🏈News: Mike LaFleur is the front-runner for the Cardinals’ HC job, offering offensive creativity, a Packers pedigree, and a prime shot to elevate a young QB—could be a slam-dunk if the deal sticks. #AZCardinals #LaFleur
🚨🔗 Full story in bio👆 https://ift.tt/RczortV
Published my first PyPI package today, called lafleur.
#lafleur is a specialized CPython JIT fuzzer that uses a coverage-guided, evolutionary approach. It executes test cases, observes their effect on the JIT's state by analyzing verbose trace logs, and uses that feedback to guide its mutations, becoming smarter at finding interesting code paths over time.
Let me know if you use it or have any questions.
https://pypi.org/project/lafleur/
https://github.com/devdanzin/lafleur
Willis had a commendable performance, throwing for 348 total yards and three touchdowns. Despite trailing 27-14 at halftime, he managed to rally the team with
It turns out that, by running it on an interpreter with ASan enabled, I was the culprit of my fuzzer lafleur using way too much memory.
That even led to me buying some DDR5 to be able to fuzz a bit more comfortably. Running without ASan reduces memory usage to 1/15. So I guess now I'll have some spare RAM, and less money, going forward :)
I'll enhance the JIT fuzzer to run on a different interpreter than the fuzzing scripts, which benefit from ASan.
lafleur, the CPython JIT fuzzer, now can compare timings for running a piece of code with JIT on and off.
The idea is that if the run with JIT on is much slower that with JIT off, we have found a performance bug.
Brandt Bucher suggested this mode. It took a while to get started on it, but it was simple to implement on top of Differential Mode.
So many modes, so little compute available...
Differential Mode has just landed in lafleur, the CPython JIT fuzzer. It runs the same code with and without the JIT, compares the result, and flags any discrepancies.
It does this while mutating the code in a feedback-guided loop, so it evolves the fuzzing scripts trying to find one where the JIT gives wrong results.
This is actually the 2nd time this feature is implemented (now better and more robust): it used to work, broke, and now is back.
Posted a topic about lafleur, the CPython JIT fuzzer I'm developing: https://discuss.python.org/t/introducing-lafleur-a-cpython-jit-fuzzer/103452
It has many details about the project, the results (4 JIT crashes so far), calls for help and ideas for the future.
If any of this interests you, please give a read. Boosts welcome :)
Greetings all, I’d like to present lafleur, a CPython JIT fuzzer. We’d love and welcome any kind of collaborations, from questions to vague ideas to working code to being part of a research project. Please start discussions, file issues or propose pull requests if you want to contribute. Summary What lafleur is a specialized fuzzer designed to find crashes in CPython’s experimental JIT, having found 4 JIT crashes so far. How It works by mutating code samples and reading JIT debug information w...
Benoît Melançon
cedisports
danzin
