Reported by heep on 3 Dec 2008 14:20

Version: 4.0b8

The "fast mode" is ~10 times slower than in OMNET-3.4b2, TKenv blocks.

Mantis Bugtracker #16

Reported by baumgart on 16 Dec 2008 09:23

Version: 4.0rc1

The basicspringembedderlayout produces strange results if some nodes are fixed.
If e.g. you have one unconnected fixed node and about 100 nodes connected nodes in circle, you'll get a huge canvas - if you remove the fixed node, everything works well.

I've taken a look at the source - there's a comment, that a layout with fixed nodes should not be scaled, so that fixed nodes keep their position. Maybe a fix would be to scale all movable nodes and keep the position of fixed nodes. Of course fixed nodes could get covered in this case - but this problem only occurs if you have a large number of nodes (in this case they currently get covered anyway). If tried the "USE_CONTRACTING_BOX" feature as well, but this seems to be broken (segmentation fault).

BTW, it would also be great to have an option to adapt the size of the canvas used for layouting to the current size of the Tk window. In this case the user could dynamically adapt the canvas size.

I've also tried the new layout mechanism - but this seems to produce worse results and takes very long to complete.

Mantis Bugtracker #38

Reported by rblenin on 5 Dec 2008 16:38

Version: 4.0b8

I tried to compile, in omnet++ 4.0b8, some of my omnet++ projects which were created and compiled successfully using omnet++ 3.3 but started getting errors for member functions (such as 'cell', 'pdf', 'cdf') of the class cDoubleHistrogram. When I looked at the API reference, I noticed changes in the member functions' names and also in the class hierarchy. I have not gone through the manual of 4.0b8 yet and I have done a quick check after installing 4.0b8.

Is there any simple way to enable the backward compatibility?

Thanks.
-Lenin.

Mantis Bugtracker #24

Reported by baumgart on 4 Dec 2008 12:14

Version: 4.0b8

The ned-path option in the ini file isn't platform-independent: On Windows XP (with MinGW) it has to contains ";" as separation character. ":" doesn't work.

Mantis Bugtracker #20

Reported by baumgart on 7 Dec 2008 15:24

Version: 4.0b8

As far as I understand, the prefer way to generate Makefiles for projects with src/ directory is as follows:

The IDE creates auto-generated Makefiles with opp_makemake in PROJDIR and PROJDIR/src. Additionally you can export a ./makemake script to generate these Makefiles without IDE.

I would prefer the way how INET does this currently: Have a Makefile in PROJDIR which calls opp_makemake to generate the Makefile in PROJDIR/src.
Then the IDE should only create the Makefile in PROJDIR/src and have an option to export the "Makefile-generation-Makefile" as PROJDIR/Makefile. Benefit: You can simply call make in the PROJDIR to compile the project from the commandline. This top-level Makefile could also be extended e.g. to generate doxygen documentation.

This even works with the current IDE already, if you manually set the build PATH to PROJDIR/src - but maybe this should be the default behaviour for new projects as well. BTW, the simulation binary (or better a symbolic link to the binary) is always generated in PROJDIR/src/ . It would be cleaner to have this in PROJDIR or PROJDIR/bin.

Mantis Bugtracker #31

Reported by ViRe on 20 Jan 2009 09:27

Version: 4.0rc1

This is an error of eclipse, but I think it is caused by my project.

There were many syntax errors in my project(more than 400, most of them are in the NED files), and I was tring to fix them in the IDE. As we know, after an syntax error is fixed, some others will gone, and the IDE will refresh the error messages in the "Problems" column. The IDE is very slowly, after about 1/3 of the errors are fixed, the error happend, as showned in the graph. I checked my computer, about 2G memory of my computer was "eaten" by the IDE.

Mantis Bugtracker #45

Reported by baumgart on 26 Nov 2008 10:57

Version: 4.0b8

omnetpp-4.0b8 is using an ugly font for ev output with Tkenv (on Ubuntu 8.04). With omnetpp-4.0b7 (and omnetpp-3.4b2) the font was a lot nicer.

Mantis Bugtracker #8

Reported by stkrause on 10 Dec 2008 13:15

Version: 4.0b8

When inspecting an object (simple or compound), the display string can be expanded by clicking the [+] field, which in turn gives access to a "base" field. Trying to expand this field by clicking [+] results in a crash in

cDisplayStringDescriptor::getFieldAsString

in line

3997 return basedesc->getFieldAsString(object,field,i,resultbuf,bufsize);

Mantis Bugtracker #33

Reported by mbergner on 24 Nov 2008 04:22

Version: 4.0b8

When you import the INET framework and build it, you get a lot of errors. These are mostly compiler warnings which should be treated as such.

I would like to see those as warnings and not as errors, since errors signal (for me) a real problem that causes the simulation to not run or impact the linking.
An error should only occur if the executable/library could not be linked or if the simulation will definitely (without any chance that this is desired) is incorrect.

Mantis Bugtracker #3

Reported by baumgart on 26 Nov 2008 11:21

Version: 4.0b8

The IDE ini editor doesn't know about parameters of dynamically created modules, because it only parses the ned files of all static modules (and submodules) in a network.

