[PATCH] D120150: Constant folding of llvm.amdgcn.trig.preop

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp:1024
+
+    if (!(Csrc && Cseg))
+      break;
----------------
demorgan this


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp:1035
+    APInt EClamp(Numbits, 1077, Signed);
+    APInt E = (Fsrc.bitcastToAPInt()).ashr(52);
+    E &= 0x7ff;
----------------
extra parentheses


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp:1042
+                     Ovflow);
+    int32_t I = (Shift.ashr(5)).getSExtValue();
+    APInt Bshift = Shift & 0x1f;
----------------
extra parentheses


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp:1054
+      Signed = true;
+      Thi = (Thi.shl(Bshift)) |
+            (Tlo.lshr(APInt(Numbits, 64, Signed).ssub_ov(Bshift, Ovflow)));
----------------
extra parentheses


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp:1055
+      Thi = (Thi.shl(Bshift)) |
+            (Tlo.lshr(APInt(Numbits, 64, Signed).ssub_ov(Bshift, Ovflow)));
+    }
----------------
extra parentheses


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D120150/new/

https://reviews.llvm.org/D120150