This is the mail archive of the cygwin mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [1.7] Updated: cygwin-1.7.0-65


On Tue, Nov 24, 2009 at 10:25:12AM +0100, Corinna Vinschen wrote:
>On Nov 23 13:20, Christopher Faylor wrote:
>> On Mon, Nov 23, 2009 at 10:07:05PM +0800, Huang Bambo wrote:
>> >There's some diffirence between cygwin and other *nix:
>> >In other *nix with this condition, those ended child process could be
>> >list by ps command with <defunc>tag, will you fix it?
>> 
>> Cygwin should produce zombie processes.  You don't see zombie processes
>> if the child has exited and the parent goes away though.  In that case
>> the process just disappears, just like on linux.
>> 
>> You *should* see a zombie when the parent is alive by not waiting for
>> children.
>
>Hmm, but you don't.  Huang's example code forks a child and the parent
>returns to the select loop without reaping the child via wait.  When the
>child exits after its 3 second sleep, the handle to the child process
>does not get closed, and the cygpid shared memory regions for the child
>pid still exists.  But the process list does not show the child process
>at all.  Nor will you find it in /proc.
>
>That's not a regression, though.  The same occurs in 1.5.25.  However,
>I have a vagure memory that we really saw zombies in the process list
>way back when.

Ok, I suspected as much.  I just wanted to squash the notion that this
wasn't possible in Windows.  It may be broken now but it is possible.

cgf

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]