Switch to DuckDuckGo Search
   February 10, 2018  
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | >

Toggle Join/Part | bottom
[00:06:27] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has quit IRC (Read error: Connection reset by peer)
[00:06:34] *** psydroid_ <psydroid_!~psydroid@ip127-8-212-87.adsl2.static.versatel.nl> has joined #oi-dev
[00:08:36] *** psydroid_ <psydroid_!~psydroid@ip127-8-212-87.adsl2.static.versatel.nl> has quit IRC (Client Quit)
[00:08:49] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has joined #oi-dev
[00:13:25] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has quit IRC (Read error: Connection reset by peer)
[00:18:52] *** andy_js <andy_js!~andy@5ec0fdb3.skybroadband.com> has quit IRC (Quit: andy_js)
[00:25:10] *** tozhu <tozhu!~Thunderbi@> has joined #oi-dev
[00:29:35] *** tozhu <tozhu!~Thunderbi@> has quit IRC (Ping timeout: 248 seconds)
[00:48:36] *** merzo <merzo!~merzo@234-39-132-95.pool.ukrtel.net> has joined #oi-dev
[00:49:02] *** tozhu <tozhu!~Thunderbi@> has joined #oi-dev
[00:49:27] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has quit IRC (Ping timeout: 260 seconds)
[00:53:11] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has joined #oi-dev
[01:03:35] *** tozhu <tozhu!~Thunderbi@> has quit IRC (Ping timeout: 260 seconds)
[01:06:37] *** tozhu <tozhu!~Thunderbi@> has joined #oi-dev
[01:07:32] *** Woodstock <Woodstock!grumpf@beiss.dessnourst.de> has joined #oi-dev
[01:10:41] *** tozhu <tozhu!~Thunderbi@> has quit IRC (Ping timeout: 248 seconds)
[01:43:42] *** Toasterson <Toasterson!~toast@2a02:1205:5072:b660:1000:6d60:4ec2:1a3b> has quit IRC (Ping timeout: 256 seconds)
[01:53:57] *** mnowak_ <mnowak_!~mnowak_@> has quit IRC (Ping timeout: 240 seconds)
[01:56:23] *** tozhu <tozhu!~Thunderbi@> has joined #oi-dev
[02:12:10] *** tozhu <tozhu!~Thunderbi@> has quit IRC (Ping timeout: 240 seconds)
[02:12:32] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has quit IRC (Ping timeout: 276 seconds)
[02:15:31] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has joined #oi-dev
[02:23:51] *** tozhu <tozhu!~Thunderbi@> has joined #oi-dev
[02:28:20] *** tozhu <tozhu!~Thunderbi@> has quit IRC (Ping timeout: 255 seconds)
[02:48:04] *** v_a_b <v_a_b!~volker@p509050B9.dip0.t-ipconnect.de> has joined #oi-dev
[02:50:25] *** v_a_b1 <v_a_b1!~volker@p509059E5.dip0.t-ipconnect.de> has quit IRC (Ping timeout: 248 seconds)
[02:53:20] *** tozhu <tozhu!~Thunderbi@> has joined #oi-dev
[02:54:00] <oibot> Project illumos-gate build #5137: SUCCESS in 1 hr 48 min: https://hipster.openindiana.org/jenkins/job/illumos-gate/5137/
[02:54:40] *** tozhu <tozhu!~Thunderbi@> has quit IRC (Client Quit)
[03:52:49] *** merzo <merzo!~merzo@234-39-132-95.pool.ukrtel.net> has quit IRC (Ping timeout: 248 seconds)
[04:39:57] *** teferi <teferi!~adam@> has quit IRC (Ping timeout: 240 seconds)
[04:40:08] *** teferi <teferi!~adam@singularity.crossproduct.net> has joined #oi-dev
[07:00:57] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has quit IRC (Ping timeout: 240 seconds)
[07:02:14] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has joined #oi-dev
[07:32:20] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has quit IRC (Ping timeout: 276 seconds)
[07:42:58] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has joined #oi-dev
[07:53:47] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has quit IRC (Ping timeout: 276 seconds)
[07:59:15] *** MarcelT <MarcelT!~marcel@tortuga.telka.sk> has joined #oi-dev
[08:10:48] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has joined #oi-dev
[08:11:44] *** leoric <leoric!~leoric@> has quit IRC (Remote host closed the connection)
[08:19:00] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has joined #oi-dev
[08:29:07] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has quit IRC (Ping timeout: 260 seconds)
[08:36:21] *** MarcelT <MarcelT!~marcel@tortuga.telka.sk> has quit IRC (Ping timeout: 264 seconds)
[08:38:11] *** MarcelT <MarcelT!~marcel@tortuga.telka.sk> has joined #oi-dev
[08:59:48] *** mnowak_ <mnowak_!~mnowak_@> has joined #oi-dev
[09:06:13] <psydroid> I've placed my orders for new VGA cables, DVI-to-HDMI adapters and some HDMI cables so hopefully the fun can start again soon
