[PATCH] D121242: [AMDGPU] gfx940 memory model

[Jay Foad via Phabricator via llvm-commits]

foad added inline comments.


================
Comment at: llvm/docs/AMDGPUUsage.rst:8673
 
+.. _amdgpu-amdhsa-memory-model-gfx940:
+
----------------
I don't know how to review 2500 lines of text. How is it different from GFX90A?


================
Comment at: llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp:1486
+
+  /// Other address spaces do not hava a cache.
+
----------------
Typo "hava"


================
Comment at: llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp:1527
+
+  /// Other address spaces do not hava a cache.
+
----------------
Typo "hava"


================
Comment at: llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp:1568
+  // be used for cache control.
+  assert(MI->mayLoad() ^ MI->mayStore());
+
----------------
Doesn't really matter but I find `!=` a bit easier to understand than `^` on booleans.


================
Comment at: llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp:1574
+  // the nontemporal attribute.
+  assert( Op == SIMemOp::LOAD || Op == SIMemOp::STORE);
+
----------------
Space after `(`


================
Comment at: llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp:1680
+
+  /// Other address spaces do not hava a cache.
+
----------------
Typo "hava"


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D121242/new/

https://reviews.llvm.org/D121242