| *** toscalix has joined #buildstream | 07:40 | |
| gitlab-br-bot | juergbi opened MR !1414 (juerg/fetch-subprojects->master: Improve subproject fetching) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1414 | 07:56 |
|---|---|---|
| *** bochecha has joined #buildstream | 08:15 | |
| *** rdale has joined #buildstream | 08:35 | |
| *** lantw44 has joined #buildstream | 08:36 | |
| *** raoul has joined #buildstream | 08:42 | |
| raoul | jennis, you happy with !1400? If so think I'll hand it over to marge | 08:58 |
| gitlab-br-bot | MR !1400: CLI options for on demand blob fetching https://gitlab.com/BuildStream/buildstream/merge_requests/1400 | 08:58 |
| jennis | raoul, looking now | 09:00 |
| raoul | Think everything is resolved since you looked on Friday | 09:00 |
| *** jonathanmaw has joined #buildstream | 09:01 | |
| jennis | raoul, how comes there is added artifacts config to the autotools example? https://gitlab.com/BuildStream/buildstream/merge_requests/1400/diffs#723fb246d6298cb56eb99c65b71dc929135bd3d0_14_15 | 09:02 |
| raoul | oops good catch, that'd be when I was poking it, will remove | 09:03 |
| benschubert | hey tristan, when you have time, could you have a look at https://gitlab.com/BuildStream/buildstream/merge_requests/1401#note_184528846 ? | 09:03 |
| jennis | Cool, also another nit with a doc comment | 09:08 |
| jennis | raoul, a few more comments there | 09:20 |
| *** lachlan has joined #buildstream | 09:39 | |
| raoul | Hmmm seems that some of the integration shell tests fail depending on how they are run. Errors seem to point to _yaml.pyx so have you got any idea benschubert? | 10:00 |
| raoul | error is: | 10:00 |
| raoul | src/buildstream/_yaml.pyx:407: in buildstream._yaml.load_data | 10:00 |
| raoul | cpdef Node load_data(str data, int file_index=_SYNTHETIC_FILE_INDEX, str file_name=None, bint copy_tree=False): | 10:00 |
| raoul | _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ | 10:00 |
| raoul | > f_info = <FileInfo> _FILE_LIST[file_index] | 10:00 |
| raoul | E IndexError: list index out of range | 10:00 |
| raoul | it's happening on master even after a git clean | 10:00 |
| benschubert | raoul: oh, do you have a link? | 10:01 |
| raoul | only seems to happen when you run a subset of tests, so I've only done it locally | 10:01 |
| benschubert | raoul: ouch, that seems weird | 10:02 |
| raoul | 'tox -- --integration -n 4 "tests/integration/shell.py"' seems to fail, but not if you don't have -n 4 | 10:02 |
| raoul | so definitely weird :/ | 10:02 |
| *** lachlan has quit IRC | 10:04 | |
| *** lachlan has joined #buildstream | 10:05 | |
| juergbi | raoul: this seems to fail on its own: tox -- --integration tests/integration/shell.py::test_isolated_no_mount | 10:06 |
| juergbi | probably accidentally relying by side-effects of other tests | 10:06 |
| juergbi | and missing test isolation | 10:06 |
| raoul | as does 'tox -- --integration "tests/integration/shell.py::test_env_assign[Pony]"', seems a few do | 10:07 |
| *** tristan has quit IRC | 10:07 | |
| benschubert | raoul: thanks for noticing, that's definitely bad, would you mind opening an issue? I'll try to have a look at it later today | 10:09 |
| raoul | will do | 10:09 |
| *** cs-shadow has joined #buildstream | 10:12 | |
| gitlab-br-bot | raoul.hidalgocharman opened issue #1051 (Some integration shell tests fail when run on their own) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1051 | 10:13 |
| *** lachlan has quit IRC | 10:14 | |
| raoul | ^^ benschubert | 10:14 |
| benschubert | thanks! | 10:15 |
| *** tristan has joined #buildstream | 10:22 | |
| *** lachlan has joined #buildstream | 10:41 | |
| gitlab-br-bot | jennis approved MR !1400 (raoul/1044-blobs-on-demand->master: CLI options for on demand blob fetching) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1400 | 10:43 |
