Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.os.linux.development.system > #460

Re: History of SO_LINGER Socket Option

Newsgroups comp.os.linux.development.system
Date 2012-09-14 02:51 -0700
References <Yx937.912$Ka2.319796@monger.newsread.com>
Message-ID <a00285b3-dc2a-4984-b49b-71ae297cfc91@googlegroups.com> (permalink)
Subject Re: History of SO_LINGER Socket Option
From lenky0401@gmail.com

Show all headers | View raw


Hi,
Does this BUG exists also in Linux kernel 2.6.x/3.x?

Lenky Gao
On Thursday, July 12, 2001 12:17:27 PM UTC+8, Phillip Soltan wrote:
> I have generally found Linux networking to be a superset of BSD networking.
> The one exception I've come across is the SO_LINGER socket option.  If
> "linger" is turned on, it is supposed to abort the TCP connection if there
> is still data left to send after the linger time has completed.  This is my
> understanding based on Richard Steven's  book "Unix Network Programming".
> Linux (RH6.2/2.2.14 kernel), on the other hand, returns from "close" and
> continues to transmit what is left in the send queue.
> 
> Does anyone know the history of the SO_LINGER socket option?  Was this a
> purposeful decision or an oversight?  Am I missing something?
> 
> Thanks,
> 
> Phillip Soltan

Back to comp.os.linux.development.system | Previous | Next | Find similar


Thread

Re: History of SO_LINGER Socket Option lenky0401@gmail.com - 2012-09-14 02:51 -0700

csiph-web