Re: [Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
Hi Thorsten, > Anyway, you maybe you can double-check if that patch really changed > anything concerning initialization. I checked it in gdb when I made the patch, the members of FGForce contained garbage, some of which was NaN. This was removed by after adding the InitMatrix calls. > If it really did, then I suspect > there must be something else going terribly wrong (memory > corruption?), which could explain why the FGMatrix33 constructors > weren't executed properly. looks very much like it. Note this was triggered by re-initing the Catalina at its mooring place (the Nasal script moves it there from the runway). The first initialization on the runway works fine. Best regards, Andreas -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
> Hi Andreas, > > had a look at the patch you suggested on the JSBSim list > (http://sourceforge.net/mailarchive/attachment.php?list_name=jsbsim-devel&message_id=4D5436B3.10606%40web.de&counter=1). > > However, I looking at the code it seems calling these "InitMatrix" in > the FGForce constructor shouldn't be necessary. The FGMatrix33 > constructor in fact does initialize the matrix to 0 (see > FGMatrix33.cpp, i.e. > FGMatrix33::FGMatrix33(void) > { > data[0] = data[1] = data[2] = data[3] = data[4] = data[5] = > data[6] = data[7] = data[8] = 0.0; > } > > . And the FGForce constructor will automatically call all its members' > constructors. > ... Oops, wasn't finished typing... :) Anyway, you maybe you can double-check if that patch really changed anything concerning initialization. If it really did, then I suspect there must be something else going terribly wrong (memory corruption?), which could explain why the FGMatrix33 constructors weren't executed properly. cheers, Thorsten -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
On Thu, Feb 10, 2011 at 8:05 PM, Andreas Gaeb wrote: > Hi Henri, > > I think I found the error, it was in JSBSim's FGForce class. I've > proposed a fix on the JSBSim-devel mailing list. > > Best regards, > Andreas Hi Andreas, had a look at the patch you suggested on the JSBSim list (http://sourceforge.net/mailarchive/attachment.php?list_name=jsbsim-devel&message_id=4D5436B3.10606%40web.de&counter=1). However, I looking at the code it seems calling these "InitMatrix" in the FGForce constructor shouldn't be necessary. The FGMatrix33 constructor in fact does initialize the matrix to 0 (see FGMatrix33.cpp, i.e. FGMatrix33::FGMatrix33(void) { data[0] = data[1] = data[2] = data[3] = data[4] = data[5] = data[6] = data[7] = data[8] = 0.0; } . And the FGForce constructor will automatically call all its members' constructors. -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
Hi, Andreas Thanks a lot. However , with the Catalina last updated version: https://sites.google.com/site/grtuxhangar/home/download/PBY-Catalina.tar.bz2?attredirects=0&d=1 On my system, the error vanished, i don't understand why. Anyhow your jsbsim fix should solve , or at least reduce the bad luck to get again such error. I 'll give it a try . 2011/2/10 Andreas Gaeb > Hi Henri, > > I think I found the error, it was in JSBSim's FGForce class. I've > proposed a fix on the JSBSim-devel mailing list. > > Best regards, >Andreas > > > -- > The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: > Pinpoint memory and threading errors before they happen. > Find and fix more than 250 security defects in the development cycle. > Locate bottlenecks in serial and parallel code that limit performance. > http://p.sf.net/sfu/intel-dev2devfeb > ___ > Flightgear-devel mailing list > Flightgear-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/flightgear-devel > -- Best regards, Henri, aka Alva Official grtux hangar maintainer -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
Hi Henri, I think I found the error, it was in JSBSim's FGForce class. I've proposed a fix on the JSBSim-devel mailing list. Best regards, Andreas -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
Hello Henri, Am 07.02.2011 18:38, schrieb henri orange: > [...] I am getting ( randomly ) the following messages > > Warning: invalid line segment passed to IntersectVisitor::addLineSegment(..) > nan nan nan nan nan nan segment ignored.. I've looked into this a bit. It seems again to have to do with a missing initialization during a trim run of JSBSim. I have not yet found the definite cause, but tracked the NaNs at least to the PQR rates in FGPropagate during the initial conditions run of the trim which is triggered by the presets-commit command in mooring.nas:230. Unfortunately, I don't know when I will have spare time to continue investigating. At the moment I'd guess the fix, if there is one, would affect the JSBSim internals, so it would go to the JSBSim devel list first. Hope that helps so far, best regards, Andreas -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
Hello Henri, Am 07.02.2011 18:38, schrieb henri orange: > [...] I am getting ( randomly ) the following messages > > Warning: invalid line segment passed to IntersectVisitor::addLineSegment(..) > nan nan nan nan nan nan segment ignored.. I've looked into this a bit. It seems again to have to do with a missing initialization during a trim run of JSBSim. I have not yet found the definite cause, but tracked the NaNs at least to the PQR rates in FGPropagate during the initial conditions run of the trim which is triggered by the presets-commit command in mooring.nas:230. Unfortunately, I don't know when I will have spare time to continue investigating. At the moment I'd guess the fix, if there is one, would affect the JSBSim internals, so it would go to the JSBSim devel list first. Hope that helps so far, best regards, Andreas -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
