| *** rdale has quit IRC | 04:53 | |
| *** ingoerler has joined #buildstream | 06:40 | |
| *** ingoerler has quit IRC | 06:42 | |
| juergbi | tlater[m]: I'm now also seeing the empty alpine base issue with bb-run-userchroot and your branch. I'll start looking into this as well unless you've already discovered the root cause | 08:00 |
|---|---|---|
| juergbi | I'll keep you posted | 08:01 |
| *** traveltissues has joined #buildstream | 08:47 | |
| gitlab-br-bot | BenjaminSchubert approved MR !1702 (chandan/fix-implicit-string-concat->master: tests/frontend/completions: Fix implicit string concatenation) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1702 | 09:20 |
| tlater[m] | Ta juergbi - not figured it our yet, no, though I may not yet have changed the tempdir to be group-writeable | 09:23 |
| *** rdale has joined #buildstream | 09:27 | |
| benschubert | Does someone knows why we have: | 09:39 |
| benschubert | ``` | 09:39 |
| benschubert | # Write down the user configuration file | 09:39 |
| benschubert | _yaml.roundtrip_dump(user_config, file=user_config_file) | 09:39 |
| benschubert | # Ensure CLI calls will use it | 09:39 |
| benschubert | cli.configure(user_config) | 09:39 |
| benschubert | ``` | 09:39 |
| benschubert | In most tests? I would expect one of them, but both? | 09:39 |
| *** tiagogomes has joined #buildstream | 09:42 | |
| traveltissues | that doesn't seem right | 09:46 |
| *** santi has joined #buildstream | 09:46 | |
| gitlab-br-bot | tpollard approved MR !1701 (bschubert/backport-1690->bst-1: app.py: Also catch SystemError with click.Abort) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1701 | 09:47 |
| gitlab-br-bot | traveltissues approved MR !1702 (chandan/fix-implicit-string-concat->master: tests/frontend/completions: Fix implicit string concatenation) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1702 | 09:56 |
| juergbi | tlater[m]: the sourcecache staging temp dir is group readable but not executable | 09:56 |
| juergbi | not sure why, though | 09:56 |
| juergbi | also, utils._group_tempdir() doesn't pass on the kwargs | 09:56 |
| juergbi | benschubert: might have been incomplete migration to cli.configure(), no idea | 09:59 |
| benschubert | juergbi: ok :) thanks! | 09:59 |
| benschubert | juergbi: do you know if there is a helper to create an artifact server that forbids 'push' for the testsuite? | 10:02 |
| juergbi | benschubert: I don't know, brief look at the code seems like it's always enable push | 10:03 |
| juergbi | should be simple enough to parametrize, though | 10:03 |
| benschubert | right :) thanks! | 10:03 |
| tpollard | you could set push to false in the user_config, but obviously that's not the server itself forbidding push | 10:04 |
| benschubert | tpollard: yeah I want to forbid the server | 10:04 |
| benschubert | because master currently will trip if you set 'push' = true but your server doesn't allow it :) | 10:04 |
| juergbi | tlater[m]: ok, found it. looks like tar's stage() method changes the permission bits of the directory | 10:06 |
| juergbi | we currently enforce at least 0440 permissions | 10:07 |
| juergbi | but that doesn't provide x for directories | 10:07 |
| juergbi | workaround shouldn't be too difficult but have to think about how to handle this in general | 10:07 |
| tlater[m] | juergbi: Gah, I missed that :| | 10:11 |
| tlater[m] | Yeah, there are a lot of workarounds about, we need something a bit more generic than that branch | 10:12 |
| juergbi | the original 0400 was also buggy in that regard | 10:13 |
| tlater[m] | Odd that we never ran into that issue, though | 10:14 |
| juergbi | it's very uncommon for tar files to have directories that are not readable/executable by the owner, I suppose | 10:14 |
| juergbi | so unless we have a test for that, easy to miss | 10:15 |
| juergbi | for group permissions it's probably more common | 10:15 |
| juergbi | for now we could say sources must respect umask and not do anything more restrictive | 10:16 |
| *** jonathanmaw has joined #buildstream | 10:16 | |
| juergbi | this wouldn't allow exploiting the new UnixMode node property, however, we don't have immediate plans for that yet anyway (#38) | 10:17 |
| juergbi | but we should probably think about the long term plan, to avoid imposing this constraint on sources now and changing that again later on | 10:18 |
| *** adds68 has joined #buildstream | 10:22 | |
| benschubert | should a server configured as `push` which doens't allow `push` be blacklisted entirely? | 10:27 |
| juergbi | not sure, I'd rather say to keep it for pull | 10:29 |
| juergbi | but at least warn | 10:29 |
| benschubert | ok, because currently it's entirely blacklisted :) | 10:30 |
| coldtom | i would expect it to just pull too | 10:31 |
| gitlab-br-bot | traveltissues opened issue #1201 (Improve error communication from buildbox) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1201 | 10:31 |
| *** lachlan has joined #buildstream | 10:32 | |
| *** bochecha has joined #buildstream | 10:47 | |
