Wind River Debugging - No such sourcefile

[djdaugherty]

The team is using Wind River with C++ for our software. We are able to download the software to cRIO and control some features. When trying to debug we are following the C/C++ Getting Started Guide and have setup the Target Server as required and see the target server entry populated with tasks. When we follow the debugging instructions we get the Debug Perspective and appear to be able to set a breakpoint. When we select "resume" we get the following error:

Error Message = No such source file position in current context...possibly caused by source not part of executable or wrong path mapping setup Back End ID=6.

Also, can someone point us to the referenced Wind River Workbench User's Guide? Thanks!

[byteit101]

I haven't tried the debugger, but the Windriver manual is http://first.wpi.edu/Images/CMS/Firs...sers_Guide.pdf and http://first.wpi.edu/Images/CMS/Firs...artedWithC.pdf

[gvarndell]

Originally Posted by djdaugherty

we get the Debug Perspective and appear to be able to set a breakpoint. When we select "resume" we get the following error:

Error Message = No such source file position in current context...possibly caused by source not part of executable or wrong path mapping setup Back End ID=6.

This means the source file you were looking at when you set the breakpoint is not the source file from which the running binary code was built.

[Trump]

Originally Posted by gvarndell

This means the source file you were looking at when you set the breakpoint is not the source file from which the running binary code was built.

We had a feeling that this was what was happening. Does anyone have any ideas on what we are doing wrong, or how to fix this issue?
Thanks.

[byteit101]

You could try rebuilding. Right click the project>rebuild and then Debug Kernel Task

[Trump]

Do we have to undeploy the robot code first, should the crio be in a certain mode? We have already cleaned, and built the code several times... Does rebuilding do something other than this?

[Trump]

Heres something else I've found. I deleted all the binary files and the build targets. Then i remade them. Now i am able to debug, but only if i use the step over/ step into commands.

If i press the resume button the process terminates and says <terminated, exit value: 0>tFRC_UserProgram_StartupLibraryInit : 0x1e92678

As a note, when i press the step over/into buttons, it shows only the assembly code.

[byteit101]

Originally Posted by Trump

Do we have to undeploy the robot code first, should the crio be in a certain mode? We have already cleaned, and built the code several times... Does rebuilding do something other than this?

no, at least not to my knowledge
did you configure the debug session? Page 12 of http://first.wpi.edu/Images/CMS/Firs...artedWithC.pdf

[Trump]

Yes, we have set up everything the way tha the manual says to, but we still can't get it to work. We ended just giveing up on the debug, and we are now useing the printf over serial to debug. Not the best of situations, but we'll have to deal.

BTW: I am just now updateing labview and the driversconsole to the version released today (feb 15).

Also, I saw somewher that you need the driverstation connected to debug?? if so, how do you connect the DS and the proggramming computer at the same time?

Thanks

[byteit101]

Originally Posted by Trump

Also, I saw somewher that you need the driverstation connected to debug?? if so, how do you connect the DS and the proggramming computer at the same time?

Thanks

If you can only connect one at a time, you have a problem somewhere. Why can't you connect both?

[Trump]

Im so confused right now. Anyways, the reason we can't connect both is basically because I don't know how to go about connecting both. Do i just plug both into the crio with a Cat5 cable?? I think i set everything up according to the manuals, if not could someone point me to the right page? (i've seen all the manuals that people have already linked to, so If im missing something its because i didn't see it in there.)

Thanks Byteit and others that are helping.

[Mike Mahar]

Originally Posted by Trump

Im so confused right now. Anyways, the reason we can't connect both is basically because I don't know how to go about connecting both. Do i just plug both into the crio with a Cat5 cable?? I think i set everything up according to the manuals, if not could someone point me to the right page? (i've seen all the manuals that people have already linked to, so If im missing something its because i didn't see it in there.)

Thanks Byteit and others that are helping.

You were given a wireless router as part of your kit of parts. To connect the cRio and the developer PC and the Driver station at the same time, plug them all into the router. You can use the wireless gaming adapter for the robot and then you don't have to plug that in. I would plug in the cRio with a cable to the router first, however.

[Trump]

Mike,
WOW, i totally forgot about the wireless . up until now, we have been doing everything wired... I'll try that at tonights meeting. I still have one question. When i debug the program, are the steps to click on the debug kernal task, then enable the robot from the DS? I finally feel like we're getting somewhere. We havn't really been able to get the code to work correctly without debug.

[byteit101]

yes, you have to click enable on the DS just after the debug starts. I would set breakpoints beforehand, so you can skip to your code.

Entry point
start debug
underlying WPLIB code
DS enable
Your code

PS. Debug is not as easy as Visual studio's Debug

[Trump]

Still not working . When i press debug kernel task, and then press resume it terminates almost instantly. I don't have enough time to go to the DS and press enable code.

However, i can't press enable code on the DS before i press resume, because it says No Code.

I have the entry point set to Frc_User_library_init, Is this correct??

I really don't know what to do now, without debug we really can't get much done. I spent almost 6 hours today trying to find errors in the code with just printf's coming through the null modem cable, but it really isn't that helpful.

If i'm missing something, please tell me, i think i've seen all the manuals related to Debug, but i might be wrong. If anyone can think of anything No matter how obvious it might seem, please bring it to my attention

Thanks