| *** juanalday has joined #buildstream | 00:18 | |
| *** juanalday has quit IRC | 00:21 | |
| *** dylan-m__ has quit IRC | 00:28 | |
| *** juanalday has joined #buildstream | 03:19 | |
| *** juanalday has quit IRC | 03:22 | |
| *** juanalday has joined #buildstream | 07:25 | |
| *** tpollard has joined #buildstream | 08:46 | |
| *** akvilebirgelyte has quit IRC | 09:21 | |
| *** akvilebirgelyte has joined #buildstream | 09:22 | |
| *** tiagogomes has joined #buildstream | 09:41 | |
| *** santi has joined #buildstream | 09:42 | |
| coldtom | hi, would there be any disagreement with creating a tmpfs at `/dev/shm` in the bwrap sandbox? bazel tries to use it, but I don't want to share the host one | 09:55 |
|---|---|---|
| *** jonathanmaw has joined #buildstream | 10:01 | |
| juergbi | coldtom: I can't think of an issue with that (ignoring excessive memory usage) | 10:22 |
| *** benschubert has joined #buildstream | 10:24 | |
| *** lachlan has joined #buildstream | 10:30 | |
| *** lachlan has quit IRC | 10:31 | |
| *** lachlan has joined #buildstream | 10:32 | |
| tpollard | sharing the host one does seem somewhat anti-sandbox | 10:37 |
| tlater[m] | coldtom: Once again, we don't use the host one! | 10:40 |
| tlater[m] | *Unless* you're running BuildStream interactively with `bst shell` | 10:40 |
| coldtom | tlater[m]: *or* unless you actually specify to include /dev/shm in the sandbox, when it gets --dev-bind'ed in too | 10:41 |
| coldtom | using either `Sandbox.mark_directory()` or adding it to DEVICES | 10:42 |
| coldtom | but without using one of these apis, /dev/shm doesn't exist in the sandbox | 10:43 |
| tlater[m] | Yes, I think the issue is that we want an API for adding /dev/shm, not that we're adding the host /dev/shm | 10:43 |
| coldtom | yeah, would this be better as an api to add a generic tmpfs or just adding /dev/shm by default? | 10:44 |
| tlater[m] | I'd prefer a generic way, though I'm not entirely sure what that would look like | 10:45 |
| juergbi | I would say the sandbox should setup /dev/shm by default | 10:46 |
| juergbi | libc shm_open() depends on this on Linux | 10:46 |
| juergbi | i.e., it's required for POSIX support on Linux | 10:46 |
| juergbi | and as this is platform-specific, we should rather not expose it to plugins | 10:47 |
| tlater[m] | juergbi: Does that work in a userchroot world? | 10:47 |
| juergbi | tlater[m]: yes, userchroot --install-devices mounts a 128MB tmpfs on dev/shm | 10:48 |
| juergbi | we should probably limit the size in our sandbox as well | 10:48 |
| juergbi | and an integration test using shm_open would make sense | 10:52 |
| gitlab-br-bot | BenjaminSchubert opened MR !1690 (bschubert/fix-children-termination->master: app.py: Also catch SystemError with click.Abort) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1690 | 11:43 |
| gitlab-br-bot | tpollard approved MR !1690 (bschubert/fix-children-termination->master: app.py: Also catch SystemError with click.Abort) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1690 | 11:44 |
| benschubert | juergbi: you also were seeing ^ on 1.4 right? So we should backport. Only on 1.4 or also 1.2? | 11:45 |
| benschubert | (I mean do we need another backport than 1.4?) | 11:45 |
| *** lachlan has quit IRC | 11:53 | |
| juergbi | benschubert: yes, I saw that on 1.4 | 11:55 |
| juergbi | does anyone still care about 1.2? no idea | 11:56 |
| juergbi | don't have a venv for it anymore | 11:56 |
| coldtom | afaik there are no fdsdk-derived projects still using 1.2 | 11:56 |
| benschubert | ok! I'll backport to 1.4 only then, thanks! | 12:03 |
