Re: [Flightgear-devel] FPE while in-flight
On 20 Dec 2010, at 18:58, ThorstenB wrote: > Personally I don't like ignoring such exceptions, since it often hides > real programming errors. So, yes, I think it'd be great if someone had > a look into these issues, finds all the bad computations and fixes > them. However, since everyone (?) else in FlightGear keeps FPEs > disabled, it may turn out to be a Sisyphean task to rid _all_ such > occurrences... I think we have to start caring - for example we can easily make FPE enabled by default for all debug and Git builds, at least on platforms where it's known to work sanely. We've already discussed a NaN trap into and out of Nasal, and I think we could do a similar NaN trap in the FDM shell, before *and* after the FDM is updated. That would get us much better accuracy on where the NaNs come from, and quick reporting. All of this can be disabled for official release builds, assuming it even has a measurable performance penalty - which I rather doubt. James -- Lotusphere 2011 Register now for Lotusphere 2011 and learn how to connect the dots, take your collaborative environment to the next level, and enter the era of Social Business. http://p.sf.net/sfu/lotusphere-d2d ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] FPE while in-flight
Hi,
I can keep coming with them. :) I just need to remove a hash and add it
some lines above in my start scripts (already introduced to forums).
Curt, I have also found another FPE in AIShip.cpp and I came up with
attached patch (works fine here). Or maybe it is the same? Can you check
that out, please?
The patch also contains some extra include-lines which make my compiler
(gcc (Debian 4.4.5-10) 4.4.5) a little more happier, else it would fail
to compile.
Roland
diff --git a/src/AIModel/AIShip.cxx b/src/AIModel/AIShip.cxx
index 9b6a047..4d8279e 100644
--- a/src/AIModel/AIShip.cxx
+++ b/src/AIModel/AIShip.cxx
@@ -624,6 +624,9 @@ void FGAIShip::ProcessFlightPlan(double dt) {
return;
}
+// Avoids a FPE
+if (dt <= 0.0) return;
+
double time_sec = getDaySeconds();
_dt_count += dt;
diff --git a/src/Input/FGEventInput.cxx b/src/Input/FGEventInput.cxx
index 29ee097..8bf9f30 100644
--- a/src/Input/FGEventInput.cxx
+++ b/src/Input/FGEventInput.cxx
@@ -24,6 +24,7 @@
# include
#endif
+#include
#include "FGEventInput.hxx"
#include
#include
diff --git a/src/Input/FGLinuxEventInput.cxx b/src/Input/FGLinuxEventInput.cxx
index 621e15f..c79a642 100644
--- a/src/Input/FGLinuxEventInput.cxx
+++ b/src/Input/FGLinuxEventInput.cxx
@@ -24,6 +24,10 @@
# include
#endif
+#include
+#include
+#include
+#include
#include "FGLinuxEventInput.hxx"
#include
diff --git a/src/Main/main.cxx b/src/Main/main.cxx
index 5fec795..c56687e 100644
--- a/src/Main/main.cxx
+++ b/src/Main/main.cxx
@@ -141,12 +141,12 @@ static void fgMainLoop( void ) {
altitude->getDoubleValue() * SG_FEET_TO_METER,
globals->get_time_params()->getJD() );
-#if ENABLE_ATCDCL
+#if ENABLE_ATCDCL
// Run ATC subsystem
if (fgGetBool("/sim/atc/enabled"))
globals->get_ATC_mgr()->update(sim_dt);
-#endif
-
+#endif
+
globals->get_subsystem_mgr()->update(sim_dt);
// Update the sound manager last so it can use the CPU while the GPU
signature.asc
Description: This is a digitally signed message part
--
Lotusphere 2011
Register now for Lotusphere 2011 and learn how
to connect the dots, take your collaborative environment
to the next level, and enter the era of Social Business.
