Jiahao XU

Signed-off-by: Jiahao XU This PR replaces dep `jobserver` with a fork [`jobslot`](https://github.com/cargo-bins/jobslot/blob/main/README.md). ## Advantages over `jobserver`? - `jobslot` contains bug fix for [Client::configure is unsafe] - `jobslot` removed use of...

New cmdline option `wasmer create-exe --self-test`

6

### Motivation `wasmer create-exe` requires `WASMER_DIR` to be set to the location of the Wasmer installation and it also requires at least compiler `cc` and linker `ld` to be present....

Perform operations on `Tensor` lazily?

2

Is there anyway to perform operations on `Tensor` lazily and records operations into the graph to maximize the performance like [this one](https://pytorch.org/blog/understanding-lazytensor-system-performance-with-pytorch-xla-on-cloud-tpu/)?

Get `&[MaybeUninit<f32>]` from `Tensor` for zero copy

2

I need to export images from graphicsmagick, which can be quite large and I'd like to avoid copying and unnecessary allocation. The [API](https://docs.rs/graphicsmagick/latest/graphicsmagick/wand/magick/struct.MagickWand.html#method.write_image_pixels_to) I use accepts `&'a mut [MaybeUninit]`, so...

[Proposal] Separating callback chain building and actual invocation

48

Currently, the callback chain building and the actual invocation is done at the same time. This means that invocation of `Generator::run` needs to check for initialization and that under debug...

Remove `hyper::body::to_bytes` or adds a params for size limit?

11

As pointed out by [this article](https://www.reddit.com/r/rust/comments/105nssz/watch_out_for_dos_when_using_rusts_popular_hyper), using `hyper::body::to_bytes` blindly might cause DoS attack since the function reserve the `Vec` based on `body.size_hint()`, which can be way larger than the actual...

- Use `--gc-sections` to remove unused symbols from `libbz2.a` - Add new feature `fat-lto` & `thin-lto` - Add new feat `thin` to use opt-level `z` Signed-off-by: Jiahao XU

This reduce compilation time of `cargo b --release --features static` from 4.83s => 1.94s on M1. Signed-off-by: Jiahao XU

I noticed a new PR on zstd https://github.com/gyscos/zstd-rs/pull/247 that optimizes `zstd::stream::Encoder` by reusing zstd context: > From profiling, I have found that creating a zstd::stream::Encoder and using it to encode...