Hi, Not only with the relocation to a mooring position, though that feature is randomly making the issue. I cannot explain why, and cannot reproduce twice the same. I guess the reset sequence/speed is the cause of the issue, scenery ? model ? fdm ? constant values ?. The relocation to a mooring position feature is to me, right, with every predefined place ( see the Boeing314-route.xml file ). 2011/2/7 Csaba Halász > On Mon, Feb 7, 2011 at 6:59 PM, Csaba Halász > wrote: > > > > That is already the end of the problem chain. Somewhere NaN is > > creeping into the system and propagates to various parts. Finding what > > uses z-accel-fps_sec, while possible, will not help you in > > troubleshooting the root cause. > > It might have something to do with the relocation to a mooring position. > I had some hint of NaN when that happened near KSFO, but had no > problems at EHAM (which doesn't have a mooring position). > > -- > Csaba/Jester > > -- Best regards, Henri, aka Alva Official grtux hangar maintainer -- The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
On Mon, Feb 7, 2011 at 6:59 PM, Csaba Halász wrote: > > That is already the end of the problem chain. Somewhere NaN is > creeping into the system and propagates to various parts. Finding what > uses z-accel-fps_sec, while possible, will not help you in > troubleshooting the root cause. It might have something to do with the relocation to a mooring position. I had some hint of NaN when that happened near KSFO, but had no problems at EHAM (which doesn't have a mooring position). -- Csaba/Jester -- The modern datacenter depends on network connectivity to access resources and provide services. The best practices for maximizing a physical server's connectivity to a physical network are well understood - see how these rules translate into the virtual world? http://p.sf.net/sfu/oracle-sfdevnlfb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
On Mon, Feb 7, 2011 at 6:38 PM, henri orange wrote: > Hello, devel-members, > > I am getting ( randomly ) the following messages > > Warning: invalid line segment passed to IntersectVisitor::addLineSegment(..) > nan nan nan nan nan nan segment ignored.. > Warning: invalid line segment passed to IntersectVisitor::addLineSegment(..) > nan nan nan nan nan nan segment ignored.. > Warning: invalid line segment passed to IntersectVisitor::addLineSegment(..) > nan nan nan nan nan nan segment ignored.. > PT_vs_hpt: ran out of layers for h=nan > PT_vs_hpt: ran out of layers for h=nan > Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN > Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN > Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN > > I would like to know, which feature, within flightgear is using "Nasal > getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0]". > I do not try to access that property, so i guess it is an internal > (automated ?) process That is already the end of the problem chain. Somewhere NaN is creeping into the system and propagates to various parts. Finding what uses z-accel-fps_sec, while possible, will not help you in troubleshooting the root cause. You might want to try using a debugger and running with the command line option --enable-fpe. -- Csaba/Jester -- The modern datacenter depends on network connectivity to access resources and provide services. The best practices for maximizing a physical server's connectivity to a physical network are well understood - see how these rules translate into the virtual world? http://p.sf.net/sfu/oracle-sfdevnlfb ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN
Hello, devel-members, I am getting ( randomly ) the following messages Warning: invalid line segment passed to IntersectVisitor::addLineSegment(..) nan nan nan nan nan nan segment ignored.. Warning: invalid line segment passed to IntersectVisitor::addLineSegment(..) nan nan nan nan nan nan segment ignored.. Warning: invalid line segment passed to IntersectVisitor::addLineSegment(..) nan nan nan nan nan nan segment ignored.. PT_vs_hpt: ran out of layers for h=nan PT_vs_hpt: ran out of layers for h=nan Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0] is NaN I would like to know, which feature, within flightgear is using "Nasal getprop: property /accelerations[0]/pilot[0]/z-accel-fps_sec[0]". I do not try to access that property, so i guess it is an internal (automated ?) process Your answer could help me to debug the model i am working on ( http://www.flightgear.org/forums/viewtopic.php?f=4&t=10992 ) OR Is it only a known bug ? Thanks -- Best regards, Henri, aka Alva Official grtux hangar maintainer -- The modern datacenter depends on network connectivity to access resources and provide services. The best practices for maximizing a physical server's connectivity to a physical network are well understood - see how these rules translate into the virtual world? http://p.sf.net/sfu/oracle-sfdevnlfb___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel