http://issues.apache.org/SpamAssassin/show_bug.cgi?id=4703
------- Additional Comments From [EMAIL PROTECTED] 2005-11-23 11:20 ------- > This patch is the one for bug 4518 adapted for 3.0. I think it better deals > with the source of the problem, avoiding leaking open file handles as well as > processes. uhm .. I cannot see a waitpid for the terminated process. Wouldn't that leave a zombie running? isn't a cleanup_kids call missing from outside the eval block? Also, if the kid is really hang, for whichever reason, the TERM would leave it there .. wouldn't it be better to add a waitpid with WNOHANG parameter and if it still is alive send a KILL signal? (an attempt to do this in my previous patch) Cheers, Carlo ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee.
