem_lock); + if (rc != MIGRATEPAGE_SUCCESS) return rc; - } - - if (atomic_written) { - struct inmem_pages *cur; - - list_for_each_entry(cur, &fi->inmem_pages, list) - if (cur->page == page) { - cur->page = newpage; - break; - } - mutex_unlock(&fi->inmem_lock); - put_page(page); - get_page(newpage); - } /* guarantee to start from no stale private field */ set_page_private(newpage, 0); diff --git a/fs/f2fs/debug.c b/fs/f2fs/debug.c index fcdf253cd211..65f0bcf498bb 100644 --- a/fs/f2fs/debug.c +++ b/fs/f2fs/debug.c @@ -91,7 +91,6 @@ static void update_general_status(struct f2fs_sb_info *sbi) si->ndirty_files = sbi->ndirty_inode[FILE_INODE]; si->nquota_files = sbi->nquota_files; si->ndirty_all = sbi->ndirty_inode[DIRTY_META]; - si->inmem_pages = get_pages(sbi, F2FS_INMEM_PAGES); si->aw_cnt = sbi->atomic_files; si->vw_cnt = atomic_read(&sbi->vw_cnt); si->max_aw_cnt = atomic_read(&sbi->max_aw_cnt); @@ -167,8 +166,6 @@ static void update_general_status(struct f2fs_sb_info *sbi) si->alloc_nids = NM_I(sbi)->nid_cnt[PREALLOC_NID]; si->io_skip_bggc = sbi->io_skip_bggc; si->other_skip_bggc = sbi->other_skip_bggc; - si->skipped_atomic_files[BG_GC] = sbi->skipped_atomic_files[BG_GC]; - si->skipped_atomic_files[FG_GC] = sbi->skipped_atomic_files[FG_GC]; si->util_free = (int)(free_user_blocks(sbi) >> sbi->log_blocks_per_seg) * 100 / (int)(sbi->user_block_count >> sbi->log_blocks_per_seg) / 2; @@ -296,7 +293,6 @@ get_cache: sizeof(struct nat_entry); si->cache_mem += NM_I(sbi)->nat_cnt[DIRTY_NAT] * sizeof(struct nat_entry_set); - si->cache_mem += si->inmem_pages * sizeof(struct inmem_pages); for (i = 0; i < MAX_INO_ENTRY; i++) si->cache_mem += sbi->im[i].ino_num * sizeof(struct ino_entry); si->cache_mem += atomic_read(&sbi->total_ext_tree) * @@ -491,10 +487,6 @@ static int stat_show(struct seq_file *s, void *v) si->bg_data_blks); seq_printf(s, " - node blocks : %d (%d)\n", si->node_blks, si->bg_node_blks); - seq_printf(s, "Skipped : atomic write %llu (%llu)\n", - si->skipped_atomic_files[BG_GC] + - si->skipped_atomic_files[FG_GC], - si->skipped_atomic_files[BG_GC]); seq_printf(s, "BG skip : IO: %u, Other: %u\n", si->io_skip_bggc, si->other_skip_bggc); seq_puts(s, "\nExtent Cache:\n"); @@ -519,9 +511,9 @@ static int stat_show(struct seq_file *s, void *v) si->flush_list_empty, si->nr_discarding, si->nr_discarded, si->nr_discard_cmd, si->undiscard_blks); - seq_printf(s, " - inmem: %4d, atomic IO: %4d (Max. %4d), " + seq_printf(s, " - atomic IO: %4d (Max. %4d), " "volatile IO: %4d (Max. %4d)\n", - si->inmem_pages, si->aw_cnt, si->max_aw_cnt, + si->aw_cnt, si->max_aw_cnt, si->vw_cnt, si->max_vw_cnt); seq_printf(s, " - compress: %4d, hit:%8d\n", si->compress_pages, si->compress_page_hit); seq_printf(s, " - nodes: %4d in %4d\n", diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index efe5e80163a8..68d299b54820 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -716,7 +716,6 @@ enum { enum { GC_FAILURE_PIN, - GC_FAILURE_ATOMIC, MAX_GC_FAILURE }; @@ -738,7 +737,6 @@ enum { FI_UPDATE_WRITE, /* inode has in-place-update data */ FI_NEED_IPU, /* used for ipu per file */ FI_ATOMIC_FILE, /* indicate atomic file */ - FI_ATOMIC_COMMIT, /* indicate the state of atomical committing */ FI_VOLATILE_FILE, /* indicate volatile file */ FI_FIRST_BLOCK_WRITTEN, /* indicate #0 data block was written */ FI_DROP_CACHE, /* drop dirty page cache */ @@ -752,7 +750,6 @@ enum { FI_EXTRA_ATTR, /* indicate file has extra attribute */ FI_PROJ_INHERIT, /* indicate file inherits projectid */ FI_PIN_FILE, /* indicate file should not be gced */ - FI_ATOMIC_REVOKE_REQUEST, /* request to drop atomic data */ FI_VERITY_IN_PROGRESS, /* building fs-verity Merkle tree */ FI_COMPRESSED_FILE, /* indicate file's data can be compressed */ FI_COMPRESS_CORRUPT, /* indicate compressed cluster is corrupted */ @@ -794,11 +791,9 @@ struct f2fs_inode_info { #endif struct list_head dirty_list; /* dirty list for dirs and files */ struct list_head gdirty_list; /* linked in global dirty list */ - struct list_head inmem_ilist; /* list for inmem inodes */ - struct list_head inmem_pages; /* inmemory pages managed by f2fs */ - struct task_struct *inmem_task; /* store inmemory task */ - struct mutex inmem_lock; /* lock for inmemory pages */ + struct task_struct *atomic_write_task; /* store atomic write task */ struct extent_tree *extent_tree; /* cached extent_tree entry */ + struct inode *cow_inode; /* copy-on-write inode for atomic write */ /* avoid racing between foreground op and gc */ struct f2fs_rwsem i_gc_rwsem[2]; @@ -1092,7 +1087,6 @@ enum count_type { F2FS_DIRTY_QDATA, F2FS_DIRTY_NODES, F2FS_DIRTY_META, - F2FS_INMEM_PAGES, F2FS_DIRTY_IMETA, F2FS_WB_CP_DATA, F2FS_WB_DATA, @@ -1122,11 +1116,7 @@ enum page_type { META, NR_PAGE_TYPE, META_FLUSH, - INMEM, /* the below types are used by tracepoints only. */ - INMEM_DROP, - INMEM_INVALIDATE, - INMEM_REVOKE, - IPU, + IPU, /* the below types are used by tracepoints only. */ OPU, }; @@ -1718,7 +1708,6 @@ struct f2fs_sb_info { /* for skip statistic */ unsigned int atomic_files; /* # of opened atomic file */ - unsigned long long skipped_atomic_files[2]; /* FG_GC and BG_GC */ unsigned long long skipped_gc_rwsem; /* FG_GC only */ /* threshold for gc trials on pinned files */ @@ -3202,11 +3191,6 @@ static inline bool f2fs_is_atomic_file(struct inode *inode) return is_inode_flag_set(inode, FI_ATOMIC_FILE); } -static inline bool f2fs_is_commit_atomic_write(struct inode *inode) -{ - return is_inode_flag_set(inode, FI_ATOMIC_COMMIT); -} - static inline bool f2fs_is_volatile_file(struct inode *inode) { return is_inode_flag_set(inode, FI_VOLATILE_FILE); @@ -3444,6 +3428,8 @@ void f2fs_handle_failed_inode(struct inode *inode); int f2fs_update_extension_list(struct f2fs_sb_info *sbi, const char *name, bool hot, bool set); struct dentry *f2fs_get_parent(struct dentry *child); +int f2fs_get_tmpfile(struct user_namespace *mnt_userns, struct inode *dir, + struct inode **new_inode); /* * dir.c @@ -3579,11 +3565,8 @@ void f2fs_destroy_node_manager_caches(void); * segment.c */ bool f2fs_need_SSR(struct f2fs_sb_info *sbi); -void f2fs_register_inmem_page(struct inode *inode, struct page *page); -void f2fs_drop_inmem_pages_all(struct f2fs_sb_info *sbi, bool gc_failure); -void f2fs_drop_inmem_pages(struct inode *inode); -void f2fs_drop_inmem_page(struct inode *inode, struct page *page); -int f2fs_commit_inmem_pages(struct inode *inode); +int f2fs_commit_atomic_write(struct inode *inode); +void f2fs_abort_atomic_write(struct inode *inode, bool clean); void f2fs_balance_fs(struct f2fs_sb_info *sbi, bool need); void f2fs_balance_fs_bg(struct f2fs_sb_info *sbi, bool from_bg); int f2fs_issue_flush(struct f2fs_sb_info *sbi, nid_t ino); @@ -3815,7 +3798,6 @@ struct f2fs_stat_info { int ext_tree, zombie_tree, ext_node; int ndirty_node, ndirty_dent, ndirty_meta, ndirty_imeta; int ndirty_data, ndirty_qdata; - int inmem_pages; unsigned int ndirty_dirs, ndirty_files, nquota_files, ndirty_all; int nats, dirty_nats, sits, dirty_sits; int free_nids, avail_nids, alloc_nids; @@ -3845,7 +3827,6 @@ struct f2fs_stat_info { int bg_node_segs, bg_data_segs; int tot_blks, data_blks, node_blks; int bg_data_blks, bg_node_blks; - unsigned long long skipped_atomic_files[2]; int curseg[NR_CURSEG_TYPE]; int cursec[NR_CURSEG_TYPE]; int curzone[NR_CURSEG_TYPE]; diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index b307d96a0a7c..7e5ec0c48b2a 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -1813,9 +1813,8 @@ static int f2fs_release_file(struct inode *inode, struct file *filp) atomic_read(&inode->i_writecount) != 1) return 0; - /* some remained atomic pages should discarded */ if (f2fs_is_atomic_file(inode)) - f2fs_drop_inmem_pages(inode); + f2fs_abort_atomic_write(inode, true); if (f2fs_is_volatile_file(inode)) { set_inode_flag(inode, FI_DROP_CACHE); filemap_fdatawrite(inode->i_mapping); @@ -1837,8 +1836,8 @@ static int f2fs_file_flush(struct file *file, fl_owner_t id) * before dropping file lock, it needs to do in ->flush. */ if (f2fs_is_atomic_file(inode) && - F2FS_I(inode)->inmem_task == current) - f2fs_drop_inmem_pages(inode); + F2FS_I(inode)->atomic_write_task == current) + f2fs_abort_atomic_write(inode, true); return 0; } @@ -2001,6 +2000,7 @@ static int f2fs_ioc_start_atomic_write(struct file *filp) struct user_namespace *mnt_userns = file_mnt_user_ns(filp); struct f2fs_inode_info *fi = F2FS_I(inode); struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + struct inode *pinode; int ret; if (!inode_owner_or_capable(mnt_userns, inode)) @@ -2023,11 +2023,8 @@ static int f2fs_ioc_start_atomic_write(struct file *filp) goto out; } - if (f2fs_is_atomic_file(inode)) { - if (is_inode_flag_set(inode, FI_ATOMIC_REVOKE_REQUEST)) - ret = -EINVAL; + if (f2fs_is_atomic_file(inode)) goto out; - } ret = f2fs_convert_inline_inode(inode); if (ret) @@ -2048,19 +2045,33 @@ static int f2fs_ioc_start_atomic_write(struct file *filp) goto out; } + /* Create a COW inode for atomic write */ + pinode = f2fs_iget(inode->i_sb, fi->i_pino); + if (IS_ERR(pinode)) { + f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + ret = PTR_ERR(pinode); + goto out; + } + + ret = f2fs_get_tmpfile(mnt_userns, pinode, &fi->cow_inode); + iput(pinode); + if (ret) { + f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); + goto out; + } + f2fs_i_size_write(fi->cow_inode, i_size_read(inode)); + spin_lock(&sbi->inode_lock[ATOMIC_FILE]); - if (list_empty(&fi->inmem_ilist)) - list_add_tail(&fi->inmem_ilist, &sbi->inode_list[ATOMIC_FILE]); sbi->atomic_files++; spin_unlock(&sbi->inode_lock[ATOMIC_FILE]); - /* add inode in inmem_list first and set atomic_file */ set_inode_flag(inode, FI_ATOMIC_FILE); - clear_inode_flag(inode, FI_ATOMIC_REVOKE_REQUEST); + set_inode_flag(fi->cow_inode, FI_ATOMIC_FILE); + clear_inode_flag(fi->cow_inode, FI_INLINE_DATA); f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]); f2fs_update_time(F2FS_I_SB(inode), REQ_TIME); - F2FS_I(inode)->inmem_task = current; + F2FS_I(inode)->atomic_write_task = current; stat_update_max_atomic_write(inode); out: inode_unlock(inode); @@ -2091,21 +2102,17 @@ static int f2fs_ioc_commit_atomic_write(struct file *filp) } if (f2fs_is_atomic_file(inode)) { - ret = f2fs_commit_inmem_pages(inode); + ret = f2fs_commit_atomic_write(inode); if (ret) goto err_out; ret = f2fs_do_sync_file(filp, 0, LLONG_MAX, 0, true); if (!ret) - f2fs_drop_inmem_pages(inode); + f2fs_abort_atomic_write(inode, false); } else { ret = f2fs_do_sync_file(filp, 0, LLONG_MAX, 1, false); } err_out: - if (is_inode_flag_set(inode, FI_ATOMIC_REVOKE_REQUEST)) { - clear_inode_flag(inode, FI_ATOMIC_REVOKE_REQUEST); - ret = -EINVAL; - } inode_unlock(inode); mnt_drop_write_file(filp); return ret; @@ -2193,15 +2200,13 @@ static int f2fs_ioc_abort_volatile_write(struct file *filp) inode_lock(inode); if (f2fs_is_atomic_file(inode)) - f2fs_drop_inmem_pages(inode); + f2fs_abort_atomic_write(inode, true); if (f2fs_is_volatile_file(inode)) { clear_inode_flag(inode, FI_VOLATILE_FILE); stat_dec_volatile_write(inode); ret = f2fs_do_sync_file(filp, 0, LLONG_MAX, 0, true); } - clear_inode_flag(inode, FI_ATOMIC_REVOKE_REQUEST); - inode_unlock(inode); mnt_drop_write_file(filp); diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 3009c0a97ab4..ba8e93e517be 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -1245,13 +1245,6 @@ static int move_data_block(struct inode *inode, block_t bidx, goto out; } - if (f2fs_is_atomic_file(inode)) { - F2FS_I(inode)->i_gc_failures[GC_FAILURE_ATOMIC]++; - F2FS_I_SB(inode)->skipped_atomic_files[gc_type]++; - err = -EAGAIN; - goto out; - } - err = f2fs_gc_pinned_control(inode, gc_type, segno); if (err) goto out; @@ -1393,12 +1386,6 @@ static int move_data_page(struct inode *inode, block_t bidx, int gc_type, goto out; } - if (f2fs_is_atomic_file(inode)) { - F2FS_I(inode)->i_gc_failures[GC_FAILURE_ATOMIC]++; - F2FS_I_SB(inode)->skipped_atomic_files[gc_type]++; - err = -EAGAIN; - goto out; - } err = f2fs_gc_pinned_control(inode, gc_type, segno); if (err) goto out; @@ -1765,8 +1752,6 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, .ilist = LIST_HEAD_INIT(gc_list.ilist), .iroot = RADIX_TREE_INIT(gc_list.iroot, GFP_NOFS), }; - unsigned long long last_skipped = sbi->skipped_atomic_files[FG_GC]; - unsigned long long first_skipped; unsigned int skipped_round = 0, round = 0; trace_f2fs_gc_begin(sbi->sb, sync, background, @@ -1780,7 +1765,6 @@ int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, cpc.reason = __get_cp_reason(sbi); sbi->skipped_gc_rwsem = 0; - first_skipped = last_skipped; gc_more: if (unlikely(!(sbi->sb->s_flags & SB_ACTIVE))) { ret = -EINVAL; @@ -1831,10 +1815,8 @@ retry: total_freed += seg_freed; if (gc_type == FG_GC) { - if (sbi->skipped_atomic_files[FG_GC] > last_skipped || - sbi->skipped_gc_rwsem) + if (sbi->skipped_gc_rwsem) skipped_round++; - last_skipped = sbi->skipped_atomic_files[FG_GC]; round++; } @@ -1860,13 +1842,6 @@ retry: segno = NULL_SEGNO; goto gc_more; } - if (first_skipped < last_skipped && - (last_skipped - first_skipped) > - sbi->skipped_gc_rwsem) { - f2fs_drop_inmem_pages_all(sbi, true); - segno = NULL_SEGNO; - goto gc_more; - } if (gc_type == FG_GC && !is_sbi_flag_set(sbi, SBI_CP_DISABLED)) ret = f2fs_write_checkpoint(sbi, &cpc); stop: diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index 02630c17da93..2fce8fa0dac8 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -745,9 +745,8 @@ void f2fs_evict_inode(struct inode *inode) nid_t xnid = F2FS_I(inode)->i_xattr_nid; int err = 0; - /* some remained atomic pages should discarded */ if (f2fs_is_atomic_file(inode)) - f2fs_drop_inmem_pages(inode); + f2fs_abort_atomic_write(inode, true); trace_f2fs_evict_inode(inode); truncate_inode_pages_final(&inode->i_data); diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c index 37bdda931e0c..c549acb52ac4 100644 --- a/fs/f2fs/namei.c +++ b/fs/f2fs/namei.c @@ -840,8 +840,8 @@ out: } static int __f2fs_tmpfile(struct user_namespace *mnt_userns, struct inode *dir, - struct dentry *dentry, umode_t mode, - struct inode **whiteout) + struct dentry *dentry, umode_t mode, bool is_whiteout, + struct inode **new_inode) { struct f2fs_sb_info *sbi = F2FS_I_SB(dir); struct inode *inode; @@ -855,7 +855,7 @@ static int __f2fs_tmpfile(struct user_namespace *mnt_userns, struct inode *dir, if (IS_ERR(inode)) return PTR_ERR(inode); - if (whiteout) { + if (is_whiteout) { init_special_inode(inode, inode->i_mode, WHITEOUT_DEV); inode->i_op = &f2fs_special_inode_operations; } else { @@ -880,21 +880,25 @@ static int __f2fs_tmpfile(struct user_namespace *mnt_userns, struct inode *dir, f2fs_add_orphan_inode(inode); f2fs_alloc_nid_done(sbi, inode->i_ino); - if (whiteout) { + if (is_whiteout) { f2fs_i_links_write(inode, false); spin_lock(&inode->i_lock); inode->i_state |= I_LINKABLE; spin_unlock(&inode->i_lock); - - *whiteout = inode; } else { - d_tmpfile(dentry, inode); + if (dentry) + d_tmpfile(dentry, inode); + else + f2fs_i_links_write(inode, false); } /* link_count was changed by d_tmpfile as well. */ f2fs_unlock_op(sbi); unlock_new_inode(inode); + if (new_inode) + *new_inode = inode; + f2fs_balance_fs(sbi, true); return 0; @@ -915,7 +919,7 @@ static int f2fs_tmpfile(struct user_namespace *mnt_userns, struct inode *dir, if (!f2fs_is_checkpoint_ready(sbi)) return -ENOSPC; - return __f2fs_tmpfile(mnt_userns, dir, dentry, mode, NULL); + return __f2fs_tmpfile(mnt_userns, dir, dentry, mode, false, NULL); } static int f2fs_create_whiteout(struct user_namespace *mnt_userns, @@ -925,7 +929,13 @@ static int f2fs_create_whiteout(struct user_namespace *mnt_userns, return -EIO; return __f2fs_tmpfile(mnt_userns, dir, NULL, - S_IFCHR | WHITEOUT_MODE, whiteout); + S_IFCHR | WHITEOUT_MODE, true, whiteout); +} + +int f2fs_get_tmpfile(struct user_namespace *mnt_userns, struct inode *dir, + struct inode **new_inode) +{ + return __f2fs_tmpfile(mnt_userns, dir, NULL, S_IFREG, false, new_inode); } static int f2fs_rename(struct user_namespace *mnt_userns, struct inode *old_dir, diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 51230cba841b..beda8cbb791d 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -90,10 +90,6 @@ bool f2fs_available_free_memory(struct f2fs_sb_info *sbi, int type) atomic_read(&sbi->total_ext_node) * sizeof(struct extent_node)) >> PAGE_SHIFT; res = mem_size < ((avail_ram * nm_i->ram_thresh / 100) >> 1); - } else if (type == INMEM_PAGES) { - /* it allows 20% / total_ram for inmemory pages */ - mem_size = get_pages(sbi, F2FS_INMEM_PAGES); - res = mem_size < (val.totalram / 5); } else if (type == DISCARD_CACHE) { mem_size = (atomic_read(&dcc->discard_cmd_cnt) * sizeof(struct discard_cmd)) >> PAGE_SHIFT; diff --git a/fs/f2fs/node.h b/fs/f2fs/node.h index 