1
0
Fork 0
mirror of https://github.com/denoland/deno.git synced 2024-12-14 03:23:17 -05:00
Commit graph

646 commits

Author SHA1 Message Date
Yoshiya Hinosawa
c1054b88b2
Merge branch 'main' into support_create_connection 2024-10-24 14:36:55 +09:00
Nathan Whitaker
27df42f659
fix(ext/node): cancel pending ipc writes on channel close (#26504)
Fixes the issue described in
https://github.com/denoland/deno/issues/23882#issuecomment-2423316362.

The parent was starting to send a message right before the process would
exit, and the channel closed in the middle of the write. Unlike with
reads, we weren't cancelling the pending writes, which resulted in a
`Broken pipe` error surfacing to the user.
2024-10-24 04:50:35 +00:00
Satya Rohith
92ed4d38db
fix(node:tls): set TLSSocket.alpnProtocol for client connections (#26476)
Towards https://github.com/denoland/deno/issues/26127
2024-10-23 07:47:43 +00:00
jiang1997
afb33b3c25
fix(ext/node): use primordials in ext/node/polyfills/https.ts (#26323)
Towards https://github.com/denoland/deno/issues/24236
2024-10-21 09:50:53 +02:00
Satya Rohith
f62bb6ff1b
Merge branch 'main' into support_create_connection 2024-10-20 05:52:02 +05:30
Divy Srivastava
0710af034f
perf: avoid multiple calls to runMicrotask (#26378)
Improves HTTP throughput by 8-9k rps on Linux:

this patch
```
Requests/sec: 145001.69
Transfer/sec:     20.74MB
```

main
```
Requests/sec: 137866.61
Transfer/sec:     19.72MB
```

The improvements comes from the reduced number of calls to
`op_run_microtask` per request. Returning `true` from a macrotask
callback already calls `op_run_microtask` so the extra call was
redundant.

Here's `--strace-ops` output for a single request:

main
```
[     4.667] op_http_wait                                       : CompletedAsync Async
[     4.667] op_run_microtasks                                  : Dispatched Slow
[     4.668] op_http_try_wait                                   : Dispatched Slow
[     4.668] op_http_try_wait                                   : Completed Slow
[     4.668] op_http_wait                                       : Dispatched Async
[     4.668] op_http_set_response_header                        : Dispatched Slow
[     4.668] op_http_set_response_header                        : Completed Slow
[     4.669] op_http_set_response_body_text                     : Dispatched Slow
[     4.669] op_http_set_response_body_text                     : Completed Slow
[     4.669] op_run_microtasks                                  : Completed Slow
[     4.669] op_has_tick_scheduled                              : Dispatched Slow
[     4.669] op_has_tick_scheduled                              : Completed Slow
[     4.669] op_run_microtasks                                  : Dispatched Slow
[     4.669] op_run_microtasks                                  : Completed Slow
[     4.669] op_run_microtasks                                  : Dispatched Slow
[     4.669] op_run_microtasks                                  : Completed Slow
```

this pr
```
[     3.726] op_http_wait                                       : CompletedAsync Async
[     3.727] op_run_microtasks                                  : Dispatched Slow
[     3.727] op_http_try_wait                                   : Dispatched Slow
[     3.727] op_http_try_wait                                   : Completed Slow
[     3.727] op_http_wait                                       : Dispatched Async
[     3.727] op_http_set_response_header                        : Dispatched Slow
[     3.728] op_http_set_response_header                        : Completed Slow
[     3.728] op_http_set_response_body_text                     : Dispatched Slow
[     3.728] op_http_set_response_body_text                     : Completed Slow
[     3.728] op_run_microtasks                                  : Completed Slow
[     3.728] op_run_microtasks                                  : Dispatched Slow
[     3.728] op_run_microtasks                                  : Completed Slow
```
2024-10-19 08:42:59 +05:30
Nathan Whitaker
d48434e91f
fix(ext/node): stub HTTPParser internal binding (#26401)
Fixes https://github.com/denoland/deno/issues/26394.
2024-10-18 21:39:32 +00:00
Nathan Whitaker
1bccf45ecb
fix(ext/node): properly map reparse point error in readlink (#26375) 2024-10-18 10:11:06 -07:00
Yoshiya Hinosawa
765c481cb8
Merge branch 'main' into support_create_connection 2024-10-18 16:17:15 +09:00
Bartek Iwańczuk
8f3eb9d0e7
fix(ext/node): add null check for kStreamBaseField (#26368)
It's not guaranteed that `kStreamBaseField` is not undefined, so
added a check for it.