[09:16:08] *** lgtaube <lgtaube!~lgtaube@213-67-21-71-no84.tbcn.telia.com> has quit IRC (Read error: Connection reset by peer)
[09:17:33] *** lgtaube <lgtaube!~lgtaube@213-67-21-71-no84.tbcn.telia.com> has joined #oi-dev
[09:27:23] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has joined #oi-dev
[09:35:02] *** pjama <pjama!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has quit IRC (Ping timeout: 260 seconds)
[09:35:34] *** pjama <pjama!~pjama@> has joined #oi-dev
[09:48:11] *** lgtaube <lgtaube!~lgtaube@213-67-21-71-no84.tbcn.telia.com> has quit IRC (Read error: Connection reset by peer)
[09:49:30] *** lgtaube <lgtaube!~lgtaube@213-67-21-71-no84.tbcn.telia.com> has joined #oi-dev
[10:39:53] *** andy_js <andy_js!~andy@5ec0fdb3.skybroadband.com> has joined #oi-dev
[11:11:28] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has quit IRC (Read error: Connection reset by peer)
[11:30:46] *** leoric <leoric!~leoric@> has joined #oi-dev
[11:34:42] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has joined #oi-dev
[11:55:09] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has quit IRC (Remote host closed the connection)
[12:06:05] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has joined #oi-dev
[12:38:33] *** psydroid_ <psydroid_!~psydroid@ip127-8-212-87.adsl2.static.versatel.nl> has joined #oi-dev
[12:42:57] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has quit IRC (Ping timeout: 264 seconds)
[14:26:09] *** merzo <merzo!~merzo@117-62-132-95.pool.ukrtel.net> has joined #oi-dev
[14:28:26] *** pjama1 <pjama1!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has joined #oi-dev
[14:29:02] *** pjama <pjama!~pjama@> has quit IRC (Ping timeout: 260 seconds)
[14:45:05] *** rk4n3 <rk4n3!~rk4n3@24-159-210-156.static.roch.mn.charter.com> has quit IRC (Ping timeout: 248 seconds)
[14:53:50] <leoric> alarcher: are we ready for gcc-6 ?
[14:55:53] *** psydroid_ <psydroid_!~psydroid@ip127-8-212-87.adsl2.static.versatel.nl> has quit IRC (Read error: Connection reset by peer)
[14:55:55] <leoric> How are we going to merge it?
[14:56:29] <leoric> I suppose we can extract a list of pre-gcc-6 changes, which can be merged first, and then gcc-6 switch
[14:56:49] <leoric> but gcc-6 branch needs some cleanup
[15:02:25] *** jollyd <jollyd!~alarcher@> has joined #oi-dev
[15:02:33] <jollyd> leoric: just saw your message
[15:02:46] <jollyd> I will remove commit from the gcc6 branch
[15:02:50] <jollyd> commits
[15:06:44] <leoric> Hi
[15:07:12] <jollyd> I meant that I will remove the spurious commits and clean up
[15:07:55] <leoric> OK, when you ready we'll review what's left for merging and discuss how should we proceed
[15:09:26] <leoric> try merging things like https://github.com/OpenIndiana/oi-userland/pull/3014/commits/4d56c24c24204ae33dedd680b6fcd1c7ad2be5b5 and https://github.com/OpenIndiana/oi-userland/pull/3014/commits/3320a1b6cc0207968f5437a90672a8003657dda0 or several "openjdk-8: fix build WIP"
[15:10:11] <jollyd> yes I squashed these just now
[15:11:00] <jollyd> I will also reorder commits by type (fix REQUIRED_PACKAGES, FTBFS)
[15:12:03] <leoric> after sleeping in the afternoon I'm ready to work in your TZ today so we could proceed further :)
[15:12:21] <jollyd> ahah :)
[15:25:19] *** rk4n3 <rk4n3!~rk4n3@24-159-210-156.static.roch.mn.charter.com> has joined #oi-dev
[15:32:36] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has joined #oi-dev
[15:47:54] *** rk4n3 <rk4n3!~rk4n3@24-159-210-156.static.roch.mn.charter.com> has quit IRC (Ping timeout: 252 seconds)
[15:48:37] *** psydroid_ <psydroid_!~psydroid@ip127-8-212-87.adsl2.static.versatel.nl> has joined #oi-dev
[15:52:20] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has quit IRC (Ping timeout: 255 seconds)
[15:56:40] <jollyd> leoric: I have reorderd the commits and I am now squashing what can be
[15:59:22] <leoric> ping me when you will need review :)
[16:12:09] <jollyd> leoric: something that I could not understand is python/coherence and ruby-23 missing some files
[16:17:55] <jollyd> leoric: github insists on not displaying the commits in the right order...
[16:18:36] <jollyd> the display on the web is not the same as git log
[16:18:42] <jollyd> very useful
[16:20:33] <leoric> coherence publishes fine here in gcc-4 zone ...