http://p.sf.net/sfu/lotusphere-d2d___
Flightgear-devel mailing list
Flightgear-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] FPE while in-flight
ThorstenB skrev 2010-12-20 19.58: > So, unfortunately, divisons-by-zero, processing with "infinity" or > "NaN" (not-a-number) values isn't really uncommon for FG. And since > the FG default is to ignore any FPE (i.e. not to use the > "--enable-fpe" option), few people care (I guess). > Personally I don't like ignoring such exceptions, since it often hides > real programming errors. So, yes, I think it'd be great if someone had > a look into these issues, finds all the bad computations and fixes > them. However, since everyone (?) else in FlightGear keeps FPEs > disabled, it may turn out to be a Sisyphean task to rid _all_ such > occurrences... It is always possible to throw in asserts (http://www.cplusplus.com/reference/clibrary/cassert/assert/) in the code and later use NDEBUG to disable the asserts for production code. Cheers, Jari -- Lotusphere 2011 Register now for Lotusphere 2011 and learn how to connect the dots, take your collaborative environment to the next level, and enter the era of Social Business. http://p.sf.net/sfu/lotusphere-d2d ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] FPE while in-flight
Hi Thorsten,
I was looking into a crash earlier today, and I think I traced it down to
"dt" being zero.
My little nasal script I was playing around used the simple d = r*t formula
rearranged to read vel = dist / dt. I was using the result of this
computation to compute an autopilot target (target rate of climb in my
case.)
So what I think was happening is that the autopilot then tried to compute an
elevator position to achieve a nan rate of climb, which yielded a nan
elevator position. This nan elevator position was fed into the flight
dynamics model (yasim in this case) and everything blew up from that point
on.
It might be interesting to baracade the flight dynamics models a bit better
against nan input values, and then as you suggest, trap errors where they
occur so we can find them and fix them.
For what it's worth I just pushed a small fix to AIShip to handle a dt=0
situation more gracefully and cured one of the fpe's I was seeing.
Regards,
Curt.
On Mon, Dec 20, 2010 at 12:58 PM, ThorstenB wrote:
> On Mon, Dec 20, 2010 at 9:12 AM, Roland Haeder wrote:
> > I got another FPE with GIT version while in-flight with the A380 (recent
> > updates of fgdata). I flew with auto-pilot turned on from EDDL to EHAM
> > and suddenly the FPE happens:
>
> I had a look into some of these FPEs. However, there's tons of places
> where FG runs into such problems. May be caused by
> incorrect/incomplete FDM configurations (e.g. affects specific
> aircraft only), due to computation on unused/uninitialized data (e.g.
> since a feature is disabled but the associated computation is still
> executed), due to plain programming bugs - or even used deliberately
> ("...we don't need a division-by-zero check here since FPEs are
> disabled anyway..."). And it affects the FG core - but also aircraft
> specific Nasal scripts.
> So, unfortunately, divisons-by-zero, processing with "infinity" or
> "NaN" (not-a-number) values isn't really uncommon for FG. And since
> the FG default is to ignore any FPE (i.e. not to use the
> "--enable-fpe" option), few people care (I guess).
> Personally I don't like ignoring such exceptions, since it often hides
> real programming errors. So, yes, I think it'd be great if someone had
> a look into these issues, finds all the bad computations and fixes
> them. However, since everyone (?) else in FlightGear keeps FPEs
> disabled, it may turn out to be a Sisyphean task to rid _all_ such
> occurrences...
>
> cheers,
> Thorsten
>
>
> --
> Lotusphere 2011
> Register now for Lotusphere 2011 and learn how
> to connect the dots, take your collaborative environment
> to the next level, and enter the era of Social Business.
> http://p.sf.net/sfu/lotusphere-d2d
> ___
> Flightgear-devel mailing list
> Flightgear-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/flightgear-devel
>
--
Curtis Olson:
http://www.atiak.com - http://aem.umn.edu/~uav/
http://www.flightgear.org -
http://www.flightgear.org/blogs/category/curt/
Re: [Flightgear-devel] FPE while in-flight
On Mon, Dec 20, 2010 at 9:12 AM, Roland Haeder wrote:
> I got another FPE with GIT version while in-flight with the A380 (recent
> updates of fgdata). I flew with auto-pilot turned on from EDDL to EHAM
> and suddenly the FPE happens:
I had a look into some of these FPEs. However, there's tons of places
where FG runs into such problems. May be caused by
incorrect/incomplete FDM configurations (e.g. affects specific
aircraft only), due to computation on unused/uninitialized data (e.g.
since a feature is disabled but the associated computation is still
executed), due to plain programming bugs - or even used deliberately
("...we don't need a division-by-zero check here since FPEs are
disabled anyway..."). And it affects the FG core - but also aircraft
specific Nasal scripts.