I've try to add a "hint" in the ned file by adding the to be dynamically created module as empty array (e.g. overlayTerminal[0]: OverlayTerminal), but this doesn't show any effect in the ini editor.

Mantis Bugtracker #9

Reported by baumgart on 7 Dec 2008 13:15

Version: 4.0b8

If you create a new OMNeT++-project with the IDE and select "Source-sink example with src folder" the generated project can't be started from the command line, because the ned-path is incomplete (started from the simulations/ directory with ../src/SimpleProject). Starting from the IDE works fine.

Additional information: Ubuntu 8.04

Mantis Bugtracker #27

Reported by baumgart on 7 Dec 2008 13:32

Version: 4.0b8

omnetpp-3.3 supported statically compiled ned-files, which allowed to build a self-contained simulation binary. As omnetpp-4.0 has dropped support for statically compiled ned-files, this is not possible anymore!

Therefor it would be important to add a feature to embed ned-files in the simulation binary. This is especially useful for the distribution of a simulation in a grid network.

Mantis Bugtracker #29

Reported by baumgart on 4 Dec 2008 11:04

Version: 4.0b8

For running scripted simulations it would be useful to have a command line option to print the names of all configurations in the ini file (maybe you could add the number of runs for each configuration in this output, too).

Example:

SomeConfiguration 4
SomeOtherConfiguration 2

BTW, there's also a small typo in the documentation when calling the simulation binary with "-h": It says the "-g, -G" option makes "-n" and "-c" more verbose. Probably the "-n" should be "-x".

Mantis Bugtracker #18

Reported by ViRe on 16 Jan 2009 06:15

Version: 4.0rc1

I built a simulation project(case) with om3.3 and INET, and want to migrate it to om4.0rc1. The mainly project(main part) has been migrated successfully,
and I want to migrate a simulation scenario(with some NED files and some INI files). The "migratened | tee ned.out" command was called, and an error happend:"./Libraries/Scenarios/NewTI/TI_NET.ned:29: Error: parse error in display string `p=50,50;i=misc/sun,,100;"
The migrating couldn't continue. I want to fix the error file in the IDE. So I find out the error file in the "Project Explorer", double cilck, the IDE crashed(quit silently).

I tried several times, every time I try to open the error file in the IDE, the bug happens.

Finally, I fixed the bug in a txt editor, and the IDE works well now.

By the way, INET is a little big, has about more 500 source files, could it be the reason?
    
Some information:  
	eclipse.buildId=M20080911-1700
	java.version=1.5.0_06
	java.vendor=Sun Microsystems Inc.
	BootLoader constants: OS=win32, ARCH=x86, WS=win32, NL=zh_CN
	Command-line arguments:  -os win32 -ws win32 -arch x86
	OS: MicroSoft Windows Professional SP2

Additional information: eclipse.buildId=M20080911-1700
java.version=1.5.0_06
java.vendor=Sun Microsystems Inc.
BootLoader constants: OS=win32, ARCH=x86, WS=win32, NL=zh_CN
Command-line arguments: -os win32 -ws win32 -arch x86
OS: MicroSoft Windows Professional SP2

Mantis Bugtracker #44

Reported by ViRe on 20 Jan 2009 10:21

Version: 4.0rc1

The "ev" do NOT work well in the initialization, that means, is an ev state is added in the initialize method, no output will show up in the GUI. But it works well in the handleMessage(). Is it a BUG?

Mantis Bugtracker #48

Reported by baumgart on 15 Dec 2008 22:05

Version: 4.0rc1

Dynamic module creation is slow in "Run mode" (many slow "initialize()" visualizations). Pressing the STOP button does not immediately stop the simulation (slow output continues).

Mantis Bugtracker #36

Reported by baumgart on 7 Dec 2008 12:18

Version: 4.0b8

The legend box for scatter plots cannot be closed. There is no close button and pressing the legend icon again produces an additional legend. Maybe this is already fixed with ~72

Additional information: Ubuntu 8.04

Mantis Bugtracker #26

Reported by baumgart on 24 Nov 2008 11:35

Version: 4.0b8

There are several valgrind errors with Tkenv::objectDeleted(), e.g.:

==13972== Conditional jump or move depends on uninitialised value(s)
==13972== at 0x47FCA28: Tkenv::objectDeleted(cObject*) (tkenv.cc:1115)
==13972== by 0x4AD7615: cObject::~cObject() (cobject.cc:31)
==13972== by 0x4AC1AC0: cRNG::~cRNG() (crng.h:58)
==13972== by 0x4AC4516: cMersenneTwister::~cMersenneTwister() (cmersennetwister.h:50)
==13972== by 0x4903B27: EnvirBase::readPerRunOptions() (envirbase.cc:1076)
==13972== by 0x47FD3AE: Tkenv::readPerRunOptions() (tkenv.cc:1055)
==13972== by 0x47FE84A: Tkenv::newRun(char const*, int) (tkenv.cc:748)
==13972== by 0x480DAEA: newRun_cmd(void*, Tcl_Interp*, int, char const**) (tkcmd.cc:301)
==13972== by 0x4F8FB6D: TclInvokeStringCommand (in /usr/lib/libtcl8.4.so.0)
==13972== by 0x4F90E55: TclEvalObjvInternal (in /usr/lib/libtcl8.4.so.0)
==13972== by 0x4FBBDE6: (within /usr/lib/libtcl8.4.so.0)
==13972== by 0x4FB9F3E: TclCompEvalObj (in /usr/lib/libtcl8.4.so.0)

