Forum Replies Created
-
AuthorPosts
-
supportKeymaster
OK, please try this version http://sysprogs.com/files/tmp/SeggerEDP.dll
It will also show the data received from JLINKARM_ReadMemEx().
supportKeymasterHi,
Thanks for checking this. Could you please try this diagnostic build: http://sysprogs.com/files/tmp/VisualGDB-5.2.4.880.msi
Then replace the %LOCALAPPDATA%\VisualGDB\EmbeddedDebugPackages\com.sysprogs.arm.segger-dmsp\SeggerEDP.dll file with this one: http://sysprogs.com/files/tmp/SeggerEDP.dll
Finally start your debug session and open the View->Other Windows->VisualGDB Diagnostics Console window. It should show what exactly the Segger memory reading API returns. Please let us know what is shown there so that we could fix this.
supportKeymasterLooks like you are still trying to use your Makefile with i586-poky-linux-gcc.exe.
In order to isolate the problem please try using the minimal Makefile we provided (with the normal GCC from the MinGW toolchain and make.exe from MinGW as well). If it works, please try explicitly using the make.exe from the MinGW toolchain instead of make.exe from the i586-poky-linux toolchain.
supportKeymasterHi,
Please update your J-Link EDP to v3.2 via Tools->Embedded Tools Manager. This should fix the problem.
supportKeymasterStrange. Could you please archive the project and send it to us via our support form? If we can reproduce it on our side, we’ll fix it very quickly.
supportKeymasterHi,
OK, please try the following Makefile from the command line:
test: gcc -DTIMESTAMP=$(shell C:/mingw/msys/1.0/bin/date.exe +%s) -E test.c -o test.E
You can use it with the following test.c file:
int timestamp = TIMESTAMP;
If this does not work, please let us know the error you get with this simple file. If it does, please compare the syntax here with your Makefile.
supportKeymasterHi,
VisualGDB does not normally *overwrite* the Makefile, it simply edits the SOURCEFILES line and a few other relevant lines, but only if it finds special tags in the file. Please let us know more details on what exactly is broken and we could suggest how to fix it.
supportKeymasterHi,
We would recommend exporting it as GccARM, unpacking the archive to any location within your source control system and then simply using the “Import Folder Recursively” command to import those files into your project.
If you run into problems with missing include files, you can try a daily build of VisualGDB 5.2 that has experimental support for repairing them automatically. Let us know if you want to try it out and we’ll post a link to it.
supportKeymasterHi,
This happens because the FreeRTOS port hardcodes the ISR vector table name variable name as ‘__isr_vector’ and the VisualGDB startup file uses a different name.
The easiest way to solve this is to replace this (in port.c):
" ldr r0, =__isr_vector \n" /* Locate the stack using __isr_vector table. */ " ldr r0, [r0] \n"
with that:
" ldr r0, =_estack \n"
Note that the “ldr ro, [r0]” line will be removed, as _estack already contains the end-of-stack address.
Let us know if you encounter further problems.
supportKeymasterHi,
Thanks for pointing this out, we will fix it in the next release of the Tiva BSP.
supportKeymasterHi,
By default mbed provides its own implementation of the _write() function that is used when you call printf().
You can replace it by a VisualGDB-provided one that will print to the Visual Studio window in 2 steps:
- Remove the retarget.cpp file from your project
- On the Embedded Frameworks page add reference to the Fast Semihosting and Profiler framework
Let us know if you encounter any problems.
supportKeymasterHi,
Please try this syntax:
$(shell C:/date.exe +%s)
supportKeymasterHi,
Normally the live variable type is saved in the <Project>-<Configuration>.vgdbsettings.<username>.user file. We have just tried creating some variables and changing their type and the type was remembered correctly.
Perhaps something about your project is confusing the live variables logic? Could you share precise repro steps that could be reproduced on a basic “LEDBlink” project?
June 17, 2016 at 21:14 in reply to: Clang-based Intillisense engine "Insert spaces" doesn't work #8446supportKeymasterHi,
Could you share a sample project that always causes the problem? We should be able to fix it easily then.
supportKeymasterHi,
We are considering it for the next update of our nRF52 BSP, but it’s not there yet. However we can easily help you get through the errors you are encountering if you share more details.
-
AuthorPosts