PRRT merge requestshttps://git.nt.uni-saarland.de/LARN/PRRT/-/merge_requests2023-09-25T13:58:14Zhttps://git.nt.uni-saarland.de/LARN/PRRT/-/merge_requests/8Feature/improved deepsharq building2023-09-25T13:58:14ZMoritz MiodekFeature/improved deepsharq building## Changes:
Add `setup.sh` to setup dependencies including recursively setting up full-search (streamline dependency installation).
Let `cmake` build `full-search` and symbolically link the generated dependencies.
We use the `libt...## Changes:
Add `setup.sh` to setup dependencies including recursively setting up full-search (streamline dependency installation).
Let `cmake` build `full-search` and symbolically link the generated dependencies.
We use the `libtensorflowlite_c.so` generated from compiling the [tflitec crate](https://crates.io/crates/tflitec) instead of compiling it ourselves.
Generic build:
1. `./setup.sh`
2. `cd build; cmake ..`
3. `make`
### Handling offline cpu cores
In some virtualized environments (i.e. LXC containers) cpu cores can be online or offline. PRRT already uses the optional `PRRT_CORES` environment variable that addresses this issue for thread pinning the sender and receiver threads. Use the same mechanism for the timer threads. This change preserves the default core assignment:
- core 0 -> timer threads
- core 1 -> sender thread
- core 2 -> receiver thread
Quick `PRRT_CORES` setup: `export PRRT_CORES=$(cat /sys/devices/system/cpu/online)`Marlene BöhmerMarlene Böhmerhttps://git.nt.uni-saarland.de/LARN/PRRT/-/merge_requests/1WIP: coding params python binding2018-06-25T20:24:08ZKai VogelgesangWIP: coding params python bindingAdd Python bindings for reading the socket's current coding configurationAdd Python bindings for reading the socket's current coding configuration