1
0
Fork 0
mirror of https://github.com/denoland/deno.git synced 2024-11-22 15:06:54 -05:00
Commit graph

158 commits

Author SHA1 Message Date
Matt Mastracci
bb1f5e4262
perf(core): async op pseudo-codegen and performance work (#18887)
Performance:

```
async_ops.js: 760k -> 1030k (!)
async_ops_deferred.js: 730k -> 770k
Deno.serve bench: 118k -> 124k
WS test w/ third_party/prebuilt/mac/load_test 100 localhost 8000 0 0: unchanged

Startup time: approx 0.5ms slower (13.7 -> 14.2ms)
```
2023-04-30 08:50:24 +00:00
Matt Mastracci
e2761df3fe
fix(ext/http): internal upgradeHttpRaw works with "Deno.serve()" API (#18859)
Fix internal "upgradeHttpRaw" API restoring capability to upgrade HTTP
connection in polyfilles "node:http" API.
2023-04-27 00:58:18 +02:00
Bartek Iwańczuk
7415aff983
bench: fix more benchmarks (#18864) 2023-04-26 21:59:32 +00:00
David Sherret
c2f5c09692
chore: fix benchmarks (#18863) 2023-04-26 19:09:28 +00:00
Bartek Iwańczuk
1b450015e7
BREAKING(unstable): remove "Deno.serve(handler, options)" overload (#18759)
In preparation to stabilization of the API this overload was decided to
be removed.
2023-04-26 14:54:03 +02:00
Matt Mastracci
bdffcb409f
feat(ext/http): Rework Deno.serve using hyper 1.0-rc3 (#18619)
This is a rewrite of the `Deno.serve` API to live on top of hyper
1.0-rc3. The code should be more maintainable long-term, and avoids some
of the slower mpsc patterns that made the older code less efficient than
it could have been.

Missing features:

- `upgradeHttp` and `upgradeHttpRaw` (`upgradeWebSocket` is available,
however).
- Automatic compression is unavailable on responses.
2023-04-22 11:48:21 -06:00
Bartek Iwańczuk
02dd09a36f
bench: add benchmark for deferred async ops (#18722)
```
./target/release/deno run cli/bench/async_ops_deferred.js
time 794 ms rate 1259445
time 786 ms rate 1272264
time 770 ms rate 1298701
time 784 ms rate 1275510
time 775 ms rate 1290322
time 786 ms rate 1272264
time 773 ms rate 1293661
time 771 ms rate 1297016
time 774 ms rate 1291989
time 767 ms rate 1303780
time 764 ms rate 1308900
time 768 ms rate 1302083
time 763 ms rate 1310615
time 761 ms rate 1314060
time 761 ms rate 1314060
time 762 ms rate 1312335
time 763 ms rate 1310615
time 759 ms rate 1317523
time 760 ms rate 1315789
time 761 ms rate 1314060
time 769 ms rate 1300390
time 763 ms rate 1310615
time 760 ms rate 1315789
time 763 ms rate 1310615
time 761 ms rate 1314060
time 759 ms rate 1317523
time 765 ms rate 1307189
time 760 ms rate 1315789
time 764 ms rate 1308900
time 763 ms rate 1310615
time 760 ms rate 1315789
time 757 ms rate 1321003
time 763 ms rate 1310615
time 759 ms rate 1317523
time 771 ms rate 1297016
time 759 ms rate 1317523
time 759 ms rate 1317523
time 763 ms rate 1310615
time 754 ms rate 1326259
time 755 ms rate 1324503
time 762 ms rate 1312335
time 752 ms rate 1329787
time 755 ms rate 1324503
time 754 ms rate 1326259
time 759 ms rate 1317523
time 754 ms rate 1326259
time 749 ms rate 1335113
time 753 ms rate 1328021
time 756 ms rate 1322751
time 753 ms rate 1328021
```

```
samply record -r 20000 target/release/deno run cli/bench/async_ops_deferred.js
```