4c1d34bfea78..3c09cae058b0 100644 --- a/fs/f2fs/node.h +++ b/fs/f2fs/node.h @@ -147,7 +147,6 @@ enum mem_type { DIRTY_DENTS, /* indicates dirty dentry pages */ INO_ENTRIES, /* indicates inode entries */ EXTENT_CACHE, /* indicates extent cache */ - INMEM_PAGES, /* indicates inmemory pages */ DISCARD_CACHE, /* indicates memory of cached discard cmds */ COMPRESS_PAGE, /* indicates memory of cached compressed pages */ BASE_CHECK, /* check kernel status */ diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index 87ff2b3cdf94..c0d7118fe171 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -30,7 +30,7 @@ static struct kmem_cache *discard_entry_slab; static struct kmem_cache *discard_cmd_slab; static struct kmem_cache *sit_entry_set_slab; -static struct kmem_cache *inmem_entry_slab; +static struct kmem_cache *revoke_entry_slab; static unsigned long __reverse_ulong(unsigned char *str) { @@ -185,304 +185,180 @@ bool f2fs_need_SSR(struct f2fs_sb_info *sbi) SM_I(sbi)->min_ssr_sections + reserved_sections(sbi)); } -void f2fs_register_inmem_page(struct inode *inode, struct page *page) +void f2fs_abort_atomic_write(struct inode *inode, bool clean) { - struct inmem_pages *new; - - set_page_private_atomic(page); - - new = f2fs_kmem_cache_alloc(inmem_entry_slab, - GFP_NOFS, true, NULL); - - /* add atomic page indices to the list */ - new->page = page; - INIT_LIST_HEAD(&new->list); + struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + struct f2fs_inode_info *fi = F2FS_I(inode); - /* increase reference count with clean state */ - get_page(page); - mutex_lock(&F2FS_I(inode)->inmem_lock); - list_add_tail(&new->list, &F2FS_I(inode)->inmem_pages); - inc_page_count(F2FS_I_SB(inode), F2FS_INMEM_PAGES); - mutex_unlock(&F2FS_I(inode)->inmem_lock); + if (f2fs_is_atomic_file(inode)) { + if (clean) + truncate_inode_pages_final(inode->i_mapping); + clear_inode_flag(fi->cow_inode, FI_ATOMIC_FILE); + iput(fi->cow_inode); + fi->cow_inode = NULL; + clear_inode_flag(inode, FI_ATOMIC_FILE); - trace_f2fs_register_inmem_page(page, INMEM); + spin_lock(&sbi->inode_lock[ATOMIC_FILE]); + sbi->atomic_files--; + spin_unlock(&sbi->inode_lock[ATOMIC_FILE]); + } } -static int __revoke_inmem_pages(struct inode *inode, - struct list_head *head, bool drop, bool recover, - bool trylock) +static int __replace_atomic_write_block(struct inode *inode, pgoff_t index, + block_t new_addr, block_t *old_addr, bool recover) { struct f2fs_sb_info *sbi = F2FS_I_SB(inode); - struct inmem_pages *cur, *tmp; - int err = 0; - - list_for_each_entry_safe(cur, tmp, head, list) { - struct page *page = cur->page; - - if (drop) - trace_f2fs_commit_inmem_page(page, INMEM_DROP); - - if (trylock) { - /* - * to avoid deadlock in between page lock and - * inmem_lock. - */ - if (!trylock_page(page)) - continue; - } else { - lock_page(page); - } - - f2fs_wait_on_page_writeback(page, DATA, true, true); - - if (recover) { - struct dnode_of_data dn; - struct node_info ni; + struct dnode_of_data dn; + struct node_info ni; + int err; - trace_f2fs_commit_inmem_page(page, INMEM_REVOKE); retry: - set_new_dnode(&dn, inode, NULL, NULL, 0); - err = f2fs_get_dnode_of_data(&dn, page->index, - LOOKUP_NODE); - if (err) { - if (err == -ENOMEM) { - memalloc_retry_wait(GFP_NOFS); - goto retry; - } - err = -EAGAIN; - goto next; - } - - err = f2fs_get_node_info(sbi, dn.nid, &ni, false); - if (err) { - f2fs_put_dnode(&dn); - return err; - } - - if (cur->old_addr == NEW_ADDR) { - f2fs_invalidate_blocks(sbi, dn.data_blkaddr); - f2fs_update_data_blkaddr(&dn, NEW_ADDR); - } else - f2fs_replace_block(sbi, &dn, dn.data_blkaddr, - cur->old_addr, ni.version, true, true); - f2fs_put_dnode(&dn); - } -next: - /* we don't need to invalidate this in the sccessful status */ - if (drop || recover) { - ClearPageUptodate(page); - clear_page_private_gcing(page); + set_new_dnode(&dn, inode, NULL, NULL, 0); + err = f2fs_get_dnode_of_data(&dn, index, LOOKUP_NODE_RA); + if (err) { + if (err == -ENOMEM) { + f2fs_io_schedule_timeout(DEFAULT_IO_TIMEOUT); + goto retry; } - detach_page_private(page); - set_page_private(page, 0); - f2fs_put_page(page, 1); - - list_del(&cur->list); - kmem_cache_free(inmem_entry_slab, cur); - dec_page_count(F2FS_I_SB(inode), F2FS_INMEM_PAGES); + return err; } - return err; -} -void f2fs_drop_inmem_pages_all(struct f2fs_sb_info *sbi, bool gc_failure) -{ - struct list_head *head = &sbi->inode_list[ATOMIC_FILE]; - struct inode *inode; - struct f2fs_inode_info *fi; - unsigned int count = sbi->atomic_files; - unsigned int looped = 0; -next: - spin_lock(&sbi->inode_lock[ATOMIC_FILE]); - if (list_empty(head)) { - spin_unlock(&sbi->inode_lock[ATOMIC_FILE]); - return; + err = f2fs_get_node_info(sbi, dn.nid, &ni, false); + if (err) { + f2fs_put_dnode(&dn); + return err; } - fi = list_first_entry(head, struct f2fs_inode_info, inmem_ilist); - inode = igrab(&fi->vfs_inode); - if (inode) - list_move_tail(&fi->inmem_ilist, head); - spin_unlock(&sbi->inode_lock[ATOMIC_FILE]); - if (inode) { - if (gc_failure) { - if (!fi->i_gc_failures[GC_FAILURE_ATOMIC]) - goto skip; + if (recover) { + /* dn.data_blkaddr is always valid */ + if (!__is_valid_data_blkaddr(new_addr)) { + if (new_addr == NULL_ADDR) + dec_valid_block_count(sbi, inode, 1); + f2fs_invalidate_blocks(sbi, dn.data_blkaddr); + f2fs_update_data_blkaddr(&dn, new_addr); + } else { + f2fs_replace_block(sbi, &dn, dn.data_blkaddr, + new_addr, ni.version, true, true); } - set_inode_flag(inode, FI_ATOMIC_REVOKE_REQUEST); - f2fs_drop_inmem_pages(inode); -skip: - iput(inode); - } - f2fs_io_schedule_timeout(DEFAULT_IO_TIMEOUT); - if (gc_failure) { - if (++looped >= count) - return; - } - goto next; -} - -void f2fs_drop_inmem_pages(struct inode *inode) -{ - struct f2fs_sb_info *sbi = F2FS_I_SB(inode); - struct f2fs_inode_info *fi = F2FS_I(inode); + } else { + blkcnt_t count = 1; - do { - mutex_lock(&fi->inmem_lock); - if (list_empty(&fi->inmem_pages)) { - fi->i_gc_failures[GC_FAILURE_ATOMIC] = 0; - - spin_lock(&sbi->inode_lock[ATOMIC_FILE]); - if (!list_empty(&fi->inmem_ilist)) - list_del_init(&fi->inmem_ilist); - if (f2fs_is_atomic_file(inode)) { - clear_inode_flag(inode, FI_ATOMIC_FILE); - sbi->atomic_files--; - } - spin_unlock(&sbi->inode_lock[ATOMIC_FILE]); + *old_addr = dn.data_blkaddr; + f2fs_truncate_data_blocks_range(&dn, 1); + dec_valid_block_count(sbi, F2FS_I(inode)->cow_inode, count); + inc_valid_block_count(sbi, inode, &count); + f2fs_replace_block(sbi, &dn, dn.data_blkaddr, new_addr, + ni.version, true, false); + } - mutex_unlock(&fi->inmem_lock); - break; - } - __revoke_inmem_pages(inode, &fi->inmem_pages, - true, false, true); - mutex_unlock(&fi->inmem_lock); - } while (1); + f2fs_put_dnode(&dn); + return 0; } -void f2fs_drop_inmem_page(struct inode *inode, struct page *page) +static void __complete_revoke_list(struct inode *inode, struct list_head *head, + bool revoke) { - struct f2fs_inode_info *fi = F2FS_I(inode); - struct f2fs_sb_info *sbi = F2FS_I_SB(inode); - struct list_head *head = &fi->inmem_pages; - struct inmem_pages *cur = NULL; - struct inmem_pages *tmp; - - f2fs_bug_on(sbi, !page_private_atomic(page)); + struct revoke_entry *cur, *tmp; - mutex_lock(&fi->inmem_lock); - list_for_each_entry(tmp, head, list) { - if (tmp->page == page) { - cur = tmp; - break; - } + list_for_each_entry_safe(cur, tmp, head, list) { + if (revoke) + __replace_atomic_write_block(inode, cur->index, + cur->old_addr, NULL, true); + list_del(&cur->list); + kmem_cache_free(revoke_entry_slab, cur); } - - f2fs_bug_on(sbi, !cur); - list_del(&cur->list); - mutex_unlock(&fi->inmem_lock); - - dec_page_count(sbi, F2FS_INMEM_PAGES); - kmem_cache_free(inmem_entry_slab, cur); - - ClearPageUptodate(page); - clear_page_private_atomic(page); - f2fs_put_page(page, 0); - - detach_page_private(page); - set_page_private(page, 0); - - trace_f2fs_commit_inmem_page(page, INMEM_INVALIDATE); } -static int __f2fs_commit_inmem_pages(struct inode *inode) +static int __f2fs_commit_atomic_write(struct inode *inode) { struct f2fs_sb_info *sbi = F2FS_I_SB(inode); struct f2fs_inode_info *fi = F2FS_I(inode); - struct inmem_pages *cur, *tmp; - struct f2fs_io_info fio = { - .sbi = sbi, - .ino = inode->i_ino, - .type = DATA, - .op = REQ_OP_WRITE, - .op_flags = REQ_SYNC | REQ_PRIO, - .io_type = FS_DATA_IO, - }; + struct inode *cow_inode = fi->cow_inode; + struct revoke_entry *new; struct list_head revoke_list; - bool submit_bio = false; - int err = 0; + block_t blkaddr; + struct dnode_of_data dn; + pgoff_t len = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE); + pgoff_t off = 0, blen, index; + int ret = 0, i; INIT_LIST_HEAD(&revoke_list); - list_for_each_entry_safe(cur, tmp, &fi->inmem_pages, list) { - struct page *page = cur->page; + while (len) { + blen = min_t(pgoff_t, ADDRS_PER_BLOCK(cow_inode), len); - lock_page(page); - if (page->mapping == inode->i_mapping) { - trace_f2fs_commit_inmem_page(page, INMEM); + set_new_dnode(&dn, cow_inode, NULL, NULL, 0); + ret = f2fs_get_dnode_of_data(&dn, off, LOOKUP_NODE_RA); + if (ret && ret != -ENOENT) { + goto out; + } else if (ret == -ENOENT) { + ret = 0; + if (dn.max_level == 0) + goto out; + goto next; + } - f2fs_wait_on_page_writeback(page, DATA, true, true); + blen = min((pgoff_t)ADDRS_PER_PAGE(dn.node_page, cow_inode), + len); + index = off; + for (i = 0; i < blen; i++, dn.ofs_in_node++, index++) { + blkaddr = f2fs_data_blkaddr(&dn); - set_page_dirty(page); - if (clear_page_dirty_for_io(page)) { - inode_dec_dirty_pages(inode); - f2fs_remove_dirty_inode(inode); - } -retry: - fio.page = page; - fio.old_blkaddr = NULL_ADDR; - fio.encrypted_page = NULL; - fio.need_lock = LOCK_DONE; - err = f2fs_do_write_data_page(&fio); - if (err) { - if (err == -ENOMEM) { - memalloc_retry_wait(GFP_NOFS); - goto retry; - } - unlock_page(page); - break; + if (!__is_valid_data_blkaddr(blkaddr)) { + continue; + } else if (!f2fs_is_valid_blkaddr(sbi, blkaddr, + DATA_GENERIC_ENHANCE)) { + f2fs_put_dnode(&dn); + ret = -EFSCORRUPTED; + goto out; } - /* record old blkaddr for revoking */ - cur->old_addr = fio.old_blkaddr; - submit_bio = true; - } - unlock_page(page); - list_move_tail(&cur->list, &revoke_list); - } - if (submit_bio) - f2fs_submit_merged_write_cond(sbi, inode, NULL, 0, DATA); + new = f2fs_kmem_cache_alloc(revoke_entry_slab, GFP_NOFS, + true, NULL); + if (!new) { + f2fs_put_dnode(&dn); + ret = -ENOMEM; + goto out; + } - if (err) { - /* - * try to revoke all committed pages, but still we could fail - * due to no memory or other reason, if that happened, EAGAIN - * will be returned, which means in such case, transaction is - * already not integrity, caller should use journal to do the - * recovery or rewrite & commit last transaction. For other - * error number, revoking was done by filesystem itself. - */ - err = __revoke_inmem_pages(inode, &revoke_list, - false, true, false); + ret = __replace_atomic_write_block(inode, index, blkaddr, + &new->old_addr, false); + if (ret) { + f2fs_put_dnode(&dn); + kmem_cache_free(revoke_entry_slab, new); + goto out; + } - /* drop all uncommitted pages */ - __revoke_inmem_pages(inode, &fi->inmem_pages, - true, false, false); - } else { - __revoke_inmem_pages(inode, &revoke_list, - false, false, false); + f2fs_update_data_blkaddr(&dn, NULL_ADDR); + new->index = index; + list_add_tail(&new->list, &revoke_list); + } + f2fs_put_dnode(&dn); +next: + off += blen; + len -= blen; } - return err; +out: + __complete_revoke_list(inode, &revoke_list, ret ? true : false); + + return ret; } -int f2fs_commit_inmem_pages(struct inode *inode) +int f2fs_commit_atomic_write(struct inode *inode) { struct f2fs_sb_info *sbi = F2FS_I_SB(inode); struct f2fs_inode_info *fi = F2FS_I(inode); int err; - f2fs_balance_fs(sbi, true); + err = filemap_write_and_wait_range(inode->i_mapping, 0, LLONG_MAX); + if (err) + return err; f2fs_down_write(&fi->i_gc_rwsem[WRITE]); - f2fs_lock_op(sbi); - set_inode_flag(inode, FI_ATOMIC_COMMIT); - - mutex_lock(&fi->inmem_lock); - err = __f2fs_commit_inmem_pages(inode); - mutex_unlock(&fi->inmem_lock); - clear_inode_flag(inode, FI_ATOMIC_COMMIT); + err = __f2fs_commit_atomic_write(inode); f2fs_unlock_op(sbi); f2fs_up_write(&fi->i_gc_rwsem[WRITE]); @@ -5360,9 +5236,9 @@ int __init f2fs_create_segment_manager_caches(void) if (!sit_entry_set_slab) goto destroy_discard_cmd; - inmem_entry_slab = f2fs_kmem_cache_create("f2fs_inmem_page_entry", - sizeof(struct inmem_pages)); - if (!inmem_entry_slab) + revoke_entry_slab = f2fs_kmem_cache_create("f2fs_revoke_entry", + sizeof(struct revoke_entry)); + if (!revoke_entry_slab) goto destroy_sit_entry_set; return 0; @@ -5381,5 +5257,5 @@ void f2fs_destroy_segment_manager_caches(void) kmem_cache_destroy(sit_entry_set_slab); kmem_cache_destroy(discard_cmd_slab); kmem_cache_destroy(discard_entry_slab); - kmem_cache_destroy(inmem_entry_slab); + kmem_cache_destroy(revoke_entry_slab); } diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h index 8fbc9f6afa55..3f277dfcb131 100644 --- a/fs/f2fs/segment.h +++ b/fs/f2fs/segment.h @@ -225,10 +225,10 @@ struct segment_allocation { #define MAX_SKIP_GC_COUNT 16 -struct inmem_pages { +struct revoke_entry { struct list_head list; - struct page *page; block_t old_addr; /* for revoking when fail to commit */ + pgoff_t index; }; struct sit_info { diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index aa51c30333d3..0900c552a16c 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -1339,9 +1339,6 @@ static struct inode *f2fs_alloc_inode(struct super_block *sb) spin_lock_init(&fi->i_size_lock); INIT_LIST_HEAD(&fi->dirty_list); INIT_LIST_HEAD(&fi->gdirty_list); - INIT_LIST_HEAD(&fi->inmem_ilist); - INIT_LIST_HEAD(&fi->inmem_pages); - mutex_init(&fi->inmem_lock); init_f2fs_rwsem(&fi->i_gc_rwsem[READ]); init_f2fs_rwsem(&fi->i_gc_rwsem[WRITE]); init_f2fs_rwsem(&fi->i_xattr_sem); @@ -1382,9 +1379,8 @@ static int f2fs_drop_inode(struct inode *inode) atomic_inc(&inode->i_count); spin_unlock(&inode->i_lock); - /* some remained atomic pages should discarded */ if (f2fs_is_atomic_file(inode)) - f2fs_drop_inmem_pages(inode); + f2fs_abort_atomic_write(inode, true); /* should remain fi->extent_tree for writepage */ f2fs_destroy_extent_node(inode); diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h index 4d1ad64d4cab..7e915dbf3674 100644 --- a/include/trace/events/f2fs.h +++ b/include/trace/events/f2fs.h @@ -15,10 +15,6 @@ TRACE_DEFINE_ENUM(NODE); TRACE_DEFINE_ENUM(DATA); TRACE_DEFINE_ENUM(META); TRACE_DEFINE_ENUM(META_FLUSH); -TRACE_DEFINE_ENUM(INMEM); -TRACE_DEFINE_ENUM(INMEM_DROP); -TRACE_DEFINE_ENUM(INMEM_INVALIDATE); -TRACE_DEFINE_ENUM(INMEM_REVOKE); TRACE_DEFINE_ENUM(IPU); TRACE_DEFINE_ENUM(OPU); TRACE_DEFINE_ENUM(HOT); @@ -59,10 +55,6 @@ TRACE_DEFINE_ENUM(CP_RESIZE); { DATA, "DATA" }, \ { META, "META" }, \ { META_FLUSH, "META_FLUSH" }, \ - { INMEM, "INMEM" }, \ - { INMEM_DROP, "INMEM_DROP" }, \ - { INMEM_INVALIDATE, "INMEM_INVALIDATE" }, \ - { INMEM_REVOKE, "INMEM_REVOKE" }, \ { IPU, "IN-PLACE" }, \ { OPU, "OUT-OF-PLACE" }) @@ -1289,20 +1281,6 @@ DEFINE_EVENT(f2fs__page, f2fs_vm_page_mkwrite, TP_ARGS(page, type) ); -DEFINE_EVENT(f2fs__page, f2fs_register_inmem_page, - - TP_PROTO(struct page *page, int type), - - TP_ARGS(page, type) -); - -DEFINE_EVENT(f2fs__page, f2fs_commit_inmem_page, - - TP_PROTO(struct page *page, int type), - - TP_ARGS(page, type) -); - TRACE_EVENT(f2fs_filemap_fault, TP_PROTO(struct inode *inode, pgoff_t index, unsigned long ret), -- cgit v1.3-8-gc7d7 From d147ea4adb969b03f7f1c7613cc6f44b70760eb3 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Fri, 6 May 2022 11:40:33 -0700 Subject: f2fs: introduce f2fs_gc_control to consolidate f2fs_gc parameters No functional change. - remove checkpoint=disable check for f2fs_write_checkpoint - get sec_freed all the time Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 11 +++++-- fs/f2fs/file.c | 30 +++++++++++++++--- fs/f2fs/gc.c | 74 +++++++++++++++++++++++++-------------------- fs/f2fs/segment.c | 8 ++++- fs/f2fs/super.c | 8 ++++- include/trace/events/f2fs.h | 18 +++++------ 6 files changed, 98 insertions(+), 51 deletions(-) (limited to 'include/trace') diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 0699d7460d5d..9920b2d6af8f 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -1265,6 +1265,14 @@ struct atgc_management { unsigned long long age_threshold; /* age threshold */ }; +struct f2fs_gc_control { + unsigned int victim_segno; /* target victim segment number */ + int init_gc_type; /* FG_GC or BG_GC */ + bool no_bg_gc; /* check the space and stop bg_gc */ + bool should_migrate_blocks; /* should migrate blocks */ + bool err_gc_skipped; /* return EAGAIN if GC skipped */ +}; + /* For s_flag in struct f2fs_sb_info */ enum { SBI_IS_DIRTY, /* dirty flag for checkpoint */ @@ -3761,8 +3769,7 @@ extern const struct iomap_ops f2fs_iomap_ops; int f2fs_start_gc_thread(struct f2fs_sb_info *sbi); void f2fs_stop_gc_thread(struct f2fs_sb_info *sbi); block_t f2fs_start_bidx_of_node(unsigned int node_ofs, struct inode *inode); -int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, bool background, bool force, - unsigned int segno); +int f2fs_gc(struct f2fs_sb_info *sbi, struct f2fs_gc_control *gc_control); void f2fs_build_gc_manager(struct f2fs_sb_info *sbi); int f2fs_resize_fs(struct f2fs_sb_info *sbi, __u64 block_count); int __init f2fs_create_garbage_collection_cache(void); diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 6da8a663de7b..d0547bef0851 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -1647,6 +1647,10 @@ static int expand_inode_data(struct inode *inode, loff_t offset, struct f2fs_map_blocks map = { .m_next_pgofs = NULL, .m_next_extent = NULL, .m_seg_type = NO_CHECK_TYPE, .m_may_create = true }; + struct f2fs_gc_control gc_control = { .victim_segno = NULL_SEGNO, + .init_gc_type = FG_GC, + .should_migrate_blocks = false, + .err_gc_skipped = true }; pgoff_t pg_start, pg_end; loff_t new_size = i_size_read(inode); loff_t off_end; @@ -1684,7 +1688,7 @@ next_alloc: if (has_not_enough_free_secs(sbi, 0, GET_SEC_FROM_SEG(sbi, overprovision_segments(sbi)))) { f2fs_down_write(&sbi->gc_lock); - err = f2fs_gc(sbi, true, false, false, NULL_SEGNO); + err = f2fs_gc(sbi, &gc_control); if (err && err != -ENODATA) goto out_err; } @@ -2344,6 +2348,9 @@ static int f2fs_ioc_gc(struct file *filp, unsigned long arg) { struct inode *inode = file_inode(filp); struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + struct f2fs_gc_control gc_control = { .victim_segno = NULL_SEGNO, + .no_bg_gc = false, + .should_migrate_blocks = false }; __u32 sync; int ret; @@ -2369,7 +2376,9 @@ static int f2fs_ioc_gc(struct file *filp, unsigned long arg) f2fs_down_write(&sbi->gc_lock); } - ret = f2fs_gc(sbi, sync, true, false, NULL_SEGNO); + gc_control.init_gc_type = sync ? FG_GC : BG_GC; + gc_control.err_gc_skipped = sync; + ret = f2fs_gc(sbi, &gc_control); out: mnt_drop_write_file(filp); return ret; @@ -2378,6 +2387,11 @@ out: static int __f2fs_ioc_gc_range(struct file *filp, struct f2fs_gc_range *range) { struct f2fs_sb_info *sbi = F2FS_I_SB(file_inode(filp)); + struct f2fs_gc_control gc_control = { + .init_gc_type = range->sync ? FG_GC : BG_GC, + .no_bg_gc = false, + .should_migrate_blocks = false, + .err_gc_skipped = range->sync }; u64 end; int ret; @@ -2405,8 +2419,8 @@ do_more: f2fs_down_write(&sbi->gc_lock); } - ret = f2fs_gc(sbi, range->sync, true, false, - GET_SEGNO(sbi, range->start)); + gc_control.victim_segno = GET_SEGNO(sbi, range->start); + ret = f2fs_gc(sbi, &gc_control); if (ret) { if (ret == -EBUSY) ret = -EAGAIN; @@ -2820,6 +2834,10 @@ static int f2fs_ioc_flush_device(struct file *filp, unsigned long arg) unsigned int start_segno = 0, end_segno = 0; unsigned int dev_start_segno = 0, dev_end_segno = 0; struct f2fs_flush_device range; + struct f2fs_gc_control gc_control = { + .init_gc_type = FG_GC, + .should_migrate_blocks = true, + .err_gc_skipped = true }; int ret; if (!capable(CAP_SYS_ADMIN)) @@ -2863,7 +2881,9 @@ static int f2fs_ioc_flush_device(struct file *filp, unsigned long arg) sm->last_victim[GC_CB] = end_segno + 1; sm->last_victim[GC_GREEDY] = end_segno + 1; sm->last_victim[ALLOC_NEXT] = end_segno + 1; - ret = f2fs_gc(sbi, true, true, true, start_segno); + + gc_control.victim_segno = start_segno; + ret = f2fs_gc(sbi, &gc_control); if (ret == -EAGAIN) ret = 0; else if (ret < 0) diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index ba8e93e517be..f3d58d154240 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -35,6 +35,9 @@ static int gc_thread_func(void *data) wait_queue_head_t *wq = &sbi->gc_thread->gc_wait_queue_head; wait_queue_head_t *fggc_wq = &sbi->gc_thread->fggc_wq; unsigned int wait_ms; + struct f2fs_gc_control gc_control = { + .victim_segno = NULL_SEGNO, + .should_migrate_blocks = false }; wait_ms = gc_th->min_sleep_time; @@ -141,8 +144,12 @@ do_gc: if (foreground) sync_mode = false; + gc_control.init_gc_type = sync_mode ? FG_GC : BG_GC; + gc_control.no_bg_gc = foreground; + gc_control.err_gc_skipped = sync_mode; + /* if return value is not zero, no victim was selected */ - if (f2fs_gc(sbi, sync_mode, !foreground, false, NULL_SEGNO)) + if (f2fs_gc(sbi, &gc_control)) wait_ms = gc_th->no_gc_sleep_time; if (foreground) @@ -1740,21 +1747,20 @@ skip: return seg_freed; } -int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, - bool background, bool force, unsigned int segno) +int f2fs_gc(struct f2fs_sb_info *sbi, struct f2fs_gc_control *gc_control) { - int gc_type = sync ? FG_GC : BG_GC; + int gc_type = gc_control->init_gc_type; + unsigned int segno = gc_control->victim_segno; int sec_freed = 0, seg_freed = 0, total_freed = 0; int ret = 0; struct cp_control cpc; - unsigned int init_segno = segno; struct gc_inode_list gc_list = { .ilist = LIST_HEAD_INIT(gc_list.ilist), .iroot = RADIX_TREE_INIT(gc_list.iroot, GFP_NOFS), }; unsigned int skipped_round = 0, round = 0; - trace_f2fs_gc_begin(sbi->sb, sync, background, + trace_f2fs_gc_begin(sbi->sb, gc_type, gc_control->no_bg_gc, get_pages(sbi, F2FS_DIRTY_NODES), get_pages(sbi, F2FS_DIRTY_DENTS), get_pages(sbi, F2FS_DIRTY_IMETA), @@ -1781,8 +1787,7 @@ gc_more: * threshold, we can make them free by checkpoint. Then, we * secure free segments which doesn't need fggc any more. */ - if (prefree_segments(sbi) && - !is_sbi_flag_set(sbi, SBI_CP_DISABLED)) { + if (prefree_segments(sbi)) { ret = f2fs_write_checkpoint(sbi, &cpc); if (ret) goto stop; @@ -1792,7 +1797,7 @@ gc_more: } /* f2fs_balance_fs doesn't need to do BG_GC in critical path. */ - if (gc_type == BG_GC && !background) { + if (gc_type == BG_GC && gc_control->no_bg_gc) { ret = -EINVAL; goto stop; } @@ -1808,45 +1813,48 @@ retry: goto stop; } - seg_freed = do_garbage_collect(sbi, segno, &gc_list, gc_type, force); - if (gc_type == FG_GC && - seg_freed == f2fs_usable_segs_in_sec(sbi, segno)) - sec_freed++; + seg_freed = do_garbage_collect(sbi, segno, &gc_list, gc_type, + gc_control->should_migrate_blocks); total_freed += seg_freed; - if (gc_type == FG_GC) { - if (sbi->skipped_gc_rwsem) - skipped_round++; - round++; - } + if (seg_freed == f2fs_usable_segs_in_sec(sbi, segno)) + sec_freed++; if (gc_type == FG_GC) sbi->cur_victim_sec = NULL_SEGNO; - if (sync) + if (gc_control->init_gc_type == FG_GC) goto stop; - if (!has_not_enough_free_secs(sbi, sec_freed, 0)) + if (!has_not_enough_free_secs(sbi, + (gc_type == FG_GC) ? sec_freed : 0, 0)) goto stop; - if (skipped_round <= MAX_SKIP_GC_COUNT || skipped_round * 2 < round) { - - /* Write checkpoint to reclaim prefree segments */ - if (free_sections(sbi) < NR_CURSEG_PERSIST_TYPE && - prefree_segments(sbi) && - !is_sbi_flag_set(sbi, SBI_CP_DISABLED)) { + /* FG_GC stops GC by skip_count */ + if (gc_type == FG_GC) { + if (sbi->skipped_gc_rwsem) + skipped_round++; + round++; + if (skipped_round > MAX_SKIP_GC_COUNT && + skipped_round * 2 >= round) { ret = f2fs_write_checkpoint(sbi, &cpc); - if (ret) - goto stop; + goto stop; } - segno = NULL_SEGNO; - goto gc_more; } - if (gc_type == FG_GC && !is_sbi_flag_set(sbi, SBI_CP_DISABLED)) + + /* Write checkpoint to reclaim prefree segments */ + if (free_sections(sbi) < NR_CURSEG_PERSIST_TYPE && + prefree_segments(sbi)) { ret = f2fs_write_checkpoint(sbi, &cpc); + if (ret) + goto stop; + } + segno = NULL_SEGNO; + goto gc_more; + stop: SIT_I(sbi)->last_victim[ALLOC_NEXT] = 0; - SIT_I(sbi)->last_victim[FLUSH_DEVICE] = init_segno; + SIT_I(sbi)->last_victim[FLUSH_DEVICE] = gc_control->victim_segno; if (gc_type == FG_GC) f2fs_unpin_all_sections(sbi, true); @@ -1864,7 +1872,7 @@ stop: put_gc_inode(&gc_list); - if (sync && !ret) + if (gc_control->err_gc_skipped && !ret) ret = sec_freed ? 0 : -EAGAIN; return ret; } diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index 47934995e2ca..8b4f2b1d2cca 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -399,8 +399,14 @@ void f2fs_balance_fs(struct f2fs_sb_info *sbi, bool need) io_schedule(); finish_wait(&sbi->gc_thread->fggc_wq, &wait); } else { + struct f2fs_gc_control gc_control = { + .victim_segno = NULL_SEGNO, + .init_gc_type = BG_GC, + .no_bg_gc = true, + .should_migrate_blocks = false, + .err_gc_skipped = false }; f2fs_down_write(&sbi->gc_lock); - f2fs_gc(sbi, false, false, false, NULL_SEGNO); + f2fs_gc(sbi, &gc_control); } } } diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 8c81dd324297..a28c27eed6d0 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -2076,8 +2076,14 @@ static int f2fs_disable_checkpoint(struct f2fs_sb_info *sbi) sbi->gc_mode = GC_URGENT_HIGH; while (!f2fs_time_over(sbi, DISABLE_TIME)) { + struct f2fs_gc_control gc_control = { + .victim_segno = NULL_SEGNO, + .init_gc_type = FG_GC, + .should_migrate_blocks = false, + .err_gc_skipped = true }; + f2fs_down_write(&sbi->gc_lock); - err = f2fs_gc(sbi, true, false, false, NULL_SEGNO); + err = f2fs_gc(sbi, &gc_control); if (err == -ENODATA) { err = 0; break; diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h index 7e915dbf3674..54ec9e543f09 100644 --- a/include/trace/events/f2fs.h +++ b/include/trace/events/f2fs.h @@ -644,19 +644,19 @@ TRACE_EVENT(f2fs_background_gc, TRACE_EVENT(f2fs_gc_begin, - TP_PROTO(struct super_block *sb, bool sync, bool background, + TP_PROTO(struct super_block *sb, int gc_type, bool no_bg_gc, long long dirty_nodes, long long dirty_dents, long long dirty_imeta, unsigned int free_sec, unsigned int free_seg, int reserved_seg, unsigned int prefree_seg), - TP_ARGS(sb, sync, background, dirty_nodes, dirty_dents, dirty_imeta, + TP_ARGS(sb, gc_type, no_bg_gc, dirty_nodes, dirty_dents, dirty_imeta, free_sec, free_seg, reserved_seg, prefree_seg), TP_STRUCT__entry( __field(dev_t, dev) - __field(bool, sync) - __field(bool, background) + __field(int, gc_type) + __field(bool, no_bg_gc) __field(long long, dirty_nodes) __field(long long, dirty_dents) __field(long long, dirty_imeta) @@ -668,8 +668,8 @@ TRACE_EVENT(f2fs_gc_begin, TP_fast_assign( __entry->dev = sb->s_dev; - __entry->sync = sync; - __entry->background = background; + __entry->gc_type = gc_type; + __entry->no_bg_gc = no_bg_gc; __entry->dirty_nodes = dirty_nodes; __entry->dirty_dents = dirty_dents; __entry->dirty_imeta = dirty_imeta; @@ -679,12 +679,12 @@ TRACE_EVENT(f2fs_gc_begin, __entry->prefree_seg = prefree_seg; ), - TP_printk("dev = (%d,%d), sync = %d, background = %d, nodes = %lld, " + TP_printk("dev = (%d,%d), gc_type = %s, no_background_GC = %d, nodes = %lld, " "dents = %lld, imeta = %lld, free_sec:%u, free_seg:%u, " "rsv_seg:%d, prefree_seg:%u", show_dev(__entry->dev), - __entry->sync, - __entry->background, + show_gc_type(__entry->gc_type), + (__entry->gc_type == BG_GC) ? __entry->no_bg_gc : -1, __entry->dirty_nodes, __entry->dirty_dents, __entry->dirty_imeta, -- cgit v1.3-8-gc7d7 From f67bed134a053663852a1a3ab1b3223bfc2104a2 Mon Sep 17 00:00:00 2001 From: Vasily Averin Date: Thu, 12 May 2022 20:23:07 -0700 Subject: percpu: improve percpu_alloc_percpu event trace Add call_site, bytes_alloc and gfp_flags fields to the output of the percpu_alloc_percpu ftrace event: mkdir-4393 [001] 169.334788: percpu_alloc_percpu: call_site=mem_cgroup_css_alloc+0xa6 reserved=0 is_atomic=0 size=2408 align=8 base_addr=0xffffc7117fc00000 off=402176 ptr=0x3dc867a62300 bytes_alloc=14448 gfp_flags=GFP_KERNEL_ACCOUNT This is required to track memcg-accounted percpu allocations. Link: https://lkml.kernel.org/r/a07be858-c8a3-7851-9086-e3262cbcf707@openvz.org Signed-off-by: Vasily Averin Acked-by: Roman Gushchin Cc: Shakeel Butt Cc: Steven Rostedt Cc: Ingo Molnar Cc: Vlastimil Babka Cc: Michal Hocko Cc: Dennis Zhou Cc: Tejun Heo Cc: Christoph Lameter Signed-off-by: Andrew Morton --- include/trace/events/percpu.h | 23 +++++++++++++++++------ mm/percpu-internal.h | 8 ++++---- mm/percpu.c | 5 +++-- 3 files changed, 24 insertions(+), 12 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/percpu.h b/include/trace/events/percpu.h index df112a64f6c9..e989cefc0def 100644 --- a/include/trace/events/percpu.h +++ b/include/trace/events/percpu.h @@ -6,15 +6,20 @@ #define _TRACE_PERCPU_H #include +#include TRACE_EVENT(percpu_alloc_percpu, - TP_PROTO(bool reserved, bool is_atomic, size_t size, - size_t align, void *base_addr, int off, void __percpu *ptr), + TP_PROTO(unsigned long call_site, + bool reserved, bool is_atomic, size_t size, + size_t align, void *base_addr, int off, + void __percpu *ptr, size_t bytes_alloc, gfp_t gfp_flags), - TP_ARGS(reserved, is_atomic, size, align, base_addr, off, ptr), + TP_ARGS(call_site, reserved, is_atomic, size, align, base_addr, off, + ptr, bytes_alloc, gfp_flags), TP_STRUCT__entry( + __field( unsigned long, call_site ) __field( bool, reserved ) __field( bool, is_atomic ) __field( size_t, size ) @@ -22,9 +27,11 @@ TRACE_EVENT(percpu_alloc_percpu, __field( void *, base_addr ) __field( int, off ) __field( void __percpu *, ptr ) + __field( size_t, bytes_alloc ) + __field( gfp_t, gfp_flags ) ), - TP_fast_assign( + __entry->call_site = call_site; __entry->reserved = reserved; __entry->is_atomic = is_atomic; __entry->size = size; @@ -32,12 +39,16 @@ TRACE_EVENT(percpu_alloc_percpu, __entry->base_addr = base_addr; __entry->off = off; __entry->ptr = ptr; + __entry->bytes_alloc = bytes_alloc; + __entry->gfp_flags = gfp_flags; ), - TP_printk("reserved=%d is_atomic=%d size=%zu align=%zu base_addr=%p off=%d ptr=%p", + TP_printk("call_site=%pS reserved=%d is_atomic=%d size=%zu align=%zu base_addr=%p off=%d ptr=%p bytes_alloc=%zu gfp_flags=%s", + (void *)__entry->call_site, __entry->reserved, __entry->is_atomic, __entry->size, __entry->align, - __entry->base_addr, __entry->off, __entry->ptr) + __entry->base_addr, __entry->off, __entry->ptr, + __entry->bytes_alloc, show_gfp_flags(__entry->gfp_flags)) ); TRACE_EVENT(percpu_free_percpu, diff --git a/mm/percpu-internal.h b/mm/percpu-internal.h index 411d1593ef23..70b1ea23f4d2 100644 --- a/mm/percpu-internal.h +++ b/mm/percpu-internal.h @@ -113,7 +113,6 @@ static inline int pcpu_chunk_map_bits(struct pcpu_chunk *chunk) return pcpu_nr_pages_to_map_bits(chunk->nr_pages); } -#ifdef CONFIG_MEMCG_KMEM /** * pcpu_obj_full_size - helper to calculate size of each accounted object * @size: size of area to allocate in bytes @@ -123,13 +122,14 @@ static inline int pcpu_chunk_map_bits(struct pcpu_chunk *chunk) */ static inline size_t pcpu_obj_full_size(size_t size) { - size_t extra_size; + size_t extra_size = 0; - extra_size = size / PCPU_MIN_ALLOC_SIZE * sizeof(struct obj_cgroup *); +#ifdef CONFIG_MEMCG_KMEM + extra_size += size / PCPU_MIN_ALLOC_SIZE * sizeof(struct obj_cgroup *); +#endif return size * num_possible_cpus() + extra_size; } -#endif /* CONFIG_MEMCG_KMEM */ #ifdef CONFIG_PERCPU_STATS diff --git a/mm/percpu.c b/mm/percpu.c index ea28db283044..3633eeefaa0d 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -1884,8 +1884,9 @@ area_found: ptr = __addr_to_pcpu_ptr(chunk->base_addr + off); kmemleak_alloc_percpu(ptr, size, gfp); - trace_percpu_alloc_percpu(reserved, is_atomic, size, align, - chunk->base_addr, off, ptr); + trace_percpu_alloc_percpu(_RET_IP_, reserved, is_atomic, size, align, + chunk->base_addr, off, ptr, + pcpu_obj_full_size(size), gfp); pcpu_memcg_post_alloc_hook(objcg, chunk, off, size); -- cgit v1.3-8-gc7d7 From fe573327ffb1deba802c91dd1d4ff41dafa97a0e Mon Sep 17 00:00:00 2001 From: Vasily Averin Date: Thu, 12 May 2022 20:23:08 -0700 Subject: tracing: incorrect gfp_t conversion Fixes the following sparse warnings: include/trace/events/*: sparse: cast to restricted gfp_t include/trace/events/*: sparse: restricted gfp_t degrades to integer gfp_t type is bitwise and requires __force attributes for any casts. Link: https://lkml.kernel.org/r/331d88fe-f4f7-657c-02a2-d977f15fbff6@openvz.org Signed-off-by: Vasily Averin Cc: Steven Rostedt Cc: Ingo Molnar Signed-off-by: Andrew Morton --- include/linux/gfp.h | 2 +- include/trace/events/btrfs.h | 4 +- include/trace/events/compaction.h | 4 +- include/trace/events/kmem.h | 12 +++--- include/trace/events/mmflags.h | 78 +++++++++++++++++++-------------------- include/trace/events/vmscan.h | 16 ++++---- mm/compaction.c | 2 +- 7 files changed, 59 insertions(+), 59 deletions(-) (limited to 'include/trace') diff --git a/include/linux/gfp.h b/include/linux/gfp.h index 2a08a3c4ba95..2d2ccae933c2 100644 --- a/include/linux/gfp.h +++ b/include/linux/gfp.h @@ -367,7 +367,7 @@ static inline int gfp_migratetype(const gfp_t gfp_flags) return MIGRATE_UNMOVABLE; /* Group based on mobility */ - return (gfp_flags & GFP_MOVABLE_MASK) >> GFP_MOVABLE_SHIFT; + return (__force unsigned long)(gfp_flags & GFP_MOVABLE_MASK) >> GFP_MOVABLE_SHIFT; } #undef