[16:21:27] <jollyd> leoric: missing dependency do you think?
[16:21:41] <leoric> checking in gcc-6 zone
[16:25:05] <leoric> coherence publishes fine in gcc-6 zone
[16:25:29] <leoric> without commenting out 'file path=usr/lib/python$(PYVER)/vendor-packages/Coherence-$(COMPONENT_VERSION)-py$(PYVER).egg-info/requires.txt'
[16:26:32] <leoric> no
[16:34:10] <leoric> jollyd: avoid changing GeoIP database checksums
[16:34:18] <leoric> this will not help build server
[16:34:36] <jollyd> do they change often?
[16:34:51] <leoric> ask xenol
[16:35:04] <leoric> at least we have cached copies on build server
[16:35:52] <leoric> also, avoid changing runtime/perl-522 to runtime/perl-524 without need in REQUIRED_PACKAGES
[16:35:53] <jollyd> leoric: when I mentioned it he replied that "this geoip thing is bullshit" but he did not explain ;)
[16:36:04] <leoric> :)
[16:36:27] <jollyd> the changes are actually detected by pkgdepend
[16:36:32] <leoric> no
[16:36:34] <leoric> look
[16:36:40] <leoric> we make REQUIRED_PACKAGES
[16:36:55] <jollyd> hmm ok
[16:37:08] <leoric> we have require type=any fmri=runtime/perl-522 fmri=runtime/perl-524
[16:37:10] <jollyd> so if REQUIRED_PACKAGES changes the version I should ignore it?
[16:37:17] <leoric> add both
[16:37:21] <jollyd> ah
[16:37:55] <leoric> on the next run it will fail to detect perl-522 dependency, as it will be searching only in REQUIRED_PACKAGES
[16:37:59] <leoric> and this is an issue
[16:38:55] <leoric> for mongodb-33, mongodb-34 - can we just remove all 'transform file -> add pkg.depend.bypass-generate 'libstdc\+\+\.so\.6'' nonsence in p5m files
[16:38:56] <leoric> ?
[16:39:36] <jollyd> I can try
[16:39:51] <jollyd> I am fixing the perl stuff now
[16:39:51] <leoric> it was a private implementation of what transforms/ignore-gcc-usr-lib do now in a general way
[16:40:09] <leoric> (or at least should do)
[16:40:48] <jollyd> also ruby-23 had missing documentation files, I tried adding locale/ja but it did not help
[16:42:06] <leoric> doesn't work in gcc-6 zone, checking in gcc-5
[16:42:09] <leoric> gcc-4
[16:42:25] <leoric> clang-40
[16:42:43] <leoric> have you checked that it likes gcc-6 c++ libraries?
[16:43:52] <leoric> Does gcc-3 build fine with gcc-6?
[16:45:40] <jollyd> I have done basic testing of clang
[16:45:50] <leoric> OK.
[16:46:02] <leoric> luajit: why do we need this -lumem ?
[16:46:23] <jollyd> luajit is partly broken anyway
[16:46:26] <leoric> (and please, convert patch to universal diff)
[16:46:29] <jollyd> need more work on it
[16:46:29] <leoric> I know :)
[16:50:45] <leoric> illumos-gcc misses build time dependency on gcc-4, but it's not critical
[16:50:57] <leoric> ruby publishes fine in gcc-4 zone...
[16:59:50] <jollyd> I just finished the perl modifications, I will push in a moment
[17:01:35] <leoric> commit 'Bump branch revision' likely has outdated message
[17:06:55] <jollyd> amended
[17:17:51] <jollyd> leoric: so I fixed the perl deps, the illumos-gcc dep to gcc49, the luajit patch, the commit msg for RELEASE_MAJOR, now looking at mongodb
[17:20:03] <leoric> I'
[17:20:10] <leoric> m looking at ruby-23
[17:20:20] <leoric> (trying to find 10 differences)
[17:25:35] <jollyd> hehe
[17:31:05] <leoric> jollyd: +COMPONENT_BUILD_ENV += LC_ALL=en_US.UTF-8
[17:31:05] <leoric> +COMPONENT_INSTALL_ENV += LC_ALL=en_US.UTF-8
[17:31:28] <jollyd> I have C
[17:31:32] <jollyd> indeed
[17:39:24] <jollyd> leoric: removing the gcc-related transforms in mongedb-33 is fine
[17:40:49] <leoric> so, I suppose, we are close to finish with gcc-6?
[17:41:24] <leoric> I'm going to take a supper and hope you'll make a final pass on gcc-6 branch
[17:41:48] <leoric> after that I'll try to create new BE on build server and update it to gcc-6 repo
[17:42:13] <leoric> then - merge PR and rebuild the world
[17:43:52] <leoric> remind me to make a backup of current package repository :)
[17:44:02] <jollyd> ;)
[17:48:14] <jollyd> leoric: how about python/coherence ?
[17:48:30] <jollyd> do you have the file usr/lib/python2.7/vendor-packages/Coherence-$(COMPONENT_VERSION)-py2.7.egg-info/requires.txt ?
