V4.3.1 proposal #5200
V4.3.1 proposal #5200
Conversation
MylesBorins
force-pushed
the
v4.3.1-proposal
branch
2 times, most recently
from
5bdfd82
to
b46ee1d
Compare
|
CI: https://ci.nodejs.org/job/node-test-pull-request/1642/ edit: ARM CI failure is known flaky test. Fix should land in next release edit: citgm is green! (known failure on ppc) |
mscdex
added
the
meta
|
Build Job for rc1: https://ci.nodejs.org/job/iojs+release/393/ |
|
LGTM if CI is green. |
|
Failures on windows Digging into npm tree changes |
|
oy.. I think this has come up before. /cc @zkat |
|
Yeaaaaah. It's tricky cause we can't just flatten the tree -- I'll talk with folks tomorrow and see if we can come up with a solution that fixes this without breaking everything else. |
Lastly, @thealphanerd, how's the cherry-picking going? I'm experiencing a little bit of pain when doing |
Lastly, not real problems with cherry-picking tbqh. I wonder if the staging branch has made things a bit simpler? |
|
+1 to being conservative with the perf related fixes. I want to make sure we start doing a RC cycle with every LTS release so we can catch issues faster but we definitely need everyone to help test to make that effective. I think the cherry-picking in v4.x is going fairly smooth because we are also avoiding the semver-minor bumps. |
|
So it looks like there are a handful of things going on with npm 4.2.4 had a similar issue but managed to avoid it via deduping The same dedupe is not working anymore. These problems are present in I have sent a PR over to gauge that updates the lodash dependency. The updated lodash modules have significantly shallower trees, doing away with the two modules that are being repeated and not deduping. This alone may fix the problem. It also appears that At this point I am ready to revert the npm updates so that we can get and RC out... but I would really like to avoid that as there are some bugs that are being fixed by the npm update /cc @nodejs/npm |
|
It looks like there is an open PR to update npmlog in node-gyp. @rvagg can you escalate this? |
|
It looks like npm-install-checks that is present in npm v2 is two versions behind. The latest version doesn't even have edit: talked to @zkat version 2 - 3 do not work with npm@2. It might be possible to release a new version of the 1.x line with an updated npmlog |
MylesBorins
force-pushed
the
v4.3.1-proposal
branch
3 times, most recently
from
c172a6a
to
4a4661a
Compare
|
So i've gone through and removed the updates to npm for now. I've also backed out c8cc179 as requested by @rvagg ci: https://ci.nodejs.org/job/node-test-pull-request/1648/ edit: edit: |
|
fwiw, working on getting a semver-minor update of node-gyp out in the next few days with that npmlog fix plus hopefully nodejs/node-gyp#877 and nodejs/node-gyp#878. |
MylesBorins
force-pushed
the
v4.3.1-proposal
branch
from
4a4661a
to
5dd0c2f
Compare
|
rc2 build in progress: https://ci.nodejs.org/job/iojs+release/397/ |
|
@nodejs/collaborators ... just a heads up... please help us test this release candidate for v4.3.1 : https://nodejs.org/download/rc/v4.3.1-rc.2/ |
|
Hey, I'd like to lobby for including #4482 in this release. We're hitting the idle socket timeout this commit fixes, and currently have to cherry pick this commit into Node v4 and build from source as a workaround. |
|
@mikemorris this commit will be coming very soon in v4.4.0 We are trying to get 4.3 as stable as possible with this update and then as quickly as possible get out a release with a few semver minor addition, including #4482 |
It's possible there is already an existing error on OpenSSL's error stack that is unrelated to the EVP_DigestInit_ex() operation we just executed. Fixes: nodejs#4221 PR-URL: nodejs#4731 Reviewed-By: Colin Ihrig <[email protected]> Reviewed-By: Fedor Indutny <[email protected]>
Fix a segmentation fault when the debug message handler was called from a context without an associated `node::Environment`. Fixes: nodejs#4261 Fixes: nodejs#4322 PR-URL: nodejs#4328 Reviewed-By: Evan Lucas <[email protected]> Reviewed-By: James M Snell <[email protected]>
PR-URL: nodejs#4194 Reviewed-By: James M Snell <[email protected]> Reviewed-By: Jeremiah Senkpiel <[email protected]>