| gitlab-br-bot | traveltissues opened issue #1202 (Test colors and no-colors outputs) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1202 | 10:50 |
| gitlab-br-bot | traveltissues approved MR !1704 (chandan/fix-no-color->master: _frontend: Simplify color handling) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1704 | 10:51 |
| benschubert | tlater[m]: with your cas work, are you refactoring parts of the tests too or not? | 11:19 |
| gitlab-br-bot | marge-bot123 merged MR !1702 (chandan/fix-implicit-string-concat->master: tests/frontend/completions: Fix implicit string concatenation) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1702 | 11:27 |
| tlater[m] | benschubert: A tiny bit. Not by much, just adding logs to them. | 11:31 |
| gitlab-br-bot | traveltissues opened issue #1203 (Performance regression in `bst show` and `bst build`) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1203 | 11:32 |
| gitlab-br-bot | BenjaminSchubert opened MR !1705 (bschubert/fix-capa-check->master: Fix capabilities check of remotes) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1705 | 11:42 |
| benschubert | ^ should also fix our overnight tests (unless something more is broken) | 11:43 |
| tlater[m] | benschubert: Is that RemoteError caught and then resolved so that one server not supporting capabilities doesn't keep all the others from being used? | 11:47 |
| benschubert | tlater[m]: I haven't changed the behavior of the RemoteError | 11:48 |
| benschubert | if tyou look at _remote.py:BaseRemote.check(), it does throw a RemoteError on error | 11:48 |
| benschubert | so if that worked, I would expect it to still work :) | 11:48 |
| tlater[m] | Right, yeah, I suspect it was missed during a refactor then. | 11:49 |
| tlater[m] | Ta benschubert :) | 11:49 |
| gitlab-br-bot | tlater approved MR !1705 (bschubert/fix-capa-check->master: Fix capabilities check of remotes) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1705 | 11:49 |
| tlater[m] | I also really like the context manager for the artifact share | 11:50 |
| tlater[m] | It's cute to spin up a grpc server that does nothing :D | 11:51 |
| gitlab-br-bot | traveltissues approved MR !1705 (bschubert/fix-capa-check->master: Fix capabilities check of remotes) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1705 | 11:52 |
| coldtom | hi, i'm trying to point a bst (master) project at a remote cache, but the remote is failing to initialise with "OS Error", I managed to get a more verbose error which tells me it's coming from casd.sock, anyone know if this is from the server or the client? | 11:53 |
| benschubert | coldtom: you should have casd logs in ~/.cache/buildstream/casd/logs/ , do you get anything there? | 11:55 |
| benschubert | tlater[m]: was the easiest way :D | 11:56 |
| benschubert | and this thing that does nothing catches two bugs :D | 11:56 |
| coldtom | ah, looks like it's at least trying to connect, thanks benschubert | 11:58 |
| benschubert | what's Benchmark_Prio-1 ? :D | 12:01 |
| lachlan | benschubert: This is initial testing of the benchmarking bot and that is a label used to signify that benchmarking should be carried out on that particular MR. | 12:06 |
| benschubert | oh sweet! | 12:06 |
| tlater[m] | lachlan: Any problem with modifying the MR while that is running? | 12:10 |
| lachlan | tlater[m]: Shouldn't be | 12:13 |
| lachlan | The note ref is fixed in the benchmarking pipeline build variable | 12:13 |
| * tlater[m] just so happens to want to pick that MR up again today :D | 12:13 | |
| lachlan | and the sha is fixed as well | 12:13 |
| lachlan | https://gitlab.com/BuildStream/buildstream/merge_requests/1660#note_243396747 | 12:14 |
| lachlan | https://gitlab.com/BuildStream/benchmarking/benchmarks/pipelines/95336710 | 12:14 |
| benschubert | tpollard: when running the WSL benchmark, do you still ahve the exact cmdline? And how was the terminal positioned? :) | 12:19 |
| tpollard | traveltissues ^ | 12:20 |
| tpollard | benschubert: as a side note, would you have an issue if I proposed moving the _multiprocessing to the root of our modules? | 12:22 |
| tpollard | as I think I should be used it when multiprocessing in my branch | 12:23 |
| benschubert | No issue at all :) | 12:23 |
| tpollard | cool! | 12:25 |
| benschubert | My aim for this module would be to abstract away all weird things we need to do with multiprocessing: Be it to support Windows, or fork, or whatever :) And I'd like to introduce in the future a _asyncio that would do the same around asyncio :) | 12:26 |
| traveltissues | tpollard, vertical second monitor fullscreeen | 12:26 |
| benschubert | so we don't need to care anywhere else about those, and stop having the 'if windows: do_this; else do_that' | 12:27 |
| benschubert | traveltissues: thanks! I'll run it the same, the embeded wsl terminal right? | 12:27 |
| traveltissues | time bst --builders 4 build base-files/base-files.bst | 12:27 |
| traveltissues | benschubert, yes | 12:27 |
