Xref: utzoo comp.unix.questions:19981 comp.unix.wizards:20636
Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uunet!ncrlnk!emdeng!btrue
From: btrue@emdeng.Dayton.NCR.COM (Barry.True)
Newsgroups: comp.unix.questions,comp.unix.wizards
Subject: Re: Zombies
Message-ID: <229@emdeng.Dayton.NCR.COM>
Date: 16 Feb 90 14:33:03 GMT
References: <2647@stl.stc.co.uk>
Reply-To: btrue@emdeng.UUCP (Barry.True)
Organization: NCR, E&M Dayton
Lines: 14

In article <2647@stl.stc.co.uk> dsr@stl.stc.co.uk (David Riches) writes:
>I have a program which spawns of a new program via the use of execl
>and communicates via pipes created using dup().
>
>When this new program dies it seems to leave a Zombied process
>with <defunct> as its name in the process table.
>
>Q:  Is the new program not dying properly?
>

No. But in order to avoid a zombie process the parent must wait on the
child. You might get around this by having a signal trap for the death of
child signal which executes an ISR which issues a wait(). When the wait()
is executed the zombie process created by the child's death will go away.