Re: [Qemu-devel] [PATCH] slirp: disable Nagle in outgoing connections

[Philippe Mathieu-Daudé]

On 03/07/2018 06:13 AM, Kamil Rytarowski wrote:
> This patch is correct. LLDB expects 1sec for reply, GDB by default 2.
> 
> Debuggers use this option to disable Nagle algorithm in order to quickly
> transfer messages between gdb-server and gdb-client. It's also fairy
> portable across systems.
> 
> On 04.01.2018 18:56, Andreas Gustafsson wrote:
>> slirp: disable Nagle in outgoing connections
>>
>> When setting up an outgoing user mode networking TCP connection,
>> disable the Nagle algorithm in the host-side connection.  Either the
>> guest is already doing Nagle, in which case there is no point in doing
>> it twice, or it has chosen to disable it, in which case we should
>> respect that choice.
>>
>> This change speeds up GDB remote debugging over TCP over user mode
>> networking (with GDB runing on the guest) by multiple orders of
>> magnitude, and has been part of the local patches applied by pkgsrc
>> since 2012 with no reported ill effects.
>>
>> Signed-off-by: Andreas Gustafsson <address@hidden>
> 
> Signed-off-by: Kamil Rytarowski <address@hidden>

I suppose you meant "Reviewed-by: Kamil Rytarowski <address@hidden>"

Reviewed-by: Philippe Mathieu-Daudé <address@hidden>

> 
>> ---
>>  slirp/tcp_subr.c | 2 ++
>>  1 file changed, 2 insertions(+)
>>
>> diff --git a/slirp/tcp_subr.c b/slirp/tcp_subr.c
>> index da0d53743f..8d0f94b75f 100644
>> --- a/slirp/tcp_subr.c
>> +++ b/slirp/tcp_subr.c
>> @@ -416,6 +416,8 @@ int tcp_fconnect(struct socket *so, unsigned short af)
>>      socket_set_fast_reuse(s);
>>      opt = 1;
>>      qemu_setsockopt(s, SOL_SOCKET, SO_OOBINLINE, &opt, sizeof(opt));
>> +    opt = 1;
>> +    qemu_setsockopt(s, IPPROTO_TCP, TCP_NODELAY, &opt, sizeof(opt));
>>  
>>      addr = so->fhost.ss;
>>      DEBUG_CALL(" connect()ing")
>>
> 
>

signature.asc
Description: OpenPGP digital signature