I'm working in Eclipse on Windows, using mingw for the compiling environment. My project has two parts: a library and an executable, both in separate directories.
I want to debug the shared library that I've created, but when I tell Eclipse to load the program for debugging, it just hangs forever. However, if I manually place the .dll file into the directory with the program, debugging proceeds normally.
I can certainly compile the static version of the library, which I'm doing now so as to be able to complete the project on time, but it offends my sensibilities that I can't get it to work right dynamically.
Any ideas?
Try to add the path to your DLL file to the environement variable "PATH"?
To add to environment variable:
Right click "My Computer" -> Property -> Advanced -> environment variable
Related
Trying to manage post-build steps in Eclipse CDT project. Trying to copy newly builded file to directory ~/destination .
I could do following:
cp ${BuildArtifactFileBaseName}.so ~/destination
But I would like to escape from hardcoded parts there. For this reason I need to know:
Artifact extension Eclipse variable
Custom defined destination path variable
Where I can define custom variable for ~/destination ?
Where I can find whole list of Eclipse variables?
Will they be different for Java and C++ projects?
The Eclipse CDT build variables can be found at Window -> Preferences -> C/C++ -> Build ->Build Variables.
To see all variables, check the Show system variables checkbox.
Use the Add button on the Build Variables screen to add a new variable.
I don't have experience with Java on Eclipse, but probably there's an equivalent menu when compiling Java.
I installed Eclipse for Parallel Applications and MPI v. 2.
I created inside Eclipse a new Hello World MPI project and run it using 4 processes. I can build and run it from Eclipse, it works.
Eclipse also recognizes MPI artifacts correctly.
I configured the includes as described here:
http://help.eclipse.org/juno/topic/org.eclipse.ptp.pldt.doc.user/html/includes.html
However I'm getting a lot of errors on MPI artifact on the source code, same errors as described in the above link:
The Indexer can't find your include file
"If the indexer can't find your MPI header file, you may see problems like the following:
If your MPI header file is located in a global area that automatically gets included in your project's include path location, such as /usr/include or /usr/local/include, then things will probably work just fine without further alteration. Building and indexing will probably both operate correctly.
If you do not use mpicc or a similar build command and need to specify include and libpath information on building and linking, you will probably want to set the PLDT preferences to point to your include path for MPI, in which case, the include and link arguments will be added to your build commands."
My code works but I think it is impossible to work with al these red lines on every MPI instruction.
So my question is: how can I configure Eclipse in order to recognize MPI functions without telling me that there are errors?
Using Ubuntu 64 bit with last version of MPI and Eclipse.
I found a fix:
Right click on project -> Properties -> C/C++ General -> Paths and Symbols
On tab "Includes" click on "GNU C" and then on "Add.."
"/usr/include/mpi"
rebuild
It sounds like Eclipse does not include directories recursively.
Source: http://tinyurl.com/nkq2gwp [pages 35-36]
The link you posted seems to have all the answers that you're going to get here. Are you sure that your "mpi.h" header file is in the location you put in your include path in Eclipse? It seems that you may have the wrong path.
I am working on an eclipse RCP product on windows which internally uses some native binaries (dlls) located in a folder called "bin" which would be available as a part of the product.
I need to make sure that the folder "bin" is present in the PATH when the product.exe file is executed. Is there a clean way to make sure that the folder "bin" is added to PATH when the eclipse RCP product is launched?
I don't want to write any script or batch file that updates the PATH and then calls the product.exe. I also don't want to update the environment variable through the windows UI or command prompt.
I know that when we try to run my product from the eclipse IDE environment, there is an option in the run-configuration, wherein one can configure Environment variables on the Environment tab. I can append the "bin" folder to the PATH by creating a new environment variable called "Path" whose value is C:\EclipseRCPProduct\bin;${env_var:PATH}.
However - I would like to do the same when I run my product directly from the Product.exe which is generated. Does anyone have any idea regarding this?
Also - if anyone has any idea regarding the eclipse code which processes the information specified on the Environment tab of the run configuration, it would be great. I could try playing around with the code to see if I could get it to work when the product is launched from outside eclipse IDE - via the generated RCP product.exe
I don't think there's a way to define system environment variables within the launcher.
The following bug seems to request just that
https://bugs.eclipse.org/bugs/show_bug.cgi?id=286463
In the above mentioned bug, a .bat is attached to achieve your goal.
That said, why do the native binaries have to be located in a bin folder?
I.e. why are they not included in a plugin/fragment?
I am a beginner in using Eclipse and PyDev (Aptana Studio 3). I am not used to and i don't understand the workflow in such big IDEs as Eclipse.
I have a simple task: i have a simple Python script, which i want to open and run in Eclipse, having its output in Eclipse console. Or debug it.
Until now i used another IDE called Eric4, which allowed me to do what i want - open a file and run immediately, without creating a project or setting up launch configurations.
Is this possible in Eclipse, or i have to create a project for each file i want to run or debug? I want to understand how it works.
I guess i understand that creating a project is needed at least for settings up the paths (PYTHONPATH), but if it's a single script - somehow to use by default the current directory?
For example i have a folder called snippets where i keep a lot of python scripts which demonstrate some functionality. How do i open these files one by one and run them?
Most of my coworkers launch python scripts in a separate console - python my_scipt.py.
You need to have at least one project with the configuration you want (i.e.: syntax type, interpreter), then, open the file you want to run and press F9.
If it's an external file -- i.e.: a file that's not under a project in Eclipse -- it'll ask you to associate a project with the launch to get the needed information for the launch, but the file doesn't really have to be in the project (note that you can drag external files from your filesystem into Eclipse to open them).
I suggest you follow the steps on the getting started: http://pydev.org/manual_101_root.html (it guides you to configuring PyDev and explains how to do a run/debug session).
I've downloaded C/C++ libraries from Cygwin, and set the environment variables according to the instructions here, in NetBeans.
I've also gone through this.
Actually I followed the same steps on a different machine and everything worked out fine.
But on my machine the associations aren't made.
for example: #include<stdio.h> says No such file or directory.
Any idea what the problem might be?
The Cygwin package I downloaded is also fine, I downloaded it twice.
If it's complaining about a #include, that's a compile issue, not a linking issue.
What happens if you create a new C/C++ application project and try to build it?
In the Build node of the project properties, is the correct Tool Collection selected? (Cygwin in your case). And the Tools -> Options, on the C/C++ tab, is that tool collection set up correctly?
You should check whether you have a file c:\cygwin\usr\include\stdio.h.