| *** lachlan has quit IRC | 10:51 | |
| *** bochecha_ has joined #buildstream | 11:17 | |
| *** bochecha has quit IRC | 11:18 | |
| *** bochecha_ is now known as bochecha | 11:18 | |
| *** lantw44 has quit IRC | 11:25 | |
| *** lantw44 has joined #buildstream | 11:25 | |
| gitlab-br-bot | marge-bot123 closed issue #1044 (Fetch missing blobs on demand) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1044 | 11:29 |
| gitlab-br-bot | marge-bot123 merged MR !1400 (raoul/1044-blobs-on-demand->master: CLI options for on demand blob fetching) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1400 | 11:29 |
| *** lachlan has joined #buildstream | 11:31 | |
| *** lantw44 has quit IRC | 11:33 | |
| *** lantw44 has joined #buildstream | 11:33 | |
| *** nielsdg_ has joined #buildstream | 11:33 | |
| *** bochecha_ has joined #buildstream | 11:44 | |
| *** bochecha has quit IRC | 11:47 | |
| *** bochecha_ is now known as bochecha | 11:47 | |
| *** raoul has quit IRC | 11:53 | |
| *** raoul has joined #buildstream | 12:07 | |
| * tlater[m] is experiencing some interesting breakage: https://hastebin.com/fusigewime | 12:40 | |
| tlater[m] | It looks like the tar plugin somehow doesn't read the full path of a tar sub-file thing | 12:40 |
| tlater[m] | The link it's complaining about should be named `etc/ssl/ct_log_list.cnf`, not `c/ssl/ct_log_list.cnf` | 12:41 |
| tlater[m] | But I've double checked the tarfile and this is not an issue there | 12:41 |
| tlater[m] | Even if it was, we shouldn't abort the whole build. | 12:42 |
| tlater[m] | I guess tar.py:132 is removing characters it shouldn't? | 13:02 |
| tlater[m] | Hm, maybe there's a more robust way to do that... | 13:04 |
| tlater[m] | Ah, seems like this is a hard link | 13:09 |
| tlater[m] | And i guess we have no test asserting that that redirectidoo-fixing code actually works | 13:09 |
| *** lachlan has quit IRC | 13:14 | |
| *** lachlan has joined #buildstream | 13:26 | |
| gitlab-br-bot | jonathanmaw opened (was WIP) MR !1409 (jonathan/frontend-separation->master: Separate frontend state handling from core state) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1409 | 13:26 |
| *** lachlan has quit IRC | 13:30 | |
| *** rdale has quit IRC | 13:30 | |
| *** rdale has joined #buildstream | 13:35 | |
| *** lachlan has joined #buildstream | 13:38 | |
| *** lachlan has quit IRC | 14:01 | |
| gitlab-br-bot | tlater opened issue #1052 (BuildStream chokes when staging tarfiles with hardlinks that point outside of a chosen base directory) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1052 | 14:01 |
| *** nielsdg_ has quit IRC | 14:09 | |
| *** lachlan has joined #buildstream | 14:14 | |
| *** lachlan has quit IRC | 14:21 | |
| tlater[m] | I noticed that the build directories for integration tests no longer seem to linger | 14:34 |
| tlater[m] | I can't figure out how to get my staging directories now, any idea how to get that back? | 14:34 |
| gitlab-br-bot | jonathanmaw opened (was WIP) MR !1409 (jonathan/frontend-separation->master: Separate frontend state handling from core state) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1409 | 14:41 |
| *** phil has joined #buildstream | 14:53 | |
| *** phildawson_ has quit IRC | 14:53 | |
| *** lachlan has joined #buildstream | 15:00 | |
| *** tristan has quit IRC | 15:05 | |
| *** lachlan has quit IRC | 15:05 | |
| *** lachlan has joined #buildstream | 15:25 | |
| *** tristan has joined #buildstream | 15:30 | |
| jjardon | Hi, any idea what is going wrong here? https://gitlab.com/BuildStream/buildstream/issues/1049 the logs I linked on the issue doesn't give a lot of info | 15:34 |
| *** bochecha has quit IRC | 15:37 | |
| *** bochecha has joined #buildstream | 15:50 | |
| *** lachlan has quit IRC | 15:56 | |
| *** lachlan has joined #buildstream | 16:10 | |
| *** phil has quit IRC | 16:19 | |
