[PATCH] D134862: [AMDGPU][GISel] Update `isCanonicalized`

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/SIISelLowering.cpp:10195
+  case AMDGPU::G_FCEIL:
+  case AMDGPU::G_FFLOOR:
+  case AMDGPU::G_FMA:
----------------
G_FRINT, G_NEARBYINT, G_INTRINSIC_FPTRUNC_ROUND,  G_INTRINSIC_TRUNC,  G_INTRINSIC_ROUNDEVEN, 


================
Comment at: llvm/lib/Target/AMDGPU/SIISelLowering.cpp:10231
+    case Intrinsic::amdgcn_rsq:
+    case Intrinsic::amdgcn_rsq_clamp:
+    case Intrinsic::amdgcn_rsq_legacy:
----------------
amdgcn_sqrt, fmed3, fmad_ftz, sin, cos, log, log_clamp


================
Comment at: llvm/lib/Target/AMDGPU/SIISelLowering.cpp:10239
+    case Intrinsic::amdgcn_cvt_pkrtz:
+    case Intrinsic::amdgcn_cubeid:
+    case Intrinsic::amdgcn_frexp_mant:
----------------
cubema, cubesc, cubetc?


================
Comment at: llvm/lib/Target/AMDGPU/SIISelLowering.cpp:10246
+      break;
+    }
+
----------------
Should do separately and also introduce for the DAG path, but we're missing some of the newer operations like _amdgcn_cvt_f32_bf8


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D134862/new/

https://reviews.llvm.org/D134862