| *** tpollard has quit IRC | 12:12 | |
| *** tpollard has joined #buildstream | 12:13 | |
| *** lachlan has joined #buildstream | 12:15 | |
| gitlab-br-bot | marge-bot123 merged MR !1690 (bschubert/fix-children-termination->master: app.py: Also catch SystemError with click.Abort) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1690 | 12:16 |
| tpollard | Thanks for catching that one benschubert | 12:18 |
| benschubert | tpollard: more to come ;) | 12:18 |
| *** lachlan has quit IRC | 12:21 | |
| gitlab-br-bot | BenjaminSchubert opened MR !1691 (bschubert/fix-children-termination->master: scheduler.py: Prevent the asyncio loop from leaking into subprocesses) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1691 | 12:22 |
| benschubert | juergbi: ^ I'd appreciate a thorough review on this one :) | 12:22 |
| benschubert | especially on the comments and such, since it's quite arcane... | 12:23 |
| benschubert | tpollard: ^ is the last blocker before I can actually fix the worker errors ;) | 12:23 |
| juergbi | will take a look | 12:23 |
| tpollard | I'll take a look too :) | 12:25 |
| benschubert | thanks! And if you want to ensure the behavior is correct: | 12:26 |
| benschubert | add 'assert False' at https://gitlab.com/BuildStream/buildstream/blob/master/src/buildstream/_scheduler/scheduler.py#L496, and try ctrl-c twice once the pipeline is running :) | 12:26 |
| benschubert | And then change the branch while keeping the 'assert False' | 12:26 |
| *** lachlan has joined #buildstream | 12:39 | |
| *** lachlan has quit IRC | 12:43 | |
| *** lachlan has joined #buildstream | 12:57 | |
| *** lachlan has quit IRC | 13:02 | |
| *** santi has quit IRC | 13:12 | |
| *** santi has joined #buildstream | 13:46 | |
| *** jonathanmaw has quit IRC | 14:04 | |
| *** lachlan has joined #buildstream | 14:56 | |
| gitlab-br-bot | BenjaminSchubert opened MR !1692 (bschubert/graceful-children-sigterm->master: Gracefully shutdown children on termination) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1692 | 15:00 |
| benschubert | tpollard: ^ :D | 15:02 |
| benschubert | please try to break it as inventively as you can ;) | 15:02 |
| *** lachlan has quit IRC | 15:05 | |
| *** slaf has quit IRC | 15:19 | |
| *** slaf has joined #buildstream | 15:19 | |
| *** slaf has joined #buildstream | 15:19 | |
| *** phil has joined #buildstream | 15:20 | |
| *** slaf has joined #buildstream | 15:20 | |
| *** slaf has joined #buildstream | 15:20 | |
| *** slaf has joined #buildstream | 15:20 | |
| *** slaf has joined #buildstream | 15:20 | |
| *** phildawson_ has quit IRC | 15:21 | |
| gitlab-br-bot | cs-shadow closed issue #1175 (buildobx-casd doesn't exit cleanly after some BuildStream errors) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1175 | 15:30 |
| gitlab-br-bot | coldtom opened issue #1197 (/dev/shm not available in sandbox (bwrap)) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1197 | 15:36 |
| gitlab-br-bot | coldtom opened MR !1694 (coldtom/allow-dev-shm->master: _sandboxbwrap.py: Create /dev/shm in the sandbox) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1694 | 15:38 |
| *** lachlan has joined #buildstream | 15:40 | |
| *** santi has quit IRC | 15:43 | |
| *** lachlan has quit IRC | 15:46 | |
| *** lachlan has joined #buildstream | 15:47 | |
| gitlab-br-bot | cs-shadow closed issue #1027 (bst build --track documentation is unclear) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1027 | 15:48 |
| cs-shadow | WSalmon: hi, mind having a look at https://gitlab.com/BuildStream/buildstream/issues/519#note_228932922 to see if it's still relevant | 15:51 |
