[llvm-branch-commits] [clang] 639db8a - [Clang][Doc][OpenCL] Release 16 notes

Anastasia Stulova via llvm-branch-commits llvm-branch-commits at lists.llvm.org
Sun Feb 19 13:46:11 PST 2023


Author: Anastasia Stulova
Date: 2023-02-19T21:39:02Z
New Revision: 639db8a90d2adf7789b9d3cec3a7c26502bf819b

URL: https://github.com/llvm/llvm-project/commit/639db8a90d2adf7789b9d3cec3a7c26502bf819b
DIFF: https://github.com/llvm/llvm-project/commit/639db8a90d2adf7789b9d3cec3a7c26502bf819b.diff

LOG: [Clang][Doc][OpenCL] Release 16 notes

Documented major OpenCL features in release 16.

Differential Revision: https://reviews.llvm.org/D143348

Added: 
    

Modified: 
    clang/docs/ReleaseNotes.rst

Removed: 
    


################################################################################
diff  --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index b32a84d1f6d3..525be8e76269 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -395,6 +395,18 @@ C2x Feature Support
   Also documents the builtin appropriately. Note, a type definition in C++
   continues to be rejected.
 
+OpenCL Kernel Language Changes
+------------------------------
+
+- Improved diagnostics for C++ templates used in kernel arguments.
+- Removed redundant pragma for the ``arm-integer-dot-product`` extension.
+- Improved support of enqueued block for AMDGPU.
+- Added ``nounwind`` attribute implicitly to all functions.
+- Improved builtin functions support:
+
+  * Allow disabling default header-based feature/extension support by passing ``-D__undef_<feature>``.
+  * Fixed conditional definition of the depth image and ``read_write image3d`` builtins.
+
 Non-comprehensive list of changes in this release
 -------------------------------------------------
 - It's now possible to set the crash diagnostics directory through


        


More information about the llvm-branch-commits mailing list