Hi,
flightgear freezes and uses 100% CPU time as soon as i click 'Activate' in the Route Manager GUI. The flightplan is shown in the attached screenshot, together with the ILS28 approach (relevant XML snippet only). The a/c loaded is the Citation X+ from Patten's Hangar.
FG is 2020.4 compiled 26/06/2021.
Last lines in the fgfs.log:
/home/sfr/flightgear/share/fgdata/Nasal/route_manager.nas:36: saw departure changed
68.48 [INFO]:nasal /home/sfr/flightgear/share/fgdata/Nasal/route_manager.nas:126: saw active flightplan cleared, deactivating
79.55 [INFO]:general /home/sfr/src/fgfs/flightgear/src/Airports/airport.cxx:591: EDTY: loading procedures from Path "/home/sfr/fgfs_addons/FMS/Airports/E/D/T/EDTY.procedures.xml"
186.37 [INFO]:nasal /home/sfr/flightgear/share/fgdata/Nasal/route_manager.nas:207: flightplan activated, default GPS to LEG mode
186.37 [INFO]:autopilot /home/sfr/src/fgfs/flightgear/src/Autopilot/route_mgr.cxx:866: route manager, current-wp is now 0
I've also attached the gdb log (taken after killing the fgfs process).
gdb cmd line used: gdb -q -n -batch -ex 'handle SIGTSTP nostop' -ex 'handle SIGCONT nostop' -ex 'handle SIGHUP nostop nopass' -ex 'handle SIGPIPE nostop' -ex r -ex 'thread apply all bt' -ex where --args ./flightgear_next/bin/fgfs --launcher
Let me know if any further information is needed.
Thanks Stefan, will take a look. Are you able to test the same issue with 2020.3.9?
Hi James,
yes sure. But which branch should I use? release/2020.3.9 does not exist
and release/2020.3 has version set to 2020.3.7. Has 'version' just not been
updated?
sfr@asterix:~/src/fgfs/flightgear$ git checkout release/2020.3
On Tue, Jul 20, 2021 at 9:45 AM James Turner jmturner@users.sourceforge.net
wrote:
Related
Tickets:
#2616of course 2 minutes after hitting 'Send' i discovered git tags. :)
compiling 2020.3.9 now.
On Tue, Jul 20, 2021 at 10:09 AM Stefan Frank sfrsfr@gmail.com wrote:
Related
Tickets:
#2616You can alos use the binary AppImage for 2020.3.9 to test, but indeed the git tags should be the same.
with 2020.3.9 it's the same behaviour as in next.
Okay great, that means it's just some werid edge case and not a regression.
Hi Stefan, just testing this on next and can't reproduce : does it still happen for you? I'm using a few-months-old Navigraph prcoedures for EDTY, but the ILS28 approach points looks the same as yours. I also used a different starting airport, but this shouldn't matter unless there is something very precise happening.
I also created some unit-tests for this case but again no hang. So, some more testing would help.
i tried it with next (compiled 1-2 days ago)
'/sim/version/flightgear: 2020.4.0
/sim/version/revision: 985abe8d431e7b72a7f69679a7cac5c52aa3fab4
FG hangs after clicking 'Activate' in the Route-Manager.
In addition to the procedures i have also Navigrpah navigation data in a custom-scenery folder
└── NavData
├── awy
│ └── awy.dat.gz
├── fix
│ └── fix.dat.gz
└── nav
└── nav.dat.gz
Uh, What's an addon-metadata file ?
I tried to replicate the setup by adding ~/fgfs-addons/FMS/Airports/E/D/T/ and inserting the file EDTY-procedures.xml and, after adding --addon=/home/xxxx/fgfs_addons/ and starting I get a
message:
FATAL: Unable to register an add-on.
( Error registering an add-on: unable to find add-on metadata file '/home/otto/fgfs_addons/addon-metadata.xml')
Anyway, putting that prodeures file in the normal scenery sub-folder I can load the RM to match the screenshot; clicking 'Activate' doesn't cause a hang but 'Clear' will not clear the RM list. After removing the RM waypoint one by one using 'Remove' it is not possible to re-load the list .
you need to define the FMS folder as a custom scenery, not as an add-on.
from what you wrote it seems putting the procedures file in the 'normal scenery sub-folder' worked for you.
OP Is there any chance you can see what happens with a (simple ) model from fgaddon ? Patten's hangar notes updates in this area within the last couple of days and from what I can see, seacging on 'patten citation x' there's been discussion , in french, about Route Manager issues ( I think ) Thanks
i now tried the C182s
(https://github.com/HHS81/c182s.git
commit eb6aa728708733ed6703ac26a45c58386c2a3e83 from August 5th)
and everything's fine. Activating the route does just that with FG not hanging.
Apologies for wasting everyones time. I'm really sorry :-(
Older versions of the Citation X always worked fine for me.
FWIW: when running into problems in FG, the first thing i try is doing the same thing in the UFO... at least as far as basic stuff goes... loading a route in the default route manager, checking scenery loading, etc... i was going to try to test this in the UFO but since it now appears to be limited to patton's citation x, i'll hold off on that...
Marking as WontFixs, but no need to apologise Stefan, these things happen and it's always better to ask.