https://share.firefox.dev/43Efvm6
2023-04-16 19:47:06 +00:00
Bartek Iwańczuk
702284dc22
perf(ops): directly respond for eager ops (#18683)
This commit changes "eager ops" to directly return a response value
instead of calling "opresponse" callback in JavaScript. This saves
one boundary crossing and has a fantastic impact on the "async_ops.js"
benchmark:

```
v1.32.4
$ deno run cli/bench/async_ops.js
time 329 ms rate 3039513
time 322 ms rate 3105590
time 307 ms rate 3257328
time 301 ms rate 3322259
time 303 ms rate 3300330
time 306 ms rate 3267973
time 300 ms rate 3333333
time 301 ms rate 3322259
time 301 ms rate 3322259
time 301 ms rate 3322259
time 302 ms rate 3311258
time 301 ms rate 3322259
time 302 ms rate 3311258
time 302 ms rate 3311258
time 303 ms rate 3300330
```

```
this branch
$ ./target/release/deno run -A cli/bench/async_ops.js
time 257 ms rate 3891050
time 248 ms rate 4032258
time 251 ms rate 3984063
time 246 ms rate 4065040
time 238 ms rate 4201680
time 227 ms rate 4405286
time 228 ms rate 4385964
time 229 ms rate 4366812
time 228 ms rate 4385964
time 226 ms rate 4424778
time 226 ms rate 4424778
time 227 ms rate 4405286
time 228 ms rate 4385964
time 227 ms rate 4405286
time 228 ms rate 4385964
time 227 ms rate 4405286
time 229 ms rate 4366812
time 228 ms rate 4385964
```

Prerequisite for https://github.com/denoland/deno/pull/18652
2023-04-13 14:32:47 +02:00
Divy Srivastava
34d596e04f
chore(cli/bench): add ws echo bench (#18595) 2023-04-05 18:31:07 +05:30
Bartek Iwańczuk
51d3fb78ad
refactor: remove "ext/flash" (#18578)
With https://github.com/denoland/deno/pull/18568 landed we no longer
need "ext/flash". 

This commit removes "deno_flash" extension completely.

This should have some impact on the binary and snapshot size.

Closes https://github.com/denoland/deno/issues/17356
2023-04-03 19:01:02 +02:00
Divy Srivastava
9ebce6e725
fix(cli/bench): look for clone3 syscalls for thread count (#18456) 2023-03-27 14:45:00 +00:00
Matt Mastracci
cebefa8783
chore(test_util): replace tempdir code w/tempdir crate (#18340) 2023-03-22 12:55:19 -06:00
Leo Kettmeir
4c1f3daa83
chore(bench): add RUNTIME_SNAPSHOT.bin (#18269) 2023-03-18 21:00:00 +01:00
David Sherret
fb021d7cef
refactor: remove usages of map_or / map_or_else (#18212)
These methods are confusing because the arguments are backwards. I feel
like they should have never been added to `Option<T>` and that clippy
should suggest rewriting to
`map(...).unwrap_or(...)`/`map(...).unwrap_or_else(|| ...)`

https://github.com/rust-lang/rfcs/issues/1025
2023-03-15 17:46:36 -04:00
Divy Srivastava
51649272bd
perf: do not depend on iana-time-zone (#18088)
Chrono's `clock` feature pulls in `iana-time-zone` which links to macOS
core_foundation. This PR itself is not enough to get rid of
CoreFoundation. Removal depends on getting rid of security framework,
see #18071
2023-03-15 07:14:22 +00:00
David Sherret
47012bd931
refactor(tests/lsp): consolidate more into test LspClient and reduce verbosity (#18100) 2023-03-09 15:09:03 -05:00
David Sherret
25d98ca289
refactor(lsp): improve test client initialization (#18015) 2023-03-08 23:15:20 +00:00
Divy Srivastava
38555a6a0f
feat(ops): reland fast zero copy string arguments (#17996)
Reland https://github.com/denoland/deno/pull/16777

The codegen is disabled in async ops and when fallback to slow call is
possible (return type is a Result) to avoid hitting this V8 bug:
https://github.com/denoland/deno/issues/17159
2023-03-03 19:04:10 +05:30
David Sherret
b4b718d6ae
fix(lint): revert no-deprecated-api for Deno.run (#17880) 2023-02-22 16:14:53 +00:00
David Sherret
b3e88e0681
refactor: deno_graph 0.43 upgrade (#17692) 2023-02-09 22:00:23 -05:00
David Sherret
f5840bdcd3
chore: upgrade to Rust 1.67 (#17548)
Co-authored-by: Bartek Iwańczuk <biwanczuk@gmail.com>
2023-01-27 10:43:16 -05:00
Bartek Iwańczuk
abd9610530
refactor: remove Deno.core (#16881)
This commit removes "Deno.core" namespace. It is strictly private API
that has no stability guarantees, we were supposed to remove it long time ago.

