diff options
author | Sang Ik Lee <sang.ik.lee@intel.com> | 2025-07-04 05:51:21 -0700 |
---|---|---|
committer | GitHub <noreply@github.com> | 2025-07-04 13:51:21 +0100 |
commit | b9b2661f72ac5f9d4f23d9bb83131aa3d46020b9 (patch) | |
tree | eb30139df9d28646f6ebf2b5f03092e55d2d48c7 /flang/lib/Frontend/CompilerInvocation.cpp | |
parent | 274e798a9832b582b29b069009dbab123070a940 (diff) | |
download | llvm-b9b2661f72ac5f9d4f23d9bb83131aa3d46020b9.zip llvm-b9b2661f72ac5f9d4f23d9bb83131aa3d46020b9.tar.gz llvm-b9b2661f72ac5f9d4f23d9bb83131aa3d46020b9.tar.bz2 |
[MLIR][Dialect] Add XeVM dialect (#144811)
XeVM is a new dialect that is designed to exposes Intel GPU hardware
features in a future proof way.
RFC is here:
https://discourse.llvm.org/t/mlir-rfc-dialect-xevm-proposal-for-new-xevm-dialect/86955
In short, XeVM is the nvvm or rocdl for Intel GPU.
The RFC includes background and challenges that XeVM is designed to
solve.
And also lists plan for upstreaming at the end.
This PR is the first of a series and it covers dialect definition and op
tests only.
Co-authored-by: Artem Kroviakov <artem.kroviakov@intel.com>
Diffstat (limited to 'flang/lib/Frontend/CompilerInvocation.cpp')
0 files changed, 0 insertions, 0 deletions