==13972== Conditional jump or move depends on uninitialised value(s)
==13972== at 0x47FCA28: Tkenv::objectDeleted(cObject*) (tkenv.cc:1115)
==13972== by 0x4AD7615: cObject::~cObject() (cobject.cc:31)
==13972== by 0x4ADA1A4: cNamedObject::~cNamedObject() (cnamedobject.cc:73)
==13972== by 0x4ADCFCB: cOwnedObject::~cOwnedObject() (cownedobject.cc:116)
==13972== by 0x8202625: cNoncopyableOwnedObject::~cNoncopyableOwnedObject() (cownedobject.h:251)
==13972== by 0x4AA4A82: cDefaultList::~cDefaultList() (cdefaultlist.cc:91)
==13972== by 0x4A9D516: cModuleType::create(char const*, cModule*, int, int) (ccomponenttype.cc:162)
==13972== by 0x4A9C671: cModuleType::create(char const*, cModule*) (ccomponenttype.cc:106)
==13972== by 0x4AEDD46: cSimulation::setupNetwork(cModuleType*) (csimulation.cc:381)
==13972== by 0x47FE94D: Tkenv::newRun(char const*, int) (tkenv.cc:762)
==13972== by 0x480DAEA: newRun_cmd(void*, Tcl_Interp*, int, char const**) (tkcmd.cc:301)
==13972== by 0x4F8FB6D: TclInvokeStringCommand (in /usr/lib/libtcl8.4.so.0)

==13972== Conditional jump or move depends on uninitialised value(s)
==13972== at 0x47FCA28: Tkenv::objectDeleted(cObject*) (tkenv.cc:1115)
==13972== by 0x4AD7615: cObject::~cObject() (cobject.cc:31)
==13972== by 0x4ADBC41: cPar::~cPar() (cpar.cc:40)
==13972== by 0x4A9ACEE: cComponent::reallocParamv(int) (ccomponent.cc:100)
==13972== by 0x4A9AE71: cComponent::addPar(cParImpl*) (ccomponent.cc:112)
==13972== by 0x4B4D62D: cNEDNetworkBuilder::doParam(cComponent*, ParamElement*, bool) (cnednetworkbuilder.cc:175)
==13972== by 0x4B4D9E9: cNEDNetworkBuilder::doParams(cComponent*, ParametersElement*, bool) (cnednetworkbuilder.cc:138)
==13972== by 0x4B4EFA0: cNEDNetworkBuilder::addParametersAndGatesTo(cComponent*, cNEDDeclaration*) (cnednetworkbuilder.cc:104)
==13972== by 0x4B49F49: cDynamicModuleType::addParametersAndGatesTo(cModule*) (cdynamicmoduletype.cc:72)
==13972== by 0x4A9D4D7: cModuleType::create(char const*, cModule*, int, int) (ccomponenttype.cc:156)
==13972== by 0x4A9C671: cModuleType::create(char const*, cModule*) (ccomponenttype.cc:106)
==13972== by 0x4B4E6D0: cNEDNetworkBuilder::addSubmodule(cModule*, SubmoduleElement*) (cnednetworkbuilder.cc:532)

Additional information: Ubuntu 8.04 with gcc 4.2.4

Mantis Bugtracker #4

Reported by stkrause on 4 Dec 2008 11:04

Version: 4.0b8

When you filter out a bigger part of the output, scrolling starts to get more and more strange: The main window will scroll for a bit, than stop for a while, and then start scrolling agin. Meanwhile the size of the scrollbar will fluctuate from really small to fairly big in a seemingly random manner.
This can get really annoing if trying to scroll through the output output of one single module in a bigger simulation.
It seems scrolling is done before filtering, i.e. it will scroll (internally) through all messages, even non-displayed ones, which leads to "stuttering".

Mantis Bugtracker #17

Reported by baumgart on 7 Dec 2008 13:24

Version: 4.0b8

Setting the display string fails sometimes.

Example to reproduce:
Create a new Sink-Source project with the IDE new project wizard.

Add the following line at the end of Sink::initialize():

gate("in")->setDisplayString("ls=red,1");

If you start the simulation the connection arrow stays black.

valgrind output:

==12553== Conditional jump or move depends on uninitialised value(s)
==12553== at 0x45857BE: (within /lib/tls/i686/cmov/libc-2.8.90.so)
==12553== by 0x4589380: vfprintf (in /lib/tls/i686/cmov/libc-2.8.90.so)
==12553== by 0x45AF0E3: vsnprintf (in /lib/tls/i686/cmov/libc-2.8.90.so)
==12553== by 0x46D88FF: opp_stringf(char const*, ...) (stringutil.cc:175)
==12553== by 0x4394BD2: cNEDDeclaration::doConnectionProperties(int, char const*) const (cneddeclaration.cc:207)
==12553== by 0x4394ED2: cNEDDeclaration::getConnectionProperties(int, char const*) const (cneddeclaration.cc:199)
==12553== by 0x439A89F: cDynamicModuleType::getConnectionProperties(int, char const*) const (cdynamicmoduletype.cc:114)
==12553== by 0x42EBB3D: cChannel::getProperties() const (cchannel.cc:170)
==12553== by 0x42EDAD8: cComponent::getDisplayString() (ccomponent.cc:185)
==12553== by 0x430E21F: cGate::getDisplayString() (cgate.cc:427)
==12553== by 0x430E232: cGate::setDisplayString(char const*) (cgate.cc:432)
==12553== by 0x804BAAA: simpleproject::Sink::initialize() (Sink.cc:18)

