Mention ${fuzzer_name}.dict and ${fuzzer_name}_seed_corpus.zip in the build pipeline and libFuzzer/AFL docs

[Dor1s]

"If you're setting up a build pipeline for in-process fuzzing, ....."

[jonathanmetzman]

Should this be documented in the build pipeline doc or the libfuzzer_and_afl doc?

[Dor1s]

I think some part of it must be in the build pipeline, at least the naming format and the requirement to archive that into the same dir with the fuzzing binary -- just because it totally belongs to the build system. More detail can be added in libfuzzer_and_afl doc, I guess. Some notes about how important those things are?

[jonathanmetzman]

just because it totally belongs to the build system.

I don't really see why, dictionary and seed corpus work just as well with custom builds.
I think it would be best for build pipeline to be solely about continuous building rather than about all things about builds (which is why I cover building targets for AFL and libFuzzer in that doc).

[Dor1s]

Good point regarding custom builds! Maybe we should have some generic sentence in the build pipeline then, like "Make sure to include all important build artifacts into the archive (e.g. dictionaries or seed corpora [link], some other example [link], etc)"

[Dor1s]

I just want to make sure people remember this stuff when they write a script / build rule for generating a build archive.

[jonathanmetzman]

OK sgtm.

[jonathanmetzman]

#96 does this.