child reaping

child reaping

[James Cloos]

Have there been any changes in the last three or four days which would
affect how emacs reaps its zombie kids?

Nothing jumped out at me from a look at the bzr log.

On a non-gtk build from a few days ago I get un-reaped zombies for any
process created by (call-process-region).

The only thing I found in the bzr log which looked related is rev
110784, but I'm working on a non-Emacs bug right now and do not want
to get too distracted.  (I thought the zombies were from that other
bug, but discovered that they were all Emacs' kids.)

-JimC
-- 
James Cloos <cloos@jhcloos.com>         OpenPGP: 1024D/ED7DAEA6

Re: child reaping

[Paul Eggert]

On 11/16/2012 05:08 PM, James Cloos wrote:
> Have there been any changes in the last three or four days which would
> affect how emacs reaps its zombie kids?

Not in the trunk, no.

> On a non-gtk build from a few days ago I get un-reaped zombies for any
> process created by (call-process-region).

I could not reproduce that bug on my platform (Fedora 17, GCC 4.7.2,
configured --with-x-toolkit=lucid).  I did a call-process-region on
a small buffer a few times, both with the minimal number of arguments
and with BUFFER being t, and never created zombies.

> The only thing I found in the bzr log which looked related is rev
> 110784

That sounds like a plausible rev, but it was installed into the
trunk two weeks ago.  Did you not do a build, or not use Emacs, for
a week or two?

If you're still having the problem, might be time for a bug report.