Benoît CORTIER
|
60bc72f873
|
test: organize integration tests into single crate
Put all integration tests in a single crate, and organized in modules.
This is similar to what was done in `cargo` repository:
https://github.com/rust-lang/cargo/pull/5022#issuecomment-364691154
```
$ rustup show
stable-x86_64-unknown-linux-gnu (default)
rustc 1.69.0 (84c898d65 2023-04-16)
```
Run on a recent high-end laptop:
12th Gen Intel(R) Core(TM) i9-12900HK
Original (multiple integration binaries):
- `cargo clean && cargo test --no-run`: 62.893s
- `cargo clean && cargo build --lib`: 54.959s
- `cargo clean && cargo build --bins`: 55.933s
- `cargo test --no-run` (after `cargo build`): 14.472s
- `cargo test` (after `cargo test --no-run`): 1.786s
- `du -hs target/`: 4.2G
After (ironrdp-testsuite):
- `cargo clean && cargo test --no-run`: 41.157s (crates with no tests are ignored)
- `cargo clean && cargo build --lib`: 53.983s
- `cargo clean && cargo build --bins`: 54.482s
- `cargo test --no-run` (after `cargo build`): 12.915s
- `cargo test` (after `cargo test --no-run`): 0.240s
- `du -hs target/`: 3.4G
Absolute diff:
- `cargo clean && cargo test --no-run`: -21.736s
- `cargo clean && cargo build --lib`: -0.976s
- `cargo clean && cargo build --bins`: -1.451s
- `cargo test --no-run` (after `cargo build`): -1.557s
- `cargo test` (after `cargo test --no-run`): -1.546s
- `du -hs target/`: -0.8G
Relative diff (%):
- `cargo clean && cargo test --no-run`: -34.5%
- `cargo clean && cargo build --lib`: -1.77%
- `cargo clean && cargo build --bins`: -2.59%
- `cargo test --no-run` (after `cargo build`): -10.7%
- `cargo test` (after `cargo test --no-run`): -86.5%
- `du -hs target/`: -19.0%
|
2023-05-11 19:41:11 -04:00 |
|