diff --git a/Compiler/instructions.py b/Compiler/instructions.py
index 9a2e88a2..63ad1173 100644
--- a/Compiler/instructions.py
+++ b/Compiler/instructions.py
@@ -913,7 +913,7 @@ class print_reg_plain(base.IOInstruction):
 class print_float_plain(base.IOInstruction):
     __slots__ = []
     code = base.opcodes['PRINTFLOATPLAIN']
-    arg_format = ['c', 'c', 'c']
+    arg_format = ['c', 'c', 'c', 'c']
 
 
 class print_char(base.IOInstruction):
diff --git a/Compiler/library.py b/Compiler/library.py
index 319f54d1..d32a7865 100644
--- a/Compiler/library.py
+++ b/Compiler/library.py
@@ -108,13 +108,7 @@ def print_str(s, *args):
             elif isinstance(val, sfix) or isinstance(val, sfloat):
                 raise CompilerError('Cannot print secret value:', args[i])
             elif isinstance(val, cfloat):
-                # Since we have only three registers, we separate the 0 case with the others
-                @if_e (val.z == 1)
-                def _():
-                    cint(0).print_reg_plain()
-                @else_
-                def _():
-                    val.print_float_plain()
+                val.print_float_plain()
             elif isinstance(val, list):
                 print_str('[' + ', '.join('%s' for i in range(len(val))) + ']', *val)
             else:
diff --git a/Compiler/types.py b/Compiler/types.py
index 3bd8ef78..8c0fa192 100644
--- a/Compiler/types.py
+++ b/Compiler/types.py
@@ -2037,21 +2037,12 @@ class sfloat(_number):
 class cfloat(object):
     # Helper class used for printing sfloats
     __slots__ = ['v', 'p', 'z', 's']
-    instruction_type = 'modp'
-    size = 1
 
     def __init__(self, v, p, z, s):
-        if not isinstance(v, cint) or not isinstance(p, cint) or not isinstance(z,cint) or not isinstance(s, cint):
-            raise CompilerError("Cfloat construction requires cints")
-        self.v = v
-        self.p = p
-        self.z = z
-        self.s = s
+        self.v, self.p, self.z, self.s = [cint.conv(x) for x in (v, p, z, s)]
 
-    @set_instruction_type
-    @vectorize
     def print_float_plain(self):
-        print_float_plain(self.v, self.p, self.s)
+        print_float_plain(self.v, self.p, self.z, self.s)
 
 _types = {
     'c': cint,
diff --git a/Math/gfp.h b/Math/gfp.h
index f35ebb8f..5ca3afb5 100644
--- a/Math/gfp.h
+++ b/Math/gfp.h
@@ -93,7 +93,8 @@ class gfp
 
 
   bool is_zero() const            { return isZero(a,ZpD); }
-  bool is_one()  const            { return isOne(a,ZpD); }
+  bool is_one()  const            { return isOne(a,ZpD); }
+  bool is_bit()  const            { return is_zero() or is_one(); }
   bool equal(const gfp& y) const  { return areEqual(a,y.a,ZpD); }
   bool operator==(const gfp& y) const { return equal(y); }
   bool operator!=(const gfp& y) const { return !equal(y); }
diff --git a/Processor/Instruction.cpp b/Processor/Instruction.cpp
index d18243f9..8f104fa6 100644
--- a/Processor/Instruction.cpp
+++ b/Processor/Instruction.cpp
@@ -8,7 +8,6 @@
 #include "Tools/time-func.h"
 
 #include <stdlib.h>
-#include <cmath>
 #include <algorithm>
 #include <sstream>
 #include <map>
@@ -36,20 +35,6 @@ int get_int(istream& s)
   return n;
 }
 
-long gfpToLong(bigint& tmp, gfp& g){
-     to_bigint(tmp, g);
-     long ret =  tmp.get_si();
-     if (ret < 0){
-        //Dirty trick to allow conversion
-         g.negate();
-         to_bigint(tmp, g);
-         ret = tmp.get_si();
-         ret *= -1;
-         g.negate();
-     }
-     return ret;
- }
-
 // Convert modp to signed bigint of a given bit length
 void to_signed_bigint(bigint& bi, const gfp& x, int len)
 {
@@ -138,7 +123,6 @@ void Instruction::parse(istream& s)
       case SUBINT:
       case MULINT:
       case DIVINT:
-      case PRINTFLOATPLAIN:
         r[0]=get_int(s);
         r[1]=get_int(s);
         r[2]=get_int(s);
@@ -296,6 +280,10 @@ void Instruction::parse(istream& s)
       case CRASH:
       case CLOSESOCKET:
 	break;
+      // instructions with 4 register operands
+      case PRINTFLOATPLAIN:
+        get_vector(4, start, s);
+        break;
       // open instructions
       case STARTOPEN:
       case STOPOPEN:
@@ -1444,17 +1432,28 @@ void Instruction::execute(Processor& Proc) const
         break;
       case PRINTFLOATPLAIN:
         if (Proc.P.my_num() == 0)
-            {
-                 gfp v = Proc.read_Cp(r[0]);
-                 gfp p = Proc.read_Cp(r[1]);
-                 gfp s = Proc.read_Cp(r[2]);
-                 long lv = gfpToLong(Proc.temp.aa, v);
-                 long lp = gfpToLong(Proc.temp.aa2, p);
-                 double res = (double)lv * pow(2.0,lp);
-                 if(!s.is_zero())
-                    res *= -1;
-                 cout << res <<flush;
-            }
+          {
+            gfp v = Proc.read_Cp(start[0]);
+            gfp p = Proc.read_Cp(start[1]);
+            gfp z = Proc.read_Cp(start[2]);
+            gfp s = Proc.read_Cp(start[3]);
+            to_bigint(Proc.temp.aa, v);
+            // MPIR can't handle more precision in exponent
+            to_signed_bigint(Proc.temp.aa2, p, 31);
+            long exp = Proc.temp.aa2.get_si();
+            mpf_class res = Proc.temp.aa;
+            if (exp > 0)
+              mpf_mul_2exp(res.get_mpf_t(), res.get_mpf_t(), exp);
+            else
+              mpf_div_2exp(res.get_mpf_t(), res.get_mpf_t(), -exp);
+            if (z.is_one())
+              res = 0;
+            if (!s.is_zero())
+              res *= -1;
+            if (not z.is_bit() or not s.is_bit())
+              throw Processor_Error("invalid floating point number");
+            cout << res << flush;
+          }
       break;
       case PRINTSTR:
         if (Proc.P.my_num() == 0)