VS 2015 support #478

rvagg · 2015-01-17T07:30:38Z

Not quite working yet, it gets to the openssl-cli compile and fails with a linking error:

apps.obj : error LNK2001: unresolved external symbol ___iob_func [C:\io.js\deps\openssl\openssl-cli.vcxproj]
C:\io.js\Release\openssl-cli.exe : fatal error LNK1120: 1 unresolved externals [C:\io.js\deps\openssl\openssl-cli.vcxproj]

This PR upgrades GYP to the latest which understands 2015, and has a lot more changes because there hasn't been an upgrade in a while I suppose.

I've also edited vcbuild.bat to do the appropriate checks.

rvagg · 2015-01-17T07:31:29Z

https://jenkins-iojs.nodesource.com/job/iojs+any-pr+multi/85/

which includes a VS2015 host

rvagg · 2015-01-17T07:49:27Z

new build: https://jenkins-iojs.nodesource.com/job/iojs+any-pr+multi/88/

I don't understand the Python stack trace for the OSX failures, anybody care to try and enlighten me on that one?

chrisdickinson · 2015-01-17T11:08:49Z

Re: python: something in here is adding a non-string-y item to the env dict (likely it's adding None), and re.sub is exploding when it's handed that value.

bnoordhuis · 2015-01-17T11:11:37Z

@rvagg Can you mention the gyp commit (e.g. r1234) it's been upgraded to? I'll see if I can reproduce the OS X failure on my 10.8 machine.

EDIT: No, seems to be working for me.

rvagg · 2015-01-17T11:19:03Z

@bnoordhuis I would have put a version in there if I knew where to find one, I went looking but alas.. Any hints on where to find a version for this?

bnoordhuis · 2015-01-17T11:23:42Z

@rvagg I don't know where you got it from but if it's from subversion, then it's the svn commit. If it's git, it's the number from git-svn-id. If this upgrade is to today's HEAD, that would make it r2024.

@chrisdickinson The stack trace suggests it's from here but it looks to me that additional_settings[k] has been sanitized at that point. Strange.

chrisdickinson · 2015-01-17T11:42:19Z

@bnoordhuis I thought so too at first, but additional_settings gets updated with the contents of env.

rvagg · 2015-01-17T11:44:22Z

I printed them out and it is failing on a None, SDKROOT is the one it's borking on.

rvagg · 2015-01-17T11:46:48Z

note this machine only has the commandline tools, not full xcode, which I'm guessing is partly to blame here, still a bug though

rvagg · 2015-01-17T13:00:41Z

figured it out, needs a new entry in common.gypi:

'SDKROOT': 'macosx'

building here: https://jenkins-iojs.nodesource.com/job/iojs+any-pr+multi/90/

bnoordhuis · 2015-01-17T22:21:48Z

LGTM. Maybe you can move the change to common.gypi into a separate commit with an explanation of why it's needed (and move it before the gyp upgrade to keep git bisect happy.)

The iojs-win2012r2-msvs2015 failure seems to be related to this. I'm not sure why OpenSSL does that but I think we can just rip it out, the only part of OpenSSL that uses stdio (in our configuration) are the apps.

@piscisaureus or @seishun Agree/disagree?

rvagg · 2015-01-19T12:17:36Z

Good guess @bnoordhuis, when I fix the define to exclude MSVS 2015 it compiles fine and tests pass with only 1 failure (a TLS failure mind you...). See 7b66043.

Re-running tests in CI here: https://jenkins-iojs.nodesource.com/job/iojs+any-pr+multi/94/nodes=iojs-win2012r2-msvs2015/console

rvagg · 2015-01-19T12:17:59Z

oh, that was an all-pass that time, neat.

bnoordhuis · 2015-01-19T12:55:10Z

@rvagg It looks like it's running the message tests only.

rvagg · 2015-01-19T12:58:54Z

rvagg · 2015-01-19T13:00:33Z

Ugh, github on phone, I shouldn't try and be productive. Will look in to this tomorrow.

seishun · 2015-01-20T13:51:21Z

I don't like the idea of floating patches to OpenSSL to support a preview version of Visual Studio. Maybe it's better to wait for OpenSSL to fix it on their end?

rvagg · 2015-01-20T23:31:41Z

Yes, point taken. I was surprised that it wasn't already fixed tbh. I'll leave this PR open so ppl know we are tracking towards it but waiting for it to be easier.

