Philipp van Kempen

@lasseUrban the current default architecture should be `RV32IMACFD` as `RISCV` is quite outdated and does not pass the RISC-V compliance tests. Also, the CSR-handling should now be generated automatically using...

This patch seems to disable the CI/CD jobs for Windows and MacOS. If this is intended we have to agree on the future of those, so would you please provide...

I checked which of the options in `base.ini` and `ETISS.ini` are actually used in the project and found the following "leftovers": ``` testing = false ETISS::CPU_quantum_ps=100000 ETISS::write_pc_trace_from_time_us=0 ETISS::write_pc_trace_until_time_us=3000000 ETISS::sim_mode=0 vp::simulation_time_us=20000000...

@DanMueGri Thanks for your input. Do we want to have the SystemC-related options in the ETISS-repo? For the bare_etiss_processor where there are currently mentioned, the will not have an effect...

> What effect did the non-const `.rodata` have, and can we also add a test case for it? I got the following linker warnings when building using a recent RISC-V...

@Lunderberg CI passed. Can we merge this?

I uploaded the model to reproduce the bug [here]( https://github.com/PhilippvK/tensorflow-operators-cmake/releases/tag/v0.1) because github does not allow attaching `.tflite` files in comments: Direct download link: https://github.com/PhilippvK/tensorflow-operators-cmake/releases/download/v0.1/model_with_multiple_outputs.tflite Model graph: 

Ooops, it is actually not a problem of multi-output models but rather multi-input ones. For the given graph, `TVMWrap_GetNumInputs` seems to return 6 instead of 3. The generated `inArgInfo` looks...

Rebased on master

After the latest commit, building etiss plugins can be done without making any changes to the cmakefiles of the plugin: Just `cmake -DETISS_DIR=/path/to/etiss/installation/lib/CMake/ETISS ..` and you should be good to...