This is the mail archive of the
mailing list for the Cygwin project.
Re: Fork issue with timerfd
On Wed, 2019-03-06 at 22:23 +0100, Corinna Vinschen wrote:
> On Mar 6 21:54, Peter Kozich (UM) wrote:
> > In response to this one:
> > Re: Fork issue with timerfd
> > From: Achim Gratz <Stromeko at nexgo dot de>
> > To: cygwin at cygwin dot com
> > Date: Tue, 05 Mar 2019 20:15:32 +0100
> > Subject: Re: Fork issue with timerfd
> > ....
> > Anyway, the latest snapshot seemingly resolved that issue as
> > well, I can no longer reproduce it.
> > Unfortunately, only seemingly. Factually, the issue persists. And I
> > don't have to do anything with emacs just shoot up, wait cca. 20
> > seconds and it inevitably crashes, and it does so in the
> > foreground,
> > too:
> > peter@D11934N ~
> > $ emacs .bash_history &
> >  489
> >  Done emacs .bash_history
> > peter@D11934N ~
> > $ 1 [main] emacs 489 C:\CygWin\bin\emacs-w32.exe: *** fatal
> > error -
> > CreateThread failed for pipesel - 0x0<0x0>, Win32 error 8
> I couldn't directly reproduce this issue, but I saw emacs using more
> more memory while the cursor was blinking. I found a resource leak
> posix timers which was probably the cause for this problem.
> I pushed a patch and uploaded new developer snapshots to
> Please test.
I tested this:
emacs was on for a long time, did not crash and as far as could judge
it did not leak memory.
But it froze a couple of times which I cannot recall I have ever
experienced before. I mean it suddenly became deaf and dumb and while I
was actively editing the text, both the text area and the menu line got
unresponsive. I could only shut it down from the task manager, because
the red X didn't work either.
On the next week I will try the debug version in the hope that it can
make a trace or something helpful. Is there a guide somewhere how to
Problem reports: http://cygwin.com/problems.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple