[Bf-blender-cvs] [e69f3d7db17] blender-v3.2-release: GPU: Updated comment about HQ normals workaround.

[Jeroen Bakker]

Commit: e69f3d7db17893c2153ec8395db1168d6fc7608e
Author: Jeroen Bakker
Date:   Wed May 25 08:03:52 2022 +0200
Branches: blender-v3.2-release
https://developer.blender.org/rBe69f3d7db17893c2153ec8395db1168d6fc7608e

GPU: Updated comment about HQ normals workaround.

===================================================================

M	source/blender/gpu/opengl/gl_backend.cc

===================================================================

diff --git a/source/blender/gpu/opengl/gl_backend.cc b/source/blender/gpu/opengl/gl_backend.cc
index f24dc86a0a7..0c796ddc765 100644
--- a/source/blender/gpu/opengl/gl_backend.cc
+++ b/source/blender/gpu/opengl/gl_backend.cc
@@ -291,7 +291,8 @@ static void detect_workarounds()
   }
   /* See T82856: AMD drivers since 20.11 running on a polaris architecture doesn't support the
    * `GL_INT_2_10_10_10_REV` data type correctly. This data type is used to pack normals and flags.
-   * The work around uses `GPU_RGBA16I`.
+   * The work around uses `GPU_RGBA16I`. In 22.?.? drivers this has been fixed for
+   * polaris platform. Keeping legacy platforms around just in case.
    */
   if (GPU_type_matches(GPU_DEVICE_ATI, GPU_OS_ANY, GPU_DRIVER_OFFICIAL)) {
     const Vector<std::string> matches = {