I'd love to see the graphs with the x axis being the number of fuzz cases. BBs per unit time obviously is really all that matters, but BB/fuzz case gives a really cool "density" benchmark. Really helpful for tracking progress in fuzzer growth without comparing perf.
New kid on the block: Redqueen. Results looks impressive. Genotype-phenotype (input-behaviorl) mapping promises to be very useful. https://www.syssec.ruhr-uni-bochum.de/media/emma/veroeffentlichungen/2018/12/17/NDSS19-Redqueen.pdf …pic.twitter.com/K0NGaHAQ3Y
-
-
-
They are slower, so more dense :P Since Vuzzer, academia don't like BB/fuzz case too much I think because they used it to kinda "hide" their inefficiencies.pic.twitter.com/utWRLHhrt2
- Još 3 druga odgovora
Novi razgovor -
-
-
Nice,
@hackerschoice was just mentioning that last night -
I'm getting behind track. Didn't check anything that's happening NDSS until now :/
Kraj razgovora
Novi razgovor -
-
-
just to "explain" the honggfuzz's results :), what was used here, was the Intel PT, which, in the absence of a more fine-grained method, is the best we have (along with things like afl-qemu/pt). But the lack of speed+int-mem-cmp-edge instrumentation if clearly visible here.
-
You think improvement from using instrumentation would be significant?
- Još 27 drugih odgovora
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.