deps: update nghttp3 to 1.0.0

[nodejs-github-bot]

This is an automated update of nghttp3 to 1.0.0.

[lpinca]

RSLGTM

[VoltrexKeyva]

RSLGTM

[nodejs-github-bot]

Commit Queue failed

- Loading data for nodejs/node/pull/47576
✔  Done loading data for nodejs/node/pull/47576
----------------------------------- PR info ------------------------------------
Title      deps: update nghttp3 to 0.10.0 (#47576)
   ⚠  Could not retrieve the email or name of the PR author's from user's GitHub profile!
Branch     nodejs-github-bot:actions/tools-update-nghttp3 -> nodejs:main
Labels     needs-ci, quic, dependencies, dont-land-on-v14.x
Commits    1
 - deps: update nghttp3 to 0.10.0
Committers 1
 - Node.js GitHub Bot 
PR-URL: https://github.com/nodejs/node/pull/47576
Reviewed-By: Luigi Pinca 
Reviewed-By: Mohammed Keyvanzadeh 
Reviewed-By: Tobias Nießen 
Reviewed-By: Marco Ippolito 
Reviewed-By: James M Snell 
------------------------------ Generated metadata ------------------------------
PR-URL: https://github.com/nodejs/node/pull/47576
Reviewed-By: Luigi Pinca 
Reviewed-By: Mohammed Keyvanzadeh 
Reviewed-By: Tobias Nießen 
Reviewed-By: Marco Ippolito 
Reviewed-By: James M Snell 
--------------------------------------------------------------------------------
   ℹ  This PR was created on Sun, 16 Apr 2023 00:24:33 GMT
   ✔  Approvals: 5
   ✔  - Luigi Pinca (@lpinca): https://github.com/nodejs/node/pull/47576#pullrequestreview-1386703737
   ✔  - Mohammed Keyvanzadeh (@VoltrexKeyva): https://github.com/nodejs/node/pull/47576#pullrequestreview-1386816166
   ✔  - Tobias Nießen (@tniessen) (TSC): https://github.com/nodejs/node/pull/47576#pullrequestreview-1387101562
   ✔  - Marco Ippolito (@marco-ippolito): https://github.com/nodejs/node/pull/47576#pullrequestreview-1387635921
   ✔  - James M Snell (@jasnell) (TSC): https://github.com/nodejs/node/pull/47576#pullrequestreview-1389053297
   ✔  Last GitHub CI successful
   ✘  No Jenkins CI runs detected
--------------------------------------------------------------------------------
   ✔  Aborted `git node land` session in /home/runner/work/node/node/.ncu

https://github.com/nodejs/node/actions/runs/4805669128

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/51475/

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/51564/

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/51866/

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/51868/

[tniessen]

Just making sure this doesn't land as-is

[fasenderos]

That's really strange. I can't explain why, cause locally never happens, but the job on 30th April that try to update ngtcp2 5e7d37e and nghttp3 69f8b22 create duplicated code.

nghttp3 on file deps/ngtcp2/nghttp3/lib/includes/nghttp3/nghttp3.h

node/deps/ngtcp2/nghttp3/lib/includes/nghttp3/nghttp3.h

Lines 2580 to 2596 in 69f8b22

	/**
	* @function
	*
	* `nghttp3_conn_is_drained` returns nonzero if
	* `nghttp3_conn_shutdown` has been called, and there is no active
	* remote streams. This function is for server use only.
	*/
	NGHTTP3_EXTERN int nghttp3_conn_is_drained(nghttp3_conn *conn);
	/**
	* @function
	*
	* `nghttp3_conn_is_drained` returns nonzero if
	* `nghttp3_conn_shutdown` has been called, and there is no active
	* remote streams. This function is for server use only.
	*/
	NGHTTP3_EXTERN int nghttp3_conn_is_drained(nghttp3_conn *conn);

ngtcp2 on file deps/ngtcp2/ngtcp2/lib/ngtcp2_cc.c (#47578)

node/deps/ngtcp2/ngtcp2/lib/ngtcp2_cc.c

Lines 364 to 382 in 5e7d37e

	cwnd_thres =
	(target * (((t + cstat->smoothed_rtt) << 10) / NGTCP2_SECONDS)) >> 10;
	if (cwnd_thres < cstat->cwnd) {
	target = cstat->cwnd;
	} else if (2 * cwnd_thres > 3 * cstat->cwnd) {
	target = cstat->cwnd * 3 / 2;
	} else {
	target = cwnd_thres;
	}
	cwnd_thres =
	(target * (((t + cstat->smoothed_rtt) << 10) / NGTCP2_SECONDS)) >> 10;
	if (cwnd_thres < cstat->cwnd) {
	target = cstat->cwnd;
	} else if (2 * cwnd_thres > 3 * cstat->cwnd) {
	target = cstat->cwnd * 3 / 2;
	} else {
	target = cwnd_thres;
	}

Of course both duplicate codes never existed in the depencies repos

Last but not least the job that in theory should have skipped the update, because locally it actually skips the update, fix these differences with a second commit for the same version (on 7th May)

[mhdawson]

@marco-ippolito wondering if you are still looking at this issue? I wonder if it might be at all related to the same branch being used by the updater each time. Have we tried closing this PR, deleting the branch and seeing if the problem re-occurs in the newly created PR?

[marco-ippolito]

I totally missed this notification, I agree we should delete it and see what happens with the update. I had look at the code and tested on my machine and it was all working fine

[mhdawson]

Closing and will delete branch, then we can see what happens

[mhdawson]

Branch deleted