| WSalmon | cs-shadow, no problem | 15:52 |
| cs-shadow | thanks :) | 15:52 |
| WSalmon | but its based on freedesktop so im not sure i can see how well it work on master, | 15:52 |
| WSalmon | i will have another look anyway | 15:53 |
| WSalmon | and remind myself what i was doing | 15:53 |
| *** phil has quit IRC | 15:54 | |
| *** lachlan has quit IRC | 15:57 | |
| coldtom | WSalmon: bschubert/freedesktop-sdk-19.08.3-bst2 branch in freedesktop-sdk works with master, so you could use that to check | 16:04 |
| gitlab-br-bot | BenjaminSchubert approved MR !1694 (coldtom/allow-dev-shm->master: _sandboxbwrap.py: Create /dev/shm in the sandbox) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1694 | 16:06 |
| benschubert | juergbi: https://gitlab.com/BuildStream/buildstream/-/jobs/345764095 have any idea? | 16:08 |
| juergbi | benschubert: doesn't look familiar | 16:10 |
| tpollard | I've never seen that one | 16:10 |
| benschubert | yeah, that seems to be coming from my changes | 16:10 |
| benschubert | but why only in this one? -_-' | 16:11 |
| benschubert | and I don't like the 'bad file descriptor' there | 16:11 |
| benschubert | guess I need more thorough testing | 16:11 |
| *** lachlan has joined #buildstream | 16:18 | |
| gitlab-br-bot | cs-shadow closed issue #691 (BuildStream does not automatically fetch a junction within a junction / displays an inaccurate helper) on buildstream https://gitlab.com/BuildStream/buildstream/issues/691 | 16:18 |
| WSalmon | should this be a thing? | 16:19 |
| WSalmon | (venvbuildstream) [will@wsx280 buildstream-kicad]$ bst source track freedesktop-sdk.bst | 16:19 |
| WSalmon | Error loading user configuration: /home/will/projects/buildstream/venvbuildstream/lib64/python3.7/site-packages/buildstream/data/userconfig.yaml: Severely malformed YAML: | 16:19 |
| WSalmon | 'NoneType' object is not callable | 16:19 |
| WSalmon | i havent played with master for a while so i checked out master and pip unintalled and re installed and git a git clean | 16:20 |
| benschubert | WSalmon: pip uninstall ruamel.yaml.clib && pip install ruamel.yaml.clib and retry please :) | 16:20 |
| WSalmon | thanks benschubert | 16:21 |
| WSalmon | i see casd has main its way in to master, well dont juergbi! et al! now to rember how to install it | 16:22 |
| juergbi | benschubert: corresponding file descriptor closed without calling set_wakeup_fd(-1)? | 16:22 |
| benschubert | juergbi: yep | 16:23 |
| benschubert | this doesn't happen when testing locally | 16:23 |
| benschubert | so I have no idea where this is happening | 16:23 |
| benschubert | also, the test passed after a retry | 16:23 |
| cs-shadow | WSalmon: https://buildstream.build/buildbox_install.html | 16:25 |
| *** lachlan has quit IRC | 16:29 | |
| *** lachlan has joined #buildstream | 16:29 | |
| WSalmon | git clone https://gitlab.com/BuildGrid/buildbox/buildbox-casd.git | 16:29 |
| WSalmon | cd buildbox-casd | 16:29 |
| WSalmon | git checkout 0.0.4 | 16:29 |
| WSalmon | cmake -DBUILD_TESTING=OFF -Bbuild | 16:29 |
| WSalmon | make -C build | 16:29 |
| WSalmon | make -C build install n | 16:29 |
| WSalmon | sorry | 16:30 |
| WSalmon | this was left in my paste buffer | 16:30 |
| tpollard | :) | 16:33 |
| *** phil has joined #buildstream | 16:34 | |
| *** lachlan has quit IRC | 16:36 | |
| *** tpollard has quit IRC | 16:43 | |
| *** lachlan has joined #buildstream | 16:49 | |
| *** lachlan has quit IRC | 16:53 | |
