foster/denoland-deno
1
0
Fork 0
mirror of https://github.com/denoland/deno.git synced 2024-12-22 15:24:46 -05:00
Code Issues Packages 1 Wiki Activity
denoland-deno/ext/ffi/Cargo.toml

33 lines
815 B
TOML
Raw Normal View History

chore: update copyright to 2024 (#21753)
2024-01-01 14:58:21 -05:00
# Copyright 2018-2024 the Deno authors. All rights reserved. MIT license.
feat: ffi to replace plugins (#11152) This commit removes implementation of "native plugins" and replaces it with FFI API. Effectively "Deno.openPlugin" API was replaced with "Deno.dlopen" API.
2021-08-06 17:28:10 -04:00
[package]
name = "deno_ffi"
chore: forward v1.45.5 release commit to main (#24818)
2024-07-31 18:14:27 -04:00
version = "0.151.0"
chore: workspace inheritance (#16343)
2022-11-22 15:07:35 -05:00
authors.workspace = true
edition.workspace = true
license.workspace = true
feat: ffi to replace plugins (#11152) This commit removes implementation of "native plugins" and replaces it with FFI API. Effectively "Deno.openPlugin" API was replaced with "Deno.dlopen" API.
2021-08-06 17:28:10 -04:00
readme = "README.md"
chore: workspace inheritance (#16343)
2022-11-22 15:07:35 -05:00
repository.workspace = true
feat: ffi to replace plugins (#11152) This commit removes implementation of "native plugins" and replaces it with FFI API. Effectively "Deno.openPlugin" API was replaced with "Deno.dlopen" API.
2021-08-06 17:28:10 -04:00
description = "Dynamic library ffi for deno"
[lib]
path = "lib.rs"
[dependencies]
chore: workspace inheritance (#16343)
2022-11-22 15:07:35 -05:00
deno_core.workspace = true
refactor: remove `PermissionsContainer` in deno_runtime (#24119) Also removes permissions being passed in for node resolution. It was completely useless because we only checked it for reading package.json files, but Deno reading package.json files for resolution is perfectly fine. My guess is this is also a perf improvement because Deno is doing less work.
2024-06-06 23:37:53 -04:00
deno_permissions.workspace = true
chore(ext/ffi): use dlopen2 crate (#21093) Closes #21046
2023-11-05 04:29:26 -05:00
dlopen2.workspace = true
feat(ext/ffi): Implement FFI fast-call trampoline with Dynasmrt (#15305)
2022-09-07 02:53:56 -04:00
dynasmrt = "1.2.3"
chore: Ensure we only end up with the clang version we want & upgrade libffi (#19421) The number of clang versions installed on the build machines is too dang high.
2023-06-08 11:16:24 -04:00
libffi = "=3.2.0"
chore: upgrade Rust to 1.70 and libffi-sys to 2.3.0 (#19639) Bump: - Rust -> 1.7.0 - libffi-sys -> 2.3.0 LLVM version won't change often, but it's slightly easier to edit now.
2023-06-29 17:25:48 -04:00
libffi-sys = "=2.3.0"
chore: enable clippy::print_stdout and clippy::print_stderr (#23732) 1. Generally we should prefer to use the `log` crate. 2. I very often accidentally commit `eprintln`s. When we should use `println` or `eprintln`, it's not too bad to be a bit more verbose and ignore the lint rule.
2024-05-08 22:45:06 -04:00
log.workspace = true
chore: workspace inheritance (#16343)
2022-11-22 15:07:35 -05:00
serde.workspace = true
fix(ext/ffi): improve error messages in FFI module (#17786) Fixes denoland#16922. The error messages in the `ffi` module are somewhat cryptic when passing functions that have invalid `parameters` or `result` type strings. While the generated serializer for the `ForeignFunction` struct correctly outputs a correct and verbose message, the user sees a far less helpful `data did not match any variant` message instead. The underlying cause appears to be the fallback message in the auto-derived deserializer for untagged enums [1] generated as a result of `ForeignSymbol` being marked as `#[serde(untagged)]` [2]. Passing an unexpected value for `NativeType` causes it to error out while attempting to deserialize both enum variants -- once because it's not a match for the `ForeignStatic` variant, and once because the `ForeignFunction` deserializer rejects the invalid type for the parameters/return type. This is currently open as [serde #773](https://github.com/serde-rs/serde/issues/773), and not a trivial exercise to fix generically. [1] https://github.com/serde-rs/serde/blob/v0.9.7/serde_derive/src/de.rs#L730 [2] https://github.com/denoland/deno/blob/main/ext/ffi/dlfcn.rs#L102 [3] https://github.com/serde-rs/serde/issues/773 Note that the auto-generated deserializer for untagged enums uses a private API to buffer deserializer content that we don't have access to. Instead, we can make use of the `serde_value` crate to buffer the values. This can likely be removed once the official buffering API lands (see [4] and [5]). In addition, this crate pulls in `serde_json` as a cheap way to test that the deserializer works properly. [4] https://github.com/serde-rs/serde/issues/741 [5] https://github.com/serde-rs/serde/pull/2348
2023-02-14 23:11:59 -05:00
serde-value = "0.7"
serde_json = "1.0"
fix(ext/ffi): formatting dlopen errors on Windows (#12301)
2021-10-05 20:13:56 -04:00
[target.'cfg(windows)'.dependencies]
chore: workspace inheritance (#16343)
2022-11-22 15:07:35 -05:00
winapi = { workspace = true, features = ["errhandlingapi", "minwindef", "ntdef", "winbase", "winnt"] }
chore: upgrade to rust 1.80 (#24778)
2024-07-29 12:58:04 -04:00
[lints.rust]
unexpected_cfgs = { level = "warn", check-cfg = ['cfg(target_aarch, values("x86_64", "aarch64"))'] }
Copy permalink