String comparison

This repository holds very simple code snippets that compare two strings in different languages .

The code is written in such a way that it is easy to measure if there is a time side-channel in the string comparison implementation.

Pico

The code in this repository is just a small part of a bigger research I'm doing on timing attacks. More info and a tool can be found in pico's repository.

Output

An easy to read summary of the data can be found in pico's wiki and in the README.md files inside each language name in this repository.

Languages

• Python 2.7.6
• Python 3.4.3
• Ruby
• Java
• C

Common steps for all languages

Install taskset

sudo apt-get install util-linux

Boot your linux kernel using these parameters:

isolcpus=0 idle=poll noht

isolcpus=0 will isolate the CPU (no processes will be assigned to it by the kernel process scheduler). Please note that using these instructions the CPU core 0 will be isolated, this will most likely fail if your workstation only has once core.

Without idle=poll the clock is stopped for a short period of time if the CPU is idle. Consequently, CPU ticks have different time durations.

noht will disable hyperthreading in all CPU cores.

To temporarily add a boot parameter to a kernel start your system and wait for the GRUB menu to show (if you don't see a GRUB menu, press and hold the left Shift key right after starting the system).

Now highlight the kernel you want to use, and press the e key. You should be able to see and edit the commands associated with the highlighted kernel. Go down to the line starting with linux and add the isolcpus=0 parameter to its end. Now press Ctrl + x to boot.

Once your workstation has started, verify the new kernel parameters are present using:

cat /proc/cmdline

Set other operating system settings:

sudo sh -c "echo "performance" > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor"

You can verify hyperthreading is disabled using utils/ht-read.sh.

Documentation

Good documentation on how to use isolcpus

Graph with Plotly

In order to post the graph to plotly you'll have to configure the credentials:

In [1]: import plotly 
In [2]: plotly.tools.set_credentials_file(username='andres.riancho', api_key='...')

Environment

The results stored in this repository were generated using:

• Ubuntu 14.04
• Linux 3.13.0-91-generic #138-Ubuntu SMP Fri Jun 24 17:00:34 UTC 2016 x86_64
• Quad Core / 64-bit CPU / AMD Phenom(tm) II X4 945 Processor
• 3.00 GHz
• ldd --version: (Ubuntu EGLIBC 2.19-0ubuntu6.9) 2.19

References

More information about this repository can be found in this pico issue.