Remove a comment that has a word 'XXX'. And add a line to output debuglog of error. PR-URL: nodejs#4690 Reviewed-By: Colin Ihrig <[email protected]> Reviewed-By: James M Snell <[email protected]>
PR-URL: nodejs#4691 Reviewed-By: Colin Ihrig <[email protected]> Reviewed-By: Evan Lucas <[email protected]> Reviewed-By: Jeremiah Senkpiel <[email protected]>
Refs: nodejs#4642 PR-URL: nodejs#4719 Reviewed-By: Colin Ihrig <[email protected]> Reviewed-By: Rich Trott <[email protected]>
Add `servername` parameter docs for `https.request()` method. Follows nodejs#4389 PR-URL: nodejs#4729 Reviewed-By: James M Snell <[email protected]> Reviewed-By: Fedor Indutny <[email protected]> Reviewed-By: Roman Reiss <[email protected]>
Previously, test-cluster-disconnect-leak had two issues: * Magic numbers: How many times to spawn a worker was determined through empirical experimentation. This means that as new platforms and new CPU/RAM configurations are tested, the magic numbers require more and more refinement. This brings us to... * Non-determinism: The test *seems* to fail all the time when the bug it tests for is present, but it's really a judgment based on sampling. "Oh, with 8 workers per CPU, it fails about 80% of the time. Let's try 16..." This revised version of the test takes a different approach. The fix for the bug that the test was written for means that the `disconnect` event will fire reliably for a single worker. So we check for that and the test still fails when the fix is not in the code base and succeeds when it is. Advantages of this approach include: * The test runs much faster. * The test now works on Windows. The previous version skipped Windows. * The test should be reliable on any new platform regardless of CPU and RAM. Ref: nodejs#4674 PR-URL: nodejs#4736 Reviewed-By: Ben Noordhuis <[email protected]> Reviewed-By: James M Snell <[email protected]> Reviewed-By: Johan Bergström <[email protected]>
Make the byteLength work correctly when input is Buffer. e.g: ```js // The incomplete unicode string Buffer.byteLength(new Buffer([0xe4, 0xb8, 0xad, 0xe6, 0x96])) ``` The old output: 9 The new output: 5 PR-URL: nodejs#4738 Reviewed-By: Ben Noordhuis <[email protected]> Reviewed-By: Brian White <[email protected]> Reviewed-By: James M Snell <[email protected]> Reviewed-By: Colin Ihrig <[email protected]>
Previously, test-cluster-disconnect-suicide-race had two issues: * Magic numbers: How many times to spawn a worker was determined through empirical experimentation. This means that as new platforms and new CPU/RAM configurations are tested, the magic numbers require more and more refinement. This brings us to... * Non-determinism: The test seems to fail all the time when the bug it tests for is present, but it's really a judgment based on sampling. "Oh, with 8 workers per CPU, it fails about 80% of the time. Let's try 16..." This revised version of the test takes a different approach. The fix for the bug that the test was written for means that the disconnect event will fire on a subsequent tick. So we check for that and the test still fails when the fix is not in the code base and succeeds when it is. Advantages of this approach include: * The test runs much faster. * The test should be reliable on any new platform regardless of CPU and RAM. PR-URL: nodejs#4739 Ref: nodejs#4674 Reviewed-By: Ben Noordhuis <[email protected]> Reviewed-By: James M Snell <[email protected]>
- Hide the scrollbar on the TOC on all browsers. It was never the intention for it to be visible with the scroll indication in place. A wrapper element with 20px padding was added to accommodate for hopefully all scrollbar widths as well as to avoid overflowing content. - Fixed the scroll indication gradient on Safari, which was caused by the wrong from-color, which now matches the to-color. - Fixed a issue in old IE where the TOC didn't render on the correct position through setting `left: 0` and `top: 0` on it. PR-URL: nodejs#4748 Reviewed-By: James M Snell <[email protected]>
PR-URL: nodejs#4753 Reviewed-By: Colin Ihrig <[email protected]> Reviewed-By: Evan Lucas <[email protected]>
Ref: http://eslint.org/docs/rules/space-in-parens.html PR-URL: nodejs#4753 Reviewed-By: Colin Ihrig <[email protected]> Reviewed-By: Evan Lucas <[email protected]>