[18:05:34] <leoric> yes
[18:07:11] *** merzo <merzo!~merzo@117-62-132-95.pool.ukrtel.net> has quit IRC (Ping timeout: 256 seconds)
[18:09:58] <jollyd> leoric: do you think I may miss a python dep?
[18:10:14] <leoric> unlikely
[18:10:24] <leoric> it builds in gcc-6 zone
[18:10:56] <leoric> and it has minimal set of packages here
[18:11:44] <leoric> pkg list '*python*' - https://pastebin.com/bPBQ6mXA
[18:12:55] <jollyd> leoric: is apache24 a build dep here https://github.com/OpenIndiana/oi-userland/pull/3014/commits/6a2c7c866c17151587a95098e25a127592356ca4 ?
[18:13:15] <leoric> yes
[18:14:01] <leoric> it uses apxs
[18:20:49] <jollyd> ok I fixed it
[18:21:53] <leoric> hm???
[18:22:08] <leoric> I still see '#file path=usr/lib/python$(PYVER)/vendor-packages/Coherence-$(COMPONENT_VERSION)-py$(PYVER).egg-info/requires.txt'
[18:22:14] <leoric> https://github.com/OpenIndiana/oi-userland/pull/3014/commits/323a1d8f851c9dc813a1a9aaa7f3a2652124d5a2
[18:24:55] <leoric> jollyd: perhaps, just don't touch it?
[18:25:17] <leoric> we'll see how it goes on the build server :)
[18:26:02] <jollyd> I meant I fixed the apache24 dep in php
[18:26:14] <jollyd> I am now bumping a few missing revisions
[18:27:31] <leoric> should I update build server already?
[18:29:52] * leoric is preparing build server
[18:33:40] <jollyd> leoric: I guess it should be fine
[18:34:16] <leoric> The plan is: disable oi-userland jenkins job (ready); update build server; git clean -dfx in oi-userland and illumos-gate jenkins jobs work directories; removing package repositories for oi-userland and illumos-gate on the build server; rebuilding illumos-gate; updating to rebuilt illumos-gate; rebuild oi-userland; updating to rebuilt oi-userland; clean up official repositories from old packages (as time permits)
[18:35:02] <richlowe> do you have an automatic tarball mirror going?
[18:35:08] <richlowe> if not, might want a snapshot before the git clean
[18:35:14] <richlowe> lesson I've learned the hard way :|
[18:35:33] <leoric> we have tarballs on dlc.oi.org and local archive on the build server
[18:36:01] <richlowe> another reason I should just give you all my packages, no matter how niche ;)
[18:36:28] <leoric> need to modify oi-userland job before running it to avoid incremental build
[18:39:48] <leoric> the only thing I'm worried about is that I have no access to KVM ;)
[18:40:15] <jollyd> leoric: I should mention that gcc-5 failed to build once with -j40 because of the autoreconf of mpfr
[18:40:40] <jollyd> I just remembered
[18:40:55] <leoric> we'll see how it goes...
[18:42:17] <jollyd> I am trying to multitask between oi-userland and fixing some parallel code, needless to say the efficiency is not stellar
[18:43:59] <leoric> :)
[18:45:04] <richlowe> I can't tell if that's setup and punchline, or just actual truth
[18:45:05] <richlowe> or both.
[18:45:06] <jollyd> ...and fixing some random bug showing up with 140 MPI processes is not my favourite way to spend Saturday evenings...
[18:45:30] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has joined #oi-dev
[18:48:00] <leoric> yes, better find out why this package fails to build with gcc-6 ;)
[18:48:16] * jollyd puts a red nose on.. *ta-da poom*
[18:48:49] <jollyd> leoric: which one?
[18:49:31] *** psydroid_ <psydroid_!~psydroid@ip127-8-212-87.adsl2.static.versatel.nl> has quit IRC (Ping timeout: 268 seconds)
[18:49:34] <leoric> I'm just suggesting hypothetic better way of spending weekend
[18:49:50] <jollyd> ahah
[18:51:45] <jollyd> yesterday I tried to invade England as William the Conqueror on Civilization V and failed to kick Denmark out of it, so week-end is doomed anyway...
[18:52:14] <leoric> heh
[18:52:50] <jollyd> :P
[18:53:10] <richlowe> it's easier in Civ than bloody CK2
[18:54:31] <jollyd> never tried
[18:54:46] <leoric> make it work in wine on OI :)
[18:55:32] <jollyd> huhu
[18:57:24] <leoric> here vlc is sometimes shows garbage in some scenes, sometimes hangs for a bit. This makes me wonder - if 480 or 720p video is too much for my pc or our vlc port needs some love....
[18:59:16] <jollyd> maybe we should move to 64bit at some point
[19:16:11] <jollyd> I hate graph partitioning...