Additional information: Ubuntu 8.04

Mantis Bugtracker #28

Reported by baumgart on 7 Dec 2008 13:56

Version: 4.0b8

opp_makemake adds a relative rpath for dynamically linked library.

Example for linking with INET:

LIBS += -Wl,-rpath,../../INET-20081105/src -Wl,-rpath,/usr/lib

If the simulation binary is started from a different directory level then PROJDIR/src (e.g. PROJDIR/out/gcc-debug/src/) the library is not found.

Possible fix: Add absolute rpath

Mantis Bugtracker #30

Reported by baumgart on 28 Nov 2008 11:31

Version: 4.0b8

When using the IDE with the default discovery profile "OmnetppGCCPerProjectProfile" the CDT indexer doesn't work, because no include paths are discovered and added to the project's include path list.

Switching to "GCC per project scanner info profile" (in "Project properties"->"C/C++ Build"->"Discovery Options"->"Discovery Profile") fixes the problem.

Additional information: Ubuntu 8.04, gcc 4.2.4

Mantis Bugtracker #11

Reported by baumgart on 28 Nov 2008 14:57

Version: 4.0b8

I'm wondering if the beautiful new results analysis tool could be used to generate fast plots out of the scenario description in the ini file.

Take for example the following ini file with an additional [Plot] section:

[General]
repeat = 2
**.lifetimeMean = ${lifetime=1000s,10000s}

[Config ChordChurn]
**.overlayType = "ChordModules"
**.stabilizeInterval = ${stab=5s,20s}

[Config KademliaChurn]
**.overlayType = "KademliaModules"

[Plot ChurnComparison]
x = lifetime
y = "Sent Total Bytes"
lines = KademliaChurn, ChordChurn-stab=5s, ChordChurn-stab=20s

The idea is, that this could generate a plot with the input parameter "lifetime" on the x axes and the scalar matching the (unambiguous) string "Sent Total Bytes" on the y axes. The plot should contain three lines taking the results from the scenarios KademliaChurn and ChordChurn. Because the scenario ChordChurn has a second input parameter "stab", it has to be specified, too.

The lines should represent the averages of all repetitions. The title and axes labels could be also generated from the plot ("ChurnComparison", "lifetime", "Sent Total Bytes").

Maybe this can already be accomplished with the GUI of the results analysis tool, but probably is a lot more complicated than the ini file approach (at least I haven't found an easy way to do this with the GUI).

Mantis Bugtracker #12

Reported by baumgart on 5 Dec 2008 16:11

Version: 4.0b8

cMessage::getArrivalTime() returns simtime_t by value. This is a performance problem especially for the FES (shiftup and insert).

After changing the method to return const simtime_t& we had performance improvements of about 10-20 percent (compiled with -O3).

Probably this should be changed for all simtime_t setters and getters in cmessage.h.

Mantis Bugtracker #22

Reported by baumgart on 15 Dec 2008 22:43

Version: 4.0rc1

The "Inspect module output" window shows several unrelated "Initializing channel" messages (probably some problem with the filter expression).

Mantis Bugtracker #37

Reported by horal on 4 Dec 2008 23:29

This option **.server.udpApp[*].waitInterval = ${1ms,0.666ms} sign ini editor like syntax error. Everything other work well.

Mantis Bugtracker #21

Reported by baumgart on 1 Dec 2008 16:15

Version: 4.0b8

./configure fails with the following warning, if JRE is not installed, although JRE is not required to compile the simulation kernel:

"WARNING: Cannot find a suitable Java Runtime Environment - The IDE requires JRE 1.5 or later. Please install it before configuring"

Mantis Bugtracker #14

Reported by baumgart on 5 Dec 2008 16:16

Version: 4.0b8

If there is a manually generated Makefile, which would be overwritten by the IDE with a automatically generated Makefile, there is a warning message that there is already a manually generated Makefile present. Afterwards the manually generated Makefile is deleted.

Mantis Bugtracker #23

Reported by baumgart on 4 Dec 2008 11:22

Version: 4.0b8

Currently the IDE simulation launcher only supports to simulate all runs in a single configuration. An option to run all configurations would be useful.

Maybe this option should also be implemented in for the Cmdenv runtime.

E.g. you could change the default behavior to simulate all configurations in a given ini file. In this case I would even exclude the General configuration (if there is only a General section run this one, and in all other cases only run all configuration sections).

Mantis Bugtracker #19

Reported by horal on 24 Nov 2008 15:21

Version: 4.0b7

It would be nice, if I'll be able to save settings of sequence charts. Not every time add vector to axis, set hide/show options, and so on.