GFP_MOVABLE_MASK #undef GFP_MOVABLE_SHIFT diff --git a/include/trace/events/btrfs.h b/include/trace/events/btrfs.h index f068ff30d654..ed92c80331ea 100644 --- a/include/trace/events/btrfs.h +++ b/include/trace/events/btrfs.h @@ -1344,13 +1344,13 @@ TRACE_EVENT(alloc_extent_state, TP_STRUCT__entry( __field(const struct extent_state *, state) - __field(gfp_t, mask) + __field(unsigned long, mask) __field(const void*, ip) ), TP_fast_assign( __entry->state = state, - __entry->mask = mask, + __entry->mask = (__force unsigned long)mask, __entry->ip = (const void *)IP ), diff --git a/include/trace/events/compaction.h b/include/trace/events/compaction.h index c6d5d70dc7a5..3313eb83c117 100644 --- a/include/trace/events/compaction.h +++ b/include/trace/events/compaction.h @@ -162,13 +162,13 @@ TRACE_EVENT(mm_compaction_try_to_compact_pages, TP_STRUCT__entry( __field(int, order) - __field(gfp_t, gfp_mask) + __field(unsigned long, gfp_mask) __field(int, prio) ), TP_fast_assign( __entry->order = order; - __entry->gfp_mask = gfp_mask; + __entry->gfp_mask = (__force unsigned long)gfp_mask; __entry->prio = prio; ), diff --git a/include/trace/events/kmem.h b/include/trace/events/kmem.h index ddc8c944f417..71c141804222 100644 --- a/include/trace/events/kmem.h +++ b/include/trace/events/kmem.h @@ -24,7 +24,7 @@ DECLARE_EVENT_CLASS(kmem_alloc, __field( const void *, ptr ) __field( size_t, bytes_req ) __field( size_t, bytes_alloc ) - __field( gfp_t, gfp_flags ) + __field( unsigned long, gfp_flags ) ), TP_fast_assign( @@ -32,7 +32,7 @@ DECLARE_EVENT_CLASS(kmem_alloc, __entry->ptr = ptr; __entry->bytes_req = bytes_req; __entry->bytes_alloc = bytes_alloc; - __entry->gfp_flags = gfp_flags; + __entry->gfp_flags = (__force unsigned long)gfp_flags; ), TP_printk("call_site=%pS ptr=%p bytes_req=%zu bytes_alloc=%zu gfp_flags=%s", @@ -75,7 +75,7 @@ DECLARE_EVENT_CLASS(kmem_alloc_node, __field( const void *, ptr ) __field( size_t, bytes_req ) __field( size_t, bytes_alloc ) - __field( gfp_t, gfp_flags ) + __field( unsigned long, gfp_flags ) __field( int, node ) ), @@ -84,7 +84,7 @@ DECLARE_EVENT_CLASS(kmem_alloc_node, __entry->ptr = ptr; __entry->bytes_req = bytes_req; __entry->bytes_alloc = bytes_alloc; - __entry->gfp_flags = gfp_flags; + __entry->gfp_flags = (__force unsigned long)gfp_flags; __entry->node = node; ), @@ -208,14 +208,14 @@ TRACE_EVENT(mm_page_alloc, TP_STRUCT__entry( __field( unsigned long, pfn ) __field( unsigned int, order ) - __field( gfp_t, gfp_flags ) + __field( unsigned long, gfp_flags ) __field( int, migratetype ) ), TP_fast_assign( __entry->pfn = page ? page_to_pfn(page) : -1UL; __entry->order = order; - __entry->gfp_flags = gfp_flags; + __entry->gfp_flags = (__force unsigned long)gfp_flags; __entry->migratetype = migratetype; ), diff --git a/include/trace/events/mmflags.h b/include/trace/events/mmflags.h index 6532119a6bf1..fcb1ffb2cc8d 100644 --- a/include/trace/events/mmflags.h +++ b/include/trace/events/mmflags.h @@ -14,48 +14,48 @@ */ #define __def_gfpflag_names \ - {(unsigned long)GFP_TRANSHUGE, "GFP_TRANSHUGE"}, \ - {(unsigned long)GFP_TRANSHUGE_LIGHT, "GFP_TRANSHUGE_LIGHT"}, \ - {(unsigned long)GFP_HIGHUSER_MOVABLE, "GFP_HIGHUSER_MOVABLE"},\ - {(unsigned long)GFP_HIGHUSER, "GFP_HIGHUSER"}, \ - {(unsigned long)GFP_USER, "GFP_USER"}, \ - {(unsigned long)GFP_KERNEL_ACCOUNT, "GFP_KERNEL_ACCOUNT"}, \ - {(unsigned long)GFP_KERNEL, "GFP_KERNEL"}, \ - {(unsigned long)GFP_NOFS, "GFP_NOFS"}, \ - {(unsigned long)GFP_ATOMIC, "GFP_ATOMIC"}, \ - {(unsigned long)GFP_NOIO, "GFP_NOIO"}, \ - {(unsigned long)GFP_NOWAIT, "GFP_NOWAIT"}, \ - {(unsigned long)GFP_DMA, "GFP_DMA"}, \ - {(unsigned long)__GFP_HIGHMEM, "__GFP_HIGHMEM"}, \ - {(unsigned long)GFP_DMA32, "GFP_DMA32"}, \ - {(unsigned long)__GFP_HIGH, "__GFP_HIGH"}, \ - {(unsigned long)__GFP_ATOMIC, "__GFP_ATOMIC"}, \ - {(unsigned long)__GFP_IO, "__GFP_IO"}, \ - {(unsigned long)__GFP_FS, "__GFP_FS"}, \ - {(unsigned long)__GFP_NOWARN, "__GFP_NOWARN"}, \ - {(unsigned long)__GFP_RETRY_MAYFAIL, "__GFP_RETRY_MAYFAIL"}, \ - {(unsigned long)__GFP_NOFAIL, "__GFP_NOFAIL"}, \ - {(unsigned long)__GFP_NORETRY, "__GFP_NORETRY"}, \ - {(unsigned long)__GFP_COMP, "__GFP_COMP"}, \ - {(unsigned long)__GFP_ZERO, "__GFP_ZERO"}, \ - {(unsigned long)__GFP_NOMEMALLOC, "__GFP_NOMEMALLOC"}, \ - {(unsigned long)__GFP_MEMALLOC, "__GFP_MEMALLOC"}, \ - {(unsigned long)__GFP_HARDWALL, "__GFP_HARDWALL"}, \ - {(unsigned long)__GFP_THISNODE, "__GFP_THISNODE"}, \ - {(unsigned long)__GFP_RECLAIMABLE, "__GFP_RECLAIMABLE"}, \ - {(unsigned long)__GFP_MOVABLE, "__GFP_MOVABLE"}, \ - {(unsigned long)__GFP_ACCOUNT, "__GFP_ACCOUNT"}, \ - {(unsigned long)__GFP_WRITE, "__GFP_WRITE"}, \ - {(unsigned long)__GFP_RECLAIM, "__GFP_RECLAIM"}, \ - {(unsigned long)__GFP_DIRECT_RECLAIM, "__GFP_DIRECT_RECLAIM"},\ - {(unsigned long)__GFP_KSWAPD_RECLAIM, "__GFP_KSWAPD_RECLAIM"},\ - {(unsigned long)__GFP_ZEROTAGS, "__GFP_ZEROTAGS"} \ + {(__force unsigned long)GFP_TRANSHUGE, "GFP_TRANSHUGE"}, \ + {(__force unsigned long)GFP_TRANSHUGE_LIGHT, "GFP_TRANSHUGE_LIGHT"}, \ + {(__force unsigned long)GFP_HIGHUSER_MOVABLE, "GFP_HIGHUSER_MOVABLE"},\ + {(__force unsigned long)GFP_HIGHUSER, "GFP_HIGHUSER"}, \ + {(__force unsigned long)GFP_USER, "GFP_USER"}, \ + {(__force unsigned long)GFP_KERNEL_ACCOUNT, "GFP_KERNEL_ACCOUNT"}, \ + {(__force unsigned long)GFP_KERNEL, "GFP_KERNEL"}, \ + {(__force unsigned long)GFP_NOFS, "GFP_NOFS"}, \ + {(__force unsigned long)GFP_ATOMIC, "GFP_ATOMIC"}, \ + {(__force unsigned long)GFP_NOIO, "GFP_NOIO"}, \ + {(__force unsigned long)GFP_NOWAIT, "GFP_NOWAIT"}, \ + {(__force unsigned long)GFP_DMA, "GFP_DMA"}, \ + {(__force unsigned long)__GFP_HIGHMEM, "__GFP_HIGHMEM"}, \ + {(__force unsigned long)GFP_DMA32, "GFP_DMA32"}, \ + {(__force unsigned long)__GFP_HIGH, "__GFP_HIGH"}, \ + {(__force unsigned long)__GFP_ATOMIC, "__GFP_ATOMIC"}, \ + {(__force unsigned long)__GFP_IO, "__GFP_IO"}, \ + {(__force unsigned long)__GFP_FS, "__GFP_FS"}, \ + {(__force unsigned long)__GFP_NOWARN, "__GFP_NOWARN"}, \ + {(__force unsigned long)__GFP_RETRY_MAYFAIL, "__GFP_RETRY_MAYFAIL"}, \ + {(__force unsigned long)__GFP_NOFAIL, "__GFP_NOFAIL"}, \ + {(__force unsigned long)__GFP_NORETRY, "__GFP_NORETRY"}, \ + {(__force unsigned long)__GFP_COMP, "__GFP_COMP"}, \ + {(__force unsigned long)__GFP_ZERO, "__GFP_ZERO"}, \ + {(__force unsigned long)__GFP_NOMEMALLOC, "__GFP_NOMEMALLOC"}, \ + {(__force unsigned long)__GFP_MEMALLOC, "__GFP_MEMALLOC"}, \ + {(__force unsigned long)__GFP_HARDWALL, "__GFP_HARDWALL"}, \ + {(__force unsigned long)__GFP_THISNODE, "__GFP_THISNODE"}, \ + {(__force unsigned long)__GFP_RECLAIMABLE, "__GFP_RECLAIMABLE"}, \ + {(__force unsigned long)__GFP_MOVABLE, "__GFP_MOVABLE"}, \ + {(__force unsigned long)__GFP_ACCOUNT, "__GFP_ACCOUNT"}, \ + {(__force unsigned long)__GFP_WRITE, "__GFP_WRITE"}, \ + {(__force unsigned long)__GFP_RECLAIM, "__GFP_RECLAIM"}, \ + {(__force unsigned long)__GFP_DIRECT_RECLAIM, "__GFP_DIRECT_RECLAIM"},\ + {(__force unsigned long)__GFP_KSWAPD_RECLAIM, "__GFP_KSWAPD_RECLAIM"},\ + {(__force unsigned long)__GFP_ZEROTAGS, "__GFP_ZEROTAGS"} \ #ifdef CONFIG_KASAN_HW_TAGS #define __def_gfpflag_names_kasan , \ - {(unsigned long)__GFP_SKIP_ZERO, "__GFP_SKIP_ZERO"}, \ - {(unsigned long)__GFP_SKIP_KASAN_POISON, "__GFP_SKIP_KASAN_POISON"}, \ - {(unsigned long)__GFP_SKIP_KASAN_UNPOISON, "__GFP_SKIP_KASAN_UNPOISON"} + {(__force unsigned long)__GFP_SKIP_ZERO, "__GFP_SKIP_ZERO"}, \ + {(__force unsigned long)__GFP_SKIP_KASAN_POISON, "__GFP_SKIP_KASAN_POISON"}, \ + {(__force unsigned long)__GFP_SKIP_KASAN_UNPOISON, "__GFP_SKIP_KASAN_UNPOISON"} #else #define __def_gfpflag_names_kasan #endif diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h index de136dbd623a..408c86244d64 100644 --- a/include/trace/events/vmscan.h +++ b/include/trace/events/vmscan.h @@ -96,14 +96,14 @@ TRACE_EVENT(mm_vmscan_wakeup_kswapd, __field( int, nid ) __field( int, zid ) __field( int, order ) - __field( gfp_t, gfp_flags ) + __field( unsigned long, gfp_flags ) ), TP_fast_assign( __entry->nid = nid; __entry->zid = zid; __entry->order = order; - __entry->gfp_flags = gfp_flags; + __entry->gfp_flags = (__force unsigned long)gfp_flags; ), TP_printk("nid=%d order=%d gfp_flags=%s", @@ -120,12 +120,12 @@ DECLARE_EVENT_CLASS(mm_vmscan_direct_reclaim_begin_template, TP_STRUCT__entry( __field( int, order ) - __field( gfp_t, gfp_flags ) + __field( unsigned long, gfp_flags ) ), TP_fast_assign( __entry->order = order; - __entry->gfp_flags = gfp_flags; + __entry->gfp_flags = (__force unsigned long)gfp_flags; ), TP_printk("order=%d gfp_flags=%s", @@ -210,7 +210,7 @@ TRACE_EVENT(mm_shrink_slab_start, __field(void *, shrink) __field(int, nid) __field(long, nr_objects_to_shrink) - __field(gfp_t, gfp_flags) + __field(unsigned long, gfp_flags) __field(unsigned long, cache_items) __field(unsigned long long, delta) __field(unsigned long, total_scan) @@ -222,7 +222,7 @@ TRACE_EVENT(mm_shrink_slab_start, __entry->shrink = shr->scan_objects; __entry->nid = sc->nid; __entry->nr_objects_to_shrink = nr_objects_to_shrink; - __entry->gfp_flags = sc->gfp_mask; + __entry->gfp_flags = (__force unsigned long)sc->gfp_mask; __entry->cache_items = cache_items; __entry->delta = delta; __entry->total_scan = total_scan; @@ -446,13 +446,13 @@ TRACE_EVENT(mm_vmscan_node_reclaim_begin, TP_STRUCT__entry( __field(int, nid) __field(int, order) - __field(gfp_t, gfp_flags) + __field(unsigned long, gfp_flags) ), TP_fast_assign( __entry->nid = nid; __entry->order = order; - __entry->gfp_flags = gfp_flags; + __entry->gfp_flags = (__force unsigned long)gfp_flags; ), TP_printk("nid=%d order=%d gfp_flags=%s", diff --git a/mm/compaction.c b/mm/compaction.c index 65970107b789..0e0ef9e44530 100644 --- a/mm/compaction.c +++ b/mm/compaction.c @@ -2569,7 +2569,7 @@ enum compact_result try_to_compact_pages(gfp_t gfp_mask, unsigned int order, unsigned int alloc_flags, const struct alloc_context *ac, enum compact_priority prio, struct page **capture) { - int may_perform_io = gfp_mask & __GFP_IO; + int may_perform_io = (__force int)(gfp_mask & __GFP_IO); struct zoneref *z; struct zone *zone; enum compact_result rc = COMPACT_SKIPPED; -- cgit v1.3-8-gc7d7 From c81d5bae404abc6b257667e84d39b9b50c7063d4 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Fri, 6 May 2022 13:34:41 -0700 Subject: f2fs: do not stop GC when requiring a free section The f2fs_gc uses a bitmap to indicate pinned sections, but when disabling chckpoint, we call f2fs_gc() with NULL_SEGNO which selects the same dirty segment as a victim all the time, resulting in checkpoint=disable failure, for example. Let's pick another one, if we fail to collect it. Reviewed-by: Chao Yu Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 1 + fs/f2fs/file.c | 12 ++++++++---- fs/f2fs/gc.c | 14 +++++++++----- fs/f2fs/segment.c | 3 ++- fs/f2fs/super.c | 3 ++- include/trace/events/f2fs.h | 11 ++++++++--- 6 files changed, 30 insertions(+), 14 deletions(-) (limited to 'include/trace') diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 9920b2d6af8f..492af5b96de1 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -1271,6 +1271,7 @@ struct f2fs_gc_control { bool no_bg_gc; /* check the space and stop bg_gc */ bool should_migrate_blocks; /* should migrate blocks */ bool err_gc_skipped; /* return EAGAIN if GC skipped */ + unsigned int nr_free_secs; /* # of free sections to do GC */ }; /* For s_flag in struct f2fs_sb_info */ diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index d0547bef0851..216081ea8c81 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -1650,7 +1650,8 @@ static int expand_inode_data(struct inode *inode, loff_t offset, struct f2fs_gc_control gc_control = { .victim_segno = NULL_SEGNO, .init_gc_type = FG_GC, .should_migrate_blocks = false, - .err_gc_skipped = true }; + .err_gc_skipped = true, + .nr_free_secs = 0 }; pgoff_t pg_start, pg_end; loff_t new_size = i_size_read(inode); loff_t off_end; @@ -2350,7 +2351,8 @@ static int f2fs_ioc_gc(struct file *filp, unsigned long arg) struct f2fs_sb_info *sbi = F2FS_I_SB(inode); struct f2fs_gc_control gc_control = { .victim_segno = NULL_SEGNO, .no_bg_gc = false, - .should_migrate_blocks = false }; + .should_migrate_blocks = false, + .nr_free_secs = 0 }; __u32 sync; int ret; @@ -2391,7 +2393,8 @@ static int __f2fs_ioc_gc_range(struct file *filp, struct f2fs_gc_range *range) .init_gc_type = range->sync ? FG_GC : BG_GC, .no_bg_gc = false, .should_migrate_blocks = false, - .err_gc_skipped = range->sync }; + .err_gc_skipped = range->sync, + .nr_free_secs = 0 }; u64 end; int ret; @@ -2837,7 +2840,8 @@ static int f2fs_ioc_flush_device(struct file *filp, unsigned long arg) struct f2fs_gc_control gc_control = { .init_gc_type = FG_GC, .should_migrate_blocks = true, - .err_gc_skipped = true }; + .err_gc_skipped = true, + .nr_free_secs = 0 }; int ret; if (!capable(CAP_SYS_ADMIN)) diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index e275b72bc65f..d5fb426e0747 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -147,6 +147,7 @@ do_gc: gc_control.init_gc_type = sync_mode ? FG_GC : BG_GC; gc_control.no_bg_gc = foreground; + gc_control.nr_free_secs = foreground ? 1 : 0; /* if return value is not zero, no victim was selected */ if (f2fs_gc(sbi, &gc_control)) @@ -1761,6 +1762,7 @@ int f2fs_gc(struct f2fs_sb_info *sbi, struct f2fs_gc_control *gc_control) unsigned int skipped_round = 0, round = 0; trace_f2fs_gc_begin(sbi->sb, gc_type, gc_control->no_bg_gc, + gc_control->nr_free_secs, get_pages(sbi, F2FS_DIRTY_NODES), get_pages(sbi, F2FS_DIRTY_DENTS), get_pages(sbi, F2FS_DIRTY_IMETA), @@ -1823,12 +1825,13 @@ retry: if (gc_type == FG_GC) sbi->cur_victim_sec = NULL_SEGNO; - if (gc_control->init_gc_type == FG_GC) - goto stop; - - if (!has_not_enough_free_secs(sbi, - (gc_type == FG_GC) ? sec_freed : 0, 0)) + if (gc_control->init_gc_type == FG_GC || + !has_not_enough_free_secs(sbi, + (gc_type == FG_GC) ? sec_freed : 0, 0)) { + if (gc_type == FG_GC && sec_freed < gc_control->nr_free_secs) + goto go_gc_more; goto stop; + } /* FG_GC stops GC by skip_count */ if (gc_type == FG_GC) { @@ -1849,6 +1852,7 @@ retry: if (ret) goto stop; } +go_gc_more: segno = NULL_SEGNO; goto gc_more; diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index 8b4f2b1d2cca..0a4180f64291 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -404,7 +404,8 @@ void f2fs_balance_fs(struct f2fs_sb_info *sbi, bool need) .init_gc_type = BG_GC, .no_bg_gc = true, .should_migrate_blocks = false, - .err_gc_skipped = false }; + .err_gc_skipped = false, + .nr_free_secs = 1 }; f2fs_down_write(&sbi->gc_lock); f2fs_gc(sbi, &gc_control); } diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index a28c27eed6d0..63daae67a9d9 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -2080,7 +2080,8 @@ static int f2fs_disable_checkpoint(struct f2fs_sb_info *sbi) .victim_segno = NULL_SEGNO, .init_gc_type = FG_GC, .should_migrate_blocks = false, - .err_gc_skipped = true }; + .err_gc_skipped = true, + .nr_free_secs = 1 }; f2fs_down_write(&sbi->gc_lock); err = f2fs_gc(sbi, &gc_control); diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h index 54ec9e543f09..16c67ede85b6 100644 --- a/include/trace/events/f2fs.h +++ b/include/trace/events/f2fs.h @@ -645,18 +645,21 @@ TRACE_EVENT(f2fs_background_gc, TRACE_EVENT(f2fs_gc_begin, TP_PROTO(struct super_block *sb, int gc_type, bool no_bg_gc, + unsigned int nr_free_secs, long long dirty_nodes, long long dirty_dents, long long dirty_imeta, unsigned int free_sec, unsigned int free_seg, int reserved_seg, unsigned int prefree_seg), - TP_ARGS(sb, gc_type, no_bg_gc, dirty_nodes, dirty_dents, dirty_imeta, + TP_ARGS(sb, gc_type, no_bg_gc, nr_free_secs, dirty_nodes, + dirty_dents, dirty_imeta, free_sec, free_seg, reserved_seg, prefree_seg), TP_STRUCT__entry( __field(dev_t, dev) __field(int, gc_type) __field(bool, no_bg_gc) + __field(unsigned int, nr_free_secs) __field(long long, dirty_nodes) __field(long long, dirty_dents) __field(long long, dirty_imeta) @@ -670,6 +673,7 @@ TRACE_EVENT(f2fs_gc_begin, __entry->dev = sb->s_dev; __entry->gc_type = gc_type; __entry->no_bg_gc = no_bg_gc; + __entry->nr_free_secs = nr_free_secs; __entry->dirty_nodes = dirty_nodes; __entry->dirty_dents = dirty_dents; __entry->dirty_imeta = dirty_imeta; @@ -679,12 +683,13 @@ TRACE_EVENT(f2fs_gc_begin, __entry->prefree_seg = prefree_seg; ), - TP_printk("dev = (%d,%d), gc_type = %s, no_background_GC = %d, nodes = %lld, " - "dents = %lld, imeta = %lld, free_sec:%u, free_seg:%u, " + TP_printk("dev = (%d,%d), gc_type = %s, no_background_GC = %d, nr_free_secs = %u, " + "nodes = %lld, dents = %lld, imeta = %lld, free_sec:%u, free_seg:%u, " "rsv_seg:%d, prefree_seg:%u", show_dev(__entry->dev), show_gc_type(__entry->gc_type), (__entry->gc_type == BG_GC) ? __entry->no_bg_gc : -1, + __entry->nr_free_secs, __entry->dirty_nodes, __entry->dirty_dents, __entry->dirty_imeta, -- cgit v1.3-8-gc7d7 From 45cb7955c180a2a34d291e68938250c4f9bd294f Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 14 Apr 2022 15:50:12 -0400 Subject: SUNRPC: Clean up svc_deferred_class trace events Replace the temporary fix from commit 4d5004451ab2 ("SUNRPC: Fix the svc_deferred_event trace class") with the use of __sockaddr and friends, which is the preferred solution (but only available in 5.18 and newer). Signed-off-by: Chuck Lever --- include/trace/events/sunrpc.h | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 3995c58a1c51..ee84e9289707 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -2015,19 +2015,18 @@ DECLARE_EVENT_CLASS(svc_deferred_event, TP_STRUCT__entry( __field(const void *, dr) __field(u32, xid) - __array(__u8, addr, INET6_ADDRSTRLEN + 10) + __sockaddr(addr, dr->addrlen) ), TP_fast_assign( __entry->dr = dr; __entry->xid = be32_to_cpu(*(__be32 *)(dr->args + (dr->xprt_hlen>>2))); - snprintf(__entry->addr, sizeof(__entry->addr) - 1, - "%pISpc", (struct sockaddr *)&dr->addr); + __assign_sockaddr(addr, &dr->addr, dr->addrlen); ), - TP_printk("addr=%s dr=%p xid=0x%08x", __entry->addr, __entry->dr, - __entry->xid) + TP_printk("addr=%pISpc dr=%p xid=0x%08x", __get_sockaddr(addr), + __entry->dr, __entry->xid) ); #define DEFINE_SVC_DEFERRED_EVENT(name) \ -- cgit v1.3-8-gc7d7 From 983084b2672c593959e3148d6a17c8b920797dde Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 6 Apr 2022 14:38:59 -0400 Subject: SUNRPC: Remove svc_rqst::rq_xprt_hlen Clean up: This field is now always set to zero. Signed-off-by: Chuck Lever --- include/linux/sunrpc/svc.h | 2 -- include/trace/events/sunrpc.h | 3 +-- net/sunrpc/svc_xprt.c | 10 ++++------ net/sunrpc/svcsock.c | 2 -- net/sunrpc/xprtrdma/svc_rdma_recvfrom.c | 1 - 5 files changed, 5 insertions(+), 13 deletions(-) (limited to 'include/trace') diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index 217711fc9cac..b73704fbd09b 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -257,7 +257,6 @@ struct svc_rqst { void * rq_xprt_ctxt; /* transport specific context ptr */ struct svc_deferred_req*rq_deferred; /* deferred request we are replaying */ - size_t rq_xprt_hlen; /* xprt header len */ struct xdr_buf rq_arg; struct xdr_stream rq_arg_stream; struct xdr_stream rq_res_stream; @@ -397,7 +396,6 @@ struct svc_deferred_req { size_t daddrlen; void *xprt_ctxt; struct cache_deferred_req handle; - size_t xprt_hlen; int argslen; __be32 args[]; }; diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index ee84e9289707..b61d9c90fa26 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -2020,8 +2020,7 @@ DECLARE_EVENT_CLASS(svc_deferred_event, TP_fast_assign( __entry->dr = dr; - __entry->xid = be32_to_cpu(*(__be32 *)(dr->args + - (dr->xprt_hlen>>2))); + __entry->xid = be32_to_cpu(*(__be32 *)dr->args); __assign_sockaddr(addr, &dr->addr, dr->addrlen); ), diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index 5b59e2103526..f9d9922f1629 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -1230,7 +1230,6 @@ static struct cache_deferred_req *svc_defer(struct cache_req *req) dr->addrlen = rqstp->rq_addrlen; dr->daddr = rqstp->rq_daddr; dr->argslen = rqstp->rq_arg.len >> 2; - dr->xprt_hlen = rqstp->rq_xprt_hlen; dr->xprt_ctxt = rqstp->rq_xprt_ctxt; rqstp->rq_xprt_ctxt = NULL; @@ -1258,22 +1257,21 @@ static noinline int svc_deferred_recv(struct svc_rqst *rqstp) trace_svc_defer_recv(dr); /* setup iov_base past transport header */ - rqstp->rq_arg.head[0].iov_base = dr->args + (dr->xprt_hlen>>2); + rqstp->rq_arg.head[0].iov_base = dr->args; /* The iov_len does not include the transport header bytes */ - rqstp->rq_arg.head[0].iov_len = (dr->argslen<<2) - dr->xprt_hlen; + rqstp->rq_arg.head[0].iov_len = dr->argslen << 2; rqstp->rq_arg.page_len = 0; /* The rq_arg.len includes the transport header bytes */ - rqstp->rq_arg.len = dr->argslen<<2; + rqstp->rq_arg.len = dr->argslen << 2; rqstp->rq_prot = dr->prot; memcpy(&rqstp->rq_addr, &dr->addr, dr->addrlen); rqstp->rq_addrlen = dr->addrlen; /* Save off transport header len in case we get deferred again */ - rqstp->rq_xprt_hlen = dr->xprt_hlen; rqstp->rq_daddr = dr->daddr; rqstp->rq_respages = rqstp->rq_pages; rqstp->rq_xprt_ctxt = dr->xprt_ctxt; svc_xprt_received(rqstp->rq_xprt); - return (dr->argslen<<2) - dr->xprt_hlen; + return dr->argslen << 2; } diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 68e46d4547b4..05452318afec 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -250,8 +250,6 @@ static ssize_t svc_tcp_read_msg(struct svc_rqst *rqstp, size_t buflen, ssize_t len; size_t t; - rqstp->rq_xprt_hlen = 0; - clear_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags); for (i = 0, t = 0; t < buflen; i++, t += PAGE_SIZE) { diff --git a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c index 864131a9fc6e..5242ad121450 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c +++ b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c @@ -831,7 +831,6 @@ int svc_rdma_recvfrom(struct svc_rqst *rqstp) goto out_err; if (ret == 0) goto out_drop; - rqstp->rq_xprt_hlen = 0; if (svc_rdma_is_reverse_direction_reply(xprt, ctxt)) goto out_backchannel; -- cgit v1.3-8-gc7d7 From 10e0f7530205799e7e971aba699a7cb3a47456de Mon Sep 17 00:00:00 2001 From: Wonhyuk Yang Date: Thu, 19 May 2022 14:08:54 -0700 Subject: mm/page_alloc: fix tracepoint mm_page_alloc_zone_locked() Currently, trace point mm_page_alloc_zone_locked() doesn't show correct information. First, when alloc_flag has ALLOC_HARDER/ALLOC_CMA, page can be allocated from MIGRATE_HIGHATOMIC/MIGRATE_CMA. Nevertheless, tracepoint use requested migration type not MIGRATE_HIGHATOMIC and MIGRATE_CMA. Second, after commit 44042b4498728 ("mm/page_alloc: allow high-order pages to be stored on the per-cpu lists") percpu-list can store high order pages. But trace point determine whether it is a refiil of percpu-list by comparing requested order and 0. To handle these problems, make mm_page_alloc_zone_locked() only be called by __rmqueue_smallest with correct migration type. With a new argument called percpu_refill, it can show roughly whether it is a refill of percpu-list. Link: https://lkml.kernel.org/r/20220512025307.57924-1-vvghjk1234@gmail.com Signed-off-by: Wonhyuk Yang Acked-by: Mel Gorman Cc: Baik Song An Cc: Hong Yeon Kim Cc: Taeung Song Cc: Cc: Steven Rostedt Cc: Ingo Molnar Signed-off-by: Andrew Morton --- include/trace/events/kmem.h | 14 +++++++++----- mm/page_alloc.c | 13 +++++-------- 2 files changed, 14 insertions(+), 13 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/kmem.h b/include/trace/events/kmem.h index 71c141804222..f76668305ac5 100644 --- a/include/trace/events/kmem.h +++ b/include/trace/events/kmem.h @@ -229,20 +229,23 @@ TRACE_EVENT(mm_page_alloc, DECLARE_EVENT_CLASS(mm_page, - TP_PROTO(struct page *page, unsigned int order, int migratetype), + TP_PROTO(struct page *page, unsigned int order, int migratetype, + int percpu_refill), - TP_ARGS(page, order, migratetype), + TP_ARGS(page, order, migratetype, percpu_refill), TP_STRUCT__entry( __field( unsigned long, pfn ) __field( unsigned int, order ) __field( int, migratetype ) + __field( int, percpu_refill ) ), TP_fast_assign( __entry->pfn = page ? page_to_pfn(page) : -1UL; __entry->order = order; __entry->migratetype = migratetype; + __entry->percpu_refill = percpu_refill; ), TP_printk("page=%p pfn=0x%lx order=%u migratetype=%d percpu_refill=%d", @@ -250,14 +253,15 @@ DECLARE_EVENT_CLASS(mm_page, __entry->pfn != -1UL ? __entry->pfn : 0, __entry->order, __entry->migratetype, - __entry->order == 0) + __entry->percpu_refill) ); DEFINE_EVENT(mm_page, mm_page_alloc_zone_locked, - TP_PROTO(struct page *page, unsigned int order, int migratetype), + TP_PROTO(struct page *page, unsigned int order, int migratetype, + int percpu_refill), - TP_ARGS(page, order, migratetype) + TP_ARGS(page, order, migratetype, percpu_refill) ); TRACE_EVENT(mm_page_pcpu_drain, diff --git a/mm/page_alloc.c b/mm/page_alloc.c index a4a8fbbd6217..10305b10fe93 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -2466,6 +2466,9 @@ struct page *__rmqueue_smallest(struct zone *zone, unsigned int order, del_page_from_free_list(page, zone, current_order); expand(zone, page, order, current_order, migratetype); set_pcppage_migratetype(page, migratetype); + trace_mm_page_alloc_zone_locked(page, order, migratetype, + pcp_allowed_order(order) && + migratetype < MIGRATE_PCPTYPES); return page; } @@ -2989,7 +2992,7 @@ __rmqueue(struct zone *zone, unsigned int order, int migratetype, zone_page_state(zone, NR_FREE_PAGES) / 2) { page = __rmqueue_cma_fallback(zone, order); if (page) - goto out; + return page; } } retry: @@ -3002,9 +3005,6 @@ retry: alloc_flags)) goto retry; } -out: - if (page) - trace_mm_page_alloc_zone_locked(page, order, migratetype); return page; } @@ -3723,11 +3723,8 @@ struct page *rmqueue(struct zone *preferred_zone, * reserved for high-order atomic allocation, so order-0 * request should skip it. */ - if (order > 0 && alloc_flags & ALLOC_HARDER) { + if (order > 0 && alloc_flags & ALLOC_HARDER) page = __rmqueue_smallest(zone, order, MIGRATE_HIGHATOMIC); - if (page) - trace_mm_page_alloc_zone_locked(page, order, migratetype); - } if (!page) { page = __rmqueue(zone, order, migratetype, alloc_flags); if (!page) -- cgit v1.3-8-gc7d7 From 2b132903de7124dd9a758be0c27562e91a510848 Mon Sep 17 00:00:00 2001 From: Vasily Averin Date: Wed, 11 May 2022 12:46:53 +0300 Subject: tracing: incorrect isolate_mote_t cast in mm_vmscan_lru_isolate Fixes following sparse warnings: CHECK mm/vmscan.c mm/vmscan.c: note: in included file (through include/trace/trace_events.h, include/trace/define_trace.h, include/trace/events/vmscan.h): ./include/trace/events/vmscan.h:281:1: sparse: warning: cast to restricted isolate_mode_t ./include/trace/events/vmscan.h:281:1: sparse: warning: restricted isolate_mode_t degrades to integer Link: https://lkml.kernel.org/r/e85d7ff2-fd10-53f8-c24e-ba0458439c1b@openvz.org Signed-off-by: Vasily Averin Acked-by: Steven Rostedt (Google) Signed-off-by: Andrew Morton --- include/trace/events/vmscan.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h index 408c86244d64..d2123dd960d5 100644 --- a/include/trace/events/vmscan.h +++ b/include/trace/events/vmscan.h @@ -297,7 +297,7 @@ TRACE_EVENT(mm_vmscan_lru_isolate, __field(unsigned long, nr_scanned) __field(unsigned long, nr_skipped) __field(unsigned long, nr_taken) - __field(isolate_mode_t, isolate_mode) + __field(unsigned int, isolate_mode) __field(int, lru) ), @@ -308,7 +308,7 @@ TRACE_EVENT(mm_vmscan_lru_isolate, __entry->nr_scanned = nr_scanned; __entry->nr_skipped = nr_skipped; __entry->nr_taken = nr_taken; - __entry->isolate_mode = isolate_mode; + __entry->isolate_mode = (__force unsigned int)isolate_mode; __entry->lru = lru; ), -- cgit v1.3-8-gc7d7 From a05754295e01f006a651eec759c5dbe682ef6cef Mon Sep 17 00:00:00 2001 From: David Howells Date: Sat, 21 May 2022 08:45:22 +0100 Subject: rxrpc: Use refcount_t rather than atomic_t Move to using refcount_t rather than atomic_t for refcounts in rxrpc. Signed-off-by: David Howells cc: Marc Dionne cc: linux-afs@lists.infradead.org Signed-off-by: David S. Miller --- include/trace/events/rxrpc.h | 2 +- net/rxrpc/af_rxrpc.c | 2 +- net/rxrpc/ar-internal.h | 18 +++++----------- net/rxrpc/call_accept.c | 4 ++-- net/rxrpc/call_object.c | 44 ++++++++++++++++++++------------------- net/rxrpc/conn_client.c | 30 ++++++++++++++------------- net/rxrpc/conn_object.c | 49 ++++++++++++++++++++++---------------------- net/rxrpc/conn_service.c | 8 ++++---- net/rxrpc/input.c | 4 +--- net/rxrpc/local_object.c | 31 ++++++++++++++-------------- net/rxrpc/peer_object.c | 40 +++++++++++++++++++----------------- net/rxrpc/proc.c | 8 ++++---- net/rxrpc/skbuff.c | 1 - 13 files changed, 119 insertions(+), 122 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index 4a3ab0ed6e06..cdb28976641b 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -583,7 +583,7 @@ TRACE_EVENT(rxrpc_client, TP_fast_assign( __entry->conn = conn ? conn->debug_id : 0; __entry->channel = channel; - __entry->usage = conn ? atomic_read(&conn->usage) : -2; + __entry->usage = conn ? refcount_read(&conn->ref) : -2; __entry->op = op; __entry->cid = conn ? conn->proto.cid : 0; ), diff --git a/net/rxrpc/af_rxrpc.c b/net/rxrpc/af_rxrpc.c index 2b5f89713e36..ceba28e9dce6 100644 --- a/net/rxrpc/af_rxrpc.c +++ b/net/rxrpc/af_rxrpc.c @@ -351,7 +351,7 @@ static void rxrpc_dummy_notify_rx(struct sock *sk, struct rxrpc_call *rxcall, */ void rxrpc_kernel_end_call(struct socket *sock, struct rxrpc_call *call) { - _enter("%d{%d}", call->debug_id, atomic_read(&call->usage)); + _enter("%d{%d}", call->debug_id, refcount_read(&call->ref)); mutex_lock(&call->user_mutex); rxrpc_release_call(rxrpc_sk(sock->sk), call); diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index ac1f88a0e120..52a23d03d694 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -15,14 +15,6 @@ #include #include "protocol.h" -#if 0 -#define CHECK_SLAB_OKAY(X) \ - BUG_ON(atomic_read((X)) >> (sizeof(atomic_t) - 2) == \ - (POISON_FREE << 8 | POISON_FREE)) -#else -#define CHECK_SLAB_OKAY(X) do {} while (0) -#endif - #define FCRYPT_BSIZE 8 struct rxrpc_crypt { union { @@ -279,7 +271,7 @@ struct rxrpc_security { struct rxrpc_local { struct rcu_head rcu; atomic_t active_users; /* Number of users of the local endpoint */ - atomic_t usage; /* Number of references to the structure */ + refcount_t ref; /* Number of references to the structure */ struct rxrpc_net *rxnet; /* The network ns in which this resides */ struct hlist_node link; struct socket *socket; /* my UDP socket */ @@ -304,7 +296,7 @@ struct rxrpc_local { */ struct rxrpc_peer { struct rcu_head rcu; /* This must be first */ - atomic_t usage; + refcount_t ref; unsigned long hash_key; struct hlist_node hash_link; struct rxrpc_local *local; @@ -406,7 +398,7 @@ enum rxrpc_conn_proto_state { */ struct rxrpc_bundle { struct rxrpc_conn_parameters params; - atomic_t usage; + refcount_t ref; unsigned int debug_id; bool try_upgrade; /* True if the bundle is attempting upgrade */ bool alloc_conn; /* True if someone's getting a conn */ @@ -427,7 +419,7 @@ struct rxrpc_connection { struct rxrpc_conn_proto proto; struct rxrpc_conn_parameters params; - atomic_t usage; + refcount_t ref; struct rcu_head rcu; struct list_head cache_link; @@ -609,7 +601,7 @@ struct rxrpc_call { int error; /* Local error incurred */ enum rxrpc_call_state state; /* current state of call */ enum rxrpc_call_completion completion; /* Call completion condition */ - atomic_t usage; + refcount_t ref; u16 service_id; /* service ID */ u8 security_ix; /* Security type */ enum rxrpc_interruptibility interruptibility; /* At what point call may be interrupted */ diff --git a/net/rxrpc/call_accept.c b/net/rxrpc/call_accept.c index 1ae90fb97936..8ae59edc2551 100644 --- a/net/rxrpc/call_accept.c +++ b/net/rxrpc/call_accept.c @@ -91,7 +91,7 @@ static int rxrpc_service_prealloc_one(struct rxrpc_sock *rx, (head + 1) & (size - 1)); trace_rxrpc_conn(conn->debug_id, rxrpc_conn_new_service, - atomic_read(&conn->usage), here); + refcount_read(&conn->ref), here); } /* Now it gets complicated, because calls get registered with the @@ -104,7 +104,7 @@ static int rxrpc_service_prealloc_one(struct rxrpc_sock *rx, call->state = RXRPC_CALL_SERVER_PREALLOC; trace_rxrpc_call(call->debug_id, rxrpc_call_new_service, - atomic_read(&call->usage), + refcount_read(&call->ref), here, (const void *)user_call_ID); write_lock(&rx->call_lock); diff --git a/net/rxrpc/call_object.c b/net/rxrpc/call_object.c index 043508fd8d8a..8764a4f19c03 100644 --- a/net/rxrpc/call_object.c +++ b/net/rxrpc/call_object.c @@ -112,7 +112,7 @@ struct rxrpc_call *rxrpc_find_call_by_user_ID(struct rxrpc_sock *rx, found_extant_call: rxrpc_get_call(call, rxrpc_call_got); read_unlock(&rx->call_lock); - _leave(" = %p [%d]", call, atomic_read(&call->usage)); + _leave(" = %p [%d]", call, refcount_read(&call->ref)); return call; } @@ -160,7 +160,7 @@ struct rxrpc_call *rxrpc_alloc_call(struct rxrpc_sock *rx, gfp_t gfp, spin_lock_init(&call->notify_lock); spin_lock_init(&call->input_lock); rwlock_init(&call->state_lock); - atomic_set(&call->usage, 1); + refcount_set(&call->ref, 1); call->debug_id = debug_id; call->tx_total_len = -1; call->next_rx_timo = 20 * HZ; @@ -299,7 +299,7 @@ struct rxrpc_call *rxrpc_new_client_call(struct rxrpc_sock *rx, call->interruptibility = p->interruptibility; call->tx_total_len = p->tx_total_len; trace_rxrpc_call(call->debug_id, rxrpc_call_new_client, - atomic_read(&call->usage), + refcount_read(&call->ref), here, (const void *)p->user_call_ID); if (p->kernel) __set_bit(RXRPC_CALL_KERNEL, &call->flags); @@ -352,7 +352,7 @@ struct rxrpc_call *rxrpc_new_client_call(struct rxrpc_sock *rx, goto error_attached_to_socket; trace_rxrpc_call(call->debug_id, rxrpc_call_connected, - atomic_read(&call->usage), here, NULL); + refcount_read(&call->ref), here, NULL); rxrpc_start_call_timer(call); @@ -372,7 +372,7 @@ error_dup_user_ID: __rxrpc_set_call_completion(call, RXRPC_CALL_LOCAL_ERROR, RX_CALL_DEAD, -EEXIST); trace_rxrpc_call(call->debug_id, rxrpc_call_error, - atomic_read(&call->usage), here, ERR_PTR(-EEXIST)); + refcount_read(&call->ref), here, ERR_PTR(-EEXIST)); rxrpc_release_call(rx, call); mutex_unlock(&call->user_mutex); rxrpc_put_call(call, rxrpc_call_put); @@ -386,7 +386,7 @@ error_dup_user_ID: */ error_attached_to_socket: trace_rxrpc_call(call->debug_id, rxrpc_call_error, - atomic_read(&call->usage), here, ERR_PTR(ret)); + refcount_read(&call->ref), here, ERR_PTR(ret)); set_bit(RXRPC_CALL_DISCONNECTED, &call->flags); __rxrpc_set_call_completion(call, RXRPC_CALL_LOCAL_ERROR, RX_CALL_DEAD, ret); @@ -442,8 +442,9 @@ void rxrpc_incoming_call(struct rxrpc_sock *rx, bool rxrpc_queue_call(struct rxrpc_call *call) { const void *here = __builtin_return_address(0); - int n = atomic_fetch_add_unless(&call->usage, 1, 0); - if (n == 0) + int n; + + if (!