[19:19:15] <leoric> btw, there's vlc 3.0.0
[19:37:10] *** psydroid_ <psydroid_!~psydroid@ip127-8-212-87.adsl2.static.versatel.nl> has joined #oi-dev
[19:39:43] *** oibot <oibot!~oibot@> has quit IRC ()
[19:39:46] <leoric> OK, trying to boot gcc 6 be
[19:41:03] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has quit IRC (Ping timeout: 248 seconds)
[19:42:14] <jollyd> fingers crossed
[19:42:26] <leoric> booted
[19:42:37] *** oibot <oibot!~oibot@> has joined #oi-dev
[19:43:15] <richlowe> leoric: did you build illumos with 6?
[19:43:17] <leoric> some packages were not updated - https://pastebin.com/Xu3KWkcQ
[19:43:19] <richlowe> leoric: I don't think tsoome jollyd backported some of my stuff to his gcc6 branch.
[19:43:22] <leoric> no
[19:43:24] <richlowe> k
[19:43:26] <leoric> we use gcc 4
[19:43:37] <richlowe> just wanted to be sure, 'cos one of the patches fixes dtrace and the debugger :)
[19:43:48] <richlowe> (well, CTF, but it's those two things that break)
[19:43:51] <leoric> :)
[19:44:20] <leoric> was it submitted for review or still not?
[19:44:34] <richlowe> from what I understand, tsoome wants to go to 7 rather than 6.
[19:44:46] <richlowe> I'm giving my changes to him, he's getting his fixes into illumos.
[19:44:51] <richlowe> I don't know how much is outstanding.
[19:45:12] <tsoome> well, thats my “feeling” based on the warnings I see, not on technical knowledge about compiler internals
[19:45:15] <richlowe> I'm currently hacking cw so that we can build with a gcc primary and a gcc shadow, among other complications.
[19:45:33] <richlowe> uh, that was meant to say "combinations"
[19:45:35] <leoric> Well, rebuilding the world to gcc-7 before 2017.04 ? I'm not sure...
[19:45:38] <richlowe> my mood is showing :)
[19:45:58] <richlowe> tsoome: well, 7 is what I have tested/am testing.
[19:46:14] <richlowe> leoric: as with currently, there's no reason you have to build everything with the compiler you use to build illumos.
[19:46:15] *** jacobs <jacobs!~jacobs@2603:300a:1618:b200:4637:e6ff:fec1:70ad> has quit IRC (Remote host closed the connection)
[19:46:29] *** jacobs <jacobs!~jacobs@2603:300a:1618:b200:4637:e6ff:fec1:70ad> has joined #oi-dev
[19:46:33] <tsoome> the chase of compiler versions is always just a bit about being sane or insane:D
[19:46:48] <richlowe> it's more about deciding you have to stop working on it and let people use it
[19:46:56] <richlowe> you can chase upgrades forever, eventually you have to stop and push it
[19:47:09] <richlowe> retpoline stuff, etc, provide a good motivation to do that.
[19:47:17] <tsoome> yes, exactly
[19:47:47] <richlowe> once I have cw done, I'm going to switch, myself.
[19:48:10] <leoric> alarcher: how do you feel about that? ;)
[19:48:30] <jollyd> leoric: anyway once gcc-6 is the main userland component I will move to building with gcc-7
[19:48:30] <leoric> I think we have other important things to do till April
[19:48:41] <richlowe> gcc7 is harder for you, because of the default compilation environment.
[19:48:45] <jollyd> I think next snapshot is more reasonable
[19:48:48] <richlowe> I think you should stay with 6 and just build illumos with what we say
[19:48:55] <jollyd> I meant 2018.10
[19:48:57] <leoric> jollyd:++
[19:49:12] <richlowe> and honestly, I expect illumos with 4.4.4 to stay around long enough you could just stick with that too.
[19:49:21] <richlowe> the only realistic plan is to support 4 + 7 shadow, 7 + 4 shadow, just 7
[19:49:28] <richlowe> not that I've discussed that with tsoome...
[19:49:30] <leoric> yes
[19:49:42] <richlowe> it's his projcet, but I'm the advocate who'd make him do what I say there anyway ;)
[19:50:42] <jollyd> but I'll start building userland right away with gcc-7, and will follow up on the (illumos-)gcc-7 stuff as soon as I get these issues at work fixed...
[19:51:09] <richlowe> I imagine userland and gcc7 to be a whole massive mess of warnings about -std=c99 v. -D_XOPEN_SOURCE=600
[19:51:23] <tsoome> :D
[19:51:28] <richlowe> perhaps gcc should make one imply the other?
[19:51:32] <richlowe> (with __EXTENSIONS__ of course)
[19:51:52] <richlowe> jacobs will have experience dealing with that, I imagine
[19:52:31] <jollyd> yeah next round I want macros to specify the standard and not bother with specifying _XOPEN_SOURCE
[19:52:40] <leoric> why should it be worse than gcc-6?
[19:52:54] <jollyd> but before dealing with gcc-6 I had actually rebuilt most of userland with gcc-7
[19:52:58] <richlowe> I thought 7 was where some defaults changed?