Mantis Bugtracker #5

Reported by rhornig on 21 Oct 2008 01:26

Version: 4.0b7

Steps to reproduce:

• Open a NED editor and switch to source editor
• Create a compound module with the folowing content:

module Module
{
types:
simple A {
gates:
inout port;
}
module mm {
submodules:
sub : A;
sub2 : A;
connections:

    }

}

• Position the cursor after "connections:" and press CTRL-SPACE
  Content assist does not list sub and sub2 submodules.

Mantis Bugtracker #1

Reported by mbergner on 9 Jan 2009 12:41

Version: 4.0rc1

It would be nice to have a function to apply where I can calculate a mean over a fixed time interval.

What I mean ist that I collect some data in a vector and then calculate a mean over every 1s or 1m or so. This is of course independent of the number of entries recorded in that time interval.

If this was already possible in the IDE, excuse this feature request and see it as a bug in the presentation/documentation (because I did not see it :)).

Mantis Bugtracker #42

Reported by baumgart on 1 Dec 2008 16:08

Version: 4.0b8

There are several warnings when trying to compile omnetpp-4.0b8 with gcc for ARM with scratchbox:

stringutil.cc: In function `std::string opp_quotestr(const char*)':
stringutil.cc:148: warning: comparison is always true due to limited range
of data type

stringutil.cc: In function `bool opp_needsquotes(const char*)':
stringutil.cc:165: warning: comparison is always true due to limited range
of data type

eventlogentry.cc: In constructor EventLogEntry::EventLogEntry()': eventlogentry.cc:35: warning: converting of negative value -0x00000000000000001' to `unsigned int'

omnetpp-4.0b8/include/cwatch.h: In member
function `virtual std::string cWatch_char::info() const':
omnetpp-4.0b8/include/cwatch.h:150: warning:
comparison is always true due to limited range of data type

cclassdescriptor.cc: In static member function `static long int
cClassDescriptor::string2enum(const char*, const char*)':
cclassdescriptor.cc:126: warning: comparison is always true due to limited
range of data type

omnetpp-4.0b8/include/cwatch.h: In member
/function `virtual std::string cWatch_char::info() const':
omnetpp-4.0b8/include/cwatch.h:150: warning:
/comparison is always true due to limited range of data type

sim_std_m.cc: In member function `virtual void*
cDensityEstBaseDescriptor::getFieldStructPointer(void*, int, int) const':
sim_std_m.cc:5610: warning: taking address of temporary

Mantis Bugtracker #13

Reported by baumgart on 1 Dec 2008 17:33

Version: 4.0b8

When encapsulating a message within a customized message (generation gap pattern, @Customize(true)) it is not displayed correctly in Tkenv.

When clicking on a message in Tkenv ("Inspect as object"), the encapsulated message is shown, but details like "creationTime" for the encapsulated message are missing. Furthermore if there is an additional encapsulated message in this message, it is neither shown.

This bug already showed up in omnetpp-3.4b2, omnetpp-3.3 is working fine).

Mantis Bugtracker #15

Reported by stkrause on 10 Dec 2008 16:34

Version: 4.0b8

In fast mode, when cComponent::bubble() is called on a recently created module, instead of showing the bubble often an error similar to "modinsp.cc#692:ptr0x9737e70 not found" is printed to stderr.

This most likely happens because "CHK(Tcl_VarEval(interp, "graphmodwin_bubble ",canvas," ",ptrToStr(mod)," {",text,"}",NULL));" is called before the module is actually displayed by Tkenv (which will only happen the next time Tkenv refreshes its view).

An easy fix might be to buffer bubbles on fast mode and show them on the next refresh.

Additional information: This error also exists in omnet 3.

Mantis Bugtracker #34

Reported by ViRe on 20 Jan 2009 10:20

Version: 4.0rc1

If a simple module has a member(such as, a queue) which is an object of class inherited from the cSimpleModule(which is inherited from cComponent), and the componenttype could be left unset. But if I try to inspect the simple module in the GUI, an error will happen. I found such codes in the tkutil.cc:
const char getObjectShortTypeName(cObject object)
{
if (dynamic_cast<cComponent>(object))
return ((cComponent)object)->getComponentType()->getName();
else
return object->getClassName();
}

Because my queue does NOT set the componenttype, so the getComponentType() returns NULL. Should some ensurance be added?

Mantis Bugtracker #47

Reported by gilles on 14 Feb 2009 21:52

Version: 4.0rc1

The function index() in the Txc9.cc, Txc10.cc, Txc11.cc and Txc12.cc must be replace by the function getIndex().

Regards,
Gilles.

Mantis Bugtracker #50

Reported by mbergner on 9 Jan 2009 12:28

Version: 4.0rc1

Sometimes the IDE crashes after or while running some experiments in batch mode (with some parameters). I get no error, the IDE just quits.

I don't know where it comes from, I attached the .metadata/log.

I will run the IDE from the commandline to see that error, when it crashes again.

