[PATCH] D29500: [DWARF][PATCH] Keep track of spilled variables in LiveDebugValues

[Adrian Prantl via Phabricator via llvm-commits]

aprantl added a comment.

Very interesting! Have you measured the performance impact of this? Is there any?



================
Comment at: lib/CodeGen/LiveDebugValues.cpp:324
+         "Spill instruction does not have exactly one memory operand?");
+  MachineInstr::mmo_iterator MMOI = MI.memoperands_begin();
+  const PseudoSourceValue *PVal = (*MMOI)->getPseudoValue();
----------------
auto ?


================
Comment at: lib/CodeGen/LiveDebugValues.cpp:405
+
+  // To identify a spill instruction, use the same criteria as in AsmPrinter.
+  if (!((TII->isStoreToStackSlotPostFE(MI, FI) ||
----------------
perhaps factor this out into a bool  isSpillInstruction()?


================
Comment at: lib/CodeGen/LiveDebugValues.cpp:411
+
+  // We are looking for a register operand that has its kill flag set.
+  unsigned Reg = 0;
----------------
It would be good if the comment could explain why we are looking for one.


================
Comment at: lib/CodeGen/LiveDebugValues.cpp:454
+      OpenRanges.insert(SpillLocID, VL.Var);
+      break;
+    }
----------------
return?


================
Comment at: test/DebugInfo/MIR/X86/live-debug-values-spill.mir:195
+  
+  attributes #0 = { nounwind uwtable "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+fxsr,+mmx,+sse,+sse2,+x87" "unsafe-fp-math"="false" "use-soft-float"="false" }
+  attributes #1 = { argmemonly nounwind }
----------------
most of the "string" attributes can probably be safely removed.


https://reviews.llvm.org/D29500