[19:53:02] <leoric> 6
[19:53:13] <jollyd> C++ has 14 as default in 7
[19:53:20] <leoric> ah, c++
[19:53:22] <richlowe> C++ will always be sadness and woe
[19:53:23] <jollyd> instead of 11 I think
[19:53:29] <jollyd> well
[19:53:45] <leoric> c99 is default even with gcc5 iirc
[19:53:45] <richlowe> I have to deal with the compiler and the link-editor, you'll never convince me otherwise :)
[19:54:23] <jollyd> hmm
[19:54:47] <jollyd> to me Python is a worst nightmare than C++
[19:55:23] <jollyd> the ugliness seems proportional to the convenience
[19:55:42] <richlowe> python isn't ugly in ways I have to maintain without ever using.
[19:56:35] <jollyd> consider all the money poured into maintaining this moving target of dependency hell...
[19:57:01] <richlowe> that is not a challenge to course python to blow up mysteriously in ld(1).
[19:57:11] <richlowe> s/course/cause/
[19:57:20] <jollyd> :)
[19:59:30] <jollyd> gcc-7 is interesting to me because of the openmp nuggets...
[20:00:27] <leoric> jollyd: do you remember what compiler should be used now for openindiana/gfx-drm ?
[20:01:45] <leoric> ah, it uses gcc 4.4
[20:01:56] <leoric> gfx-drm/usr/src/Makefile.master:GCC_ROOT= /opt/gcc/4.4.4
[20:02:18] <leoric> one more gate to switch to gcc -6 /7 ;)
[20:11:48] <jollyd> I just rebooted my compute node instead of the zone....
[20:11:54] <jollyd> :S
[20:13:29] <jollyd> leoric: what is developer/java/jdk ?
[20:14:25] <jollyd> leoric: ipython-34 does not seem to exist in oi-userland -> ipython-34.p5m.notyet
[20:16:15] *** jollyd <jollyd!~alarcher@> has quit IRC (Quit: Leaving.)
[20:17:26] <leoric> jollyd: cool, how then we have this - http://pkg.openindiana.org/hipster/en/search.shtml?token=ipython-34&action=Search ? ;)
[20:18:03] <leoric> components/meta-packages/history/history:developer/java/jdk at 0 dot 5.11,5.11-2017.0.0.0 developer/java/openjdk8 at 1 dot 8.112,5.11-2016.1.0.0
[20:21:59] *** rk4n3 <rk4n3!~rk4n3@24-159-210-156.static.roch.mn.charter.com> has joined #oi-dev
[20:22:18] *** alarcher-bis <alarcher-bis!~alarcher@> has joined #oi-dev
[20:23:20] <alarcher-bis> leoric: right ... for metapackages I should have cleaned up manually since clobber does not work...
[20:23:43] <leoric> xenol: do you wish to kindly fix ipython? ;)
[20:26:08] *** psydroid_ <psydroid_!~psydroid@ip127-8-212-87.adsl2.static.versatel.nl> has quit IRC (Read error: Connection reset by peer)
[20:26:48] *** psydroid_ <psydroid_!~psydroid@ip127-8-212-87.adsl2.static.versatel.nl> has joined #oi-dev
[20:30:49] <alarcher-bis> leoric: were ssh-askpass* and server_install not supposed to be republished with slim_source and install-types?
[20:31:28] <leoric> server_install comes from install-types
[20:32:42] <alarcher-bis>
[20:32:42] <leoric> ssh-askpass is components/network/ssh-askpass
[20:33:04] <leoric> hm...
[20:36:31] <alarcher-bis> erf it does not clobber
[20:36:33] <oibot> Project illumos-gate build #5138: FAILURE in 24 sec: https://hipster.openindiana.org/jenkins/job/illumos-gate/5138/
[20:37:04] <leoric> ha-ha
[20:37:14] <alarcher-bis> :D
[20:37:19] <leoric> build server even doesn't have gcc-6 installed
[20:40:00] <leoric> and without gcc-6 it can't build userland gate tools to start building illumos-gate
[20:41:12] <alarcher-bis> only gcc-7 is in build-essential right?
[20:43:48] <leoric> yes
[20:43:57] <leoric> should switch 7 to 6
[20:44:23] <leoric> and perhaps clean up it a bit
[20:44:29] <leoric> to make really essential
[20:44:56] <leoric> due to REQUIRED_PACKAGES we don't need to collect there everything required to build oi-userland
[20:45:19] <alarcher-bis> do you mean like splitting lint?
[20:45:29] <alarcher-bis> ;)
[20:45:44] <leoric> lint ? don't know :)
[20:45:44] <alarcher-bis> oh you merged the PR
[20:45:56] <alarcher-bis> damn, that was a big one
[20:47:57] <alarcher-bis> I republished ssh-askpass
[20:48:09] <alarcher-bis> not that it is essential
[20:50:10] <leoric> alarcher: I suppose I can close https://github.com/OpenIndiana/oi-userland/pull/3883 and https://github.com/OpenIndiana/oi-userland/pull/3867?
