Re: [PATCH net] tcp: update skb->skb_mstamp more carefully

2017-09-15 Thread Eric Dumazet
On Fri, 2017-09-15 at 14:37 -0700, David Miller wrote: > From: Eric Dumazet > Date: Wed, 13 Sep 2017 20:30:39 -0700 > > > From: Eric Dumazet > > > > liujian reported a problem in TCP_USER_TIMEOUT processing with a patch > > in tcp_probe_timer() : > > https://www.spinics.net/lists/netdev/m

Re: [PATCH net] tcp: update skb->skb_mstamp more carefully

2017-09-15 Thread David Miller
From: Eric Dumazet Date: Wed, 13 Sep 2017 20:30:39 -0700 > From: Eric Dumazet > > liujian reported a problem in TCP_USER_TIMEOUT processing with a patch > in tcp_probe_timer() : > https://www.spinics.net/lists/netdev/msg454496.html > > After investigations, the root cause of the problem

Re: [PATCH net] tcp: update skb->skb_mstamp more carefully

2017-09-14 Thread Soheil Hassas Yeganeh
On Thu, Sep 14, 2017 at 12:32 PM, Yuchung Cheng wrote: > On Thu, Sep 14, 2017 at 6:57 AM, Neal Cardwell wrote: >> On Wed, Sep 13, 2017 at 11:30 PM, Eric Dumazet >> wrote: >>> >>> From: Eric Dumazet >>> >>> liujian reported a problem in TCP_USER_TIMEOUT processing with a patch >>> in tcp_probe_

Re: [PATCH net] tcp: update skb->skb_mstamp more carefully

2017-09-14 Thread Yuchung Cheng
On Thu, Sep 14, 2017 at 6:57 AM, Neal Cardwell wrote: > On Wed, Sep 13, 2017 at 11:30 PM, Eric Dumazet wrote: >> >> From: Eric Dumazet >> >> liujian reported a problem in TCP_USER_TIMEOUT processing with a patch >> in tcp_probe_timer() : >> https://www.spinics.net/lists/netdev/msg454496.ht

Re: [PATCH net] tcp: update skb->skb_mstamp more carefully

2017-09-14 Thread Neal Cardwell
On Wed, Sep 13, 2017 at 11:30 PM, Eric Dumazet wrote: > > From: Eric Dumazet > > liujian reported a problem in TCP_USER_TIMEOUT processing with a patch > in tcp_probe_timer() : > https://www.spinics.net/lists/netdev/msg454496.html > > After investigations, the root cause of the problem is t

Re: [PATCH net] tcp: update skb->skb_mstamp more carefully

2017-09-14 Thread liujian
On 2017/9/14 11:30, Eric Dumazet worte: > From: Eric Dumazet > > liujian reported a problem in TCP_USER_TIMEOUT processing with a patch > in tcp_probe_timer() : > https://www.spinics.net/lists/netdev/msg454496.html > > After investigations, the root cause of the problem is that we update

[PATCH net] tcp: update skb->skb_mstamp more carefully

2017-09-13 Thread Eric Dumazet
From: Eric Dumazet liujian reported a problem in TCP_USER_TIMEOUT processing with a patch in tcp_probe_timer() : https://www.spinics.net/lists/netdev/msg454496.html After investigations, the root cause of the problem is that we update skb->skb_mstamp of skbs in write queue, even if the att