Gregg, I reached the same conclusion Wednesday when I stumbled on a dirty mod_fcgid checkout here.
I'll review your comments and then determine how to proceed (commit the feature which could be useful, port the feature to apr, or concur it isn't needed.) Because of the way the Windows Service Control manager works, it may still have value. On 3/2/2012 4:34 PM, Gregg Smith wrote: > Hi, > > First, http://svn.apache.org/viewvc?view=revision&revision=1236319 > > The patch in PR 51078 was never commited. However, I am -1 on committing this > right now > because I believe the root cause of this orphan described in PR 51078 is what > has been > fixed now by the commit of PR 50309. Without the PR 50309 fix I have run into > this orphan > after stopping httpd and having to go into processes and kill it manually. > It's the > problem with graceful restarts, the fcgid process never got killed and it > would stop the > httpd parent because it was holding onto the port and logs. PR 50309 most > likely will fix > PR 48949 as well, since Mario describes the same issue on Ubuntu and the same > patch worked > for him on it. > > > Second, http://svn.apache.org/viewvc?view=revision&revision=1234178 > > http://svn.apache.org/viewvc?view=revision&revision=1213454 > The patch applied is from PR 51020, so you have killed 2 PRs with the one > stone, 50120 & > 51520 > > The latter's a nit pick I know :) > > Gregg >