Closes https://github.com/denoland/deno/issues/26363
2024-10-17 22:57:05 +00:00
Nathan Whitaker
9fde5cb5e0
fix(node/fs): copyFile with COPYFILE_EXCL should not throw if the destination doesn't exist (#26360)
Fixes #26313.

We were checking for the NotFound error, but still calling the callback
with the error / throwing.
2024-10-17 19:51:15 +00:00
Satya Rohith
996eec1276
remove stale code 2024-10-18 00:08:14 +05:30
Satya Rohith
a3c6964bd9
fix flow control 2024-10-17 23:44:16 +05:30
Satya Rohith
8c0f6b02c2
chore: make race condition more visible by removing op_node_http_wait_for_connection 2024-10-17 20:53:49 +05:30
Yoshiya Hinosawa
d6af2c918c
Merge branch 'main' into support_create_connection
Signed-off-by: Yoshiya Hinosawa <stibium121@gmail.com>
2024-10-17 22:18:36 +09:00
Yoshiya Hinosawa
781eb00dfe
fix race condition of _bodyWriter creation and end() call 2024-10-17 22:12:07 +09:00
Nathan Whitaker
a61ba3c699
fix(net): don't try to set nodelay on upgrade streams (#26342)
Fixes https://github.com/denoland/deno/issues/26341.

We try to call `op_set_nodelay` on an `UpgradeStream`, which doesn't
support that operation.
2024-10-17 03:56:57 +00:00
Yoshiya Hinosawa
33a852ea2b
fix lint errors 2024-10-17 12:33:11 +09:00
Yoshiya Hinosawa
50e66dacfa
fmt 2024-10-17 12:23:04 +09:00
Yoshiya Hinosawa
4915f342b0
remove debugs 2024-10-17 12:15:27 +09:00
Yoshiya Hinosawa
16be0202ca
Merge branch 'main' into support_create_connection
Signed-off-by: Yoshiya Hinosawa <stibium121@gmail.com>
2024-10-17 12:00:57 +09:00
Nathan Whitaker
167f674c7c
fix: don't warn on ignored signals on windows (#26332)
Closes #26183.

The warnings are super noisy and not actionable for the user
2024-10-16 19:58:11 -07:00
Nathan Whitaker
458d6278d2
fix(node/http): normalize header names in ServerResponse (#26339)
Fixes https://github.com/denoland/deno/issues/26115.

We weren't normalizing the headers to lower case, so code that attempted
to delete the `Content-Length` header (but used a different case) wasn't
actually removing the header.
2024-10-17 00:42:15 +00:00
Satya Rohith
be6a1baa06
Merge remote-tracking branch 'upstream/main' into support_create_connection 2024-10-17 01:32:10 +05:30
Nathan Whitaker
f7dba52133
fix(child_process): map node --no-warnings flag to --quiet (#26288)
Closes https://github.com/denoland/deno/issues/25899
2024-10-16 18:25:25 +00:00
Yoshiya Hinosawa
d59599fc18
fix(ext/node): fix dns.lookup result ordering (#26264)
partially unblocks #25470

This PR aligns the resolution of `localhost` hostname to Node.js
behavior.

In Node.js `dns.lookup("localhost", (_, addr) => console.log(addr))`
prints ipv6 address `::1`, but it prints ipv4 address `127.0.0.1` in
Deno. That difference causes some errors in the work of enabling
`createConnection` option in `http.request` (#25470). This PR fixes the
issue by aligning `dns.lookup` behavior to Node.js.

