| commit | c49f08c1b9f28db4444aae6fbb3a6a35866a6d1f | [log] [tgz] |
|---|---|---|
| author | Geoffrey Martin-Noble <gcmn@google.com> | Wed Mar 08 09:32:07 2023 -0800 |
| committer | GitHub <noreply@github.com> | Wed Mar 08 09:32:07 2023 -0800 |
| tree | a96f4a412fbaf733e77431f0263ce67be3a4e91f | |
| parent | d440aa021928b00b319ea11a7ddd2aa868d7fd84 [diff] |
Add `-ftrivial-auto-var-init=pattern` to ASAN build (#12447) This option initializes automatic variables with a recognizable pattern likely to cause helpful crashes rather than silent undefined behavior. See the extensive description of this option to the patch that first introduced it to clang: https://reviews.llvm.org/rG14daa20be1ad89639ec209d969232d19cf698845 This isn't technically asan, but it guards against the same type of memory errors. It's also much lower overhead, so we could consider enabling it in more builds. For release builds we might consider `-ftrivial-auto-var-init=zero`, which is more likely to mask bugs but less likely to access uninitialized memory at runtime, resulting in security vulnerabilities. For now, turning it on in one build is a good step. Example failure (now fixed): https://github.com/openxla/iree/actions/runs/4308619652/jobs/7515099139 Thanks to @Maximus- for introducing us to this option and flagging the above failure.
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.