| benschubert | with BST_PROFILE=SUBPROCESS right? | 12:27 |
| tpollard | note the times listed on the MR are without the profile running, that will add it's own overheads ofc | 12:27 |
| traveltissues | no | 12:28 |
| benschubert | ok! So I'll run it twice with and without profile, cheers | 12:28 |
| tpollard | thanks | 12:31 |
| traveltissues | benschubert, i'm launching the ubuntu 18.04 LTS app | 12:31 |
| gitlab-br-bot | marge-bot123 closed issue #1200 (`bst init` does not respect `--no-colors`) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1200 | 13:01 |
| gitlab-br-bot | marge-bot123 merged MR !1704 (chandan/fix-no-color->master: _frontend: Simplify color handling) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1704 | 13:01 |
| *** santi has quit IRC | 13:39 | |
| *** santi has joined #buildstream | 13:49 | |
| *** traveltissues has quit IRC | 14:27 | |
| gitlab-br-bot | marge-bot123 closed issue #1178 (BuildStream doesn't check that the artifact remotes has the required capabilities) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1178 | 14:31 |
| gitlab-br-bot | marge-bot123 merged MR !1705 (bschubert/fix-capa-check->master: Fix capabilities check of remotes) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1705 | 14:31 |
| gitlab-br-bot | BenjaminSchubert closed issue #1198 (overnight tests are failing: "tests/check-platform.bst: Failed to pull artifact: Method not found!") on buildstream https://gitlab.com/BuildStream/buildstream/issues/1198 | 14:36 |
| *** cs-shadow has joined #buildstream | 14:45 | |
| gitlab-br-bot | cs-shadow opened (was WIP) MR !1703 (chandan/black->master: Use Black to format code) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1703 | 14:45 |
| * tpollard looks forward to the diff | 14:48 | |
| bochecha | hi, anything I can do to get this merged? https://gitlab.com/BuildStream/bst-external/merge_requests/109 | 15:08 |
| *** lachlan has quit IRC | 15:22 | |
| *** lachlan has joined #buildstream | 15:24 | |
| *** santi has quit IRC | 15:45 | |
| *** santi has joined #buildstream | 15:45 | |
| *** lachlan has quit IRC | 15:47 | |
| bochecha | jjardon: ha, thanks! | 16:05 |
| benschubert | bochecha: would it be possible to also make a PR against bst-plugins-experimental for this? | 16:06 |
| bochecha | benschubert: I was waiting for the first one to be accepted, so I didn't have review feedback in two places at the same time | 16:07 |
| bochecha | I'm doing the -experimental right now :) | 16:07 |
| benschubert | great! | 16:07 |
| bochecha | (you had already asked me last week I think? :P) | 16:07 |
| benschubert | Gah, growing old it seems :'D | 16:07 |
| bochecha | hehe | 16:07 |
| bochecha | same for me, I would have completely forgotten about the MR itself if I hadn't kept the tab open in Firefox | 16:08 |
| bochecha | benschubert: there you go: https://gitlab.com/BuildStream/bst-plugins-experimental/merge_requests/51 | 16:10 |
| benschubert | thanks! | 16:11 |
| *** tpollard has quit IRC | 16:23 | |
| *** tpollard has joined #buildstream | 16:25 | |
| *** lachlan has joined #buildstream | 16:34 | |
| *** lachlan has quit IRC | 16:42 | |
| *** santi has quit IRC | 16:44 | |
| *** lachlan has joined #buildstream | 16:54 | |
| *** lachlan has quit IRC | 16:58 | |
| *** santi has joined #buildstream | 16:58 | |
| tpollard | hmm, I think some of the colouring of the UI has gone since 1704 | 16:58 |
| *** lachlan has joined #buildstream | 17:03 | |
| gitlab-br-bot | tpollard opened issue #1204 (Colourisation lost in Status UI) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1204 | 17:13 |
| *** jonathanmaw has quit IRC | 17:28 | |
| tlater[m] | cs-shadow: I don't want to scope creep that MR too much, I'd just like to catch all files now if it's not too much effort. Current exceptions are mostly accidental, with the exception of proto/fuse stuff, it would be nice to clean that up. | 17:32 |
| cs-shadow | tlater[m]: I get that. As I said in the last comment, I was just not sure whether anyone was planning on `sphinx-quickstart` ever again | 17:33 |
| * cs-shadow thinks not | 17:33 | |
| tlater[m] | Hah | 17:34 |
| *** testy has joined #buildstream | 17:41 | |
| *** lachlan has quit IRC | 17:55 | |
| *** santi has quit IRC | 18:02 | |
| *** lachlan has joined #buildstream | 18:16 | |
| *** testy has quit IRC | 18:26 | |
| *** phoenix has joined #buildstream | 18:39 | |
| *** testy has joined #buildstream | 18:42 | |
| *** lachlan has quit IRC | 19:04 | |
| gitlab-br-bot | cs-shadow opened issue #1205 (Improve messaging around state of junction elements) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1205 | 20:15 |
| *** phoenix has quit IRC | 20:23 | |
| *** bochecha_ has joined #buildstream | 21:30 | |
| *** cs-shadow has quit IRC | 21:37 | |
| *** lantw44 has quit IRC | 22:10 | |
| *** lantw44 has joined #buildstream | 22:11 | |
| *** rdale has quit IRC | 23:10 | |
| *** bochecha_ has quit IRC | 23:51 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!