Mantis Bugtracker #41

Reported by baumgart on 26 Nov 2008 08:12

Version: 4.0b8

IDE crashed with SEGV when trying to open a ned file:

#0 0xb7f63410 in __kernel_vsyscall ()
#1 0xb7e05085 in raise () from /lib/tls/i686/cmov/libc.so.6
#2 0xb7e06a01 in abort () from /lib/tls/i686/cmov/libc.so.6
#3 0xb7e3db7c in ?? () from /lib/tls/i686/cmov/libc.so.6
#4 0xb7e45a85 in ?? () from /lib/tls/i686/cmov/libc.so.6
#5 0xb7e494f0 in free () from /lib/tls/i686/cmov/libc.so.6
#6 0xb5003b11 in operator delete () from /usr/lib/libstdc++.so.6
#7 0xb4fdf79d in std::string::_Rep::_M_destroy () from /usr/lib/libstdc++.so.6
#8 0xb4fe1571 in std::basic_string<char, std::char_traits<char>, std::allocator<char> >::~basic_string () from /usr/lib/libstdc++.so.6
#9 0xb1ef7b6c in NED2Generator::generate ()
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9959/libopplibs.so
#10 0xb1e44f8a in Java_org_omnetpp_ned_engine_nedxmlJNI_NED2Generator_1generate_1_1SWIG_11 ()
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9959/libopplibs.so
#11 0xb5c02e9d in ?? ()
#12 0xb5bfbd77 in ?? ()
#13 0xb5bfbd77 in ?? ()
#14 0xb5bfbd77 in ?? ()
#15 0xb5bfc253 in ?? ()
#16 0xb5bfbd77 in ?? ()
#17 0xb5bfc253 in ?? ()
#18 0xb5bfbedd in ?? ()
#19 0xb5bfbedd in ?? ()
#20 0xb5bfbedd in ?? ()
#21 0xb5bfc3b9 in ?? ()
#22 0xb5bfbedd in ?? ()
#23 0xb5bfbedd in ?? ()
#24 0xb5bfbedd in ?? ()
#25 0xb5bfbedd in ?? ()
#26 0xb5bfc3b9 in ?? ()
#27 0xb5bfbedd in ?? ()
#28 0xb5bfbedd in ?? ()
#29 0xb5bfbedd in ?? ()
#30 0xb5bfbedd in ?? ()
#31 0xb5bfc3b9 in ?? ()
#32 0xb5bfbedd in ?? ()
#33 0xb5bfc3b9 in ?? ()
#34 0xb5bfbedd in ?? ()
#35 0xb5bfc3b9 in ?? ()
#36 0xb5bfbedd in ?? ()
#37 0xb5bfc3b9 in ?? ()
#38 0xb5bfbd77 in ?? ()
#39 0xb5bfbd77 in ?? ()
#40 0xb5bfbedd in ?? ()
#41 0xb5bfc3b9 in ?? ()
#42 0xb5f16188 in ?? ()
#43 0xb5bfbedd in ?? ()
#44 0xb5bfbedd in ?? ()
#45 0xb5bfbedd in ?? ()
#46 0xb5bfbedd in ?? ()
#47 0xb5bfbedd in ?? ()
#48 0xb5bfbedd in ?? ()
#49 0xb5bfbedd in ?? ()
#50 0xb5bfbedd in ?? ()
#51 0xb5bfbedd in ?? ()
#52 0xb5bfbedd in ?? ()
#53 0xb5bfbedd in ?? ()
#54 0xb5bfbd77 in ?? ()
#55 0xb5bfbd77 in ?? ()
#56 0xb5bfbd77 in ?? ()
#57 0xb5bfc3b9 in ?? ()
#58 0xb5bfbedd in ?? ()
#59 0xb5bfbd77 in ?? ()
---Type <return> to continue, or q <return> to quit---
#60 0xb5bfbd77 in ?? ()
#61 0xb5bfc253 in ?? ()
#62 0xb5bfbd77 in ?? ()
#63 0xb5bfc3b9 in ?? ()
#64 0xb5bfc3b9 in ?? ()
#65 0xb5bfbedd in ?? ()
#66 0xb5bfbd77 in ?? ()
#67 0xb5bfbd77 in ?? ()
#68 0xb5bfbd77 in ?? ()
#69 0xb5bfbd77 in ?? ()
#70 0xb5bfbd77 in ?? ()
#71 0xb5bfbd77 in ?? ()
#72 0xb5bfc3b9 in ?? ()
#73 0xb5bfbedd in ?? ()
#74 0xb5bfbd77 in ?? ()
#75 0xb5bfbd77 in ?? ()
#76 0xb5bfc253 in ?? ()
#77 0xb5bfbd77 in ?? ()
#78 0xb5bfbd77 in ?? ()
#79 0xb5bfbd77 in ?? ()
#80 0xb5bfc3b9 in ?? ()
#81 0xb5bfc253 in ?? ()
#82 0xb5bfbd77 in ?? ()
#83 0xb5bfbd77 in ?? ()
#84 0xb5bfc253 in ?? ()
#85 0xb5bfbcb1 in ?? ()
#86 0xb5bfbcb1 in ?? ()
#87 0xb5bfbedd in ?? ()
#88 0xb5bfbedd in ?? ()
#89 0xb5bfbedd in ?? ()
#90 0xb5dc3e33 in ?? ()
#91 0xb6093d8b in ?? ()
#92 0xb5bfbedd in ?? ()
#93 0xb5bfbedd in ?? ()
#94 0xb5bfbcb1 in ?? ()
#95 0xb5bfbdb9 in ?? ()
#96 0xb5bfbdb9 in ?? ()
#97 0xb5f1a466 in ?? ()
#98 0xb601629a in ?? ()
#99 0xb5fa2111 in ?? ()
#100 0xb5bf9249 in ?? ()
#101 0x0621c63d in JavaCalls::call_helper () from /usr/lib/jvm/java-6-sun-1.6.0.07/jre/bin/../lib/i386/client/libjvm.so
#102 0x063107b8 in os::os_exception_wrapper () from /usr/lib/jvm/java-6-sun-1.6.0.07/jre/bin/../lib/i386/client/libjvm.so
#103 0x0621c4d0 in JavaCalls::call () from /usr/lib/jvm/java-6-sun-1.6.0.07/jre/bin/../lib/i386/client/libjvm.so
#104 0x06245d67 in jni_invoke_nonstatic () from /usr/lib/jvm/java-6-sun-1.6.0.07/jre/bin/../lib/i386/client/libjvm.so
#105 0x0622a7aa in jni_CallIntMethodV () from /usr/lib/jvm/java-6-sun-1.6.0.07/jre/bin/../lib/i386/client/libjvm.so
#106 0xb48ca45f in callback () from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/configuration/org.eclipse.osgi/bundles/113/1/.cp/libswt-gtk-3449.so
#107 0xb486c295 in ?? ()
#108 0xb4cd38d4 in ?? () from /usr/lib/libgtk-x11-2.0.so.0
#109 0xb54aa749 in g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
#110 0xb54bef7b in ?? () from /usr/lib/libgobject-2.0.so.0
#111 0xb54c0937 in g_signal_emit_valist () from /usr/lib/libgobject-2.0.so.0
#112 0xb54c0f59 in g_signal_emit () from /usr/lib/libgobject-2.0.so.0
#113 0xb4df2667 in ?? () from /usr/lib/libgtk-x11-2.0.so.0
#114 0xb4e03df0 in gtk_window_propagate_key_event () from /usr/lib/libgtk-x11-2.0.so.0
#115 0xb4e06f5c in ?? () from /usr/lib/libgtk-x11-2.0.so.0
#116 0xb4cd38d4 in ?? () from /usr/lib/libgtk-x11-2.0.so.0
#117 0xb54a9069 in ?? () from /usr/lib/libgobject-2.0.so.0
#118 0xb54aa749 in g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
#119 0xb54bf10a in ?? () from /usr/lib/libgobject-2.0.so.0
#120 0xb54c0937 in g_signal_emit_valist () from /usr/lib/libgobject-2.0.so.0
#121 0xb54c0f59 in g_signal_emit () from /usr/lib/libgobject-2.0.so.0

