Chris Hesik

Thanks for reporting this. We'll take a look.

Yes, this isn't a usage model that we considered. We'll take a look to see what can be done to make this experience better.

I've just pushed a few changes to the CMake files. WIth these changes, we've been able to include GPA in another project by simply using `add_subdirectory`. There is still a...

It sounds like you forgot to execute the UpdateCommon.py script (as mentioned here: https://github.com/GPUOpen-Tools/GPA/blob/master/BUILD.md). Can you make sure you run that script before trying to build GPA?

It sounds like you forgot to execute the UpdateCommon.py script (as mentioned here: https://github.com/GPUOpen-Tools/GPA/blob/master/BUILD.md). Can you make sure you run that script before trying to build GPA?

We'll take a look at this, but in the meantime, can you try adding "-W-no-cast-function-type" to this line in the SConscript file, and see if that allows you to compile...

AMDTActivityLogger isn't configured to build by itself. It is built as part of the RCP build process. If you make the change I suggested in the clone the that is...

When collecting performance counters, the profiler will introduce serialization to try to ensure that only one kernel is executing at a time. There is no option for this, as it...

Serialization is only done when collecting performance counters (which is the mode you would use to analyze performance of individual kernels). No additional serialization is introduced when collecting a trace...

Use the "-k isa" switch (or "-k all") However, this is not currently supported for ROCm profiling (or for OpenCL profiling on the ROCm stack).