[support]

Please try this build: http://sysprogs.com/files/tmp/VisualGDB-5.2.15.1406.msi

We have added an option under Tools->Options->VisualGDB->Embedded->Other that controls whether VisualGDB stops on unsupported syscalls.

[support]

Hi,

Yes, spaces in toolchain paths can cause strange problems, so we generally recommend avoiding them.

[support]

Hi,

Thanks, looks like our bug. Please try this build: http://sysprogs.com/files/tmp/VisualGDB-5.2.15.1403.msi

[support]

Hi,

The Makefile-based projects are designed to be built outside VisualGDB as well, so they don’t support using VisualGDB variables in settings that are stored in Makefiles. Please use relative paths (e.g. ../MathLibrary/Debug) instead.

[support]

Hi,

The Keil toolchain output should not bypass the regex filters, but it could use some special unicode characters that could interfere with the regexes.

Please try running “make > build.log” from the command line and then double-check using a hex editor that the lines generated by Keil look exactly the same as the lines printed from the Makefile. You can also send the log file to our support email so that we could check it for you (please attach it instead of pasting to avoid automatic character transformation).

We have not tested it with Keil as we don’t officially support it out-of-the-box yet, sorry.

[support]

Hi,

Thanks for letting us know. We will consider making the button state more explicit for the dark theme.

[support]

Hi,

This could indicate a corruption of the Clang IntelliSense database files. Please try closing the project and removing the CodeDB folders under the the project directory.

[support]

Hi,

This looks like an unhandled semihosting call (mechanism used by the firmware to send printf() output to the debugger). Please ensure that you have semihosting enabled via VisualGDB Project Properties -> Debug. Either way, this message can be just ignored as well as it does not indicate an error.

[support]

Hi,

JAYLINK_ERR_DEV might indicate a driver issue. Are you able to use J-Link with OpenOCD (not Segger gdb stub) to debug normal ARM devices? If yes, it could be wiring problem instead.

[support]

Hi,

Thanks, we are aware of this. As those warning don’t really have any side effects and don’t appear unless the extra warnings are enabled, we recommend simply suppressing them via per-file settings.

If you don’t want to do that, you can send us the patch adding the necessary attributes and we will include it in our release branch, but we don’t have any plans of addressing them on our side, as we prioritize other features over it. Sorry.

[support]

Hi,

In order to keep VisualGDB price competitive, we only directly support the most popular devices and offer various customization mechanisms to allow our customers to easily add support for the devices that are not directly supported.

The easiest way to create a project for an unsupported device is to follow our legacy device tutorial.

Our BSP generators on github have a very basic Wiki page that outlines the basic concepts, but generating a full-blown BSP would much harder than just setting up a manual project (and exporting it to a template).

We also provide a service for creating BSPs for rare devices on demand, so if you don’t want to go through this process yourself, feel free to contact our sales to get a quote for a custom BSP.

[support]

Hi,

Sorry, that was Ctrl-Space, not Ctrl-Tab.

[support]

Hi,

Please try pressing ‘Ctrl-Tab’ while typing your code to activate IntelliSense popup explicitly. If it helps, please click the lightning icon at the bottom of the popup to enable the “show suggestions as you type” mode.

[support]

Hi,

OK, we’ve tried reproducing this with the following extra line in Makefile:

ifeq ($(TARGETTYPE),APP)
$(BINARYDIR)/$(TARGETNAME): $(all_objs) $(EXTERNAL_LIBS)
    $(LD) -o $@ $(LDFLAGS) $(START_GROUP) $(all_objs) $(LIBRARY_LDFLAGS) $(END_GROUP)
    @echo "LEDBlink.cpp", line 10 : warning : #177-D: function "naughty_function"  was declared but never referenced
endif

The vgdbsettings file contains the following element:

  <BuildMessageTemplateFile>BuildMessageRegexes.xml</BuildMessageTemplateFile>

The BuildMessageRegexes.xml defines GCCMessage as follows:

  <GCCMessage>"(.*)", line ([0-9]+) : (error|warning) : (.*)$</GCCMessage>

This does actually produce the expected results (the message is converted to VS format in the Build Output window and VS recognizes it):

1>e:\projects\temp\EmbeddedProject19\LEDBlink.cpp(10): warning : #177-D: function "naughty_function"  was declared but never referenced

Please try checking your code against the examples shown here (note that this forum engine converts the ” character to a slightly different one and it won’t work if you simply copy-paste it). If it does not work, please try reducing the problem to a GCC project with an ‘echo’ line injected into the Makefile and send us the entire repro project so that we could help you troubleshoot this.

[support]

Hi,

The solution you have found should work. You can also override the _estack symbol name by editing the .vgdbsettings file (search for the ‘EndOfStackSymbol’ element).

[Author]

Posts