home *** CD-ROM | disk | FTP | other *** search
- From: mintz@hpindda.cup.hp.com (Ken Mintz)
- Date: Fri, 18 Dec 1992 19:48:29 GMT
- Subject: Re: HP-UX 8.07 bug causes crash (& related info on SO_LINGER)
- Message-ID: <4310330@hpindda.cup.hp.com>
- Organization: HP Information Networks, Cupertino, CA
- Path: sparky!uunet!haven.umd.edu!darwin.sura.net!tulane!spool.mu.edu!sdd.hp.com!col.hp.com!news.dtc.hp.com!hpscit.sc.hp.com!hplextra!hpfcso!hpcss01!hpindda!mintz
- Newsgroups: comp.sys.hp
- References: <1992Dec17.233412.28542@icaen.uiowa.edu>
- Lines: 28
-
- > BTW, this is unrelated, but SO_LINGER doesn't work as the man pages claim it
- > does (but then neither does the BSD networking code on which HP-UX's is based)
-
- socket.h and I believe the man pages were cleaned up for 9.0 to try to
- correctly reflect the operation of SO_LINGER. I think you summarized
- some or all of it correctly, but let me just reiterate.
-
- (BTW, SO_LINGER is defined deferently for 7.0 than for 8.0 and 9.0. The
- following is for 8.0 and 9.0.)
-
- If l_onoff is zero, close(2) returns immediately, but any remaining data
- in the send buffer will be sent before the TCP connection is closed.
-
- If l_onoff is non-zero and l_linger is zero, close(2) returns immediately,
- and any remaining data in the send buffer is flushed (not sent).
-
- If l_onoff is non-zero and l_linger is non-zero, close(2) will not return
- until all remaining data in the send buffer is sent.
-
- Note that l_linger is __not__ a measure of time, as it is not in 4.3BSD.
- 4.4BSD does implement it as a measure of time, but it measures "ticks",
- not seconds.
-
- -- Ken Mintz
-
- DISCLAIMER: This response does not represent the official position of, or
- statement by, the Hewlett-Packard Company. This response is provided for
- informational purposes only and is supplied without warranty of any kind.
-