[PATCH net-next 2/2] tcp: handle TCP_TIME_WAIT/TCP_NEW_SYN_RECV in tcp_set_state tracepoint

2017-11-09 Thread Yafang Shao
When TCP connetion in TCP_TIME_WAIT or TCP_NEW_SYN_RECV state, it can't get the sport/dport/saddr/daddr from inet_sock. trace_tcp_set_state may be called when the oldstate in these two states. Signed-off-by: Yafang Shao --- include/trace/events/tcp.h | 33 ++--- 1 fi

Re: [PATCH net-next 2/2] tcp: handle TCP_TIME_WAIT/TCP_NEW_SYN_RECV in tcp_set_state tracepoint

2017-11-09 Thread Eric Dumazet
On Thu, 2017-11-09 at 14:26 +, Yafang Shao wrote: > When TCP connetion in TCP_TIME_WAIT or TCP_NEW_SYN_RECV state, it can't > get the sport/dport/saddr/daddr from inet_sock. > > trace_tcp_set_state may be called when the oldstate in these two states. > > Signed-off-by: Yafang Shao > --- > i

Re: [PATCH net-next 2/2] tcp: handle TCP_TIME_WAIT/TCP_NEW_SYN_RECV in tcp_set_state tracepoint

2017-11-09 Thread Eric Dumazet
On Thu, 2017-11-09 at 06:52 -0800, Eric Dumazet wrote: > Wow. > > > Since all three variants of sockets (full sockets, request sockets, > timewait sockets) are all hashed into ehash table these days, they all > have the fields at the same offset > > For IPv4, that would be : > > __sk_common.sk

Re: [PATCH net-next 2/2] tcp: handle TCP_TIME_WAIT/TCP_NEW_SYN_RECV in tcp_set_state tracepoint

2017-11-09 Thread Yafang Shao
2017-11-09 22:58 GMT+08:00 Eric Dumazet : > On Thu, 2017-11-09 at 06:52 -0800, Eric Dumazet wrote: > >> Wow. >> >> >> Since all three variants of sockets (full sockets, request sockets, >> timewait sockets) are all hashed into ehash table these days, they all >> have the fields at the same offset >

Re: [PATCH net-next 2/2] tcp: handle TCP_TIME_WAIT/TCP_NEW_SYN_RECV in tcp_set_state tracepoint

2017-11-09 Thread Eric Dumazet
On Thu, 2017-11-09 at 23:11 +0800, Yafang Shao wrote: > I'm also very sad that I'm still using IPv4 in 2017 : ( > > Okay then another issue, > shoule we reduce the complexity in the function tcp4_seq_show() ? This is irrelevant really. I do not see how tcp4_seq_show() could avoid testing sk_st

Re: [PATCH net-next 2/2] tcp: handle TCP_TIME_WAIT/TCP_NEW_SYN_RECV in tcp_set_state tracepoint

2017-11-12 Thread kbuild test robot
Hi Yafang, Thank you for the patch! Yet something to improve: [auto build test ERROR on net-next/master] url: https://github.com/0day-ci/linux/commits/Yafang-Shao/net-tcp-track-all-TCP-IP-state-transition-in-tcp_set_state/20171112-203643 config: i386-randconfig-i1-201746 (attached as .config

Re: [PATCH net-next 2/2] tcp: handle TCP_TIME_WAIT/TCP_NEW_SYN_RECV in tcp_set_state tracepoint

2017-11-12 Thread kbuild test robot
Hi Yafang, Thank you for the patch! Yet something to improve: [auto build test ERROR on net-next/master] url: https://github.com/0day-ci/linux/commits/Yafang-Shao/net-tcp-track-all-TCP-IP-state-transition-in-tcp_set_state/20171112-203643 config: mips-malta_kvm_defconfig (attached as .config)