[LLVMdev] Eliminating redundant loads

[Dibyendu Majumdar]

Hi,

I am generating following code:

  %base = getelementptr inbounds %ravi.CallInfo* %6, i32 0, i32 4, i32 0
  %7 = load %ravi.TValue** %base
  %8 = bitcast %ravi.TValue* %7 to i8*
  %9 = bitcast %ravi.TValue* %5 to i8*
  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %8, i8* %9, i32 16, i32 8, i1 false)
  %10 = load %ravi.CallInfo** %L_ci
  %base1 = getelementptr inbounds %ravi.CallInfo* %10, i32 0, i32 4, i32 0
  %11 = load %ravi.TValue** %base1
  %12 = getelementptr inbounds %ravi.TValue* %11, i32 1
  %13 = getelementptr inbounds %ravi.TValue* %5, i32 1
  %14 = bitcast %ravi.TValue* %12 to i8*
  %15 = bitcast %ravi.TValue* %13 to i8*
  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %14, i8* %15, i32 16, i32 8, i1 false
)
  %16 = load %ravi.CallInfo** %L_ci
  %base2 = getelementptr inbounds %ravi.CallInfo* %16, i32 0, i32 4, i32 0
  %17 = load %ravi.TValue** %base2

Now base, base2 and base1 are really the same - i.e., nothing's
happened to change the pointer held at this location. So should I
expect the redundant getelementptr and load instructions to be
eliminated during optimization phase?

I am using the optimisation passes as described in Kaleidoscope
tutorial but it seems that the redundant loads are not being removed.

Is it the call to memcpy that's preventing this?

Thanks and Regards
Dibyendu