[support]

Hi,

This is somewhat of a limitation of the way VisualGDB stores the remote directories. As they can be different for different hosts anyway, we would recommend opening the Makefile Settings page of VisualGDB Project Properties, changing something there (e.g. adding and removing space to CFLAGS) and pressing ‘Apply’. This will automatically redetect and re-configure the IntelliSense directories.

You can also try our new MSBuild backend in VisualGDB 5.2 (currently in Beta). It has a more automated way of dealing with remote include directories (you will still need to to click “repair MSBuild settings” after you change the alias target).

[support]

Looks like the Raspicam library is picking up the Windows build of OpenCV.
If you want to use it with OpenCV, please consider following this tutorial for build it for Raspberry Pi. If not, please edit the CMakeLists.txt of the raspicam library to remove references to OpenCV.

The step 22 looks different because you have created your main project with CMake, not GNU Make. Both should work though, so it should not be a problem.

[support]

Hi,

Looks like your library somehow ended up missing those method definitions. According to your build log, your application is linked against the library:

g++ -o LinuxConApp/RevLibConApp -Wl,-gc-sections -Wl,--start-group LinuxConApp/main.o LinuxConApp/ProgressMonitor.o LinuxConApp/RevLibTest.o ../RevLib/wrk/LinuxConApp/RevLib.so -Wl,--rpath='$ORIGIN' -Wl,--rpath='$ORIGIN/../../RevLib/wrk/LinuxConApp/' -Wl,--end-group

You can check the list of functions exported from a library by running the following command on the Linux side:

nm <full path to .so file> | c++filt

If it does not mention anything similar to RevLib::GetRevLibObjectInterface(), please locate the .cpp file that defines it, find a matching .o file on your Linux side and run ‘nm’ on it. If the .o file does not contain the method, please double-check your source code. If the .o file does not exist, please double-check the Library Makefile.

[support]

Hi,

Normally all you need to do is add a reference to the .so project in the properties of your main project (right-click in Solution Explorer, Add Reference).

[support]

Hi,

Yes, you can always download the older BSPs here: http://visualgdb.com/hwsupport/

Normally though a BSP upgrade should not affect stand-alone projects (you can try simply renaming the BSP directory and confirming that the project still builds), so perhaps the upgrade coincided with something else?

[support]

Hi,

Thanks very much for sharing this, it is indeed a VisualGDB bug. We have fixed it in this build: http://sysprogs.com/files/tmp/VisualGDB-5.2.8.1132.msi

Please open your VisualGDB Project Properties and change the Include Directories field arbitrarily (e.g. add and remove a space), then press OK. VisualGDB will then update the CMakeLists.txt correctly.

[support]

Hi,

According to your log, there are 2 definitions of the posix_memalign() function:

1. C:\Users\akhils\AppData\Local\VisualGDB\RemoteSourceCache\192.168.0.54\0001\include\mm_malloc.h(34,64)
2. C:\Users\akhils\AppData\Local\VisualGDB\RemoteSourceCache\192.168.0.54\0000\include\stdlib.h(503,12)

This normally means that your system is too old and does not support C++11 properly. You can try patching the headers to have matching definitions at your own risk, but you may encounter further problems with C++11 on that system.

[support]

Hi,

This one can be safely ignored and should be solved in the upcoming Beta 1.

If everything else works, we would recommend simply updating to Beta 1 once it is released.

[support]

Hi,

The “127.0.0.1” address always refers to the “current machine”, so if you try to connect to 127.0.0.1 from VisualGDB, it will try to connect to the Windows machine itself and will not work. Please use the “ifconfig” command on the Linux machine to find our its actual IP address and use it instead. Please also ensure that your VM network settings are configured in a way that allows the two machine to communicate with each other (e.g. bridged connection to your local network).

[support]

Hi,

Oops, his looks like a bug. Does the C:/root-directory-of-application/=/opt/libName/include path appear on the command line, or does it only appear in the verbose GCC output? Perhaps some other setting or a small typo somewhere is interfering with it? Could you share the entire build log showing the command lines and the diagnostic output from gcc?

[support]

Please try downloading the latest version of msp430.dll from TI and replace the one in the msp430 toolchain directory with it.

We also recommend updating the firmware on your USB programmer to the latest version as sometimes old firmware causes strange problems.

[support]

Hi,

You can build the OpenOCD fork that we use from the sources here: https://github.com/sysprogs/openocd

You can use this MinGW environment that has all the necessary libraries: http://sysprogs.com/files/tmp/MinGW.7z

Simply search the source code for the “Cannot identify target as a STM32L4 family” line and you will see the table that checks the device IDs and selects the FLASH size based on them. Adding your device there should solve the problem.

[support]

Hi,

This happens because the latest SDK from ST expects your projects to define this variable in the HAL configuration file (e.g. stm32f4xx_hal_conf.h). As the file is a part of your project and may have been modified, VisualGDB does not update it automatically. The easiest solution would be to simply create a new project using the new BSP and copy the default LSE_STARTUP_TIMEOUT definition from it.

[support]

We have answered the question about the memalign() function in this thread.

Regarding the <vector> file, can you confirm that it is physically present somewhere under C:\Users\akhils\AppData\Local\VisualGDB\RemoteSourceCache\192.168.0.54? You can also try reloading the cached directories via VisualGDB Project Properties to ensure you get the latest versions of all files.

[support]

Hi,

Looks like your system headers may be too old and don’t support C++11 properly. Please examine the build output and try to understand why they have 2 different definitions of posix_memalign(). Once you locate both definitions, you can try modifying them to ensure they both match.

[Author]

Posts