Additional information: Ubuntu 8.04

Mantis Bugtracker #7

Reported by rhornig on 20 Dec 2008 17:45

Version: 4.0rc1

If you export a makefilegeneration makefile the IDE will writed a makefile which contain variables like $(CONFIGNAME), however these veriables must be double escaped (once for the makefile, once for the shell). I.e the correct format should be:
'$$(CONFIGNAME)' instaead of $(CONFIGNAME)

Also it seems that the include directories (-I) are not correctly exported into the makefile... (i.e. several include dirs are missing which are present in the IDE generated makefile) (see: mobilityfw2.0p3-o4 testSuite dir)

Mantis Bugtracker #40

Reported by mbergner on 8 Dec 2008 10:42

Version: 4.0b8

After applying todays updates (this may or may not be important), the IDE does not allow editing files in old projects anymore.

When starting the IDE with LC_ALL=C omnetpp (because of the floating point scalar thing, the error does appear also without LC_ALL=C), I immediately get an Internal Error during Initial Indexing, details reveal a NullPointerException.

I can not edit files in my custom simulations that have worked this morning. I disabled Indexing in the General preferences but the error is still there. I can open new Projects, work there flawlessly but once I close the IDE and reopen it, I can not edit files anymore. Clicking on the small arrow next to the Project to open it makes the arrow disappear but does not show any directories or files.

Running the simulations with Right-click run as and selecting the binaries and ini files works though.

When trying to edit currently open NED Files (C files can't be edited) the programm crashes with a double free and a backtrace which I will attach.

Opening and closing the IDE without doing anything gives a java backtrace which I will also attach.

This error currently renders the IDE useless.

Mantis Bugtracker #32

Reported by ViRe on 20 Jan 2009 10:17

Version: 4.0rc1

DEBUG: Break Point
There is something seems very strange.
When I run my simulation under DEBUG mode, and use "Run until..." to stop at somewhere.
Then, I can't set breakpoint in my code file.

But, If set the breakpoint before the launching of the simulation, the breakpoint will work.
And if I cancel the breakpoint at the first stop, us "Run until..." to stop at somewhere, then enable the brakpoint, it will work too.

What a strange thing!

Could it be a bug of OM4.0?