[20:50:51] <alarcher-bis> I would expect so
[20:55:42] <leoric> it seems I have about 1.5 hrs to sleep...
[20:58:03] <xenol> leoric: what's the problem with ipython?
[20:58:24] <leoric> 34 version is not published (missing deps? )
[20:59:25] *** psychicist__ <psychicist__!~psychicis@ip127-8-212-87.adsl2.static.versatel.nl> has quit IRC (Ping timeout: 260 seconds)
[21:01:08] *** psychicist__ <psychicist__!~psychicis@ip127-8-212-87.adsl2.static.versatel.nl> has joined #oi-dev
[21:08:43] *** psydroid_ <psydroid_!~psydroid@ip127-8-212-87.adsl2.static.versatel.nl> has quit IRC (Read error: Connection reset by peer)
[21:09:43] *** psydroid_ <psydroid_!~psydroid@ip127-8-212-87.adsl2.static.versatel.nl> has joined #oi-dev
[21:27:18] <oibot> Project illumos-gate build #5139: STILL FAILING in 41 min: https://hipster.openindiana.org/jenkins/job/illumos-gate/5139/
[21:38:00] *** psydroid_ <psydroid_!~psydroid@ip127-8-212-87.adsl2.static.versatel.nl> has quit IRC (Read error: Connection reset by peer)
[21:38:32] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has joined #oi-dev
[21:44:17] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has quit IRC (Read error: Connection reset by peer)
[21:44:25] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has joined #oi-dev
[21:46:19] *** psydroid <psydroid!~psydroid@unaffiliated/psydroid> has quit IRC (Read error: Connection reset by peer)
[21:50:50] <alarcher-bis> leoric: holy shit snacks... check_rtime
[21:52:17] <alarcher-bis> I had bluntly disabled it
[22:20:29] *** pjama1 <pjama1!~pjama@c211-31-53-64.kelvn3.qld.optusnet.com.au> has quit IRC (Ping timeout: 256 seconds)
[22:23:01] *** pjama <pjama!~pjama@> has joined #oi-dev
[22:35:23] <leoric> alarcher: I suppose, we should do the same
[22:35:35] <leoric> and upstream this to illumos-gate
[22:36:28] <leoric> alarcher: ^&
[22:37:35] *** Meths <Meths!~meths@> has quit IRC (*.net *.split)
[22:37:35] *** JT-EC <JT-EC!~JT-EC@> has quit IRC (*.net *.split)
[22:37:36] *** xenol <xenol!~xenol@unaffiliated/xenol> has quit IRC (*.net *.split)
[22:38:14] *** IvanR_ <IvanR_!~ivan@dns1.seppuku.net> has quit IRC (*.net *.split)
[22:39:16] <leoric> alarcher-bis: ^
[22:40:06] <leoric> I suppose, it's -r ?
[22:42:16] *** IvanR_ <IvanR_!~ivan@dns1.seppuku.net> has joined #oi-dev
[22:45:31] *** xenol <xenol!~xenol@unaffiliated/xenol> has joined #oi-dev
[22:45:31] *** Meths <Meths!~meths@> has joined #oi-dev
[22:45:31] *** JT-EC <JT-EC!~JT-EC@> has joined #oi-dev
[22:45:31] *** karatkievich.freenode.net sets mode: +o Meths
[22:45:33] <leoric> tsoome: ^ ?
[22:49:56] <richlowe> check_rtime fixes and such would be appreciated
[22:50:14] <richlowe> especially the syntax stuff that seems to make you no longer have to know where mach-specific libs are going to be and list them.
[22:51:46] <leoric> so far I just disable -r and will re-run the job...
[22:51:59] <richlowe> probably for the best
[22:52:03] <richlowe> I believe alanc said he did it to keep X clean
[22:52:04] <leoric> richlowe: the issue is that it's really the problem
[22:52:11] <richlowe> and had emphatically not actually fixed userland at all.
[22:52:19] <richlowe> leoric: hm?
[22:52:49] <leoric> if we use gcc-6 for userland and gcc-7 for the gate, we'll have /usr/gcc/7/lib runpath in one binaries and /usr/gcc/6/lib in other
[22:53:03] <richlowe> I swear, this fucking runpath is going to be etched on my tomb
[22:53:32] <richlowe> leoric: but you do anyway?
[22:53:41] <richlowe> or more likely, you don't, because illumos-gcc specifically doesn't set them?
[22:54:05] <richlowe> so you'd have ON binaries that say "Wherever you find your libraries"
[22:54:09] <leoric> if it doesn't, binaries, looking for gcc runtime, will find wrong one
[22:54:36] <leoric> now it's not a big problem
[22:54:49] <leoric> with gcc-7 can be an issue
[22:54:52] <richlowe> but it would become one if illumos's gcc was the newer?