rust-lang/prev.rust-lang.org#288 landed in the Rust repo so it seems like a good idea to just bring the updated list in. We also received a request to do this in nodejs/inclusivity#82 so this should resolve that. Thanks to [@Charlotteis](https://github.com/Charlotteis) for bringing up the original issue. Reviewed-By: Jeremiah Senkpiel <[email protected]> Reviewed-By: Myles Borins <[email protected]> Reviewed-By: James M Snell <[email protected]> Reviewed-By: Сковорода Никита Андреевич <[email protected]> PR-URL: nodejs#4801 Fixes: nodejs/inclusivity#82
Replace grep with awk to add support for subkeys PR-URL: nodejs#4807 Reviewed-By: Rod Vagg <[email protected]>
In some conditions it can happen that the client-side socket is destroyed before the server-side socket has gracefully closed, thus causing a 'ECONNRESET' error in this socket. To solve this, wait in the client-side socket for the 'end' event before closing it. PR-URL: nodejs#4043 Reviewed-By: James M Snell <[email protected]> Reviewed-By: Ben Noordhuis <[email protected]>
Currently a debug context is created for various calls to util. If the node debugger is being run the main context is the debug context. In this case node_contextify was freeing the debug context and causing everything to explode. This change moves around the logic and no longer frees the context. There is a concern about the dangling pointer The regression test was adapted from code submitted by @3y3 in nodejs#4815 Fixes: nodejs#4440 Fixes: nodejs#4815 Fixes: nodejs#4597 Fixes: nodejs#4952 PR-URL: nodejs#4815 Reviewed-By: Fedor Indutny <[email protected]> Reviewed-By: Ben Noordhuis <[email protected]> Reviewed-By: Rich Trott <[email protected]>
MylesBorins
force-pushed
the
v4.3.1-proposal
branch
from
5dd0c2f
to
7ee5216
Compare
|
CI: https://ci.nodejs.org/job/node-test-pull-request/1670/ edit: CI failure is infra related |
MylesBorins
force-pushed
the
v4.3.1-proposal
branch
from
7ee5216
to
b321f13
Compare
|
rc.3 build: https://ci.nodejs.org/job/iojs+release/398/ |
|
As we have not had any problems with any of the rc's I am going to move forward with this release in the next two hours. Last chance to raise any concerns @nodejs/collaborators |
Notable changes: * buffer: make byteLength work with Buffer correctly (Jackson Tian) - nodejs#4738 * debugger: guard against call from non-node context (Ben Noordhuis) - nodejs#4328 * node_contextify: do not incept debug context (Myles Borins) - nodejs#4819 * deps: update to http-parser 2.5.2 (James Snell) - nodejs#5238
MylesBorins
force-pushed
the
v4.3.1-proposal
branch
from
b321f13
to
ef37a46
Compare
|
@thealphanerd #5168 is a segfault fix, but it's a rare case and perhaps not worth the rush or delaying the release. |
|
@ChALkeR I feel like that commit is still too fresh. I'll make sure to get it in the v4.4.0-rc |
|
@thealphanerd I shortened the workspace of release machines from |
|
@joaocgreis thanks for getting to that! I'm going to avoid rolling the npm updates into this release as we are not assured that the path issues are solved after install. I'll roll all those updates into the next RC though! |
|
Release Build: https://ci.nodejs.org/job/iojs+release/404/ |
Notable changes: * buffer: make byteLength work with Buffer correctly (Jackson Tian) - #4738 * debugger: guard against call from non-node context (Ben Noordhuis) - #4328 * node_contextify: do not incept debug context (Myles Borins) - #4819 * deps: update to http-parser 2.5.2 (James Snell) - #5238 PR-URL: #5200 (comment)
Notable changes
Commits
748d2b4de1] - buffer: make byteLength work with Buffer correctly (Jackson Tian) #4738fb615bdaf4] - buffer: remove unnecessary TODO comments (Peter Geiss) #4719b8213ba7e1] - cluster: ignore queryServer msgs on disconnection (Santiago Gimeno) #4465f8a676ed59] - cluster: fix race condition setting suicide prop (Santiago Gimeno) #43499d4a226dad] - crypto: clear error stack in ECDH::Initialize (Fedor Indutny) #4689583f3347d8] - debugger: remove variable redeclarations (Rich Trott) #4633667f7a7ab3] - debugger: guard against call from non-node context (Ben Noordhuis) #4328188cff3c31] - deps: update to http-parser 2.5.2 (James Snell) #52386e829b44e3] - dgram: prevent disabled optimization of bind() (Brian