[PATCH] D58957: [AMDGPU] Add an experimental buffer fat pointer address space.

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added a comment.

I think this will be accepted right now as a no-op addrspacecast to any of the global-like address spaces



================
Comment at: docs/AMDGPUUsage.rst:294-295
      6                  Constant 32-bit
-     ================== =================
+     7                  Buffer Fat Pointer (experimental)
+     ================== =================================
+
----------------
Might as well reserve more for 256-bit descriptors. We'll probably need several of these eventually 


================
Comment at: lib/Target/AMDGPU/AMDGPUTargetMachine.cpp:297
+  // flat, non-integral buffer fat pointers.
     return "e-p:64:64-p1:64:64-p2:32:32-p3:32:32-p4:64:64-p5:32:32-p6:32:32"
          "-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128"
----------------
Shouldn't this add p7:128?


================
Comment at: lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp:257
+      AddrSpace == AMDGPUAS::CONSTANT_ADDRESS_32BIT ||
+      AddrSpace == AMDGPUAS::BUFFER_FAT_POINTER) {
     return 512;
----------------
This probably shouldn't be included now, at least without a test


================
Comment at: lib/Target/AMDGPU/SIISelLowering.cpp:1050
+      AS == AMDGPUAS::CONSTANT_ADDRESS_32BIT ||
+      AS == AMDGPUAS::BUFFER_FAT_POINTER) {
     // If the offset isn't a multiple of 4, it probably isn't going to be
----------------
Ditto


Repository:
  rL LLVM

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D58957/new/

https://reviews.llvm.org/D58957