JPScore issueshttps://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues2017-05-19T08:58:39+02:00https://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/222Compiling error2017-05-19T08:58:39+02:00Mohcine Chraibim.chraibi@fz-juelich.deCompiling errorHello,
I get the attached error when compiling on opensuse 13.1 with gcc 4.8.3.
It would be great, if you have a suggestion how to fix this.
Thank you!
Best Regards
Fabian
[log_error](/uploads/5e176a33e7fee5eb60ae2da3e8bd276b/log_erro...Hello,
I get the attached error when compiling on opensuse 13.1 with gcc 4.8.3.
It would be great, if you have a suggestion how to fix this.
Thank you!
Best Regards
Fabian
[log_error](/uploads/5e176a33e7fee5eb60ae2da3e8bd276b/log_error)
[CMakeCache.txt](/uploads/b68c1dffba1dbc9dd55ebaba290948fa/CMakeCache.txt)https://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/221Router Strategy combination2018-06-13T16:41:32+02:00Mohcine Chraibim.chraibi@fz-juelich.deRouter Strategy combinationGive a warning if case
- Router is "ff_global_shortest"
- and Strategy is NOT 8Give a warning if case
- Router is "ff_global_shortest"
- and Strategy is NOT 8https://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/217Reduce timeout threshold for Jenkins test2018-06-13T16:41:31+02:00Mohcine Chraibim.chraibi@fz-juelich.deReduce timeout threshold for Jenkins testFor build 1073, runtest_juelich_3 took 3h until it timed out. The successful run in build 1074 only took 29s. Adjusting when a test in terminated would yield sooner results and reduce the load on the Jenkins machine.For build 1073, runtest_juelich_3 took 3h until it timed out. The successful run in build 1074 only took 29s. Adjusting when a test in terminated would yield sooner results and reduce the load on the Jenkins machine.Mohcine Chraibim.chraibi@fz-juelich.deMohcine Chraibim.chraibi@fz-juelich.dehttps://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/212Corret: Jenkins BOOST path2017-05-19T08:58:40+02:00Mohcine Chraibim.chraibi@fz-juelich.deCorret: Jenkins BOOST pathMohcine Chraibim.chraibi@fz-juelich.deMohcine Chraibim.chraibi@fz-juelich.dehttps://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/211Source thread hangs2018-05-01T08:23:25+02:00Mohcine Chraibim.chraibi@fz-juelich.deSource thread hangsMake sure, that the source thread terminates before the main thread.
To reproduce this case:
- set max sim time to a small number (10)
- set `v0` to 0
- set max agents in source to something high for example 50Make sure, that the source thread terminates before the main thread.
To reproduce this case:
- set max sim time to a small number (10)
- set `v0` to 0
- set max agents in source to something high for example 50https://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/209Sources Voronoi distribution2018-06-13T16:41:32+02:00Mohcine Chraibim.chraibi@fz-juelich.deSources Voronoi distributionWhen distributing sources using the Voronoi-method `ComputeBestPositionVoronoiBoost()`
agents are not distributed in the whole room, but somehow in its left half.
![myanimation](/uploads/302e873ec77588db7537e30c82e2bd74/myanimation.gif)When distributing sources using the Voronoi-method `ComputeBestPositionVoronoiBoost()`
agents are not distributed in the whole room, but somehow in its left half.
![myanimation](/uploads/302e873ec77588db7537e30c82e2bd74/myanimation.gif)https://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/208:memo: moving parameters to ini file (from being hardcoded): floorfield2018-06-13T16:41:31+02:00Mohcine Chraibim.chraibi@fz-juelich.de:memo: moving parameters to ini file (from being hardcoded): floorfield### This issue is a log of the refactoring: hardcoded parameters move to ini file### This issue is a log of the refactoring: hardcoded parameters move to ini fileArne GrafArne Grafhttps://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/207Segmentation faults with sources2018-06-13T16:41:32+02:00Mohcine Chraibim.chraibi@fz-juelich.deSegmentation faults with sources- Using sources lead to Segmentation faults
**Reproduction Steps:**
```
./bin/jpscore inputfiles/voronoi/voronoi_hall_ini.xml
```
**JuPedSim version:** 0.8, Branch develop
**OS and version:** Mac OS X Yosemite
**Compiler and version...- Using sources lead to Segmentation faults
**Reproduction Steps:**
```
./bin/jpscore inputfiles/voronoi/voronoi_hall_ini.xml
```
**JuPedSim version:** 0.8, Branch develop
**OS and version:** Mac OS X Yosemite
**Compiler and version:** Clang 7.0.2.7000181
Problem can not be produced in Branch `voronoi`.https://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/205bug in test juelich_2 / all models fail juelich_22018-06-13T16:41:30+02:00Mohcine Chraibim.chraibi@fz-juelich.debug in test juelich_2 / all models fail juelich_2An abs is missing in runtest_juelich_2.py, see diff:
--- a/Utest/juelich_tests/test_2/runtest_juelich_2.py
+++ b/Utest/juelich_tests/test_2/runtest_juelich_2.py
@@ -16,7 +16,7 @@ def runtest2(inifile, trajfile):
fps...An abs is missing in runtest_juelich_2.py, see diff:
--- a/Utest/juelich_tests/test_2/runtest_juelich_2.py
+++ b/Utest/juelich_tests/test_2/runtest_juelich_2.py
@@ -16,7 +16,7 @@ def runtest2(inifile, trajfile):
fps, n, traj = parse_file(trajfile)
evac_time = (max(traj[:, 1]) - min(traj[:, 1])) / float(fps)
tolerance = 0.01
- if (evac_time- must_time) > tolerance:
+ if abs(evac_time- must_time) > tolerance:
logging.info("%s exits with FAILURE evac_time = %f (!= %f)"%(argv[0], evac_time, must_time))
exit(FAILURE)
else:
When patched all models fail the juelich_2 test, since the evacuation time is significantly lower than 10s (~ 9.5s for GCFM, ~ 5.5s for Tordeaux)https://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/204Merge Branch 1812018-06-13T16:41:32+02:00Mohcine Chraibim.chraibi@fz-juelich.deMerge Branch 181Mohcine Chraibim.chraibi@fz-juelich.deMohcine Chraibim.chraibi@fz-juelich.dehttps://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/203Properly deleting pedestrians2018-06-13T16:41:32+02:00Mohcine Chraibim.chraibi@fz-juelich.deProperly deleting pedestriansWhen peds get stuck, they have to be removed.
