[llvm] fc4210f - AMDGPU/Docs: Fix target properties for gfx9-4-generic (#125593)

via llvm-commits llvm-commits at lists.llvm.org
Tue Feb 4 18:47:46 PST 2025


Author: Konstantin Zhuravlyov
Date: 2025-02-04T21:47:43-05:00
New Revision: fc4210fb6c5a42b3838091a97a00ed1fba042ef0

URL: https://github.com/llvm/llvm-project/commit/fc4210fb6c5a42b3838091a97a00ed1fba042ef0
DIFF: https://github.com/llvm/llvm-project/commit/fc4210fb6c5a42b3838091a97a00ed1fba042ef0.diff

LOG: AMDGPU/Docs: Fix target properties for gfx9-4-generic (#125593)

gfx9-4-generic has architected flat scratch, not absolute

Added: 
    

Modified: 
    llvm/docs/AMDGPUUsage.rst

Removed: 
    


################################################################################
diff  --git a/llvm/docs/AMDGPUUsage.rst b/llvm/docs/AMDGPUUsage.rst
index b646621d12eb0d..84980d0c31d4f9 100644
--- a/llvm/docs/AMDGPUUsage.rst
+++ b/llvm/docs/AMDGPUUsage.rst
@@ -583,11 +583,11 @@ Generic processor code objects are versioned. See :ref:`amdgpu-generic-processor
                                                                                                   - ``v_dot2_f32_f16``
 
 
-     ``gfx9-4-generic``   ``amdgcn``     - ``gfx940``      - xnack            - Absolute flat   FP8 and BF8 instructions,
-                                         - ``gfx941``      - sramecc            scratch         FP8 and BF8 conversion instructions,
-                                         - ``gfx942``                                           as well as instructions with XF32 format support
-                                         - ``gfx950``                                           are not available.
-
+     ``gfx9-4-generic``   ``amdgcn``     - ``gfx940``      - sramecc          - Architected     FP8 and BF8 instructions,
+                                         - ``gfx941``      - tgsplit            flat scratch    FP8 and BF8 conversion
+                                         - ``gfx942``      - xnack            - Packed          instructions, as well as
+                                         - ``gfx950``      - kernarg preload    work-item       instructions with XF32 format
+                                                                                IDs             support are not available.
 
      ``gfx10-1-generic``  ``amdgcn``     - ``gfx1010``     - xnack            - Absolute flat   - The following instructions are
                                          - ``gfx1011``     - wavefrontsize64    scratch           not available on ``gfx1011``


        


More information about the llvm-commits mailing list