If you are wondering what it takes to get published in phrack click through to the CFP for areas of interest but tl:dr
the requirements are:
offensive security research
10-20pgs deep dive on internals/theory
Proven practical demo and source code snapshot. submissions@phrack.org
Perhaps they mean to cite Thayer et. al. "Software Reliability"? It has some claims like this.
Love visualizable #fuzzer progress :)
100,000 fuzz iterations. I'd say this part of the system works :)
Wanna learn more about #fuzzing? The AFL++ community has moved to Zulip: fuzz.zulipchat.com
Come join us!
whoever did this deserves a medal or something
Andreas Zeller and PhD students
About time: A multi-celebration for becoming a member of Academia Europaea, my SIGSOFT Influential Educator Award, my 60th birthday, becoming an IEEE Fellow, _and_ getting the 2026 IEEE Harlan D. Mills Award. With cake and fizzy drinks!
Here are some other resources:
- www.fuzzingbook.org
- appsec.guide/docs/fuzzing/
Honestly there's a lot of different strategies. I really need to look at the current Arrow fuzzing properly at some point, it sounds like an interesting target.
Some good alternative oracles for fuzzing:
- differential fuzzing: compare two impls
- property fuzzing: check that a property holds
- metamorphic fuzzing: check that two semantically equivalent inputs have same outputs (requires metamorphic relations): www.jacarte.me/assets/pdf/w...
I'm hoping to soon start up a running list of fuzzer failures and lessons learned at fuzz.fail and maybe make some more guides on reasoning about fuzzer behaviour. Time will tell! I will be wrapping up my PhD in a few months and will take a break to work on such side projects. :)
Hello, I'm a PhD candidate at CISPA and I work on understanding theoretical limitations of fuzzing and improving empirical analysis. Most recently I've been working on analyzing input coverage strategies, new methods for comparing fuzzers, and grammar-based generation optimizations.
Alright fuzzing friends-
good time to reconnect and help each other fuzz harder better faster stronger
Reply with how long you’ve been fuzzing, preferred fuzzing framework, how you approach writing a harness and validating results
#Fuzzing #PropertyTesting to help find each other
Ah, fuzzer findings are logarithmic with time: mboehme.github.io/paper/FSE20....
Making custom fuzzers and running them for ~30m intermittently is probably sufficient. This is, for example, what PCRE2 does with differential fuzzers: github.com/PCRE2Project...
Diff is not enabled in OSS-Fuzz.
OSS-Fuzz is... fine, but there are theoretical limitations to using the same set of fuzzers for all targets. Hopefully my mail answers this in greater detail.
I'm guessing I have to subscribe first, THEN send the email? I tried emailing without subscribing and I guess it piped them to null...
I sent a longer-form reply to this email list, hopefully it appears :)
Consider using LibAFL or building custom fuzzers for this kind of problem :) There are applications where program output is considered, e.g.: ieeexplore.ieee.org/document/103...
This is available in long form: queensu.scholaris.ca/server/api/c...
See (for similar strategies): www.fuzzingbook.org
Yes, there are many such metrics! But these are mostly in the literature as, so far, very few have been made that both meaningfully increases the insight we gain into the program and don't obliterate execution speed. e.g., value profile and "features" described in libFuzzer and Centipede.
Heyo. You can find the fuzzer here: github.com/addisoncrump...
The TL;DR is that (1) there is a _mutation_ space reduction (i.e., no mutations that can't work or are destructive) and (2) input space reduction (only crate position + available crate movement); also: snapshot fuzzing!
⏱️ 9 days until submission deadline (Dec 11, 23:59 AoE).
Organized by: @yannicnoller.bsky.social, @rohan.padhye.org, @ruijiemeng.bsky.social, and Laszlo (@lszekeres.bsky.social) Szekeres.
"every function accepts and returns the entire program state" type beat
LibAFL 0.15.4 has just been released 🎉
Of the 30 Contributers for this release, almost half are new faces <3
github.com/AFLplusplus/...
#Fuzzing #LibAFL #AFLplusplus
Plakat mit Text: "Dieses Jahr werden hunderte Millionen Computer dem geplanten Verschließ zum Opfer fallen." Poster, translated to English: "This year, hundreds of millions of computers become victim to planned obsolescence."
Plakat mit Text: "Software-Obsoleszenz ist das Thema des diesjährigen Internationalen Reparaturtags, der im Saarland zum neunten Mal stattfindet. Gemeinsam mit Expert:innen aus der Industrie möchten wir die ökologischen und sozialen Konsequenzen von absichtlich veralteten Produkten diskutieren, und überlegen was wir dagegen tun können." Translated to English: "Software Obsolescence is the topic of this year's International Repair Day, celebrated in Saarland for the 9th time. Together with industry experts, we'll discuss the ecological and social impacts of products with intentionally shortened lifespans and what we can do about it." At the bottom, the address is provided: H'eck Rathausstraße 18 66125 Saarbrücken And the date: 11 November, 2025 1700-1900
Cool event in #saarland about #software #obsolescence that my partner is helping to run :) Speakers from #KDE, #StopKillingGames, etc. The event will be mostly in German, but folks who aren't so confident in German should come along; most folks attending/speaking are also fluent in English.
Oh! Totally misread this tone lmao
Okay, but also look at how these folks have been treated in the past :p I'm not saying it's a good solution but it's a damn good way to get people who depend on libxml2 for financial gain to actually contribute back/assist in remediation rather than just yelling at the devs to do so
What is shallow to grammar fuzzers might not be shallow to byte mutator fuzzers e.g.
Ah yeah I'm just being argumentative lol
shallow to whom? 🧐
sadly, yes
Though, time-to-bug metrics are still super sparse and not necessarily reliable...
Must-read for fuzzing folks (read: tooling/algorithms/academia) by Addison Crump
addisoncrump.info/research/wha...