So, unfortunately, divisons-by-zero, processing with "infinity" or
"NaN" (not-a-number) values isn't really uncommon for FG. And since
the FG default is to ignore any FPE (i.e. not to use the
"--enable-fpe" option), few people care (I guess).
Personally I don't like ignoring such exceptions, since it often hides
real programming errors. So, yes, I think it'd be great if someone had
a look into these issues, finds all the bad computations and fixes
them. However, since everyone (?) else in FlightGear keeps FPEs
disabled, it may turn out to be a Sisyphean task to rid _all_ such
occurrences...
cheers,
Thorsten
--
Lotusphere 2011
Register now for Lotusphere 2011 and learn how
to connect the dots, take your collaborative environment
to the next level, and enter the era of Social Business.
http://p.sf.net/sfu/lotusphere-d2d
___
Flightgear-devel mailing list
Flightgear-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] Increasing FOV
Hi, Tried to set the OTW view to a horizontal FOV of 66 degrees, browsing properties indicates /sim/current-view/field-of-view is set to 66 and even tried it at 90 degrees, but actual view does not change. Left edge shows parked 737 and portion of the checkered ILS shack. Tried a lower value of 40 degrees with no change. Am I mistinterpreting the intent of the --fov=x.xx option? Need to increase the horizontal FOV to around 66 degrees. Reading the mini-doc on multiscreen setups and the sample xml files... quote: The displays are arranged in a shallow arc with the left and right displays at a 45.3 degree angle to the center display because, at the assumed screen dimensions, the horizontal field of view of one display is 45.3 degrees. endquote: Does that mean increasing the FOV requires changes in the parameters for the frustum in the xml file or would that be handled by FG at init time? Thanks Jack -- Lotusphere 2011 Register now for Lotusphere 2011 and learn how to connect the dots, take your collaborative environment to the next level, and enter the era of Social Business. http://p.sf.net/sfu/lotusphere-d2d ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] message: Warning: TangentSpaceGenerator: unknown primitive mode 9 why?
On 12/20/2010 05:42 AM, henri orange wrote: > At FG load, i get a lot of these warning messages: > > TangentSpaceGenerator: unknown primitive mode 9 > > Is it just me ? is there any possibility to avoid it ? It's not just you. There are at least two or three bugs involved here. 1) It is a bug that one part of OSG would generate a primitive mode that cannot be handled by other parts of OSG. Indeed, a comparison of .../include/osg/PrimitiveSet with .../src/osgUtil/TangentSpaceGenerator.cpp suggests that there are several primitive modes that could be generated but not handled. With very slightly cleverer coding, the OSG folks could have detected this situation at compile time, and could have fixed it a long time ago. 2) It is a bug in OSG that the error message is remarkably uninformative. The message doesn't say what caused the problem. It doesn't say how serious the problem is, or what might be the consequences of ignoring the problem. It gives no clue how to avoid the problem ... unless you want to download the source and reverse-engineer a large, complex, and virtually uncommented software system. In particular, the message is presumably provoked by some object in the scene, but the message gives no clue as to what the identity of the offending object. 3) It's not clear to me whether this can be considered a bug in FG, or whether changing FG so as to avoid provoking the OSG bug would be worth the trouble. Changing FG would be a royal pain, because of item (2) above. On 04/22/2010 03:12 PM, Tim Moore wrote: >> It would be best to get the polygons out of the loaded geometry. The >> INDEX_MESH optimizer that was recently added to OSG will do this. How recently was it added? How soon will it fix the problem? Installing the latest OSG SVN does not appear to be sufficient to fix the problem. -- Lotusphere 2011 Register now for Lotusphere 2011 and learn how to connect the dots, take your collaborative environment to the next level, and enter the era of Social Business. http://p.sf.net/sfu/lotusphere-d2d ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] Flight Gear