This PR also changes the following behaviors (resolving TODOs):
- `http.createServer` now listens on ipv6 address `[::]` by default on
linux/mac
- `net.createServer` now listens on ipv6 address `[::]` by default on
linux/mac

These changes are also alignments to Node.js behaviors.
2024-10-16 20:58:44 +09:00
Divy Srivastava
21fa953f32
fix(ext/node): timingSafeEqual account for AB byteOffset (#26292)
Fixes https://github.com/denoland/deno/issues/26276
2024-10-16 14:27:28 +05:30
Kenta Moriuchi
7211028f1e
fix(ext/node): use primordials in ext/node/polyfills/internal/buffer.mjs (#24993)
Towards #24236
2024-10-15 20:32:27 +02:00
Toby Ealden
c5a7f98d82
fix(ext/node): handle http2 server ending stream (#26235)
Closes #24845
2024-10-15 23:35:10 +05:30
Satya Rohith
881fe46d3f
fix double request initiation 2024-10-15 16:46:09 +05:30
Divy Srivastava
c7153838ec
fix(ext/node): implement TCP.setNoDelay (#26263)
Fixes https://github.com/denoland/deno/issues/26177

The significant delay was caused by Nagel's algorithm + delayed ACKs in
Linux kernels. Here's the [kernel
patch](https://lwn.net/Articles/502585/) which added 40ms
`tcp_default_delack_min`

```
$ deno run -A pg-bench.mjs # main
Tue Oct 15 2024 12:27:22 GMT+0530 (India Standard Time): 42ms

$ target/release/deno run -A pg-bench.mjs # this patch
Tue Oct 15 2024 12:28:02 GMT+0530 (India Standard Time): 1ms
```

```js
import { Buffer } from "node:buffer";

import pg from 'pg'
const { Client } = pg
const client = new Client({
    connectionString: 'postgresql://postgres:postgres@127.0.0.1:5432/postgres'
})
await client.connect()

async function fetch() {
    const startPerf = performance.now();
    const res = await client.query(`select
        $1::int as int,
        $2 as string,
        $3::timestamp with time zone as timestamp,
        $4 as null,
        $5::bool as boolean,
        $6::bytea as bytea,
        $7::jsonb as json
      `, [
        1337,
        'wat',
        new Date().toISOString(),
        null,
        false,
        Buffer.from('awesome'),
        JSON.stringify([{ some: 'json' }, { array: 'object' }])
    ])
    console.log(`${new Date()}: ${Math.round(performance.now() - startPerf)}ms`)
}

for(;;) await fetch();
```
2024-10-15 14:47:12 +05:30
Satya Rohith
d19c5a73e1
wip 2024-10-14 19:28:59 +05:30
Satya Rohith
3c9ac413c7
https post request work 2024-10-14 19:28:53 +05:30
Divy Srivastava
68b388a93a
fix(ext/node): allow writing to tty columns (#26201)
Behave similar to Node.js where modifying `stdout.columns` doesn't
really resize the terminal. Ref
https://github.com/nodejs/node/issues/17529

Fixes https://github.com/denoland/deno/issues/26196
2024-10-14 14:00:02 +05:30
Nathan Whitaker
4f89225f76
fix(node/util): export styleText from node:util (#26194)
Fixes #26184.

It was added but not publicly exported.
2024-10-12 19:36:23 +00:00
Satya Rohith
e8ee448e23
Merge branch 'main' into support_create_connection 2024-10-12 17:45:53 +05:30
Marvin Hagemeister
9117a9a43c
fix(node): make process.stdout.isTTY writable (#26130)
Fixes https://github.com/denoland/deno/issues/26123
2024-10-11 19:14:10 +02:00
Satya Rohith
ff4e682ff9
fix(ext/node): internal buffer length in readSync (#26064)
Closes https://github.com/denoland/deno/issues/26054
2024-10-08 10:41:32 +00:00
Satya Rohith
9ec31c6714
Merge branch 'main' into support_create_connection 2024-10-08 10:38:39 +05:30
Nathan Whitaker
dd8cbf5e29
fix(node): fix worker_threads issues blocking Angular support (#26024)
Fixes #22995. Fixes #23000.

There were a handful of bugs here causing the hang (each with a
corresponding minimized test):

- We were canceling recv futures when `receiveMessageOnPort` was called,
but this caused the "receive loop" in the message port to exit. This was
due to the fact that `CancelHandle`s are never reset (i.e., once you
`cancel` a `CancelHandle`, it remains cancelled). That meant that after
`receieveMessageOnPort` was called, the subsequent calls to
`op_message_port_recv_message` would throw `Interrupted` exceptions, and
we would exit the loop.