Mantis Bugtracker #46

Reported by baumgart on 26 Nov 2008 08:10

Version: 4.0b8

Received a SEGV after opening a ini file with the IDE:

#0 0xb7f91410 in __kernel_vsyscall ()
#1 0xb7e33085 in raise () from /lib/tls/i686/cmov/libc.so.6
#2 0xb7e34a01 in abort () from /lib/tls/i686/cmov/libc.so.6
#3 0xb7e6bb7c in ?? () from /lib/tls/i686/cmov/libc.so.6
#4 0xb7e73a85 in ?? () from /lib/tls/i686/cmov/libc.so.6
#5 0xb7e774f0 in free () from /lib/tls/i686/cmov/libc.so.6
#6 0xb5031b11 in operator delete () from /usr/lib/libstdc++.so.6
#7 0xb500d79d in std::string::_Rep::_M_destroy () from /usr/lib/libstdc++.so.6
#8 0xb500efd2 in std::string::reserve () from /usr/lib/libstdc++.so.6
#9 0xb500f0cf in std::string::append () from /usr/lib/libstdc++.so.6
#10 0xb215e205 in UnitConversion::readUnit ()
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9 959/libopplibs.so
#11 0xb215ee3d in UnitConversion::parseQuantity ()
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9 959/libopplibs.so
#12 0xb209f550 in createQuantityLiteral ()
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9 959/libopplibs.so
#13 0xb2069b2b in ?? ()
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9 959/libopplibs.so
#14 0xb207493a in ?? ()
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9 959/libopplibs.so
#15 0xb20756b3 in ned2yyparse ()
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9 959/libopplibs.so
#16 0xb2076893 in doParseNED2 ()
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9 959/libopplibs.so
#17 0xb20677ce in NEDParser::parseNED ()
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9 959/libopplibs.so
#18 0xb20678ee in NEDParser::parseNEDText ()
---Type <return> to continue, or q <return> to quit---
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9959/libopplibs.so
#19 0xb20679c5 in NEDParser::parseNEDExpression ()
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9959/libopplibs.so
#20 0xb1fd0a43 in Java_org_omnetpp_ned_engine_nedxmlJNI_NEDParser_1parseNEDExpression ()
from /home/baumgart/scalenet/svn/Entwicklung/omnetpp-4.0b8/ide/plugins/org.omnetpp.ide.nativelibs.linux.x86_4.0.0.0811141943-9959/libopplibs.so

Additional information: Ubuntu 8.04

Mantis Bugtracker #6

Reported by baumgart on 28 Nov 2008 11:07

Version: 4.0b8

opp_run -h config shows obsolete parameter perform-gc

Mantis Bugtracker #10

Reported by rhornig on 19 Dec 2008 19:34

Version: 4.0rc1

A rework is needed, how the -rpath linker option is handled in the kernel makefiles and in opp_makemake generated makefiles. The support for rpath option is detected in the configure script but opp_makemake adds -rpath to the linker command line even if it was not detected. This leads to link errors in queueinglibext sample and in any other project that depends on an other dynamically linked model. -rpath is not supported on the OS X 10.4 default compiler. Additionally -rpath is working differently on 10.5. where we should have specified the @rpath/libname as an install location to work it correctly. This means that it is not working correctly on 10.5 either (altough) does not cause any problem.

Correct behaviour wold be: completely eliminate it on windows and mac os x. on os x either a correct install path should be specified, or we should link against the libraries against their final location (i.e. omnetpp/lib and not omnetpp/lib/gcc)

Mantis Bugtracker #39

Reported by Korko on 22 Jan 2009 10:14

The code in the txc1.cc file on http://www.omnetpp.org/doc/tictoc-tutorial/part1.html is wrong. The error lies in line 23.
it is:
if (strcmp("tic", name()) == 0)

it should be rather:

if (strcmp("tic", getName()) == 0)

Regards,
Lukas

Mantis Bugtracker #49

Reported by baumgart on 10 Dec 2008 22:16

Version: 4.0b8

In this example the display string is not correctly merged:

moduleinterface Overlay
{
parameters:
@display("i=block/network2");
}

module SimpleOverlayHost
{
submodules:
overlay: <overlayType> like Overlay {
parameters:
@display("p=370,216");
}
}

The resulting displayString is "p=370,216" and no icon is shown in Tkenv. In the IDE NED editor the icon is shown correctly.

Mantis Bugtracker #35

Reported by horal on 12 Nov 2008 04:06

Version: 4.0b7

I don;t know if it is bug. When I finished my simulations I see problems with my statistics data.

Wrong file: invalid statistics data, file ...General-0.vci, line 69

Mantis Bugtracker #2

Reported by baumgart on 5 Dec 2008 16:40

Version: 4.0b8

If you start with an almost empty network and dynamically create a new module, this often gets placed way beyond windows borders, even if there is still a lot
of space available in the currently shown pane.

I haven't noticed this with the omnetpp-3.4b2 layout mechanism.

Mantis Bugtracker #25

Reported by baumgart on 13 Jan 2009 13:40

Version: 4.0rc1

truncnormal() sometimes returns infinity

With a propability of 2^32 genk_dblrand(rng) returns 0. In this case normal() returns infinity (log(0)).

This probably effects other distribution functions with use log() as well.

Mantis Bugtracker #43