Fixes builds where environment variable has backslashes

xforce · 2015-02-04T11:46:44Z

It is fixed in OpenSSL 1.0.2. Patch is from 2 July 2014 http://git.openssl.org/gitweb/?p=openssl.git;a=commitdiff;h=aab3560b65b9254d17770bb6fe3ca7edd7451429

Fishrock123 · 2015-02-04T17:25:43Z

OpenSSL 1.0.2 discussion: #589

Make the order of name, email and other additions simpler to copy paste and/or match with git commit messages. Useful when working with `Reviewed-By`. PR-URL: nodejs#1966 Reviewed-By: Christian Tellnes <christian@tellnes.no> Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>

tmpdir creation only happens for tests that need it. So failure to refresh the temporary directory should result in a failed test. PR-URL: nodejs#1976 Reviewed-By: Rod Vagg <rod@vagg.org> Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>

Futher discussion at nodejs/node-v0.x-archive#25294 PR-URL: nodejs#1926 Reviewed-By: Rod Vagg <rod@vagg.org> Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>

The accessors run inside an implicit HandleScope, there is no need to create a new one. PR-URL: nodejs#2001 Reviewed-By: Trevor Norris <trev.norris@gmail.com>

Re-add the wrapper class id to AsyncWrap instances so they can be tracked directly in a heapdump. Previously the class id was given without setting the heap dump wrapper class info provider. Causing a segfault when a heapdump was taken. This has been added, and the label_ set to the given provider name so each instance can be identified. The id will not be set of the passed object has no internal field count. As the class pointer cannot be retrieved from the object. In order to properly report the allocated size of each class, the new pure virtual method self_size() has been introduces. PR-URL: nodejs#1896 Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>

PR-URL: nodejs#1989 Reviewed-By: Trevor Norris <trev.norris@gmail.com>

PR-URL: nodejs#1943 Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl> Reviewed-By: Rod Vagg <rod@vagg.org> Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>

PR-URL: nodejs#1994 Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

Defer reading until user-land has a chance to add listeners. This allows the TLS wrapper to listen for _tlsError and trigger a clientError event if the socket already has data that could trigger. Fixes: nodejs#1114 PR-URL: nodejs#1496 Reviewed-By: Chris Dickinson <christopher.s.dickinson@gmail.com> Reviewed-By: Fedor Indutny <fedor.indutny@gmail.com>

require() now checks that the path exists before searching further in it. PR-URL: nodejs#1920 Reviewed-By: Isaac Z. Schlueter <i@izs.me> Reviewed-By: Trevor Norris <trev.norris@gmail.com> Reviewed-By: Chris Dickinson <christopher.s.dickinson@gmail.com> Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>

Add link to Jenkins server in Collaborator Guide. PR-URL: nodejs#1995 Reviewed-By: Chris Dickinson <christopher.s.dickinson@gmail.com> Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>

PR-URL: nodejs#2000 Reviewed-By: Yosuke Furukawa <yosuke.furukawa@gmail.com> Reviewed-By: Evan Lucas <evanlucas@me.com> Reviewed-By: Johan Bergström <bugs@bergstroem.nu>

targos · 2015-06-18T18:55:08Z

rvagg@0e54add that works for me (VS 2015 RC on latest Windows 10 build).

But v8 cannot compile because of a bug in VS. There is a workaround since v4.4.42.

ghost · 2015-06-18T21:40:50Z

@targos, I would not consider that as a bug since in the linked Connect bug report, both OP (David) and VC team member (Jonathan) are in agreement that there is no violation of standard.

