Release Release v3.1.0 · iree-org/iree

Notable changes

The iree-import-onnx tool now includes options that improve handling of large models: #19217. See the ONNX framework documentation at https://iree.dev/guides/ml-frameworks/onnx/.
The PJRT plugin code at integrations/pjrt/ has been refreshed and is now tested continuously. Thank you @PragmaTwice !

Compiler support for "tuning" continues to improve. See the new documentation at https://iree.dev/reference/tuning/ and the discussion on #19214.
Support for ukernels in GPU targets has evolved:
- AMDGPU ukernels were ported from HIP to self-contained C in #19194.
- AMDGPU ukernels for select targets are now bundled with the compiler and are selected, loaded, and linked in during compilation thanks to #19274, #19323, and #19440. Externally-defined ukernels can also be provided: #19211.
Compilation time for large models has improved substantially with #19601.

There were breaking changes to some iree_hal_* APIs in #19160:
- iree_hal_buffer_subspan now requires an iree_allocator_t host_allocator that was previously implicit.
- iree_hal_subspan_buffer_initialize was removed as it was not safe.
- iree_hal_deferred_buffer_t was removed as placement checks without an allocated buffer reference are not possible.
- iree_hal_heap_buffer_wrap and all other buffer creation now requires a placement.
- iree_hal_buffer_initialize used by HAL implementations now requires a placement.
The HIP runtime can now create a single logical device backed by multiple physical devices: #18790, #19555.
Improved support for loading large programs in the HIP HAL: #19545, #19542.

The iree.build package now has reference documentation at https://iree-python-api.readthedocs.io/en/latest/compiler/build.html, has more robust HTTP fetching (#19330), and supports out of process concurrency (#19291).
Development versions of Python packages are now easier to install using build_tools/pkgci/setup_venv.py and a new package bisect script is available at build_tools/pkgci/bisect.