On Mon, Dec 02, 2013 at 01:47:09PM +0800, George M. Florendo wrote: >On 11/29/13, Christopher Faylor wrote: >> On Thu, Nov 28, 2013 at 10:55:04AM -0500, Christopher Faylor wrote: >>>On Thu, Nov 28, 2013 at 02:08:13PM +0100, Corinna Vinschen wrote: >>>>On Nov 28 15:20, George M. Florendo wrote: >>>>> Hi, >>>>> >>>>> I run a non cygwin apache process named httpd.exe. Checking it with >>>>> "ps -W" shows a listing of the same process (with PID 4560 in this >>>>> case) twice. >>>>> >>>>> $ ps -W | grep httpd >>>>> 4560 1604 4560 5304 pty0 1000 09:55:21 >>>>> /home/georgeflorendo/wamp/bin/apache/Apache2.4.4/bin/httpd >>>>> 4560 1604 4560 5304 pty0 1000 09:55:21 >>>>> /home/georgeflorendo/wamp/bin/apache/Apache2.4.4/bin/httpd >>>>> 7508 0 0 7508 ? 0 15:09:50 >>>>> C:\cygwin\home\georgeflorendo\wamp\bin\apache\Apache2.4.4\bin\httpd.exe >>>>> >>>>> Doing a "ps aux" shows it only once. >>>>> >>>>> $ ps aux >>>>> PID PPID PGID WINPID TTY UID STIME COMMAND >>>>> 1340 7856 1340 5720 pty5 1000 15:13:50 /usr/bin/ps >>>>> 1604 4008 1604 2292 pty0 1000 09:17:01 >>>>> /usr/bin/bash >>>>> 4008 1 4008 4008 ? 1000 09:17:00 >>>>> /usr/bin/mintty >>>>> 4560 1604 4560 5304 pty0 1000 09:55:21 >>>>> /home/georgeflorendo/wamp/bin/apache/Apache2.4.4/bin/httpd >>>>> >>>>> <snipped> >>>>> >>>>> Is it correct that ps -W shows the same process twice? >>>> >>>>Not exactly but it could happen. Can you try the latest snapshot Cygwin >>>>DLL from http://cygwin.com/snapshots/ and see if it fixes the problem? >>> >>>I don't see anything in the ChangeLog which would affect this and I'd >>>be concerned if there was. >>> >>>There are scenarios where two different windows pids can be associated >>>with the same Cygwin pid. You should only see it when you do a "ps -W". >>> >>>So, AFAICT, the correct answer to the "Is it correct" is "Yes'. >> >> Oops. Sorry. Just noticed that the pids weren't "different". There >> are still pathological situations where a pid can show up twice when >> doing a "ps -W". I've seen the issue before but fixing it would involve >> a global process lock which would slow down Cygwin for the benefit of >> a more accurate "ps -W". > >Oh, I'm sorry too. Didn't get to read your first sentence earlier. >I'd rather have a faster cygwin more than a more accurate "ps -W". > >Thanks for all your help.
Thanks for understanding. I will try to revisit this code sometime in the near future to see if I can do something about this case. I had a glimmer of an idea about how to fix this in the thinking room recently. 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
