Outline upcoming changes to the VFIO infrastructure, including: - Deprecation of group-based APIs in favor of container device assignment - Removal of separate no-IOMMU checks in favor of unified VFIO "modes" - Elimination of group-based functions from the vDPA driver API
Signed-off-by: Anatoly Burakov <[email protected]> --- doc/guides/rel_notes/deprecation.rst | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+) diff --git a/doc/guides/rel_notes/deprecation.rst b/doc/guides/rel_notes/deprecation.rst index 1e60ac4e82..f54a0fe89f 100644 --- a/doc/guides/rel_notes/deprecation.rst +++ b/doc/guides/rel_notes/deprecation.rst @@ -158,3 +158,29 @@ Deprecation Notices without relying on flow API. Currently there is no alternative API providing the same functionality as with ``repr_matching_en`` set to 0. + +* eal: Several VFIO APIs have changed their signatures and/or return values to improve error reporting and consistency: + - ``rte_vfio_setup_device`` and ``rte_vfio_release_device`` now return 0 on success and <0 on failure, with ``rte_errno`` set. + + Previously, if a device was not managed by VFIO (e.g., not bound to the VFIO driver), + ``rte_vfio_setup_device`` would return 1 to indicate the device should be skipped. + Now, this is treated as an error: the function returns -1 and sets ``rte_errno`` to ``ENODEV``. + + - ``rte_vfio_get_device_info`` now takes a device file descriptor and device info pointer, + and returns 0 on success, <0 on failure. Previously, the function was accepting a path to device, and would set the device up if it wasn't set up. + It will no longer do so, and users of this function must call ``rte_vfio_setup_device`` prior to using this function. + + - VFIO API has been updated with new error codes and return values, all error codes and return values are now documented in the API headers. + + Applications using these APIs should review the updated function signatures and error handling. + +* eal: The following legacy VFIO group-related APIs are deprecated and will be removed in a future release: + ``rte_vfio_clear_group``, ``rte_vfio_get_group_fd``, ``rte_vfio_container_group_bind``, and ``rte_vfio_container_group_unbind``. + These APIs are replaced by the new container device assign lifecycle: + ``rte_vfio_container_assign_device`` → ``rte_vfio_setup_device`` → ``rte_vfio_release_device``. + Users should no longer need to manage IOMMU groups directly; all group management is now handled internally. + +* vdpa: vDPA drivers have been updated to use the new VFIO device assignment APIs. + The deprecated VFIO group-based APIs (``rte_vfio_get_group_num``, ``rte_vfio_container_group_bind``, ``rte_vfio_container_group_unbind``) + are no longer used. vDPA drivers now use ``rte_vfio_container_assign_device`` instead, and no longer need to track VFIO group file descriptors. + This change aligns vDPA drivers with the new VFIO API that abstracts away IOMMU group management. -- 2.47.3
