Messages in this thread | | | From | Jason Xing <> | Date | Tue, 27 Feb 2024 14:38:38 +0800 | Subject | Re: [PATCH] net/ipv4: add tracepoint for icmp_send |
| |
On Tue, Feb 27, 2024 at 1:49 PM Eric Dumazet <edumazet@google.com> wrote: > > On Tue, Feb 27, 2024 at 3:50 AM <xu.xin16@zte.com.cn> wrote: > > > > From: xu xin <xu.xin16@zte.com.cn> > > > > Introduce a tracepoint for icmp_send, which can help users to get more > > detail information conveniently when icmp abnormal events happen. > > > > 1. Giving an usecase example: > > ============================= > > When an application experiences packet loss due to an unreachable UDP > > destination port, the kernel will send an exception message through the > > icmp_send function. By adding a trace point for icmp_send, developers or > > system administrators can obtain the detailed information easily about the > > UDP packet loss, including the type, code, source address, destination > > address, source port, and destination port. This facilitates the > > trouble-shooting of packet loss issues especially for those complicated > > network-service applications. > > > > 2. Operation Instructions: > > ========================== > > Switch to the tracing directory. > > cd /sys/kernel/debug/tracing > > Filter for destination port unreachable. > > echo "type==3 && code==3" > events/icmp/icmp_send/filter > > Enable trace event. > > echo 1 > events/icmp/icmp_send/enable > > > > 3. Result View: > > ================ > > udp_client_erro-11370 [002] ...s.12 124.728002: icmp_send: > > icmp_send: type=3, code=3.From 127.0.0.1:41895 to 127.0.0.1:6666 ulen=23 > > skbaddr=00000000589b167a > > > > Signed-off-by: He Peilin <he.peilin@zte.com.cn> > > Reviewed-by: xu xin <xu.xin16@zte.com.cn> > > Reviewed-by: Yunkai Zhang <zhang.yunkai@zte.com.cn> > > Cc: Yang Yang <yang.yang29@zte.com.cn> > > Cc: Liu Chun <liu.chun2@zte.com.cn> > > Cc: Xuexin Jiang <jiang.xuexin@zte.com.cn> > > --- > > include/trace/events/icmp.h | 57 +++++++++++++++++++++++++++++++++++++++++++++ > > net/ipv4/icmp.c | 4 ++++ > > 2 files changed, 61 insertions(+) > > create mode 100644 include/trace/events/icmp.h > > > > diff --git a/include/trace/events/icmp.h b/include/trace/events/icmp.h > > new file mode 100644 > > index 000000000000..3d9af5769bc3 > > --- /dev/null > > +++ b/include/trace/events/icmp.h > > @@ -0,0 +1,57 @@ > > +/* SPDX-License-Identifier: GPL-2.0 */ > > +#undef TRACE_SYSTEM > > +#define TRACE_SYSTEM icmp > > + > > +#if !defined(_TRACE_ICMP_H) || defined(TRACE_HEADER_MULTI_READ) > > +#define _TRACE_ICMP_H > > + > > +#include <linux/icmp.h> > > +#include <linux/tracepoint.h> > > + > > +TRACE_EVENT(icmp_send, > > + > > + TP_PROTO(const struct sk_buff *skb, int type, int code), > > + > > + TP_ARGS(skb, type, code), > > + > > + TP_STRUCT__entry( > > + __field(__u16, sport) > > + __field(__u16, dport) > > + __field(unsigned short, ulen) > > + __field(const void *, skbaddr) > > + __field(int, type) > > + __field(int, code) > > + __array(__u8, saddr, 4) > > + __array(__u8, daddr, 4) > > + ), > > + > > + TP_fast_assign( > > + // Get UDP header > > + struct udphdr *uh = udp_hdr(skb); > > + struct iphdr *iph = ip_hdr(skb); > > + __be32 *p32; > > + > > + __entry->sport = ntohs(uh->source); > > + __entry->dport = ntohs(uh->dest); > > + __entry->ulen = ntohs(uh->len); > > + __entry->skbaddr = skb; > > + __entry->type = type; > > + __entry->code = code; > > + > > + p32 = (__be32 *) __entry->saddr; > > + *p32 = iph->saddr; > > + > > + p32 = (__be32 *) __entry->daddr; > > + *p32 = iph->daddr; > > + ), > > + > > FYI, ICMP can be generated for many other protocols than UDP. > > > + TP_printk("icmp_send: type=%d, code=%d. From %pI4:%u to %pI4:%u ulen=%d skbaddr=%p", > > + __entry->type, __entry->code, > > + __entry->saddr, __entry->sport, __entry->daddr, > > + __entry->dport, __entry->ulen, __entry->skbaddr) > > +); > > + > > +#endif /* _TRACE_ICMP_H */ > > + > > +/* This part must be outside protection */ > > +#include <trace/define_trace.h> > > diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c > > index e63a3bf99617..437bdb7e2650 100644 > > --- a/net/ipv4/icmp.c > > +++ b/net/ipv4/icmp.c > > @@ -92,6 +92,8 @@ > > #include <net/inet_common.h> > > #include <net/ip_fib.h> > > #include <net/l3mdev.h> > > +#define CREATE_TRACE_POINTS > > +#include <trace/events/icmp.h> > > > > /* > > * Build xmit assembly blocks > > @@ -599,6 +601,8 @@ void __icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info, > > struct net *net; > > struct sock *sk; > > > > + trace_icmp_send(skb_in, type, code); > > I think you missed many sanity checks between lines 622 and 676 [...] > > Honestly, a kprobe BPF based solution would be less risky, and less > maintenance for us. >
I agreed. I wonder if we can remove some trace_* at the very beginning of its caller function since they can be easily replaced with bpf tools and then we make less effort to maintain them, say, trace_tcp_probe(), trace_tcp_rcv_space_adjust, etc.
Thanks, Jason
| |