Hey guys, I have had flight gear for about 3 months now, I really enjoy playing offline but only just recently ive had the urge to want to play online, i try connecting online to on my Mac but the fgfs app in the exec app in the dock closes immediatly but flight gear stays open, this does not happen when i play offline, can any one help please? email me back if you need other information regarding this Cheers Grant -- Lotusphere 2011 Register now for Lotusphere 2011 and learn how to connect the dots, take your collaborative environment to the next level, and enter the era of Social Business. http://p.sf.net/sfu/lotusphere-d2d ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] FPE while in-flight
> You can have gdb write a core file by issuing the command > "generate-core-file". Core files can be reloaded into gdb preferably > as long as there were no modifications to the binary or the source. Okay, will do it next time. I have shut it down already. Roland signature.asc Description: This is a digitally signed message part -- Lotusphere 2011 Register now for Lotusphere 2011 and learn how to connect the dots, take your collaborative environment to the next level, and enter the era of Social Business. http://p.sf.net/sfu/lotusphere-d2d___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] FPE while in-flight
On Mon, Dec 20, 2010 at 9:12 AM, Roland Haeder wrote: > Hi all, > > I got another FPE with GIT version while in-flight with the A380 (recent > updates of fgdata). I flew with auto-pilot turned on from EDDL to EHAM > and suddenly the FPE happens: > > http://pastebin.com/ixxjBv25 Looks like effective_range_m is 0, so division by zero. > I keep the gdb console a little open so I can print some variables for > you. :) You can have gdb write a core file by issuing the command "generate-core-file". Core files can be reloaded into gdb preferably as long as there were no modifications to the binary or the source. -- Csaba/Jester -- Lotusphere 2011 Register now for Lotusphere 2011 and learn how to connect the dots, take your collaborative environment to the next level, and enter the era of Social Business. http://p.sf.net/sfu/lotusphere-d2d ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] message: Warning: TangentSpaceGenerator: unknown primitive mode 9 why?
Hi, project members At FG load, i get a lot of these warning messages: TangentSpaceGenerator: unknown primitive mode 9 Is it just me ? is there any possibility to avoid it ? Thanks. Best regards, Alva, Official grtux hangar maintainer. -- Lotusphere 2011 Register now for Lotusphere 2011 and learn how to connect the dots, take your collaborative environment to the next level, and enter the era of Social Business. http://p.sf.net/sfu/lotusphere-d2d___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] FPE while in-flight
Hi again, forgot to add call parameters: /opt/bin/fgfs --ai-scenario=vinson_demo --callsign=Quix0r --multiplay=out,10,mpserver08.flightgear.org,5000 --multiplay=in,10,,5002 --config=$HOME.fgfs/cockpit-view.xml --airport=EDDL --config=$HOMEfgfs/fgfs-base/fgdata/pyromaniac.xml --control=joystick --aircraft=A380 --fg-scenery=$HOME.fgfs/kcdw: $HOME.fgfs/gothenburg:$HOME.fgfs/scotland:$HOME.fgfs/Scenery: $HOMEfgfs/fgfs-base/fgdata/Scenery-Extras: $HOMEfgfs/fgfs-base/fgdata/Scenery-Airspace: $HOMEfgfs/fgfs-base/fgdata/Scenery --log-level=warn --enable-clouds3d --enable-horizon-effect --enable-real-weather-fetch --time-match-local --enable-textures --disable-game-mode --enable-specular-highlight --enable-fullscreen --enable-distance-attenuation --enable-enhanced-lighting --disable-ai-models --fog-disable --enable-anti-alias-hud --enable-save-on-exit --enable-splash-screen --shading-smooth --time-match-local --enable-fpe If you need that paste a little longer, I can upload it to my server. Regards, Roland signature.asc Description: This is a digitally signed message part -- Lotusphere 2011 Register now for Lotusphere 2011 and learn how to connect the dots, take your collaborative environment to the next level, and enter the era of Social Business. http://p.sf.net/sfu/lotusphere-d2d___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] FPE while in-flight
Hi all, I got another FPE with GIT version while in-flight with the A380 (recent updates of fgdata). I flew with auto-pilot turned on from EDDL to EHAM and suddenly the FPE happens: http://pastebin.com/ixxjBv25 I keep the gdb console a little open so I can print some variables for you. :) Roland signature.asc Description: This is a digitally signed message part -- Lotusphere 2011 Register now for Lotusphere 2011 and learn how to connect the dots, take your collaborative environment to the next level, and enter the era of Social Business. http://p.sf.net/sfu/lotusphere-d2d___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