| *** phil has joined #buildstream | 16:21 | |
| tlater[m] | Are script elements the only elements that can decide where to stage their dependencies? | 16:25 |
| tlater[m] | I.e., use "layout" in the config? | 16:25 |
| juergbi | tlater[m]: `layout` is indeed a script element feature | 16:38 |
| juergbi | other element plugins can do this as well but it's not available via YAML for non-script plugins | 16:39 |
| tlater[m] | Hm, that's a little annoying, it's a nice and comfortable way to shove everything into a directory in a stack element, for example. | 16:40 |
| *** ChanServ sets mode: +o tristan | 16:42 | |
| tristan | tlater[m], I don't think it's suitable for stack, but I think compose could benefit from that | 16:42 |
| *** lachlan has quit IRC | 16:42 | |
| tlater[m] | Compose does make more sense | 16:43 |
| tristan | stack doesnt really do anything, just symbolic dependency grouping | 16:43 |
| benschubert | Hey tristan, before I start changing all of this, would you mind having a look at https://gitlab.com/BuildStream/buildstream/merge_requests/1401#note_184696903 ? :) | 16:43 |
| gitlab-br-bot | marge-bot123 merged MR !1406 (jennis/notify_reverse_deps->master: Notify direct reverse dependencies when an Element becomes ready for runtime and cached) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1406 | 16:47 |
| jennis | :D | 16:47 |
| tristan | benschubert, done :) | 16:48 |
| gitlab-br-bot | tlater opened issue #1053 (Compose elements should allow changing the staging layout) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1053 | 16:48 |
| *** lachlan has joined #buildstream | 16:48 | |
| benschubert | tristan: great thanks! | 16:49 |
| tristan | tlater[m], interestingly that is going to be a bit tricky... note that integration commands can only be supported for things staged at / | 16:49 |
| tlater[m] | Oh, right | 16:50 |
| tristan | tlater[m], however, one can always use multiple compose elements if one really wants relocated integrated sysroots | 16:50 |
| tristan | just a detail to keep in mind | 16:51 |
| tristan | I think we have a weird hack in fdsdk where we use a script element without any script commands especially just to leverage this feature | 16:51 |
| tlater[m] | Yep, I was about to do that for the freedesktop-sdk tarball | 16:51 |
| tlater[m] | But it doesn't sit right with me, so... | 16:52 |
| *** lachlan has quit IRC | 16:52 | |
| tristan | yeah when I suggested that we use that feature... I kind of expected someone to just roll out a project local plugin for that purpose | 16:53 |
| tristan | but script element got used and it does look confusing | 16:53 |
| *** jonathanmaw has quit IRC | 16:56 | |
| *** shibu has joined #buildstream | 17:02 | |
| *** raoul has quit IRC | 17:10 | |
| cs-shadow | jennis: around? | 17:32 |
| *** lachlan has joined #buildstream | 17:33 | |
| *** lachlan has quit IRC | 17:38 | |
| *** shibu has quit IRC | 17:59 | |
| *** lachlan has joined #buildstream | 18:04 | |
| gitlab-br-bot | LaurenceUrhegyi opened MR !1415 (laurence/update-contributing-guide->master: Adds a Committer Policy section to CONTRIBUTING.rst) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1415 | 18:09 |
| *** lachlan has quit IRC | 18:17 | |
| *** lachlan has joined #buildstream | 18:32 | |
| *** lachlan has quit IRC | 19:39 | |
| *** lachlan has joined #buildstream | 19:40 | |
| *** toscalix has quit IRC | 19:52 | |
| *** lachlan has quit IRC | 20:01 | |
| *** swick has quit IRC | 21:14 | |
| *** slaf has quit IRC | 21:15 | |
| *** swick has joined #buildstream | 21:25 | |
| *** slaf has joined #buildstream | 21:45 | |
| *** slaf has joined #buildstream | 21:45 | |
| *** slaf has joined #buildstream | 21:46 | |
| *** slaf has joined #buildstream | 21:46 | |
| *** slaf has joined #buildstream | 21:46 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!