diff --git a/README.md b/README.md
index da2cd42..5b7829a 100644
--- a/README.md
+++ b/README.md
@@ -157,7 +157,7 @@ Both `nanoBench.sh` and `kernel-nanoBench.sh` support the following command-line
 | `-cpu <n>`                   | Pins the measurement thread to CPU n. `[Default: Pin the thread to the CPU it is currently running on.]` |
 | `-verbose`                   | Outputs the results of all performance counter readings. In the user-space version, the results are printed to stdout. The output of the kernel module can be accessed using `dmesg`. |
 
-<sup id="syntax">1</sup> As an extension, the tool also supports statements of the form `|n` (with 1&le;n&le;15) that are translated to n-byte NOPs.
+<sup id="syntax">1</sup> As an extension, the tool also supports statements of the form `|n` (with 1&le;n&le;15) that are translated to n-byte NOPs, and statements of the form `n*|x|` that unroll x n times (nesting is not supported).
 
 The following parameters are only supported by `nanoBench.sh`.
 
diff --git a/kernelNanoBench.py b/kernelNanoBench.py
index a08c911..85ed75c 100644
--- a/kernelNanoBench.py
+++ b/kernelNanoBench.py
@@ -1,5 +1,6 @@
 import atexit
 import os
+import re
 import subprocess
 import sys
 
@@ -36,7 +37,7 @@ def assemble(code, objFile, asmFile='/tmp/ramdisk/asm.s'):
          code = code.replace('|3',  '.byte 0x0f,0x1f,0x00;')
          code = code.replace('|2',  '.byte 0x66,0x90;')
          code = code.replace('|1',  'nop;')
-         code = code.replace('|',   '')
+         code = re.sub(r'(\d*)\*\|(.*?)\|', lambda m: int(m.group(1))*(m.group(2)+';'), code)
       code = '.intel_syntax noprefix;' + code + ';1:;.att_syntax prefix\n'
       with open(asmFile, 'w') as f:
          f.write(code);
diff --git a/utils.sh b/utils.sh
index 2da727b..6d9dee3 100644
--- a/utils.sh
+++ b/utils.sh
@@ -1,9 +1,8 @@
 assemble() {
     asm=$1
     filename=$2
-    echo ".intel_syntax noprefix" > asm-tmp.s
-    echo "$asm" >> asm-tmp.s
-    sed -i "
+
+    asm=`sed "
         s/|15/.byte 0x66,0x66,0x66,0x66,0x66,0x66,0x2e,0x0f,0x1f,0x84,0x00,0x00,0x00,0x00,0x00;/g
         s/|14/.byte 0x66,0x66,0x66,0x66,0x66,0x2e,0x0f,0x1f,0x84,0x00,0x00,0x00,0x00,0x00;/g
         s/|13/.byte 0x66,0x66,0x66,0x66,0x2e,0x0f,0x1f,0x84,0x00,0x00,0x00,0x00,0x00;/g
@@ -19,8 +18,12 @@ assemble() {
         s/|3/.byte 0x0f,0x1f,0x00;/g
         s/|2/.byte 0x66,0x90;/g
         s/|1/nop;/g
-        s/|//g
-    " asm-tmp.s
+    " <<< "$asm"`
+
+    asm=`python3 -c 'import sys, re; print(re.sub(r"(\d*)\*\|(.*?)\|", lambda m: int(m.group(1))*(m.group(2)+";"), sys.argv[1]))' "$asm"`
+
+    echo ".intel_syntax noprefix" > asm-tmp.s
+    echo "$asm" >> asm-tmp.s
     as asm-tmp.s -o asm-tmp.o || exit
     objcopy -j .text -O binary asm-tmp.o "$filename"
     rm asm-tmp.*