__refcount_inc_not_zero(&call->ref, &n)) return false; if (rxrpc_queue_work(&call->processor)) trace_rxrpc_call(call->debug_id, rxrpc_call_queued, n + 1, @@ -459,7 +460,7 @@ bool rxrpc_queue_call(struct rxrpc_call *call) bool __rxrpc_queue_call(struct rxrpc_call *call) { const void *here = __builtin_return_address(0); - int n = atomic_read(&call->usage); + int n = refcount_read(&call->ref); ASSERTCMP(n, >=, 1); if (rxrpc_queue_work(&call->processor)) trace_rxrpc_call(call->debug_id, rxrpc_call_queued_ref, n, @@ -476,7 +477,7 @@ void rxrpc_see_call(struct rxrpc_call *call) { const void *here = __builtin_return_address(0); if (call) { - int n = atomic_read(&call->usage); + int n = refcount_read(&call->ref); trace_rxrpc_call(call->debug_id, rxrpc_call_seen, n, here, NULL); @@ -486,11 +487,11 @@ void rxrpc_see_call(struct rxrpc_call *call) bool rxrpc_try_get_call(struct rxrpc_call *call, enum rxrpc_call_trace op) { const void *here = __builtin_return_address(0); - int n = atomic_fetch_add_unless(&call->usage, 1, 0); + int n; - if (n == 0) + if (!__refcount_inc_not_zero(&call->ref, &n)) return false; - trace_rxrpc_call(call->debug_id, op, n, here, NULL); + trace_rxrpc_call(call->debug_id, op, n + 1, here, NULL); return true; } @@ -500,9 +501,10 @@ bool rxrpc_try_get_call(struct rxrpc_call *call, enum rxrpc_call_trace op) void rxrpc_get_call(struct rxrpc_call *call, enum rxrpc_call_trace op) { const void *here = __builtin_return_address(0); - int n = atomic_inc_return(&call->usage); + int n; - trace_rxrpc_call(call->debug_id, op, n, here, NULL); + __refcount_inc(&call->ref, &n); + trace_rxrpc_call(call->debug_id, op, n + 1, here, NULL); } /* @@ -527,10 +529,10 @@ void rxrpc_release_call(struct rxrpc_sock *rx, struct rxrpc_call *call) struct rxrpc_connection *conn = call->conn; bool put = false; - _enter("{%d,%d}", call->debug_id, atomic_read(&call->usage)); + _enter("{%d,%d}", call->debug_id, refcount_read(&call->ref)); trace_rxrpc_call(call->debug_id, rxrpc_call_release, - atomic_read(&call->usage), + refcount_read(&call->ref), here, (const void *)call->flags); ASSERTCMP(call->state, ==, RXRPC_CALL_COMPLETE); @@ -619,14 +621,14 @@ void rxrpc_put_call(struct rxrpc_call *call, enum rxrpc_call_trace op) struct rxrpc_net *rxnet = call->rxnet; const void *here = __builtin_return_address(0); unsigned int debug_id = call->debug_id; + bool dead; int n; ASSERT(call != NULL); - n = atomic_dec_return(&call->usage); + dead = __refcount_dec_and_test(&call->ref, &n); trace_rxrpc_call(debug_id, op, n, here, NULL); - ASSERTCMP(n, >=, 0); - if (n == 0) { + if (dead) { _debug("call %d dead", call->debug_id); ASSERTCMP(call->state, ==, RXRPC_CALL_COMPLETE); @@ -716,7 +718,7 @@ void rxrpc_destroy_all_calls(struct rxrpc_net *rxnet) list_del_init(&call->link); pr_err("Call %p still in use (%d,%s,%lx,%lx)!\n", - call, atomic_read(&call->usage), + call, refcount_read(&call->ref), rxrpc_call_states[call->state], call->flags, call->events); diff --git a/net/rxrpc/conn_client.c b/net/rxrpc/conn_client.c index 8120138dac01..3c9eeb5b750c 100644 --- a/net/rxrpc/conn_client.c +++ b/net/rxrpc/conn_client.c @@ -102,7 +102,7 @@ void rxrpc_destroy_client_conn_ids(void) if (!idr_is_empty(&rxrpc_client_conn_ids)) { idr_for_each_entry(&rxrpc_client_conn_ids, conn, id) { pr_err("AF_RXRPC: Leaked client conn %p {%d}\n", - conn, atomic_read(&conn->usage)); + conn, refcount_read(&conn->ref)); } BUG(); } @@ -122,7 +122,7 @@ static struct rxrpc_bundle *rxrpc_alloc_bundle(struct rxrpc_conn_parameters *cp, if (bundle) { bundle->params = *cp; rxrpc_get_peer(bundle->params.peer); - atomic_set(&bundle->usage, 1); + refcount_set(&bundle->ref, 1); spin_lock_init(&bundle->channel_lock); INIT_LIST_HEAD(&bundle->waiting_calls); } @@ -131,7 +131,7 @@ static struct rxrpc_bundle *rxrpc_alloc_bundle(struct rxrpc_conn_parameters *cp, struct rxrpc_bundle *rxrpc_get_bundle(struct rxrpc_bundle *bundle) { - atomic_inc(&bundle->usage); + refcount_inc(&bundle->ref); return bundle; } @@ -144,10 +144,13 @@ static void rxrpc_free_bundle(struct rxrpc_bundle *bundle) void rxrpc_put_bundle(struct rxrpc_bundle *bundle) { unsigned int d = bundle->debug_id; - unsigned int u = atomic_dec_return(&bundle->usage); + bool dead; + int r; - _debug("PUT B=%x %u", d, u); - if (u == 0) + dead = __refcount_dec_and_test(&bundle->ref, &r); + + _debug("PUT B=%x %d", d, r); + if (dead) rxrpc_free_bundle(bundle); } @@ -169,7 +172,7 @@ rxrpc_alloc_client_connection(struct rxrpc_bundle *bundle, gfp_t gfp) return ERR_PTR(-ENOMEM); } - atomic_set(&conn->usage, 1); + refcount_set(&conn->ref, 1); conn->bundle = bundle; conn->params = bundle->params; conn->out_clientflag = RXRPC_CLIENT_INITIATED; @@ -195,7 +198,7 @@ rxrpc_alloc_client_connection(struct rxrpc_bundle *bundle, gfp_t gfp) key_get(conn->params.key); trace_rxrpc_conn(conn->debug_id, rxrpc_conn_new_client, - atomic_read(&conn->usage), + refcount_read(&conn->ref), __builtin_return_address(0)); atomic_inc(&rxnet->nr_client_conns); @@ -966,14 +969,13 @@ void rxrpc_put_client_conn(struct rxrpc_connection *conn) { const void *here = __builtin_return_address(0); unsigned int debug_id = conn->debug_id; - int n; + bool dead; + int r; - n = atomic_dec_return(&conn->usage); - trace_rxrpc_conn(debug_id, rxrpc_conn_put_client, n, here); - if (n <= 0) { - ASSERTCMP(n, >=, 0); + dead = __refcount_dec_and_test(&conn->ref, &r); + trace_rxrpc_conn(debug_id, rxrpc_conn_put_client, r - 1, here); + if (dead) rxrpc_kill_client_conn(conn); - } } /* diff --git a/net/rxrpc/conn_object.c b/net/rxrpc/conn_object.c index b2159dbf5412..03c7f2269151 100644 --- a/net/rxrpc/conn_object.c +++ b/net/rxrpc/conn_object.c @@ -104,7 +104,7 @@ struct rxrpc_connection *rxrpc_find_connection_rcu(struct rxrpc_local *local, goto not_found; *_peer = peer; conn = rxrpc_find_service_conn_rcu(peer, skb); - if (!conn || atomic_read(&conn->usage) == 0) + if (!conn || refcount_read(&conn->ref) == 0) goto not_found; _leave(" = %p", conn); return conn; @@ -114,7 +114,7 @@ struct rxrpc_connection *rxrpc_find_connection_rcu(struct rxrpc_local *local, */ conn = idr_find(&rxrpc_client_conn_ids, sp->hdr.cid >> RXRPC_CIDSHIFT); - if (!conn || atomic_read(&conn->usage) == 0) { + if (!conn || refcount_read(&conn->ref) == 0) { _debug("no conn"); goto not_found; } @@ -263,11 +263,12 @@ void rxrpc_kill_connection(struct rxrpc_connection *conn) bool rxrpc_queue_conn(struct rxrpc_connection *conn) { const void *here = __builtin_return_address(0); - int n = atomic_fetch_add_unless(&conn->usage, 1, 0); - if (n == 0) + int r; + + if (!__refcount_inc_not_zero(&conn->ref, &r)) return false; if (rxrpc_queue_work(&conn->processor)) - trace_rxrpc_conn(conn->debug_id, rxrpc_conn_queued, n + 1, here); + trace_rxrpc_conn(conn->debug_id, rxrpc_conn_queued, r + 1, here); else rxrpc_put_connection(conn); return true; @@ -280,7 +281,7 @@ void rxrpc_see_connection(struct rxrpc_connection *conn) { const void *here = __builtin_return_address(0); if (conn) { - int n = atomic_read(&conn->usage); + int n = refcount_read(&conn->ref); trace_rxrpc_conn(conn->debug_id, rxrpc_conn_seen, n, here); } @@ -292,9 +293,10 @@ void rxrpc_see_connection(struct rxrpc_connection *conn) struct rxrpc_connection *rxrpc_get_connection(struct rxrpc_connection *conn) { const void *here = __builtin_return_address(0); - int n = atomic_inc_return(&conn->usage); + int r; - trace_rxrpc_conn(conn->debug_id, rxrpc_conn_got, n, here); + __refcount_inc(&conn->ref, &r); + trace_rxrpc_conn(conn->debug_id, rxrpc_conn_got, r, here); return conn; } @@ -305,11 +307,11 @@ struct rxrpc_connection * rxrpc_get_connection_maybe(struct rxrpc_connection *conn) { const void *here = __builtin_return_address(0); + int r; if (conn) { - int n = atomic_fetch_add_unless(&conn->usage, 1, 0); - if (n > 0) - trace_rxrpc_conn(conn->debug_id, rxrpc_conn_got, n + 1, here); + if (__refcount_inc_not_zero(&conn->ref, &r)) + trace_rxrpc_conn(conn->debug_id, rxrpc_conn_got, r + 1, here); else conn = NULL; } @@ -333,12 +335,11 @@ void rxrpc_put_service_conn(struct rxrpc_connection *conn) { const void *here = __builtin_return_address(0); unsigned int debug_id = conn->debug_id; - int n; + int r; - n = atomic_dec_return(&conn->usage); - trace_rxrpc_conn(debug_id, rxrpc_conn_put_service, n, here); - ASSERTCMP(n, >=, 0); - if (n == 1) + __refcount_dec(&conn->ref, &r); + trace_rxrpc_conn(debug_id, rxrpc_conn_put_service, r - 1, here); + if (r - 1 == 1) rxrpc_set_service_reap_timer(conn->params.local->rxnet, jiffies + rxrpc_connection_expiry); } @@ -351,9 +352,9 @@ static void rxrpc_destroy_connection(struct rcu_head *rcu) struct rxrpc_connection *conn = container_of(rcu, struct rxrpc_connection, rcu); - _enter("{%d,u=%d}", conn->debug_id, atomic_read(&conn->usage)); + _enter("{%d,u=%d}", conn->debug_id, refcount_read(&conn->ref)); - ASSERTCMP(atomic_read(&conn->usage), ==, 0); + ASSERTCMP(refcount_read(&conn->ref), ==, 0); _net("DESTROY CONN %d", conn->debug_id); @@ -392,8 +393,8 @@ void rxrpc_service_connection_reaper(struct work_struct *work) write_lock(&rxnet->conn_lock); list_for_each_entry_safe(conn, _p, &rxnet->service_conns, link) { - ASSERTCMP(atomic_read(&conn->usage), >, 0); - if (likely(atomic_read(&conn->usage) > 1)) + ASSERTCMP(refcount_read(&conn->ref), >, 0); + if (likely(refcount_read(&conn->ref) > 1)) continue; if (conn->state == RXRPC_CONN_SERVICE_PREALLOC) continue; @@ -405,7 +406,7 @@ void rxrpc_service_connection_reaper(struct work_struct *work) expire_at = idle_timestamp + rxrpc_closed_conn_expiry * HZ; _debug("reap CONN %d { u=%d,t=%ld }", - conn->debug_id, atomic_read(&conn->usage), + conn->debug_id, refcount_read(&conn->ref), (long)expire_at - (long)now); if (time_before(now, expire_at)) { @@ -418,7 +419,7 @@ void rxrpc_service_connection_reaper(struct work_struct *work) /* The usage count sits at 1 whilst the object is unused on the * list; we reduce that to 0 to make the object unavailable. */ - if (atomic_cmpxchg(&conn->usage, 1, 0) != 1) + if (!refcount_dec_if_one(&conn->ref)) continue; trace_rxrpc_conn(conn->debug_id, rxrpc_conn_reap_service, 0, NULL); @@ -442,7 +443,7 @@ void rxrpc_service_connection_reaper(struct work_struct *work) link); list_del_init(&conn->link); - ASSERTCMP(atomic_read(&conn->usage), ==, 0); + ASSERTCMP(refcount_read(&conn->ref), ==, 0); rxrpc_kill_connection(conn); } @@ -470,7 +471,7 @@ void rxrpc_destroy_all_connections(struct rxrpc_net *rxnet) write_lock(&rxnet->conn_lock); list_for_each_entry_safe(conn, _p, &rxnet->service_conns, link) { pr_err("AF_RXRPC: Leaked conn %p {%d}\n", - conn, atomic_read(&conn->usage)); + conn, refcount_read(&conn->ref)); leak = true; } write_unlock(&rxnet->conn_lock); diff --git a/net/rxrpc/conn_service.c b/net/rxrpc/conn_service.c index e1966dfc9152..6e6aa02c6f9e 100644 --- a/net/rxrpc/conn_service.c +++ b/net/rxrpc/conn_service.c @@ -9,7 +9,7 @@ #include "ar-internal.h" static struct rxrpc_bundle rxrpc_service_dummy_bundle = { - .usage = ATOMIC_INIT(1), + .ref = REFCOUNT_INIT(1), .debug_id = UINT_MAX, .channel_lock = __SPIN_LOCK_UNLOCKED(&rxrpc_service_dummy_bundle.channel_lock), }; @@ -99,7 +99,7 @@ conn_published: return; found_extant_conn: - if (atomic_read(&cursor->usage) == 0) + if (refcount_read(&cursor->ref) == 0) goto replace_old_connection; write_sequnlock_bh(&peer->service_conn_lock); /* We should not be able to get here. rxrpc_incoming_connection() is @@ -132,7 +132,7 @@ struct rxrpc_connection *rxrpc_prealloc_service_connection(struct rxrpc_net *rxn * the rxrpc_connections list. */ conn->state = RXRPC_CONN_SERVICE_PREALLOC; - atomic_set(&conn->usage, 2); + refcount_set(&conn->ref, 2); conn->bundle = rxrpc_get_bundle(&rxrpc_service_dummy_bundle); atomic_inc(&rxnet->nr_conns); @@ -142,7 +142,7 @@ struct rxrpc_connection *rxrpc_prealloc_service_connection(struct rxrpc_net *rxn write_unlock(&rxnet->conn_lock); trace_rxrpc_conn(conn->debug_id, rxrpc_conn_new_service, - atomic_read(&conn->usage), + refcount_read(&conn->ref), __builtin_return_address(0)); } diff --git a/net/rxrpc/input.c b/net/rxrpc/input.c index dc201363f2c4..853b869b026a 100644 --- a/net/rxrpc/input.c +++ b/net/rxrpc/input.c @@ -1154,8 +1154,6 @@ static void rxrpc_post_packet_to_local(struct rxrpc_local *local, */ static void rxrpc_reject_packet(struct rxrpc_local *local, struct sk_buff *skb) { - CHECK_SLAB_OKAY(&local->usage); - if (rxrpc_get_local_maybe(local)) { skb_queue_tail(&local->reject_queue, skb); rxrpc_queue_local(local); @@ -1413,7 +1411,7 @@ int rxrpc_input_packet(struct sock *udp_sk, struct sk_buff *skb) } } - if (!call || atomic_read(&call->usage) == 0) { + if (!call || refcount_read(&call->ref) == 0) { if (rxrpc_to_client(sp) || sp->hdr.type != RXRPC_PACKET_TYPE_DATA) goto bad_message; diff --git a/net/rxrpc/local_object.c b/net/rxrpc/local_object.c index ea6f338bd131..96ecb7356c0f 100644 --- a/net/rxrpc/local_object.c +++ b/net/rxrpc/local_object.c @@ -79,7 +79,7 @@ static struct rxrpc_local *rxrpc_alloc_local(struct rxrpc_net *rxnet, local = kzalloc(sizeof(struct rxrpc_local), GFP_KERNEL); if (local) { - atomic_set(&local->usage, 1); + refcount_set(&local->ref, 1); atomic_set(&local->active_users, 1); local->rxnet = rxnet; INIT_HLIST_NODE(&local->link); @@ -265,10 +265,10 @@ addr_in_use: struct rxrpc_local *rxrpc_get_local(struct rxrpc_local *local) { const void *here = __builtin_return_address(0); - int n; + int r; - n = atomic_inc_return(&local->usage); - trace_rxrpc_local(local->debug_id, rxrpc_local_got, n, here); + __refcount_inc(&local->ref, &r); + trace_rxrpc_local(local->debug_id, rxrpc_local_got, r + 1, here); return local; } @@ -278,12 +278,12 @@ struct rxrpc_local *rxrpc_get_local(struct rxrpc_local *local) struct rxrpc_local *rxrpc_get_local_maybe(struct rxrpc_local *local) { const void *here = __builtin_return_address(0); + int r; if (local) { - int n = atomic_fetch_add_unless(&local->usage, 1, 0); - if (n > 0) + if (__refcount_inc_not_zero(&local->ref, &r)) trace_rxrpc_local(local->debug_id, rxrpc_local_got, - n + 1, here); + r + 1, here); else local = NULL; } @@ -297,10 +297,10 @@ void rxrpc_queue_local(struct rxrpc_local *local) { const void *here = __builtin_return_address(0); unsigned int debug_id = local->debug_id; - int n = atomic_read(&local->usage); + int r = refcount_read(&local->ref); if (rxrpc_queue_work(&local->processor)) - trace_rxrpc_local(debug_id, rxrpc_local_queued, n, here); + trace_rxrpc_local(debug_id, rxrpc_local_queued, r + 1, here); else rxrpc_put_local(local); } @@ -312,15 +312,16 @@ void rxrpc_put_local(struct rxrpc_local *local) { const void *here = __builtin_return_address(0); unsigned int debug_id; - int n; + bool dead; + int r; if (local) { debug_id = local->debug_id; - n = atomic_dec_return(&local->usage); - trace_rxrpc_local(debug_id, rxrpc_local_put, n, here); + dead = __refcount_dec_and_test(&local->ref, &r); + trace_rxrpc_local(debug_id, rxrpc_local_put, r, here); - if (n == 0) + if (dead) call_rcu(&local->rcu, rxrpc_local_rcu); } } @@ -405,7 +406,7 @@ static void rxrpc_local_processor(struct work_struct *work) bool again; trace_rxrpc_local(local->debug_id, rxrpc_local_processing, - atomic_read(&local->usage), NULL); + refcount_read(&local->ref), NULL); do { again = false; @@ -461,7 +462,7 @@ void rxrpc_destroy_all_locals(struct rxrpc_net *rxnet) mutex_lock(&rxnet->local_mutex); hlist_for_each_entry(local, &rxnet->local_endpoints, link) { pr_err("AF_RXRPC: Leaked