[Bloat] bufferbloat.net server having troubles?

[Eric Dumazet]

On 03/31/2018 12:54 PM, Jonathan Morton wrote:
>> On 31 Mar, 2018, at 10:50 pm, Toke Høiland-Jørgensen <toke at toke.dk> wrote:
>>
>> Yeah, the box running the web server is having some issues with NULL
>> pointer dereferences in tcp_push() in the kernel crashing processes
>> running TCP. Haven't been able to figure out why :/
> 
> Maybe build/install a new kernel and reboot?
> 
> Possibility exists of hardware failure, too.  Less likely, perhaps, but if you don't have ECC...
> 

Nope, known bug on stable kernels.

Please upgrade or downgrade.

$ git log --oneline v4.14.31..v4.14.32  -- net/ipv4/tcp.c
e44c1733059c tcp: purge write queue upon aborting the connection

commit e44c1733059c69868e81f82eb09fcb6bbc492050
Author: Soheil Hassas Yeganeh <soheil at google.com>
Date:   Tue Mar 6 17:15:12 2018 -0500

    tcp: purge write queue upon aborting the connection
    
    
    [ Upstream commit e05836ac07c77dd90377f8c8140bce2a44af5fe7 ]
    
    When the connection is aborted, there is no point in
    keeping the packets on the write queue until the connection
    is closed.
    
    Similar to a27fd7a8ed38 ('tcp: purge write queue upon RST'),
    this is essential for a correct MSG_ZEROCOPY implementation,
    because userspace cannot call close(fd) before receiving
    zerocopy signals even when the connection is aborted.
    
    Fixes: f214f915e7db ("tcp: enable MSG_ZEROCOPY")
    Signed-off-by: Soheil Hassas Yeganeh <soheil at google.com>
    Signed-off-by: Neal Cardwell <ncardwell at google.com>
    Reviewed-by: Eric Dumazet <edumazet at google.com>
    Signed-off-by: Yuchung Cheng <ycheng at google.com>
    Signed-off-by: David S. Miller <davem at davemloft.net>
    Signed-off-by: Greg Kroah-Hartman <gregkh at linuxfoundation.org>