This leads to undefined behavior.
See comment [here.](https://cst.version.fz-juelich.de/jupedsim/jpscore/commit/f0d57090411513fd0ff5338f49d2729482ccf0cc#note_1792)When peds get stuck, they have to be removed.
This leads to undefined behavior.
See comment [here.](https://cst.version.fz-juelich.de/jupedsim/jpscore/commit/f0d57090411513fd0ff5338f49d2729482ccf0cc#note_1792)Mohcine Chraibim.chraibi@fz-juelich.deMohcine Chraibim.chraibi@fz-juelich.dehttps://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/202Memory leaks2018-06-13T16:41:29+02:00Mohcine Chraibim.chraibi@fz-juelich.deMemory leaksSeveral memory leaks should be fixed.
- [x]
![Screen_Shot_2016-06-30_at_09.26.26](/uploads/fa6eb6e8f9a8091457bdfd009afa60ab/Screen_Shot_2016-06-30_at_09.26.26.png)
- [x]
![Screen_Shot_2016-06-30_at_09.26.12](/uploads/c3b40de16078003ba...Several memory leaks should be fixed.
- [x]
![Screen_Shot_2016-06-30_at_09.26.26](/uploads/fa6eb6e8f9a8091457bdfd009afa60ab/Screen_Shot_2016-06-30_at_09.26.26.png)
- [x]
![Screen_Shot_2016-06-30_at_09.26.12](/uploads/c3b40de16078003ba522ece655bf2ccb/Screen_Shot_2016-06-30_at_09.26.12.png)
- [x]
![Screen_Shot_2016-06-30_at_09.27.02](/uploads/5f2b7a057739f527b68fe3cf737f660a/Screen_Shot_2016-06-30_at_09.27.02.png)
- [x]
![Screen_Shot_2016-06-30_at_09.27.02](/uploads/ef6acc5285e43d9cda33bd4212dcb2b7/Screen_Shot_2016-06-30_at_09.27.02.png)
- [x]
![Screen_Shot_2016-06-30_at_09.27.02](/uploads/48768a966adf269fa3e0ffcae794e07e/Screen_Shot_2016-06-30_at_09.27.02.png)
- [x]
![Screen_Shot_2016-06-30_at_09.27.13](/uploads/b1c417144d7096c4d336b9483aa6240f/Screen_Shot_2016-06-30_at_09.27.13.png)
- [x]
![Screen_Shot_2016-06-30_at_09.27.23](/uploads/9b3d9535775472487af0bc3065c60274/Screen_Shot_2016-06-30_at_09.27.23.png)
- [x]
![Screen_Shot_2016-06-30_at_09.27.32](/uploads/773c8ba43750db3419eb354659960df1/Screen_Shot_2016-06-30_at_09.27.32.png)
- [x]
![Screen_Shot_2016-06-30_at_09.27.41](/uploads/11aaadb676ca2f7cf2a0d3c2298cc801/Screen_Shot_2016-06-30_at_09.27.41.png)
- [x]
![Screen_Shot_2016-06-30_at_09.27.48](/uploads/fb34c80148b3ca00dcc1d89ad030cc03/Screen_Shot_2016-06-30_at_09.27.48.png)https://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/201Data race: Line2018-06-13T16:41:32+02:00Mohcine Chraibim.chraibi@fz-juelich.deData race: LineThe OpenMP parallelization is not thread-safe.
Here is a data race detected by Intel Inspector:
![Screen_Shot_2016-06-30_at_07.43.42](/uploads/a82578a3ea83d30cde90763716e15ad6/Screen_Shot_2016-06-30_at_07.43.42.png)The OpenMP parallelization is not thread-safe.