Co-authored-by: Yoshiya Hinosawa <stibium121@gmail.com>
2023-01-24 18:54:10 +01:00
Divy Srivastava
d5634164cb
chore: use rustfmt imports_granularity option (#17421)
Closes https://github.com/denoland/deno/issues/2699
Closes https://github.com/denoland/deno/issues/2347

Uses unstable rustfmt features. Since dprint invokes `rustfmt` we do not
need to switch the cargo toolchain to nightly. Do we care about
formatting stability of our codebase across Rust versions? (I don't)
2023-01-14 23:18:58 -05:00
Max Coplan
b23b4e231c
refactor(bench/http): Use optional chaining instead of || fallthrough (#17317) 2023-01-14 00:51:57 +01:00
Yiyu Lin
a00e432297
chore: add copyright_checker tool and add the missing copyright (#17285) 2023-01-13 16:51:32 +09:00
Yiyu Lin
896dd56b7a
refactor(cli,core,ext,rt): remove some unnecessary clone or malloc (#17274) 2023-01-05 14:29:50 -05:00
David Sherret
10e4b2e140
chore: update copyright year to 2023 (#17247)
Yearly tradition of creating extra noise in git.
2023-01-02 21:00:42 +00:00
linbingquan
f46df3e359
chore: update to Rust 1.66.0 (#17078) 2022-12-17 23:20:15 +01:00
Bartek Iwańczuk
585ec1218f
Revert "feat(ops): Fast zero copy string arguments (#16777)" (#17063)
This reverts commit 9b2b8df927.

Closes https://github.com/dsherret/ts-morph/issues/1372
Closes https://github.com/denoland/deno/issues/16979
2022-12-15 15:26:10 +00:00
Bartek Iwańczuk
cb6700fa5a
unstable: remove Deno.spawn, Deno.spawnSync, Deno.spawnChild APIs (#16893)
This commit removes three unstable Deno APIs:
- "Deno.spawn()"
- "Deno.spawnSync()"
- "Deno.spawnChild()"

These APIs were replaced by a unified "Deno.Command" API.
2022-12-09 16:43:36 +01:00
Divy Srivastava
6982c74e11
perf: use fast api for io read/write sync (#15863)
```
$ dd if=/dev/zero bs=65536 count=500000 | ./stdio # C baseline

500000+0 records in
500000+0 records out
32768000000 bytes transferred in 4.126087 secs (7941664827 bytes/sec)
c: size 32768000000 reads 500000 blocksize 65536
```

```
$ dd if=/dev/zero bs=65536 count=500000 | deno run stdio.js # Deno

500000+0 records in
500000+0 records out
32768000000 bytes transferred in 4.279032 secs (7657806719 bytes/sec)
deno: size 32768000000 reads 500000 blocksize 65536
```
2022-12-02 11:35:18 +05:30
Divy Srivastava
9b2b8df927
feat(ops): Fast zero copy string arguments (#16777)
Uses SeqOneByteString optimization to do zero-copy `&str` arguments in
fast calls.

- [x] Depends on https://github.com/denoland/rusty_v8/pull/1129
- [x] Depends on
https://chromium-review.googlesource.com/c/v8/v8/+/4036884
- [x] Disable in async ops
- [x] Make it work with owned `String` with an extra alloc in fast path.
- [x] Support `Cow<'_, str>`. Owned for slow case, Borrowed for fast
case

```rust
#[op]
fn op_string_len(s: &str) -> u32 { 
  str.len() as u32 
}
```
2022-12-02 05:29:15 +00:00
Aaron O'Mullan
238590aa9f
chore: use Rust 1.65.0 (#16688) 2022-11-18 02:59:10 +01:00
Marcos Casagrande
0832ba1deb
perf(runtime/spawn): collect output using op_read_all (#16596)
**This patch**
```
benchmark      time (avg)             (min … max)       p75       p99      p995
------------------------------------------------- -----------------------------
echo deno   23.99 ms/iter   (22.51 ms … 33.61 ms)  23.97 ms  33.61 ms  33.61 ms
cat 16kb    24.27 ms/iter    (22.5 ms … 35.21 ms)   24.2 ms  35.21 ms  35.21 ms
cat 1mb     25.88 ms/iter   (25.04 ms … 30.28 ms)  26.12 ms  30.28 ms  30.28 ms
cat 15mb    38.41 ms/iter       (35.7 ms … 50 ms)  38.31 ms     50 ms     50 ms
```

**main**
```
benchmark      time (avg)             (min … max)       p75       p99      p995
------------------------------------------------- -----------------------------
echo deno   35.66 ms/iter   (34.53 ms … 41.84 ms)  35.79 ms  41.84 ms  41.84 ms
cat 16kb    35.99 ms/iter   (34.52 ms … 44.94 ms)  36.05 ms  44.94 ms  44.94 ms
cat 1mb     38.68 ms/iter   (36.67 ms … 50.44 ms)  37.95 ms  50.44 ms  50.44 ms
cat 15mb     48.4 ms/iter   (46.19 ms … 58.41 ms)  49.16 ms  58.41 ms  58.41 ms
```
2022-11-15 14:06:52 +01:00
Divy Srivastava
e18950284f
Reland "perf(core): generate inlined wrappers for async ops" (#16455)
Reland https://github.com/denoland/deno/pull/16428
2022-10-28 16:50:17 +05:30
Divy Srivastava
115cc1e6ae
Revert "perf(core): generate inlined wrappers for async ops (#16428)" (#16443) 2022-10-27 16:58:27 +02:00
Divy Srivastava
02187966c1
perf(core): generate inlined wrappers for async ops (#16428)
V8's JIT can do a better job knowing the argument count and also enable
fast call path (in future).