White) #4613c3956d05b1] - doc: update list of personal traits in CoC (Kat Marchán) #480139cb69ca21] - doc: style fixes for the TOC (Roman Reiss) #4748cb5986da81] - doc: addservernameparameter docs (Alexander Makarenko) #472991066b5f34] - doc: update branch-diff arguments in release doc (Rod Vagg) #46919ca24de41d] - doc: add docs for more stream options (zoubin) #4639437d0e336d] - doc: mention that http.Server inherits from net.Server (Ryan Sobol) #4455393e569160] - doc: copyedit setTimeout() documentation (Rich Trott) #4434e2a682ecc3] - doc: fix formatting in process.markdown (Rich Trott) #443375b0ea85bd] - doc: add path property to Write/ReadStream in fs.markdown (Claudio Rodriguez) #436848c2783421] - doc: add docs working group (Bryan English) #4244c0432e9f56] - doc: restore ICU third-party software licenses (Richard Lau) #476236a4159dab] - doc: rebuild LICENSE using tools/license-builder.sh (Rod Vagg) #4194a2998a1bce] - gitignore: never ignore debug module (Michaël Zasso) #2286661b2557d9] - http: remove variable redeclaration (Rich Trott) #46121bb2967d48] - http: fix non-string header value concatenation (Brian White) #446015ed64e34c] - lib: fix style issues after eslint update (Michaël Zasso) nodejs/io.js#22862e92a1a6b4] - module: move unnecessary work for early return (Andres Suarez) #357940c8e6d75d] - net: remove hot path comment from connect (Evan Lucas) #46488ed0c1c22c] - net: fix dns lookup for android (Josh Dague) #458015fa555204] - net, doc: fix line wrapping lint in net.js (James M Snell) #45881b070e48e0] - node_contextify: do not incept debug context (Myles Borins) #48154fbcb47fe9] - readline: Remove XXX and output debuglog (Kohei TAKATA) #469026f02405d0] - repl: make sure historyPath is trimmed (Evan Lucas) #45395990ba2a0a] - src: remove redeclarations of variables (Rich Trott) #4605c41ed59dbc] - src: don't check failure with ERR_peek_error() (Ben Noordhuis) #4731d71f9992f9] - stream: remove useless if test in transform (zoubin) #4617f205e9920e] - test: fix tls-no-rsa-key flakiness (Santiago Gimeno) #4043447347cd62] - test: fix issues for space-in-parens ESLint rule (Roman Reiss) #4753be8274508c] - test: improve test-cluster-disconnect-suicide-race (Rich Trott) #47390178001163] - test: make test-cluster-disconnect-leak reliable (Rich Trott) #4736d615757da2] - test: fix flaky test-net-socket-local-address (cjihrig) #4650baa0a3dff5] - test: fix race in test-net-server-pause-on-connect (Rich Trott) #4637909b5167cb] - test: remove 1 second delay from test (Rich Trott) #46168ea76608ed] - test: move resource intensive tests to sequential (Rich Trott) #46157afcdd358e] - test: require common module only once (Rich Trott) #46110e02eb0bbe] - test: only include http module once (Rich Trott) #460634d9e48bb6] - test: fixhttp-upgrade-clientflakiness (Santiago Gimeno) #4602556703d531] - test: fix flaky unrefed timers test (Rich Trott) #45993d5bc69796] - test: fixhttp-upgrade-agentflakiness (Santiago Gimeno) #4520ec24d3767b] - test: fix flaky test-cluster-shared-leak (Rich Trott) #4510a256790327] - test: fix flaky cluster-net-send (Brian White) #44446809c2be1a] - test: fix flaky child-process-fork-regr-Segfault in node::StreamBase::GetFD #2847 (Brian White) #4442e6448aa36b] - test: use addon.md block headings as test dir names (Rod Vagg) #4412305d340fca] - test: test each block in addon.md contains js & cc (Rod Vagg) #4411f213406575] - test: fix tls-multi-key race condition (Santiago Gimeno) #3966607f545568] - test: fix style issues after eslint update (Michaël Zasso) nodejs/io.js#2286aefb20a94f] - tls: copy client CAs and cert store on CertCb (Fedor Indutny) #35377821b3e305] - tls_legacy: do not read on OpenSSL's stack (Fedor Indutny) #4624b66db49f94] - tools: add support for subkeys in release tools (Myles Borins) #4807837ebd1985] - tools: enable space-in-parens ESLint rule (Roman Reiss) #4753066d5e7da2] - tools: fix style issue after eslint update (Michaël Zasso) nodejs/io.js#2286b20ea69f46] - tools: update eslint config (Michaël Zasso) nodejs/io.js#22862e0352d50c] - tools: update eslint to v1.10.3 (Michaël Zasso) nodejs/io.js#2286c96800a432] - tools: fix license-builder.sh for ICU (Richard Lau) #4762720b03dca7] - tools: add license-builder.sh to construct LICENSE (Rod Vagg) #4194