encoding_rs: initial integration

[manunio]

Hi, encoding_rs an implementation of the (non-JavaScript parts of) the Encoding Standard written in Rust.

• It has 38 million+ downloads as per crates.io.
• It being used by projects like:
  • deno
  • tauri
  • wasmtime
  • reqwest
  • sqlx
  • ruffle
  • aws-lambda-rust-runtime
  • quick-xml
  • actix-web

[hsivonen]

fuzz_labels is extremely unlikely to find anything at this point and isn't a good use of resources to run.

While the other two might find something, and if they did, that would be very notable and useful, my expectation level of fuzzing finding something at this point is low enough that I feel bad about using resources for continuous fuzzing of encoding_rs.

I don't object to integration, though, if Google considers this a good use of the resources of OSS-Fuzz regardless.

[manunio]

fuzz_labels is extremely unlikely to find anything at this point and isn't a good use of resources to run.

Hi thanks for your response, I have removed fuzz_labels.

While the other two might find something, and if they did, that would be very notable and useful, my expectation level of fuzzing finding something at this point is low enough that I feel bad about using resources for continuous fuzzing of encoding_rs.

I don't object to integration, though, if Google considers this a good use of the resources of OSS-Fuzz regardless.

Given the scale of this project's usage, I think using resources for this project would be worth it, but will have to wait for panel's decision.

@jonathanmetzman received permission from maintainer, please review this pr.

[jonathanmetzman]

@jonathanmetzman received permission from maintainer, please review this pr.

Let me ask our team what we think about this since this is pretty unenthusiastic :-)

[manunio]

@jonathanmetzman received permission from maintainer, please review this pr.

Let me ask our team what we think about this since this is pretty unenthusiastic :-)

Hi @jonathanmetzman any updates ?

[hsivonen]

One more thing: --features simd-accel is more complex and, therefore, more likely to have bugs than the non-SIMD config. Therefore, if encoding_rs is fuzzed, fuzzing it with SIMD enabled is likely to be more valuable.

[manunio]

@jonathanmetzman Friendly ping :)

[manunio]

Hi, @Navidem can you please tell us panel's decision for this project.

[manunio]

@jonathanmetzman ping :)

[manunio]

@jonathanmetzman ping :)

[jonathanmetzman]

@jonathanmetzman ping :)

I've asked the team about fuzzing this. I personally do not support it but if others do I will merge.

[manunio]

@jonathanmetzman ping :)

I've asked the team about fuzzing this. I personally do not support it but if others do I will merge.

okay thanks :)

[DonggeLiu]

Our team find ClusterFuzzLite more suitable for this integration, would you please consider that instead? Thanks!
I will close this for mow : )