HackerReposA FIELD GUIDE TO OPEN SOURCE
ProSponsor the shelf♧ 0
← Back to top all-time
aoh
radamsa
Testing & Quality2016–2016
aoh / radamsa

radamsa

a general-purpose fuzzer

3Distinct people
3Mentions
0Explicit recommenders
1.3KGitHub stars
REPOSITORY AT A GLANCE

What to know before exploring

What it is

a general-purpose fuzzer

GitHub

Where it is used

No documented product use has been verified yet.

What people are saying

No recommendation phrase was found in the indexed comments.

No classified warning is attached to this page. Read the discussions before treating silence as approval.

Also in these discussions

No other repository is repeatedly linked in the same discussions.

This week on Hacker News

Nov 7–Nov 14 UTC · Direct repository links and separately labeled related project stories. Indexed through 11/14/2026, 3:56:07 PM UTC.

No direct repository links were found in the indexed HN discussions over the past seven days. Older discussions are available below.

Mentions over time

HN mentions by week3 dated mentions

Observed mentions: 2016-07-12 15:22:19 UTC to 2016-09-10 20:46:34 UTC. Weeks start Monday, UTC.

0122016-07-11 00:00:00 UTC to 2016-07-18 00:00:00 UTC (exclusive): 1 mentions; 0 submissions, 1 commentsJul 112016-07-18 00:00:00 UTC to 2016-07-25 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2016-07-25 00:00:00 UTC to 2016-08-01 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2016-08-01 00:00:00 UTC to 2016-08-08 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2016-08-08 00:00:00 UTC to 2016-08-15 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 commentsAug 82016-08-15 00:00:00 UTC to 2016-08-22 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2016-08-22 00:00:00 UTC to 2016-08-29 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2016-08-29 00:00:00 UTC to 2016-09-05 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2016-09-05 00:00:00 UTC to 2016-09-12 00:00:00 UTC (exclusive): 2 mentions; 1 submissions, 1 commentsSep 5

Counts come from the repository’s dated source evidence. Index cutoff: 2026-11-02 08:16:59 UTC.

View mention counts and date ranges
UTC ranges include their start and exclude their end. Select a period to filter the source evidence.
PeriodStart (UTC)End (exclusive, UTC)MentionsSubmissionsComments
Jul 112016-07-11 00:00:00 UTC2016-07-18 00:00:00 UTC101
Jul 182016-07-18 00:00:00 UTC2016-07-25 00:00:00 UTC000
Jul 252016-07-25 00:00:00 UTC2016-08-01 00:00:00 UTC000
Aug 12016-08-01 00:00:00 UTC2016-08-08 00:00:00 UTC000
Aug 82016-08-08 00:00:00 UTC2016-08-15 00:00:00 UTC000
Aug 152016-08-15 00:00:00 UTC2016-08-22 00:00:00 UTC000
Aug 222016-08-22 00:00:00 UTC2016-08-29 00:00:00 UTC000
Aug 292016-08-29 00:00:00 UTC2016-09-05 00:00:00 UTC000
Sep 52016-09-05 00:00:00 UTC2016-09-12 00:00:00 UTC211

What Hacker News says

3 sources
I worked on the cyber reasoning system (CRS) at Trail of Bits for our entry into the Cyber Grand Challenge [1]. Some slides describing the system are here [2]. Specifically, I implemented our fuzzer. I created a dynamic binary translator [3] that emulated the DECREE [4] operating system and x86 arhcitecture. It had the Radamsa [5] mutator built-in, along with a number of other simpler mutators. I think our fuzzer out-performed our competitors, though I am biased ;-) The fuzzer was single-threaded, but could perform more than a million fuzz/mutate-execute (with coverage) iterations every two hours. Before I optimized it, it beat the pants off PIN [6]. We ran many such fuzzer processes concurrently. They would saturate the CPUs, and actually performed no I/O because I emulated all I/O in memory ;-) This was key to us achieving such high-th
Read the rest
roughput. Our fuzzer wasn't super smart (though Radamsa is), but it benefited a lot from a feedback loop with our symbolic executors [7]. The symbolic executors would produce inputs that would then get fuzzed. These inputs could feed back into the symbolic executors, etc. That added more brains to our system. All in all, we ran the CRS across something like 300 large EC2 nodes (across three availability zones). Per node, 8 or so fuzzers processes were running constantly for 24 hours. I'd ballpark that as 28.8 billion mutate+execute cycles. In conclusion, the key for us was to make a production-quality, high-throughput fuzzer that did only one thing really well and really fast, then complement it with other more sophisticated tools like symbolic executors. [1] https://blog.trailofbits.com/2015/07/15/how-we-fared-in-the-... [2] http://infiltratecon.com/archives/Slides_Artem_Dinaburg.pdf [3] https://en.wikipedia.org/wiki/Binary_translation [4] https://github.com/CyberGrandChallenge/libcgc [5] https://github.com/aoh/radamsa [6] https://software.intel.com/en-us/articles/pin-a-dynamic-bina... [7] https://en.wikipedia.org/wiki/Symbolic_execution
In Ask HN: What is the emerging state of the art in fuzzing techniques?

MAKE YOURSELF AT HOME

A few shortcuts

/
Search the collection
?
Open this guide
Esc
Close dialogs and navigation

Your saved repositories and interests stay in this browser.