I have one application that will cause eclipse itself to crash when starting via eclipse. If I export this application as a jar and run it, it works great. But if I started (or debugged) this application seems to start (according to ps) and start; but eclipse itself will hang and report as a "stopped" program without the use of a processor or memory. I put breakpoints on the first line of this application, and it didn't even get to the point where the eclipse would end. If I wake up an eclipse from a stopped state, it will work; but he will also lose touch with the program that I started. My program, which I want to debug, will continue to work, but eclipse will not be able to control or kill it after the resumption of the eclipse.
I can run many other applications without problems from eclipse. Oddly enough, I used to have this problem, then I could run the application for a day, and now I return to the original problem. I do not know what has changed between those that matter.
Can anyone suggest what might cause this and how to restore it?
UPDATE:
I did some more linux magic. It seems that the eclipse has stopped waiting for the command: sh -c stty -lcanon min 1
It also seems that before that there was a sh command (nonexistent) that also hung without being received for several minutes, which, I think, made eclipse work correctly; sh (nonexistent) finally leaves if I wait long enough; but then the sh command that I just contacted appears. I do not know what the original non-existent SH command was; I can't make ps fast enough to catch him until he disappears. Both problems occur only with eclipse; As a jar file, this program works fine.
My guess is that eclipse is not getting or handling sigchild correctly? which would at least explain the sh (non-existent) application. It does not explain the current SH command, which does not appear as non-existent; even though it is something that needs to be completed in seconds?
UPDATE 2.0:
I found this link: http://linux.about.com/od/srl_howto/a/hwtsrl13t04_3.htm it is mostly known that stty hangs when it uses the <syntax / dev / tty; therefore, the syntax is deprecated and replaced with the new syntax. I am sure this is a problem. Unfortunately, I cannot figure out which library the obsolete command is used in. I think it all started with building ConsoleReader; but who knows what code actually executed the command that freezes? also, it seems that if it were that someone working with the consoleReader from eclipse in a linux environment would have the same problem; which, I think, is safe to assume is not the case, or it will be documented throughout the network; so maybe my understanding is still not working?