Note that the other bug in RC is fixed in recent compiler revision (build 19.00.23008, June 10, 2015) http://webcompiler.cloudapp.net/; it compiles with this code:

struct User {
  void *operator new(size_t, unsigned);
  void operator delete(void *);
};

int main()
{
    auto q = new (0) User;
    return 0;
}

targos · 2015-06-18T21:58:37Z

@jasonwilliams200OK I was indeed talking about that other bug. They say it's fixed for RTM but is there a way to get the latest version of the C++ compiler before the release?

ghost · 2015-06-18T22:08:58Z

I think no (that web-compiler is emerged as part of a research project http://rise4fun.com/), but if I venture an educated guess, VS RTM will be available few weeks before Windows 10 release (July 29, 2015), which is about any day now. :)
They are probably also considering to make cl (+link+libs) releases independent of VS release lifecycle, which means we could have win32 (cli-only) server-core running in Docker with cl installed for build stuff.

vcbuild.bat calls python configure before setting GYP_MSVS_VERSION, so SelectVisualStudioVersion (tools\gyp\pylib\gyp\MSVSVersion.py) defaults to 'auto' and selects VS 2005. vcbuild sets the environment in the current shell, so this issue would manifest itself only on the first invocation of the script in any given shell windows. Reviewed-By: Julien Gilli <jgilli@fastmail.fm> PR-URL: nodejs/node-v0.x-archive#20109

rvagg · 2015-06-19T01:06:26Z

pulled in nodejs/node-v0.x-archive@a58b174 as 396b4c6

CI run https://jenkins-iojs.nodesource.com/job/libuv+any-pr+multi/111/

rvagg · 2015-06-19T01:07:23Z

derp https://jenkins-iojs.nodesource.com/job/iojs+any-pr+multi/51/

orangemocha · 2015-06-22T13:52:56Z

Did something get messed up with this PR? GitHub shows 315 commits and 6686 files changed.

rvagg · 2015-06-22T14:03:56Z

yeah, it's really old and targetting the v1.x branch, hence the superfluous commits. It's just the head one that matters.

targos · 2015-06-22T14:04:02Z

that's because the PR targets nodejs:v1.x and it is not possible to change the target branch afterwards. Only the last 2 commits are relevant.

srl295 · 2015-07-30T21:09:11Z

@rvagg Intl won't work under VS2015 because of the same breakage (ICU bug).

rvagg force-pushed the msvs-2015 branch 2 times, most recently from 03614ae to 9e8f118 Compare January 17, 2015 07:47

rvagg force-pushed the msvs-2015 branch from 9e8f118 to 0c77da2 Compare January 17, 2015 11:27

rvagg force-pushed the msvs-2015 branch 3 times, most recently from 54f093d to 21dd523 Compare January 17, 2015 12:57

rvagg force-pushed the msvs-2015 branch from 21dd523 to af453fa Compare January 19, 2015 11:28

rvagg closed this Jan 19, 2015

rvagg reopened this Jan 19, 2015

am11 referenced this pull request in microsoft/openssl Jan 29, 2015

Merge pull request #3 from seecrypt/fix-comment-backslashes

c20688d

Fixes builds where environment variable has backslashes

jbergstroem and others added 12 commits June 17, 2015 10:07

tools: make test-npm work without global npm

91dfb5e

Futher discussion at nodejs/node-v0.x-archive#25294 PR-URL: nodejs#1926 Reviewed-By: Rod Vagg <rod@vagg.org> Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>

vm: remove unnecessary HandleScopes

5d0cee4

The accessors run inside an implicit HandleScope, there is no need to create a new one. PR-URL: nodejs#2001 Reviewed-By: Trevor Norris <trev.norris@gmail.com>

test: add Buffer slice UTF-8 test

0abcf44

PR-URL: nodejs#1989 Reviewed-By: Trevor Norris <trev.norris@gmail.com>

doc: add TSC meeting notes 2015-06-10

30638b1

PR-URL: nodejs#1943 Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl> Reviewed-By: Rod Vagg <rod@vagg.org> Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>