This also lets us call async ops without `opAsync`:

```js
const { ops } = Deno.core;
await ops.op_void_async();
```

this patch: 4405286 ops/sec
main: 3508771 ops/sec
2022-10-27 19:10:48 +05:30
Bartek Iwańczuk
ab0c33ebf8
feat: Stabilize Deno.consoleSize() API (#15933)
This commit stabilizes "Deno.consoleSize()" API. 

There is one change compared to previous unstable API,
in that the API doesn't accept any arguments. Console size
is established by querying syscalls for stdio streams at fd
0, 1 and 2.
2022-10-26 00:23:21 +02:00
Bartek Iwańczuk
60dd84a5a0
bench: don't send data to grafana (#16324)
ci / bench release ubuntu-20.04-xl (pull_request) went from 24m to 23m
2022-10-17 17:49:03 +02:00
Bartek Iwańczuk
eee323ff0d
bench: run HTTP benches for 10s (#16322)
ci / bench release ubuntu-20.04-xl (pull_request) went from 27m to 24m
2022-10-17 16:55:15 +02:00
Bartek Iwańczuk
e54a894917
bench: don't show output (#16288)
Benchmarks are extremely noisy and it makes it impossible to
inspect what's going on in GitHub action logger.
2022-10-15 15:54:39 +02:00
Bartek Iwańczuk
8283d37c51
bench: avoid port collision (#16285) 2022-10-15 13:35:04 +02:00
Luke Channings
a2488ae792
fix(npm): support compiling on linux/aarch64 (#16208)
Changes introduced in #13633 have broken the ability to compile for
linux/aarch64 - specifically the use of a `i8` as a char type, which is
an `u8` on linux/aarch64.

This PR:
- Replaces instances of `i8` with the architecture-aware wrapper type
`c_char`
- Skips the use of `--export-dynamic-symbol` on linux-aarch64, because
the target environments often rely on older libc/binutils versions
2022-10-10 07:59:31 +05:30
Divy Srivastava
a622c5df27
perf(ext/crypto): optimize getRandomValues (#16212) 2022-10-09 15:38:38 +05:30
Divy Srivastava
e136bd86b3
perf(napi): optimize primitive napi functions (#16163)
This optimization applies on `napi_get_undefined`, `napi_get_null` &
`napi_get_boolean`.

```
# main

benchmark               time (avg)             (min … max)       p75       p99      p995
---------------------------------------------------------- -----------------------------
warmup              482.55 ps/iter   (462.5 ps … 15.67 ns)    475 ps    525 ps  829.1 ps
napi_get_undefined   25.07 ns/iter   (24.03 ns … 36.87 ns)  25.37 ns  27.09 ns  34.85 ns
```

```
# This patch

benchmark               time (avg)             (min … max)       p75       p99      p995
---------------------------------------------------------- -----------------------------
warmup              484.78 ps/iter    (462.5 ps … 14.4 ns)    475 ps  554.1 ps  583.3 ps
napi_get_undefined   15.52 ns/iter   (15.35 ns … 22.14 ns)  15.41 ns  17.18 ns  20.02 ns
```
2022-10-07 16:24:01 +05:30
Marcos Casagrande
569287b15b
perf(ext/fetch): consume body using ops (#16038)
This commit adds a fast path to `Request` and `Response` that
make consuming request bodies much faster when using `Body#text`,
`Body#arrayBuffer`, and `Body#blob`, if the body is a FastStream.
Because the response bodies for `fetch` are FastStream, this speeds up
consuming `fetch` response bodies significantly.
2022-10-04 15:48:50 +02:00
Divy Srivastava
e64af6260a
feat(ext/flash): add reuseport option on Linux (#16022) 2022-09-28 23:16:29 +05:30
Satya Rohith
b312279e58
feat: implement Web Cache API (#15829) 2022-09-28 17:41:12 +05:30
Marcos Casagrande
980d65b4d0
fix(cli/bench): strace numeric format (#16055) 2022-09-27 21:33:17 +02:00
Divy Srivastava
f02f2425d5
feat(ext/crypto): add x25519 and Ed25519 CFRG curves (#14119) 2022-09-27 17:43:42 +05:30