[PATCH] D156408: [AMDGPU] Update amdgpu_cs_chain_preserve docs. NFC

[Diana Picus via Phabricator via llvm-commits]

rovka updated this revision to Diff 551965.
rovka added a comment.

Fixed typo.

Does this make sense now? Usually these functions will just forward the arguments they have received, maybe with some minor processing, so the callee shouldn't receive more VGPR arguments than the caller had.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D156408/new/

https://reviews.llvm.org/D156408

Files:
  llvm/docs/AMDGPUUsage.rst


Index: llvm/docs/AMDGPUUsage.rst
===================================================================
--- llvm/docs/AMDGPUUsage.rst
+++ llvm/docs/AMDGPUUsage.rst
@@ -1196,6 +1196,8 @@
                                      uniform control flow.
 
      ``amdgpu_cs_chain_preserve``    Same as ``amdgpu_cs_chain``, but active lanes for VGPRs starting at v8 are preserved.
+                                     Calls to ``amdgpu_gfx`` functions are not allowed, and any calls to ``llvm.amdgcn.cs.chain``
+                                     must not pass more VGPR arguments than the caller's VGPR function parameters.
 
      ``amdgpu_es``                   Used for AMDPAL shader stage before geometry shader if geometry is in
                                      use. So either the domain (= tessellation evaluation) shader if


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D156408.551965.patch
Type: text/x-patch
Size: 828 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230821/2f5b9a5d/attachment.bin>