[http://sourceforge.net/p/ufoai/bugs/3240 Item 3240] imported from sourceforge.net tracker on 2013-01-28 20:05:19

Compiled myself, today 2012-04-17. git branch
* ufoai_2.4

git log shows
commit 2d6933be41a250efe957621be8845faa8ea5802a
Author: mcr2010 <mcr2010@users.sourceforge.net>
Date: Mon Apr 16 19:40:07 2012 +0200

* Aliencontainment: Fixed bar sizes

Debian GNU/Linux 6.0.4

Mission is "A mansion near the coast". Entering it makes display freeze and ufo takes all CPU. Have to kill process to recover.

sysman@alice:/opt/Varasto/games/ufoai$ cat /opt/home/sysman/.ufoai/2.4-dev/base/ufoconsole.log
2012/04/17 15:37:44 FS_RemoveFile: remove /opt/home/sysman/.ufoai/2.4-dev/base/keys.cfg
2012/04/17 15:37:44 There are still 1 opened files
sysman@alice:/opt/Varasto/games/ufoai$

I can provide the save file, it is just prior to entering the mission. Looks like game freezees every time, tried it 3 times. Other missions with other maps have worked.
===== Comments Ported from Sourceforge =====

====== *anonymous (2012-04-17 13:07:05) ======

Did kill -KILL to the process, got this crashdump file.
====== *anonymous (2012-04-17 13:19:20) ======

Save file at http://taleman.fi/slot2.savx
====== *anonymous (2012-04-17 13:32:36) ======

Seems I pasted old ufoconsole.log. This file is from right after kill -TERM the process.

http://paste.debian.net/163521/
====== mcr2010 (2012-04-17 17:50:53) ======

+mansion should work, but the RMA assembly takes a long time for that map.
Please try to wait a bit longer for it to assemble.
====== shipit (2012-04-18 04:58:28) ======

I played the map several times when testing 2.4, it takes somewhat like two minutes to start up but should work. The problem (beside the map being very large) is that rerouting takes place before the loading screen. So the screens gets black and it looks like nothing happens while the cpu is busy with the pathfinding. Duke and Sandro were talking about a shortcut for this, but not sure yet if it is possible.
====== *anonymous (2012-04-18 07:25:31) ======

Shipit, you are right. If I just wait long enough, the map does load. Othe maps have loaded in less than 5 seconds, so I was just too hasty.

Even this map has loaded previously. Perhaps some random configurations of the map take much longer to load than others? Let us see if I can finish this map with 2 Gbytes of memory. Looks like all of memory is in use now.