[22:54:56] <leoric> yes
[22:54:59] <richlowe> to be fair, I don't think any of us ever expected that to happen :)
[22:55:13] <richlowe> we'd always assumed we'd lag
[22:55:22] <leoric> :)
[22:55:42] <richlowe> were I you, I would think about it hard (or not)
[22:55:48] <richlowe> and email me, and I'll think about it and whatnot.
[22:55:50] <leoric> so it's a bit unfortunate that we targeted different versions
[22:55:52] <richlowe> but right now I'm pretty much ready to murder.
[22:56:02] <richlowe> so it's not a great time to ask me about gcc runpaths :)
[22:56:20] <richlowe> leoric: I would expect, given the speed illumos moves, it won't actually be a problem.
[22:56:23] <richlowe> but we'll see
[22:56:55] <leoric> what about illumos-gate c++ code ?
[22:57:06] <leoric> does it depend on g++ runtime?
[22:57:35] <richlowe> Ah, we've had this discussion before
[22:57:36] <richlowe> Yes it does
[22:57:41] <richlowe> but very few people depend on _it_
[22:57:45] <igork> illumos make depend on c++
[22:57:53] <leoric> yes, make
[22:58:06] <igork> and couple of others
[22:58:07] <richlowe> the make illumos uses should be built with the host compiler and environment.
[22:58:21] <richlowe> and should be fine in the case you're worried about, unless _everything_ is busted.
[22:58:35] <richlowe> but as I said, it is very likely illumos won't switch the big default until you have too
[22:58:39] <richlowe> just because illumos is slow and argumentative
[22:58:56] <richlowe> if it looks like we may, we can come up with something.
[22:59:18] <leoric> I'm more interested to see if it fails right now
[22:59:30] <richlowe> building a primary gcc7 illumos?
[22:59:31] <leoric> (for example, because it finds wrong runtime)
[22:59:37] <richlowe> on hipster?
[22:59:41] <leoric> yes
[22:59:42] <richlowe> no
[22:59:47] <richlowe> how do you think I'm testing it? :)
[22:59:52] <leoric> :)
[23:00:10] <leoric> I mean, using gcc-6 userland && gcc-4-built-make
[23:00:10] *** alanc <alanc!~alanc@hqdc-proxy-mwg015-o.oracle.com> has quit IRC (Remote host closed the connection)
[23:00:28] *** alanc <alanc!~alanc@hqdc-proxy-mwg015-o.oracle.com> has joined #oi-dev
[23:00:29] *** ChanServ sets mode: +o alanc
[23:00:33] <richlowe> haven't tried that, I don't think
[23:01:31] <leoric> yes, elfdump components/openindiana/illumos-gate/illumos-gate/proto/root_i386/usr/bin/make |grep RPA
[23:01:31] <leoric> [35] SUNW_STRPAD 0x200
[23:01:40] <leoric> likely it didn't get it
[23:01:48] <leoric> luckily
[23:02:11] <igork> leoric: RPATH ?
[23:02:42] <igork> ldd can show dependencies
[23:02:45] <leoric> yes, luckily it doesn't use libs from outside the gate, so didn't get RPATH set
[23:02:53] <leoric> and so uses correct RUNPATH
[23:03:00] <leoric> but it's a matter of luck
[23:03:10] <richlowe> I'd expect make to need libstdc++ and libgcc_s
[23:03:19] <richlowe> but to expect to find them in /usr/lib
[23:03:21] <igork> yes
[23:03:22] <leoric> yes
[23:03:30] <igork> depend on gcc bits
[23:03:33] <richlowe> it's not luck with illumos-gcc, illumos-gcc is specifically not patched to set the runpath.
[23:03:55] <leoric> but if it was linked to host lib, which is dependent on g++ runtime...
[23:04:00] <igork> i'm using 64bits defaults
[23:04:04] <leoric> and it would be gcc 6 runtime
[23:04:21] <richlowe> right, every solution sucks
[23:04:23] <igork> and illumos make can be built in userland - outside illumos tree
[23:04:31] <richlowe> that's why we spent years reaching this one
[23:04:39] <igork> i do it now on dilos
[23:05:41] <igork> https://bitbucket.org/dilos/du2/src/5cb0fe8e766afc360a41433866e0adb9696c7013/components/sunmake/?at=master
[23:05:44] <igork> for example
[23:06:18] <igork> https://bitbucket.org/dilos/du2/src/5cb0fe8e766a/components/sunmake/debian/?at=master / rules contain rules for builds, sources are available under dilos apt
[23:07:53] <igork> sources https://bitbucket.org/dilos/dilos-sunmake/commits/all
[23:08:18] <igork> they are contain original sources + patched from illumos tree
[23:14:48] *** kohju <kohju!~kohju@gw.justplayer.com> has quit IRC (Ping timeout: 240 seconds)
[23:18:13] *** kohju <kohju!~kohju@gw.justplayer.com> has joined #oi-dev
[23:18:14] *** ChanServ sets mode: +o kohju

   February 10, 2018  
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | >