Skip to content

v0.22.0 — Ubuntu 26.04 / ROS 2 Lyrical / OptiX 9.x

Latest
Latest

Choose a tag to compare

View all tags
@doisyg doisyg released this 29 May 22:24
v0.22.0
09644d0

Ubuntu 26.04 / ROS 2 Lyrical / OptiX 8.x–9.x support

Added

  • Support for Ubuntu 26.04 and ROS 2 Lyrical (Kilted also accepted), alongside Humble and Jazzy
  • Support for building against NVIDIA OptiX 8.x and 9.x (previously 7.2 only) — optixModuleCreate / removed OptixPipelineLinkOptions::debugLevel guarded by OPTIX_VERSION, so OptiX 7.2 still builds

Changed

  • setup.py auto-selects a compatible CUDA host compiler on Linux when the system default is too new for the installed CUDA toolkit (e.g. GCC 15 on Ubuntu 26.04 with CUDA 12.x), keeping the system libstdc++ at link time. No-op on Ubuntu 22/24 or CUDA 13+.

The prebuilt artifact is built against NVIDIA OptiX 9.0 (OPTIX_VERSION 90000, function-table ABI 105). Verified on Ubuntu 26.04 + CUDA 12.4 + GCC 13: clean core ./setup.py build, 395/396 core tests pass on driver R595 (1 pre-existing skip needing RGL-blobs data, 0 failures). The source also builds with --with-ros2 and against OptiX 9.1.

ℹ️ The prebuilt libRobotecGPULidar.so is built against NVIDIA OptiX 9.0, so it loads on NVIDIA driver R570+. Building against OptiX 9.1 instead requires R590+.

Release artifact

libRobotecGPULidar.so (and RGL-core-linux-x64.zip) is the core build — no PCL/ROS2 extensions — built against OptiX 9.0 and linked with RUNPATH=$ORIGIN. It is fully relocatable and depends only on the NVIDIA driver / CUDA runtime, which is what RGLGazeboPlugin consumes. For RGL's ROS2/PCL features, build from source (e.g. ./setup.py --with-ros2-standalone --with-pcl-system --lib-rpath '$ORIGIN').

Assets 4
Loading