The cancellation, however, isn't actually necessary.
`op_message_port_recv_message` only borrows the underlying port for long
enough to poll the receiver, so the borrow there could never overlap
with `op_message_port_recv_message_sync`.

- Calling `MessagePort.unref()` caused the "receive loop" in the message
port to exit. This was because we were setting
`messageEventListenerCount` to 0 on unref. Not only does that break the
counter when multiple `MessagePort`s are present in the same thread, but
we also exited the "receive loop" whenever the listener count was 0. I
assume this was to prevent the recv promise from keeping the event loop
open.

Instead of this, I chose to just unref the recv promise as needed to
control the event loop.

- The last bug causing the hang (which was a doozy to debug) ended up
being an unfortunate interaction between how we implement our
messageport "receive loop" and a pattern found in `npm:piscina` (which
angular uses). The gist of it is that piscina uses an atomic wait loop
along with `receiveMessageOnPort` in its worker threads, and as the
worker is getting started, the following incredibly convoluted series of
events occurs:
   1. Parent sends a MessagePort `p` to worker
   2. Parent sends a message `m` to the port `p`
3. Parent notifies the worker with `Atomics.notify` that a new message
is available
   4. Worker receives message, adds "message" listener to port `p`
   5. Adding the listener triggers `MessagePort.start()` on `p`
6. Receive loop in MessagePort.start receives the message `m`, but then
hits an await point and yields (before dispatching the "message" event)
7. Worker continues execution, starts the atomic wait loop, and
immediately receives the existing notification from the parent that a
message is available
8. Worker attempts to receive the new message `m` with
`receiveMessageOnPort`, but this returns `undefined` because the receive
loop already took the message in 6
9. Atomic wait loop continues to next iteration, waiting for the next
message with `Atomic.wait`
10. `Atomic.wait` blocks the worker thread, which prevents the receive
loop from continuing and dispatching the "message" event for the
received message
11. The parent waits for the worker to respond to the first message, and
waits
12. The thread can't make any more progress, and the whole process hangs

The fix I've chosen here (which I don't particularly love, but it works)
is to just delay the `MessagePort.start` call until the end of the event
loop turn, so that the atomic wait loop receives the message first. This
prevents the hang.

---

Those were the main issues causing the hang. There ended up being a few
other small bugs as well, namely `exit` being emitted multiple times,
and not patching up the message port when it's received by
`receiveMessageOnPort`.
2024-10-04 09:26:32 -07:00
Yoshiya Hinosawa
e95d7dfe37
fix lint 2024-10-03 22:32:05 +09:00
Yoshiya Hinosawa
7471848dc4
refactor: dedupe op_node_http_request_with_conn and op_node_http_request_with_tls_conn 2024-10-03 21:42:51 +09:00
Yoshiya Hinosawa
a372ed518b
refactor: use _encrypted field 2024-10-03 19:04:45 +09:00
Yoshiya Hinosawa
6b6669f0df
chore: fix lint 2024-10-03 18:42:33 +09:00
Satya Rohith
007b5265b8
fix: support https requests
Co-authored-by: Yoshiya Hinosawa <stibium121@gmail.com>
2024-10-03 11:28:23 +05:30
Yoshiya Hinosawa
8714e718bc
enable another case 2024-10-03 13:31:59 +09:00
Yoshiya Hinosawa
cf6ae67a31
refactor: reduce duplication 2024-10-03 12:05:04 +09:00
Satya Rohith
32c1278736
feat(ext/node): buffer.transcode() (#25972)
Closes https://github.com/denoland/deno/issues/25911
2024-10-02 08:23:14 +00:00
Divy Srivastava
620e6b43a6
fix(ext/node): remove unimplemented promiseHook stubs (#25979)
`temporalio` sdk [will try to
use](faa64225a7/packages/worker/src/workflow/vm-shared.ts (L199-L202))
promiseHook if it is found. This patch removes the unimplemented stubs.

```ts
    if (promiseHooks) {
      // Node >=16.14 only
      this.stopPromiseHook = promiseHooks.createHook({
        init: (promise: Promise<unknown>, parent: Promise<unknown>) => {
```

Fixes https://github.com/denoland/deno/issues/25977
2024-10-02 12:52:05 +05:30
Yoshiya Hinosawa
e60e727e5e
enable another test case 2024-10-02 00:40:56 +09:00