local %p {%d}\n", - local, atomic_read(&local->usage)); + local, refcount_read(&local->ref)); } mutex_unlock(&rxnet->local_mutex); BUG(); diff --git a/net/rxrpc/peer_object.c b/net/rxrpc/peer_object.c index 0298fe2ad6d3..26d2ae9baaf2 100644 --- a/net/rxrpc/peer_object.c +++ b/net/rxrpc/peer_object.c @@ -121,7 +121,7 @@ static struct rxrpc_peer *__rxrpc_lookup_peer_rcu( hash_for_each_possible_rcu(rxnet->peer_hash, peer, hash_link, hash_key) { if (rxrpc_peer_cmp_key(peer, local, srx, hash_key) == 0 && - atomic_read(&peer->usage) > 0) + refcount_read(&peer->ref) > 0) return peer; } @@ -140,7 +140,7 @@ struct rxrpc_peer *rxrpc_lookup_peer_rcu(struct rxrpc_local *local, peer = __rxrpc_lookup_peer_rcu(local, srx, hash_key); if (peer) { _net("PEER %d {%pISp}", peer->debug_id, &peer->srx.transport); - _leave(" = %p {u=%d}", peer, atomic_read(&peer->usage)); + _leave(" = %p {u=%d}", peer, refcount_read(&peer->ref)); } return peer; } @@ -216,7 +216,7 @@ struct rxrpc_peer *rxrpc_alloc_peer(struct rxrpc_local *local, gfp_t gfp) peer = kzalloc(sizeof(struct rxrpc_peer), gfp); if (peer) { - atomic_set(&peer->usage, 1); + refcount_set(&peer->ref, 1); peer->local = rxrpc_get_local(local); INIT_HLIST_HEAD(&peer->error_targets); peer->service_conns = RB_ROOT; @@ -378,7 +378,7 @@ struct rxrpc_peer *rxrpc_lookup_peer(struct rxrpc_sock *rx, _net("PEER %d {%pISp}", peer->debug_id, &peer->srx.transport); - _leave(" = %p {u=%d}", peer, atomic_read(&peer->usage)); + _leave(" = %p {u=%d}", peer, refcount_read(&peer->ref)); return peer; } @@ -388,10 +388,10 @@ struct rxrpc_peer *rxrpc_lookup_peer(struct rxrpc_sock *rx, struct rxrpc_peer *rxrpc_get_peer(struct rxrpc_peer *peer) { const void *here = __builtin_return_address(0); - int n; + int r; - n = atomic_inc_return(&peer->usage); - trace_rxrpc_peer(peer->debug_id, rxrpc_peer_got, n, here); + __refcount_inc(&peer->ref, &r); + trace_rxrpc_peer(peer->debug_id, rxrpc_peer_got, r + 1, here); return peer; } @@ -401,11 +401,11 @@ struct rxrpc_peer *rxrpc_get_peer(struct rxrpc_peer *peer) struct rxrpc_peer *rxrpc_get_peer_maybe(struct rxrpc_peer *peer) { const void *here = __builtin_return_address(0); + int r; if (peer) { - int n = atomic_fetch_add_unless(&peer->usage, 1, 0); - if (n > 0) - trace_rxrpc_peer(peer->debug_id, rxrpc_peer_got, n + 1, here); + if (__refcount_inc_not_zero(&peer->ref, &r)) + trace_rxrpc_peer(peer->debug_id, rxrpc_peer_got, r + 1, here); else peer = NULL; } @@ -436,13 +436,14 @@ void rxrpc_put_peer(struct rxrpc_peer *peer) { const void *here = __builtin_return_address(0); unsigned int debug_id; - int n; + bool dead; + int r; if (peer) { debug_id = peer->debug_id; - n = atomic_dec_return(&peer->usage); - trace_rxrpc_peer(debug_id, rxrpc_peer_put, n, here); - if (n == 0) + dead = __refcount_dec_and_test(&peer->ref, &r); + trace_rxrpc_peer(debug_id, rxrpc_peer_put, r - 1, here); + if (dead) __rxrpc_put_peer(peer); } } @@ -455,11 +456,12 @@ void rxrpc_put_peer_locked(struct rxrpc_peer *peer) { const void *here = __builtin_return_address(0); unsigned int debug_id = peer->debug_id; - int n; + bool dead; + int r; - n = atomic_dec_return(&peer->usage); - trace_rxrpc_peer(debug_id, rxrpc_peer_put, n, here); - if (n == 0) { + dead = __refcount_dec_and_test(&peer->ref, &r); + trace_rxrpc_peer(debug_id, rxrpc_peer_put, r - 1, here); + if (dead) { hash_del_rcu(&peer->hash_link); list_del_init(&peer->keepalive_link); rxrpc_free_peer(peer); @@ -481,7 +483,7 @@ void rxrpc_destroy_all_peers(struct rxrpc_net *rxnet) hlist_for_each_entry(peer, &rxnet->peer_hash[i], hash_link) { pr_err("Leaked peer %u {%u} %pISp\n", peer->debug_id, - atomic_read(&peer->usage), + refcount_read(&peer->ref), &peer->srx.transport); } } diff --git a/net/rxrpc/proc.c b/net/rxrpc/proc.c index 8a8f776f91ae..8967201fd8e5 100644 --- a/net/rxrpc/proc.c +++ b/net/rxrpc/proc.c @@ -107,7 +107,7 @@ static int rxrpc_call_seq_show(struct seq_file *seq, void *v) call->cid, call->call_id, rxrpc_is_service_call(call) ? "Svc" : "Clt", - atomic_read(&call->usage), + refcount_read(&call->ref), rxrpc_call_states[call->state], call->abort_code, call->debug_id, @@ -189,7 +189,7 @@ print: conn->service_id, conn->proto.cid, rxrpc_conn_is_service(conn) ? "Svc" : "Clt", - atomic_read(&conn->usage), + refcount_read(&conn->ref), rxrpc_conn_states[conn->state], key_serial(conn->params.key), atomic_read(&conn->serial), @@ -239,7 +239,7 @@ static int rxrpc_peer_seq_show(struct seq_file *seq, void *v) " %3u %5u %6llus %8u %8u\n", lbuff, rbuff, - atomic_read(&peer->usage), + refcount_read(&peer->ref), peer->cong_cwnd, peer->mtu, now - peer->last_tx_at, @@ -357,7 +357,7 @@ static int rxrpc_local_seq_show(struct seq_file *seq, void *v) seq_printf(seq, "UDP %-47.47s %3u %3u\n", lbuff, - atomic_read(&local->usage), + refcount_read(&local->ref), atomic_read(&local->active_users)); return 0; diff --git a/net/rxrpc/skbuff.c b/net/rxrpc/skbuff.c index 0348d2bf6f7d..580a5acffee7 100644 --- a/net/rxrpc/skbuff.c +++ b/net/rxrpc/skbuff.c @@ -71,7 +71,6 @@ void rxrpc_free_skb(struct sk_buff *skb, enum rxrpc_skb_trace op) const void *here = __builtin_return_address(0); if (skb) { 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, rxrpc_skb(skb)->rx_flags, here); -- cgit v1.3-8-gc7d7 From dc9fd093b2ebabc25a5d7d0dd5bdfdd6c878e0bf Mon Sep 17 00:00:00 2001 From: David Howells Date: Sat, 21 May 2022 08:45:35 +0100 Subject: rxrpc: Automatically generate trace tag enums Automatically generate trace tag enums from the symbol -> string mapping tables rather than having the enums as well, thereby reducing duplicated data. Signed-off-by: David Howells cc: Marc Dionne cc: linux-afs@lists.infradead.org Signed-off-by: David S. Miller --- include/trace/events/rxrpc.h | 261 +++++++------------------------------------ 1 file changed, 42 insertions(+), 219 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index cdb28976641b..66915b872a44 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -13,215 +13,6 @@ #include #include -/* - * Define enums for tracing information. - * - * These should all be kept sorted, making it easier to match the string - * mapping tables further on. - */ -#ifndef __RXRPC_DECLARE_TRACE_ENUMS_ONCE_ONLY -#define __RXRPC_DECLARE_TRACE_ENUMS_ONCE_ONLY - -enum rxrpc_skb_trace { - 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, - rxrpc_skb_unshared, - rxrpc_skb_unshared_nomem, -}; - -enum rxrpc_local_trace { - rxrpc_local_got, - rxrpc_local_new, - rxrpc_local_processing, - rxrpc_local_put, - rxrpc_local_queued, -}; - -enum rxrpc_peer_trace { - rxrpc_peer_got, - rxrpc_peer_new, - rxrpc_peer_processing, - rxrpc_peer_put, -}; - -enum rxrpc_conn_trace { - rxrpc_conn_got, - rxrpc_conn_new_client, - rxrpc_conn_new_service, - rxrpc_conn_put_client, - rxrpc_conn_put_service, - rxrpc_conn_queued, - rxrpc_conn_reap_service, - rxrpc_conn_seen, -}; - -enum rxrpc_client_trace { - rxrpc_client_activate_chans, - rxrpc_client_alloc, - rxrpc_client_chan_activate, - rxrpc_client_chan_disconnect, - rxrpc_client_chan_pass, - rxrpc_client_chan_wait_failed, - rxrpc_client_cleanup, - rxrpc_client_discard, - rxrpc_client_duplicate, - rxrpc_client_exposed, - rxrpc_client_replace, - rxrpc_client_to_active, - rxrpc_client_to_idle, -}; - -enum rxrpc_call_trace { - rxrpc_call_connected, - rxrpc_call_error, - rxrpc_call_got, - rxrpc_call_got_kernel, - rxrpc_call_got_timer, - rxrpc_call_got_userid, - rxrpc_call_new_client, - rxrpc_call_new_service, - rxrpc_call_put, - rxrpc_call_put_kernel, - rxrpc_call_put_noqueue, - rxrpc_call_put_notimer, - rxrpc_call_put_timer, - rxrpc_call_put_userid, - rxrpc_call_queued, - rxrpc_call_queued_ref, - rxrpc_call_release, - rxrpc_call_seen, -}; - -enum rxrpc_transmit_trace { - rxrpc_transmit_await_reply, - rxrpc_transmit_end, - rxrpc_transmit_queue, - rxrpc_transmit_queue_last, - rxrpc_transmit_rotate, - rxrpc_transmit_rotate_last, - rxrpc_transmit_wait, -}; - -enum rxrpc_receive_trace { - rxrpc_receive_end, - rxrpc_receive_front, - rxrpc_receive_incoming, - rxrpc_receive_queue, - rxrpc_receive_queue_last, - rxrpc_receive_rotate, -}; - -enum rxrpc_recvmsg_trace { - rxrpc_recvmsg_cont, - rxrpc_recvmsg_data_return, - rxrpc_recvmsg_dequeue, - rxrpc_recvmsg_enter, - rxrpc_recvmsg_full, - rxrpc_recvmsg_hole, - rxrpc_recvmsg_next, - rxrpc_recvmsg_requeue, - rxrpc_recvmsg_return, - rxrpc_recvmsg_terminal, - rxrpc_recvmsg_to_be_accepted, - rxrpc_recvmsg_wait, -}; - -enum rxrpc_rtt_tx_trace { - rxrpc_rtt_tx_cancel, - rxrpc_rtt_tx_data, - rxrpc_rtt_tx_no_slot, - rxrpc_rtt_tx_ping, -}; - -enum rxrpc_rtt_rx_trace { - rxrpc_rtt_rx_cancel, - rxrpc_rtt_rx_lost, - rxrpc_rtt_rx_obsolete, - rxrpc_rtt_rx_ping_response, - rxrpc_rtt_rx_requested_ack, -}; - -enum rxrpc_timer_trace { - rxrpc_timer_begin, - rxrpc_timer_exp_ack, - rxrpc_timer_exp_hard, - rxrpc_timer_exp_idle, - rxrpc_timer_exp_keepalive, - rxrpc_timer_exp_lost_ack, - rxrpc_timer_exp_normal, - rxrpc_timer_exp_ping, - rxrpc_timer_exp_resend, - rxrpc_timer_expired, - rxrpc_timer_init_for_reply, - rxrpc_timer_init_for_send_reply, - rxrpc_timer_restart, - rxrpc_timer_set_for_ack, - rxrpc_timer_set_for_hard, - rxrpc_timer_set_for_idle, - rxrpc_timer_set_for_keepalive, - rxrpc_timer_set_for_lost_ack, - rxrpc_timer_set_for_normal, - rxrpc_timer_set_for_ping, - rxrpc_timer_set_for_resend, - rxrpc_timer_set_for_send, -}; - -enum rxrpc_propose_ack_trace { - rxrpc_propose_ack_client_tx_end, - rxrpc_propose_ack_input_data, - rxrpc_propose_ack_ping_for_check_life, - rxrpc_propose_ack_ping_for_keepalive, - rxrpc_propose_ack_ping_for_lost_ack, - rxrpc_propose_ack_ping_for_lost_reply, - rxrpc_propose_ack_ping_for_params, - rxrpc_propose_ack_processing_op, - rxrpc_propose_ack_respond_to_ack, - rxrpc_propose_ack_respond_to_ping, - rxrpc_propose_ack_retry_tx, - rxrpc_propose_ack_rotate_rx, - rxrpc_propose_ack_terminal_ack, -}; - -enum rxrpc_propose_ack_outcome { - rxrpc_propose_ack_subsume, - rxrpc_propose_ack_update, - rxrpc_propose_ack_use, -}; - -enum rxrpc_congest_change { - rxrpc_cong_begin_retransmission, - rxrpc_cong_cleared_nacks, - rxrpc_cong_new_low_nack, - rxrpc_cong_no_change, - rxrpc_cong_progress, - rxrpc_cong_retransmit_again, - rxrpc_cong_rtt_window_end, - rxrpc_cong_saw_nack, -}; - -enum rxrpc_tx_point { - rxrpc_tx_point_call_abort, - rxrpc_tx_point_call_ack, - rxrpc_tx_point_call_data_frag, - rxrpc_tx_point_call_data_nofrag, - rxrpc_tx_point_call_final_resend, - rxrpc_tx_point_conn_abort, - rxrpc_tx_point_rxkad_challenge, - rxrpc_tx_point_rxkad_response, - rxrpc_tx_point_reject, - rxrpc_tx_point_version_keepalive, - rxrpc_tx_point_version_reply, -}; - -#endif /* end __RXRPC_DECLARE_TRACE_ENUMS_ONCE_ONLY */ - /* * Declare tracing information enums and their string mappings for display. */ @@ -451,6 +242,36 @@ enum rxrpc_tx_point { EM(rxrpc_tx_point_version_keepalive, "VerKeepalive") \ E_(rxrpc_tx_point_version_reply, "VerReply") +/* + * Generate enums for tracing information. + */ +#ifndef __NETFS_DECLARE_TRACE_ENUMS_ONCE_ONLY +#define __NETFS_DECLARE_TRACE_ENUMS_ONCE_ONLY + +#undef EM +#undef E_ +#define EM(a, b) a, +#define E_(a, b) a + +enum rxrpc_call_trace { rxrpc_call_traces } __mode(byte); +enum rxrpc_client_trace { rxrpc_client_traces } __mode(byte); +enum rxrpc_congest_change { rxrpc_congest_changes } __mode(byte); +enum rxrpc_conn_trace { rxrpc_conn_traces } __mode(byte); +enum rxrpc_local_trace { rxrpc_local_traces } __mode(byte); +enum rxrpc_peer_trace { rxrpc_peer_traces } __mode(byte); +enum rxrpc_propose_ack_outcome { rxrpc_propose_ack_outcomes } __mode(byte); +enum rxrpc_propose_ack_trace { rxrpc_propose_ack_traces } __mode(byte); +enum rxrpc_receive_trace { rxrpc_receive_traces } __mode(byte); +enum rxrpc_recvmsg_trace { rxrpc_recvmsg_traces } __mode(byte); +enum rxrpc_rtt_rx_trace { rxrpc_rtt_rx_traces } __mode(byte); +enum rxrpc_rtt_tx_trace { rxrpc_rtt_tx_traces } __mode(byte); +enum rxrpc_skb_trace { rxrpc_skb_traces } __mode(byte); +enum rxrpc_timer_trace { rxrpc_timer_traces } __mode(byte); +enum rxrpc_transmit_trace { rxrpc_transmit_traces } __mode(byte); +enum rxrpc_tx_point { rxrpc_tx_points } __mode(byte); + +#endif /* end __RXRPC_DECLARE_TRACE_ENUMS_ONCE_ONLY */ + /* * Export enum symbols via userspace. */ @@ -459,21 +280,21 @@ enum rxrpc_tx_point { #define EM(a, b) TRACE_DEFINE_ENUM(a); #define E_(a, b) TRACE_DEFINE_ENUM(a); -rxrpc_skb_traces; -rxrpc_local_traces; -rxrpc_conn_traces; -rxrpc_client_traces; rxrpc_call_traces; -rxrpc_transmit_traces; +rxrpc_client_traces; +rxrpc_congest_changes; +rxrpc_congest_modes; +rxrpc_conn_traces; +rxrpc_local_traces; +rxrpc_propose_ack_outcomes; +rxrpc_propose_ack_traces; rxrpc_receive_traces; rxrpc_recvmsg_traces; -rxrpc_rtt_tx_traces; rxrpc_rtt_rx_traces; +rxrpc_rtt_tx_traces; +rxrpc_skb_traces; rxrpc_timer_traces; -rxrpc_propose_ack_traces; -rxrpc_propose_ack_outcomes; -rxrpc_congest_modes; -rxrpc_congest_changes; +rxrpc_transmit_traces; rxrpc_tx_points; /* @@ -1574,6 +1395,8 @@ TRACE_EVENT(rxrpc_rx_discard_ack, __entry->call_ackr_prev) ); +#undef EM +#undef E_ #endif /* _TRACE_RXRPC_H */ /* This part must be outside protection */ -- cgit v1.3-8-gc7d7 From 9a3dedcf18096e8f7f22b8777d78c4acfdea1651 Mon Sep 17 00:00:00 2001 From: David Howells Date: Sat, 21 May 2022 09:03:31 +0100 Subject: rxrpc: Fix decision on when to generate an IDLE ACK Fix the decision on when to generate an IDLE ACK by keeping a count of the number of packets we've received, but not yet soft-ACK'd, and the number of packets we've processed, but not yet hard-ACK'd, rather than trying to keep track of which DATA sequence numbers correspond to those points. We then generate an ACK when either counter exceeds 2. The counters are both cleared when we transcribe the information into any sort of ACK packet for transmission. IDLE and DELAY ACKs are skipped if both counters are 0 (ie. no change). Fixes: 805b21b929e2 ("rxrpc: Send an ACK after every few DATA packets we receive") Signed-off-by: David Howells cc: Marc Dionne cc: linux-afs@lists.infradead.org Signed-off-by: David S. Miller --- include/trace/events/rxrpc.h | 2 +- net/rxrpc/ar-internal.h | 4 ++-- net/rxrpc/input.c | 11 +++++++++-- net/rxrpc/output.c | 18 +++++++++++------- net/rxrpc/recvmsg.c | 8 +++----- 5 files changed, 26 insertions(+), 17 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index 4a3ab0ed6e06..1c714336b863 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -1509,7 +1509,7 @@ TRACE_EVENT(rxrpc_call_reset, __entry->call_serial = call->rx_serial; __entry->conn_serial = call->conn->hi_serial; __entry->tx_seq = call->tx_hard_ack; - __entry->rx_seq = call->ackr_seen; + __entry->rx_seq = call->rx_hard_ack; ), TP_printk("c=%08x %08x:%08x r=%08x/%08x tx=%08x rx=%08x", diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index 8465985a4cb6..dce056adb78c 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -680,8 +680,8 @@ struct rxrpc_call { u8 ackr_reason; /* reason to ACK */ rxrpc_serial_t ackr_serial; /* serial of packet being ACK'd */ rxrpc_seq_t ackr_highest_seq; /* Higest sequence number received */ - rxrpc_seq_t ackr_consumed; /* Highest packet shown consumed */ - rxrpc_seq_t ackr_seen; /* Highest packet shown seen */ + atomic_t ackr_nr_unacked; /* Number of unacked packets */ + atomic_t ackr_nr_consumed; /* Number of packets needing hard ACK */ /* RTT management */ rxrpc_serial_t rtt_serial[4]; /* Serial number of DATA or PING sent */ diff --git a/net/rxrpc/input.c b/net/rxrpc/input.c index 2e61545ad8ca..1145cb14d86f 100644 --- a/net/rxrpc/input.c +++ b/net/rxrpc/input.c @@ -412,8 +412,8 @@ 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 j, nr_subpackets; - rxrpc_serial_t serial = sp->hdr.serial, ack_serial = 0; + unsigned int j, nr_subpackets, nr_unacked = 0; + rxrpc_serial_t serial = sp->hdr.serial, ack_serial = serial; rxrpc_seq_t seq0 = sp->hdr.seq, hard_ack; bool immediate_ack = false, jumbo_bad = false; u8 ack = 0; @@ -569,6 +569,8 @@ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb) sp = NULL; } + nr_unacked++; + if (last) { set_bit(RXRPC_CALL_RX_LAST, &call->flags); if (!ack) { @@ -588,9 +590,14 @@ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb) } call->rx_expect_next = seq + 1; } + if (!ack) + ack_serial = serial; } ack: + if (atomic_add_return(nr_unacked, &call->ackr_nr_unacked) > 2 && !ack) + ack = RXRPC_ACK_IDLE; + if (ack) rxrpc_propose_ACK(call, ack, ack_serial, immediate_ack, true, diff --git a/net/rxrpc/output.c b/net/rxrpc/output.c index 46aae9b7006f..9683617db704 100644 --- a/net/rxrpc/output.c +++ b/net/rxrpc/output.c @@ -74,11 +74,18 @@ static size_t rxrpc_fill_out_ack(struct rxrpc_connection *conn, u8 reason) { rxrpc_serial_t serial; + unsigned int tmp; rxrpc_seq_t hard_ack, top, seq; int ix; u32 mtu, jmax; u8 *ackp = pkt->acks; + tmp = atomic_xchg(&call->ackr_nr_unacked, 0); + tmp |= atomic_xchg(&call->ackr_nr_consumed, 0); + if (!tmp && (reason == RXRPC_ACK_DELAY || + reason == RXRPC_ACK_IDLE)) + return 0; + /* Barrier against rxrpc_input_data(). */ serial = call->ackr_serial; hard_ack = READ_ONCE(call->rx_hard_ack); @@ -223,6 +230,10 @@ int rxrpc_send_ack_packet(struct rxrpc_call *call, bool ping, n = rxrpc_fill_out_ack(conn, call, pkt, &hard_ack, &top, reason); spin_unlock_bh(&call->lock); + if (n == 0) { + kfree(pkt); + return 0; + } iov[0].iov_base = pkt; iov[0].iov_len = sizeof(pkt->whdr) + sizeof(pkt->ack) + n; @@ -259,13 +270,6 @@ int rxrpc_send_ack_packet(struct rxrpc_call *call, bool ping, ntohl(pkt->ack.serial), false, true, rxrpc_propose_ack_retry_tx); - } else { - spin_lock_bh(&call->lock); - if (after(hard_ack, call->ackr_consumed)) - call->ackr_consumed = hard_ack; - if (after(top, call->ackr_seen)) - call->ackr_seen = top; - spin_unlock_bh(&call->lock); } rxrpc_set_keepalive(call); diff --git a/net/rxrpc/recvmsg.c b/net/rxrpc/recvmsg.c index eca6dda26c77..250f23bc1c07 100644 --- a/net/rxrpc/recvmsg.c +++ b/net/rxrpc/recvmsg.c @@ -260,11 +260,9 @@ static void rxrpc_rotate_rx_window(struct rxrpc_call *call) rxrpc_end_rx_phase(call, serial); } else { /* Check to see if there's an ACK that needs sending. */ - if (after_eq(hard_ack, call->ackr_consumed + 2) || - after_eq(top, call->ackr_seen + 2) || - (hard_ack == top && after(hard_ack, call->ackr_consumed))) - rxrpc_propose_ACK(call, RXRPC_ACK_DELAY, serial, - true, true, + if (atomic_inc_return(&call->ackr_nr_consumed) > 2) + rxrpc_propose_ACK(call, RXRPC_ACK_IDLE, serial, + true, false, rxrpc_propose_ack_rotate_rx); if (call->ackr_reason && call->ackr_reason != RXRPC_ACK_DELAY) rxrpc_send_ack_packet(call, false, NULL); -- cgit v1.3-8-gc7d7 From 185194f19134716f32b2ab04ac76cf93ea85706a Mon Sep 17 00:00:00 2001 From: Vasily Averin Date: Tue, 17 May 2022 09:58:14 +0300 Subject: include/trace/events/mmflags.h: cleanup for "tracing: incorrect gfp_t conversion" Redefines __def_gfpflag_names array according to akpm@, willy@ and Joe Perches recommendations. Link: https://lkml.kernel.org/r/6f811e19-41c6-f3e8-fca6-23a19a62e313@openvz.org Fixes: fe573327ffb1 ("tracing: incorrect gfp_t conversion") Signed-off-by: Vasily Averin Cc: Matthew Wilcox Cc: Joe Perches Cc: Steven Rostedt Signed-off-by: Andrew Morton --- include/trace/events/mmflags.h | 84 +++++++++++++++++++++--------------------- 1 file changed, 43 insertions(+), 41 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/mmflags.h b/include/trace/events/mmflags.h index fcb1ffb2cc8d..e87cb2b80ed3 100644 --- a/include/trace/events/mmflags.h +++ b/include/trace/events/mmflags.h @@ -13,49 +13,51 @@ * Thus most bits set go first. */ -#define __def_gfpflag_names \ - {(__force unsigned long)GFP_TRANSHUGE, "GFP_TRANSHUGE"}, \ - {(__force unsigned long)GFP_TRANSHUGE_LIGHT, "GFP_TRANSHUGE_LIGHT"}, \ - {(__force unsigned long)GFP_HIGHUSER_MOVABLE, "GFP_HIGHUSER_MOVABLE"},\ - {(__force unsigned long)GFP_HIGHUSER, "GFP_HIGHUSER"}, \ - {(__force unsigned long)GFP_USER, "GFP_USER"}, \ - {(__force unsigned long)GFP_KERNEL_ACCOUNT, "GFP_KERNEL_ACCOUNT"}, \ - {(__force unsigned long)GFP_KERNEL, "GFP_KERNEL"}, \ - {(__force unsigned long)GFP_NOFS, "GFP_NOFS"}, \ - {(__force unsigned long)GFP_ATOMIC, "GFP_ATOMIC"}, \ - {(__force unsigned long)GFP_NOIO, "GFP_NOIO"}, \ - {(__force unsigned long)GFP_NOWAIT, "GFP_NOWAIT"}, \ - {(__force unsigned long)GFP_DMA, "GFP_DMA"}, \ - {(__force unsigned long)__GFP_HIGHMEM, "__GFP_HIGHMEM"}, \ - {(__force unsigned long)GFP_DMA32, "GFP_DMA32"}, \ - {(__force unsigned long)__GFP_HIGH, "__GFP_HIGH"}, \ - {(__force unsigned long)__GFP_ATOMIC, "__GFP_ATOMIC"}, \ - {(__force unsigned long)__GFP_IO, "__GFP_IO"}, \ - {(__force unsigned long)__GFP_FS, "__GFP_FS"}, \ - {(__force unsigned long)__GFP_NOWARN, "__GFP_NOWARN"}, \ - {(__force unsigned long)__GFP_RETRY_MAYFAIL, "__GFP_RETRY_MAYFAIL"}, \ - {(__force unsigned long)__GFP_NOFAIL, "__GFP_NOFAIL"}, \ - {(__force unsigned long)__GFP_NORETRY, "__GFP_NORETRY"}, \ - {(__force unsigned long)__GFP_COMP, "__GFP_COMP"}, \ - {(__force unsigned long)__GFP_ZERO, "__GFP_ZERO"}, \ - {(__force unsigned long)__GFP_NOMEMALLOC, "__GFP_NOMEMALLOC"}, \ - {(__force unsigned long)__GFP_MEMALLOC, "__GFP_MEMALLOC"}, \ - {(__force unsigned long)__GFP_HARDWALL, "__GFP_HARDWALL"}, \ - {(__force unsigned long)__GFP_THISNODE, "__GFP_THISNODE"}, \ - {(__force unsigned long)__GFP_RECLAIMABLE, "__GFP_RECLAIMABLE"}, \ - {(__force unsigned long)__GFP_MOVABLE, "__GFP_MOVABLE"}, \ - {(__force unsigned long)__GFP_ACCOUNT, "__GFP_ACCOUNT"}, \ - {(__force unsigned long)__GFP_WRITE, "__GFP_WRITE"}, \ - {(__force unsigned long)__GFP_RECLAIM, "__GFP_RECLAIM"}, \ - {(__force unsigned long)__GFP_DIRECT_RECLAIM, "__GFP_DIRECT_RECLAIM"},\ - {(__force unsigned long)__GFP_KSWAPD_RECLAIM, "__GFP_KSWAPD_RECLAIM"},\ - {(__force unsigned long)__GFP_ZEROTAGS, "__GFP_ZEROTAGS"} \ +#define gfpflag_string(flag) {(__force unsigned long)flag, #flag} + +#define __def_gfpflag_names \ + gfpflag_string(GFP_TRANSHUGE), \ + gfpflag_string(GFP_TRANSHUGE_LIGHT), \ + gfpflag_string(GFP_HIGHUSER_MOVABLE), \ + gfpflag_string(GFP_HIGHUSER), \ + gfpflag_string(GFP_USER), \ + gfpflag_string(GFP_KERNEL_ACCOUNT), \ + gfpflag_string(GFP_KERNEL), \ + gfpflag_string(GFP_NOFS), \ + gfpflag_string(GFP_ATOMIC), \ + gfpflag_string(GFP_NOIO), \ + gfpflag_string(GFP_NOWAIT), \ + gfpflag_string(GFP_DMA), \ + gfpflag_string(__GFP_HIGHMEM), \ + gfpflag_string(GFP_DMA32), \ + gfpflag_string(__GFP_HIGH), \ + gfpflag_string(__GFP_ATOMIC), \ + gfpflag_string(__GFP_IO), \ + gfpflag_string(__GFP_FS), \ + gfpflag_string(__GFP_NOWARN), \ + gfpflag_string(__GFP_RETRY_MAYFAIL), \ + gfpflag_string(__GFP_NOFAIL), \ + gfpflag_string(__GFP_NORETRY), \ + gfpflag_string(__GFP_COMP), \ + gfpflag_string(__GFP_ZERO), \ + gfpflag_string(__GFP_NOMEMALLOC), \ + gfpflag_string(__GFP_MEMALLOC), \ + gfpflag_string(__GFP_HARDWALL), \ + gfpflag_string(__GFP_THISNODE), \ + gfpflag_string(__GFP_RECLAIMABLE), \ + gfpflag_string(__GFP_MOVABLE), \ + gfpflag_string(__GFP_ACCOUNT), \ + gfpflag_string(__GFP_WRITE), \ + gfpflag_string(__GFP_RECLAIM), \ + gfpflag_string(__GFP_DIRECT_RECLAIM), \ + gfpflag_string(__GFP_KSWAPD_RECLAIM), \ + gfpflag_string(__GFP_ZEROTAGS) #ifdef CONFIG_KASAN_HW_TAGS -#define __def_gfpflag_names_kasan , \ - {(__force unsigned long)__GFP_SKIP_ZERO, "__GFP_SKIP_ZERO"}, \ - {(__force unsigned long)__GFP_SKIP_KASAN_POISON, "__GFP_SKIP_KASAN_POISON"}, \ - {(__force unsigned long)__GFP_SKIP_KASAN_UNPOISON, "__GFP_SKIP_KASAN_UNPOISON"} +#define __def_gfpflag_names_kasan , \ + gfpflag_string(__GFP_SKIP_ZERO), \ + gfpflag_string(__GFP_SKIP_KASAN_POISON), \ + gfpflag_string(__GFP_SKIP_KASAN_UNPOISON) #else #define __def_gfpflag_names_kasan #endif -- cgit v1.3-8-gc7d7 From e5c3f619a04d9e191de66d5f8e069367f73a6487 Mon Sep 17 00:00:00 2001 From: Vasily Averin Date: Tue, 17 May 2022 09:55:51 +0300 Subject: include/trace/events/percpu.h: cleanup for "percpu: improve percpu_alloc_percpu event trace" Fix sparse warning about incorrect gfp_t cast. Link: https://lkml.kernel.org/r/001979f3-e978-0998-cbed-61a4a2ac87b8@openvz.org Fixes: f67bed134a05 ("percpu: improve percpu_alloc_percpu event trace") Signed-off-by: Vasily Averin Signed-off-by: Andrew Morton --- include/trace/events/percpu.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/percpu.h b/include/trace/events/percpu.h index e989cefc0def..5b8211ca8950 100644 --- a/include/trace/events/percpu.h +++ b/include/trace/events/percpu.h @@ -28,7 +28,7 @@ TRACE_EVENT(percpu_alloc_percpu, __field( int, off ) __field( void __percpu *, ptr ) __field( size_t, bytes_alloc ) - __field( gfp_t, gfp_flags ) + __field( unsigned long, gfp_flags ) ), TP_fast_assign( __entry->call_site = call_site; @@ -40,7 +40,7 @@ TRACE_EVENT(percpu_alloc_percpu, __entry->off = off; __entry->ptr = ptr; __entry->bytes_alloc = bytes_alloc; - __entry->gfp_flags = gfp_flags; + __entry->gfp_flags = (__force unsigned long)gfp_flags; ), TP_printk("call_site=%pS reserved=%d is_atomic=%d size=%zu align=%zu base_addr=%p off=%d ptr=%p bytes_alloc=%zu gfp_flags=%s", -- cgit v1.3-8-gc7d7 From 873a400938b31a1e443c4d94b560b78300787540 Mon Sep 17 00:00:00 2001 From: Wonhyuk Yang Date: Wed, 4 May 2022 11:32:03 +0900 Subject: workqueue: Fix type of cpu in trace event The trace event "workqueue_queue_work" use unsigned int type for req_cpu, cpu. This casue confusing cpu number like below log. $ cat /sys/kernel/debug/tracing/trace cat-317 [001] ...: workqueue_queue_work: ... req_cpu=8192 cpu=4294967295 So, change unsigned type to signed type in the trace event. After applying this patch, cpu number will be printed as -1 instead of 4294967295 as folllows. $ cat /sys/kernel/debug/tracing/trace cat-1338 [002] ...: workqueue_queue_work: ... req_cpu=8192 cpu=-1 Cc: Baik Song An Cc: Hong Yeon Kim Cc: Taeung Song Cc: linuxgeek@linuxgeek.io Signed-off-by: Wonhyuk Yang Signed-off-by: Tejun Heo --- include/trace/events/workqueue.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/workqueue.h b/include/trace/events/workqueue.h index 6154a2e72bce..262d52021c23 100644 --- a/include/trace/events/workqueue.h +++ b/include/trace/events/workqueue.h @@ -22,7 +22,7 @@ struct pool_workqueue; */ TRACE_EVENT(workqueue_queue_work, - TP_PROTO(unsigned int req_cpu, struct pool_workqueue *pwq, + TP_PROTO(int req_cpu, struct pool_workqueue *pwq, struct work_struct *work), TP_ARGS(req_cpu, pwq, work), @@ -31,8 +31,8 @@ TRACE_EVENT(workqueue_queue_work, __field( void *, work ) __field( void *, function) __string( workqueue, pwq->wq->name) - __field( unsigned int, req_cpu ) - __field( unsigned int, cpu ) + __field( int, req_cpu ) + __field( int, cpu ) ), TP_fast_assign( @@ -43,7 +43,7 @@ TRACE_EVENT(workqueue_queue_work, __entry->cpu = pwq->pool->cpu; ), - TP_printk("work struct=%p function=%ps workqueue=%s req_cpu=%u cpu=%u", + TP_printk("work struct=%p function=%ps workqueue=%s req_cpu=%d cpu=%d", __entry->work, __entry->function, __get_str(workqueue), __entry->req_cpu, __entry->cpu) ); -- cgit v1.3-8-gc7d7 From 540a92bfe6dab7310b9df2e488ba247d784d0163 Mon Sep 17 00:00:00 2001 From: Edward Wu Date: Fri, 17 Jun 2022 11:32:20 +0800 Subject: ata: libata: add qc->flags in ata_qc_complete_template tracepoint Add flags value to check the result of ata completion Fixes: 255c03d15a29 ("libata: Add tracepoints") Cc: stable@vger.kernel.org Signed-off-by: Edward Wu Signed-off-by: Damien Le Moal --- include/trace/events/libata.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/trace') diff --git a/include/trace/events/libata.h b/include/trace/events/libata.h index d4e631aa976f..6025dd8ba4aa 100644 --- a/include/trace/events/libata.h +++ b/include/trace/events/libata.h @@ -288,6 +288,7 @@ DECLARE_EVENT_CLASS(ata_qc_complete_template, __entry->hob_feature = qc->result_tf.hob_feature; __entry->nsect = qc->result_tf.nsect; __entry->hob_nsect = qc->result_tf.hob_nsect; + __entry->flags = qc->flags; ), TP_printk("ata_port=%u ata_dev=%u tag=%d flags=%s status=%s " \ -- cgit v1.3-8-gc7d7 From e70b64a3f28b9f54602ae3e706b1dc1338de3df7 Mon Sep 17 00:00:00 2001 From: Dylan Yudaken Date: Thu, 23 Jun 2022 01:37:43 -0700 Subject: io_uring: move io_uring_get_opcode out of TP_printk The TP_printk macro's are not supposed to use custom code ([1]) or else tools such as perf cannot use these events. Convert the opcode string representation to use the __string wiring that the event framework provides ([2]). [1]: https://lwn.net/Articles/379903/ [2]: https://lwn.net/Articles/381064/ Fixes: 033b87d24f72 ("io_uring: use the text representation of ops in trace") Signed-off-by: Dylan Yudaken Link: https://lore.kernel.org/r/20220623083743.2648321-1-dylany@fb.com [axboe: fixup spurious removal of sq_thread assignment] Signed-off-by: Jens Axboe --- include/trace/events/io_uring.h | 42 ++++++++++++++++++++++++++++++++++------- 1 file changed, 35 insertions(+), 7 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/io_uring.h b/include/trace/events/io_uring.h index 66fcc5a1a5b1..aa2f951b07cd 100644 --- a/include/trace/events/io_uring.h +++ b/include/trace/events/io_uring.h @@ -158,6 +158,8 @@ TRACE_EVENT(io_uring_queue_async_work, __field( unsigned int, flags ) __field( struct io_wq_work *, work ) __field( int, rw ) + + __string( op_str, io_uring_get_opcode(opcode) ) ), TP_fast_assign( @@ -168,11 +170,13 @@ TRACE_EVENT(io_uring_queue_async_work, __entry->opcode = opcode; __entry->work = work; __entry->rw = rw; + + __assign_str(op_str, io_uring_get_opcode(opcode)); ), TP_printk("ring %p, request %p, user_data 0x%llx, opcode %s, flags 0x%x, %s queue, work %p", __entry->ctx, __entry->req, __entry->user_data, - io_uring_get_opcode(__entry->opcode), + __get_str(op_str), __entry->flags, __entry->rw ? "hashed" : "normal", __entry->work) ); @@ -198,6 +202,8 @@ TRACE_EVENT(io_uring_defer, __field( void *, req ) __field( unsigned long long, data ) __field( u8, opcode ) + + __string( op_str, io_uring_get_opcode(opcode) ) ), TP_fast_assign( @@ -205,11 +211,13 @@ TRACE_EVENT(io_uring_defer, __entry->req = req; __entry->data = user_data; __entry->opcode = opcode; + + __assign_str(op_str, io_uring_get_opcode(opcode)); ), TP_printk("ring %p, request %p, user_data 0x%llx, opcode %s", __entry->ctx, __entry->req, __entry->data, - io_uring_get_opcode(__entry->opcode)) + __get_str(op_str)) ); /** @@ -298,6 +306,8 @@ TRACE_EVENT(io_uring_fail_link, __field( unsigned long long, user_data ) __field( u8, opcode ) __field( void *, link ) + + __string( op_str, io_uring_get_opcode(opcode) ) ), TP_fast_assign( @@ -306,11 +316,13 @@ TRACE_EVENT(io_uring_fail_link, __entry->user_data = user_data; __entry->opcode = opcode; __entry->link = link; + + __assign_str(op_str, io_uring_get_opcode(opcode)); ), TP_printk("ring %p, request %p, user_data 0x%llx, opcode %s, link %p", __entry->ctx, __entry->req, __entry->user_data, - io_uring_get_opcode(__entry->opcode), __entry->link) + __get_str(op_str), __entry->link) ); /** @@ -390,6 +402,8 @@ TRACE_EVENT(io_uring_submit_sqe, __field( u32, flags ) __field( bool, force_nonblock ) __field( bool, sq_thread ) + + __string( op_str, io_uring_get_opcode(opcode) ) ), TP_fast_assign( @@ -400,11 +414,13 @@ TRACE_EVENT(io_uring_submit_sqe, __entry->flags = flags; __entry->force_nonblock = force_nonblock; __entry->sq_thread = sq_thread; + + __assign_str(op_str, io_uring_get_opcode(opcode)); ), TP_printk("ring %p, req %p, user_data 0x%llx, opcode %s, flags 0x%x, " "non block %d, sq_thread %d", __entry->ctx, __entry->req, - __entry->user_data, io_uring_get_opcode(__entry->opcode), + __entry->user_data, __get_str(op_str), __entry->flags, __entry->force_nonblock, __entry->sq_thread) ); @@ -435,6 +451,8 @@ TRACE_EVENT(io_uring_poll_arm, __field( u8, opcode ) __field( int, mask ) __field( int, events ) + + __string( op_str, io_uring_get_opcode(opcode) ) ), TP_fast_assign( @@ -444,11 +462,13 @@ TRACE_EVENT(io_uring_poll_arm, __entry->opcode = opcode; __entry->mask = mask; __entry->events = events; + + __assign_str(op_str, io_uring_get_opcode(opcode)); ), TP_printk("ring %p, req %p, user_data 0x%llx, opcode %s, mask 0x%x, events 0x%x", __entry->ctx, __entry->req, __entry->user_data, - io_uring_get_opcode(__entry->opcode), + __get_str(op_str), __entry->mask, __entry->events) ); @@ -474,6 +494,8 @@ TRACE_EVENT(io_uring_task_add, __field( unsigned long long, user_data ) __field( u8, opcode ) __field( int, mask ) + + __string( op_str, io_uring_get_opcode(opcode) ) ), TP_fast_assign( @@ -482,11 +504,13 @@ TRACE_EVENT(io_uring_task_add, __entry->user_data = user_data; __entry->opcode = opcode; __entry->mask = mask; + + __assign_str(op_str, io_uring_get_opcode(opcode)); ), TP_printk("ring %p, req %p, user_data 0x%llx, opcode %s, mask %x", __entry->ctx, __entry->req, __entry->user_data, - io_uring_get_opcode(__entry->opcode), + __get_str(op_str), __entry->mask) ); @@ -523,6 +547,8 @@ TRACE_EVENT(io_uring_req_failed, __field( u64, pad1 ) __field( u64, addr3 ) __field( int, error ) + + __string( op_str, io_uring_get_opcode(sqe->opcode) ) ), TP_fast_assign( @@ -542,6 +568,8 @@ TRACE_EVENT(io_uring_req_failed, __entry->pad1 = sqe->__pad2[0]; __entry->addr3 = sqe->addr3; __entry->error = error; + + __assign_str(op_str, io_uring_get_opcode(sqe->opcode)); ), TP_printk("ring %p, req %p, user_data 0x%llx, " @@ -550,7 +578,7 @@ TRACE_EVENT(io_uring_req_failed, "personality=%d, file_index=%d, pad=0x%llx, addr3=%llx, " "error=%d", __entry->ctx, __entry->req, __entry->user_data, - io_uring_get_opcode(__entry->opcode), + __get_str(op_str), __entry->flags, __entry->ioprio, (unsigned long long)__entry->off, (unsigned long long) __entry->addr, __entry->len, -- cgit v1.3-8-gc7d7