commit | c2776ae281a796dcd36dd7a799a655a8fd845254 | [log] [tgz] |
---|---|---|
author | Ben Vanik <ben.vanik@gmail.com> | Tue Jun 14 22:22:38 2022 -0700 |
committer | Ben Vanik <ben.vanik@gmail.com> | Mon Jun 27 08:19:40 2022 -0700 |
tree | 04708aeddf01fd9a9cca33e9b6314010290040dc | |
parent | 04788d1016bed3c111b182ff08d15eb69e67bc55 [diff] |
Adding support for wait frames in the VM stack. These special frames sit on the stack during a wait operation. The code that wants to wait will push a wait frame to the stack, populate it with the wait information, and return as deferred. The scheduler can peek at the wait frame and perform the wait before then resuming the waiting code, which can pop the wait frame and get back the wait status. Since we ref count everything on the stack it's safe to yield back to the scheduler with wait sources dependent on program resources.
IREE (Intermediate Representation Execution Environment, pronounced as “eerie”) is an MLIR-based end-to-end compiler and runtime that lowers Machine Learning (ML) models to a unified IR that scales up to meet the needs of the datacenter and down to satisfy the constraints and special considerations of mobile and edge deployments.
See our website for project details, user guides, and instructions on building from source.
IREE is still in its early phase. We have settled down on the overarching infrastructure and are actively improving various software components as well as project logistics. It is still quite far from ready for everyday use and is made available without any support at the moment. With that said, we welcome any kind of feedback on any communication channels!
See our website for more information.
IREE is licensed under the terms of the Apache 2.0 License with LLVM Exceptions. See LICENSE for more information.