Here is a data race detected by Intel Inspector:
![Screen_Shot_2016-06-30_at_07.43.42](/uploads/a82578a3ea83d30cde90763716e15ad6/Screen_Shot_2016-06-30_at_07.43.42.png)https://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/200bug with high num_threads (1/2): simulation not deterministic anymore2017-05-19T08:58:40+02:00Mohcine Chraibim.chraibi@fz-juelich.debug with high num_threads (1/2): simulation not deterministic anymoreI've encountered severe issues, when increasing num_threads to higher orders e.g. 24.
An ini-file with no changes at all including constant seed does not yield reproducible results. See screenshot.
![image](/uploads/cccb9501c07762896cc7...I've encountered severe issues, when increasing num_threads to higher orders e.g. 24.
An ini-file with no changes at all including constant seed does not yield reproducible results. See screenshot.
![image](/uploads/cccb9501c07762896cc705f57ead2072/image.png)
@Arne @OSchmidts @chraibi Does someone of you have an idea about that? FYI: The Hermes Cluster is available again and could be used for reproducing that. I've attached exemplary files in #199.Mohcine Chraibim.chraibi@fz-juelich.deMohcine Chraibim.chraibi@fz-juelich.dehttps://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/199bug with high num_threads (2/2): Wrong 'index' in CBuiling::GetRoom() Room ID2018-03-16T15:01:02+01:00Mohcine Chraibim.chraibi@fz-juelich.debug with high num_threads (2/2): Wrong 'index' in CBuiling::GetRoom() Room IDI've encountered severe issues, when increasing num_threads to higher orders e.g. 24.
Despite I have carefully checked my geometry file, I randomly get the following error:
```Wrong 'index' in CBuiling::GetRoom() Room ID: 576157584 size...I've encountered severe issues, when increasing num_threads to higher orders e.g. 24.
Despite I have carefully checked my geometry file, I randomly get the following error:
```Wrong 'index' in CBuiling::GetRoom() Room ID: 576157584 size: 20
Control your rooms ID and make sure they are in the order 0, 1, 2,..```
The very concerning thing is that this somehow depends on num_threads. When using num_threads=24 it works. With 12 it does not.
@Arne @OSchmidts @chraibi Does someone of you have an idea about that? FYI: The Hermes Cluster is available again and could be used for reproducing that. I've attached exemplary files.
[room_id.zip](/uploads/9a79fdcf65b24033849a3f9ca29108f4/room_id.zip)Mohcine Chraibim.chraibi@fz-juelich.deMohcine Chraibim.chraibi@fz-juelich.dehttps://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/198Segfault demos/scenario_5_events2018-06-13T16:41:31+02:00Mohcine Chraibim.chraibi@fz-juelich.deSegfault demos/scenario_5_eventsEvents are not working properly?
Running `demos/scenario_5_events` lead to a Segmentation fault.Events are not working properly?
Running `demos/scenario_5_events` lead to a Segmentation fault.https://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/197ProgressBar slows down the simulation2018-05-01T08:24:57+02:00Mohcine Chraibim.chraibi@fz-juelich.deProgressBar slows down the simulationProgressBar slows down the simulation severly in some cases.
We test alternatives at the momentProgressBar slows down the simulation severly in some cases.
We test alternatives at the momenthttps://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/196log file now mandatory?2018-03-16T15:01:02+01:00Mohcine Chraibim.chraibi@fz-juelich.delog file now mandatory?I encountered problems with calculating some of the demo inis.
However, I've found out that ```<logfile>log</logfile>``` may neither be commented out nor left empty.
Is this a desired behaviour?
If yes, we should update the ini files....I encountered problems with calculating some of the demo inis.
However, I've found out that ```<logfile>log</logfile>``` may neither be commented out nor left empty.
Is this a desired behaviour?
If yes, we should update the ini files.
If not, has anyone changed something that may related to that issue?Mohcine Chraibim.chraibi@fz-juelich.deMohcine Chraibim.chraibi@fz-juelich.dehttps://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/195random seeds not considered anymore2018-03-16T15:01:03+01:00Mohcine Chraibim.chraibi@fz-juelich.derandom seeds not considered anymoreDear @all,
the recent major changes on JPScore develop resulted in a non-consideration of the random seeds. @laemmel (@OSchmidts) could you please have a look that immediately?
This bug is likely to delay my work here in the US sinc...Dear @all,
the recent major changes on JPScore develop resulted in a non-consideration of the random seeds. @laemmel (@OSchmidts) could you please have a look that immediately?
This bug is likely to delay my work here in the US since I wanted to start the first simulation ensembles for ORPHEUS today.
...appreciate more branch-focussed work...https://gitlab.jsc.fz-juelich.de/jupedsim/jpscore/-/issues/194num of threads is not set properly2018-06-13T16:41:32+02:00Mohcine Chraibim.chraibi@fz-juelich.denum of threads is not set properly