diff --git a/performance.md b/performance.md
--- a/performance.md
+++ b/performance.md
@@ -3,7 +3,7 @@
 
 Let's assume we have a secret of length _m_. The splitting takes _n_ evaluations of a polynomial of order _k_ (over Galois field 256) for each byte, leading to _O(n\*k\*m)_ finite field multiplications. Reconstruction of the constant parameters during joining first precomputes parts of the Lagrange polynomial and then reuses them for each byte, taking _O(k\*k + k\*m)_ multiplications.
 
-Benchmark results. The times for split and join mean _seconds per byte_ of the secret length:
+Benchmark results. Measured on a mid-end laptop made in 2020. The times for split and join mean _seconds per byte_ of the secret length:
 <table>
     <tr>
         <th>Revision</th>
@@ -36,4 +36,28 @@ Benchmark results. The times for split a
         <td>0.00741</td>
         <td>0.00156</td>
     </tr>
-</table>
\ No newline at end of file
+    <tr>
+        <td rowspan="2">0957647049ef</td>
+        <td rowspan="2">splitting with FFT</td>
+        <td>2 / 3</td>
+        <td>1.26e-05</td>
+        <td>-</td>
+    </tr>
+    <tr>
+        <td>254 / 254</td>
+        <td>0.00828</td>
+        <td>-</td>
+    </tr>
+    <tr>
+        <td rowspan="2">d5f60adc56c0</td>
+        <td rowspan="2">splitting with FFT,<br> caching gfmul(), gfpow(), precompute_x()</td>
+        <td>2 / 3</td>
+        <td>7.88e-06</td>
+        <td>-</td>
+    </tr>
+    <tr>
+        <td>254 / 254</td>
+        <td>0.00183</td>
+        <td>-</td>
+    </tr>
+</table>
diff --git a/src/shamira/fft.py b/src/shamira/fft.py
--- a/src/shamira/fft.py
+++ b/src/shamira/fft.py
@@ -1,3 +1,5 @@
+# GNU GPLv3, see LICENSE
+
 import math
 import cmath
 import itertools