doc: add TOC links to Collaborator Guide

7a3006e

PR-URL: nodejs#1994 Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

doc: mention CI in Collaborator Guide

0cf94e6

Add link to Jenkins server in Collaborator Guide. PR-URL: nodejs#1995 Reviewed-By: Chris Dickinson <christopher.s.dickinson@gmail.com> Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>

doc: change the info to the same as in gitconfig

1f93b63

PR-URL: nodejs#2000 Reviewed-By: Yosuke Furukawa <yosuke.furukawa@gmail.com> Reviewed-By: Evan Lucas <evanlucas@me.com> Reviewed-By: Johan Bergström <bugs@bergstroem.nu>

orangemocha and others added 2 commits June 19, 2015 11:04

build: add MSVS 2015 support

552664b

rvagg force-pushed the msvs-2015 branch from 0e54add to 552664b Compare June 19, 2015 01:04

rvagg closed this Jun 23, 2015

rvagg mentioned this pull request Jun 23, 2015

VS 2015 support #2036

Merged

srl295 mentioned this pull request Jul 30, 2015

intl: VS2015: workaround ICU VS2015 failure? #2279

Closed

joaocgreis mentioned this pull request Aug 13, 2015

Support for Visual Studio 2015 nodejs/node-v0.x-archive#25857

Closed

snyk-bot mentioned this pull request Apr 26, 2023

[Snyk] Fix for 4 vulnerabilities aliscco/alisco-node#232

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

VS 2015 support #478

VS 2015 support #478

rvagg commented Jan 17, 2015

rvagg commented Jan 17, 2015

rvagg commented Jan 17, 2015

chrisdickinson commented Jan 17, 2015

bnoordhuis commented Jan 17, 2015

rvagg commented Jan 17, 2015

bnoordhuis commented Jan 17, 2015

chrisdickinson commented Jan 17, 2015

rvagg commented Jan 17, 2015

rvagg commented Jan 17, 2015

rvagg commented Jan 17, 2015

bnoordhuis commented Jan 17, 2015

rvagg commented Jan 19, 2015

rvagg commented Jan 19, 2015

bnoordhuis commented Jan 19, 2015

rvagg commented Jan 19, 2015

rvagg commented Jan 19, 2015

seishun commented Jan 20, 2015

rvagg commented Jan 20, 2015

xforce commented Feb 4, 2015

Fishrock123 commented Feb 4, 2015

targos commented Jun 18, 2015

ghost commented Jun 18, 2015

targos commented Jun 18, 2015

ghost commented Jun 18, 2015

rvagg commented Jun 19, 2015

rvagg commented Jun 19, 2015

orangemocha commented Jun 22, 2015

rvagg commented Jun 22, 2015

targos commented Jun 22, 2015

srl295 commented Jul 30, 2015

VS 2015 support #478

VS 2015 support #478

Conversation

rvagg commented Jan 17, 2015

rvagg commented Jan 17, 2015

rvagg commented Jan 17, 2015

chrisdickinson commented Jan 17, 2015

bnoordhuis commented Jan 17, 2015

rvagg commented Jan 17, 2015

bnoordhuis commented Jan 17, 2015

chrisdickinson commented Jan 17, 2015

rvagg commented Jan 17, 2015

rvagg commented Jan 17, 2015

rvagg commented Jan 17, 2015

bnoordhuis commented Jan 17, 2015

rvagg commented Jan 19, 2015

rvagg commented Jan 19, 2015

bnoordhuis commented Jan 19, 2015

rvagg commented Jan 19, 2015

rvagg commented Jan 19, 2015

seishun commented Jan 20, 2015

rvagg commented Jan 20, 2015

xforce commented Feb 4, 2015

Fishrock123 commented Feb 4, 2015

targos commented Jun 18, 2015

ghost commented Jun 18, 2015

targos commented Jun 18, 2015

ghost commented Jun 18, 2015

rvagg commented Jun 19, 2015

rvagg commented Jun 19, 2015

orangemocha commented Jun 22, 2015

rvagg commented Jun 22, 2015

targos commented Jun 22, 2015

srl295 commented Jul 30, 2015