| WSalmon | is this https://hastebin.com/ipudohilok.sql just me? fells like we would rather have this have a better error message even if its my bad yaml | 16:53 |
| benschubert | WSalmon: please use gitlab snippets, I can't access this | 16:55 |
| WSalmon | oh ok | 16:55 |
| gitlab-br-bot | cs-shadow opened MR !1695 (chandan/stack-no-sources->master: plugins/elements/stack: Disallow sources) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1695 | 17:03 |
| gitlab-br-bot | BenjaminSchubert approved MR !1695 (chandan/stack-no-sources->master: plugins/elements/stack: Disallow sources) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1695 | 17:04 |
| juergbi | WSalmon: this means the plugin is not compatible with the buildstream version | 17:05 |
| juergbi | not invalid yaml | 17:05 |
| *** santi has joined #buildstream | 17:05 | |
| WSalmon | benschubert, https://gitlab.com/willsalmon/buildstream-kicad/snippets/1911549 | 17:07 |
| cs-shadow | WSalmon: seems like you need a newer version of bst-plugins-exeperimental | 17:07 |
| WSalmon | fair | 17:08 |
| WSalmon | but should we also have a nicer error message? | 17:08 |
| cs-shadow | it'll be nice from a UX perspective, but I'm not sure if it's worth maintaining such compatiability across development releases | 17:09 |
| WSalmon | I have seen that error before for other things, should it not be wrapped a i little? | 17:13 |
| gitlab-br-bot | marge-bot123 merged MR !1695 (chandan/stack-no-sources->master: plugins/elements/stack: Disallow sources) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1695 | 17:19 |
| WSalmon | any one under stand what is going on hear? https://gitlab.com/willsalmon/buildstream-kicad/snippets/1911552 | 17:23 |
| *** santi has quit IRC | 17:24 | |
| gitlab-br-bot | cs-shadow closed issue #699 (invalid buildstream.conf format returns stacktrace) on buildstream https://gitlab.com/BuildStream/buildstream/issues/699 | 17:24 |
| benschubert | WSalmon: that should be https://gitlab.com/BuildStream/buildstream/issues/1178 right? | 17:31 |
| *** juanalday has quit IRC | 17:34 | |
| *** phildawson_ has joined #buildstream | 17:35 | |
| *** phil has quit IRC | 17:35 | |
| WSalmon | benschubert, yep, but why dose your branch of free desktop not remove the caches or dose it get round it a diffrent way? | 17:39 |
| benschubert | WSalmon: BuilDStream used to check for capabilities, and it's currently broken on master | 17:41 |
| WSalmon | ah | 17:42 |
| benschubert | so I don't want to remove it because it will hide the fact that this is broken and intend on fixing this | 17:42 |
| WSalmon | yep | 17:42 |
| *** lachlan has joined #buildstream | 17:53 | |
| *** lachlan has quit IRC | 17:57 | |
| gitlab-br-bot | BenjaminSchubert opened MR !1696 (bschubert/expand-path-configs->master: _remote.py: Expand user in certificates paths) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1696 | 17:57 |
| *** bochecha has quit IRC | 17:58 | |
| *** lachlan has joined #buildstream | 18:20 | |
| gitlab-br-bot | cs-shadow opened MR !1697 (chandan/no-command-general-elements->master: Indicate that compose & stack plugins do not run commands) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1697 | 18:28 |
| *** lachlan has quit IRC | 18:28 | |
| *** lachlan has joined #buildstream | 18:33 | |
| *** lachlan has quit IRC | 18:49 | |
| *** tiagogomes has quit IRC | 18:50 | |
| *** lachlan has joined #buildstream | 18:57 | |
| *** nhbh^ has joined #buildstream | 19:37 | |
| *** lachlan has quit IRC | 20:37 | |
| *** cs-shadow has quit IRC | 20:46 | |
| *** narispo has quit IRC | 22:07 | |
| *** juanalday has joined #buildstream | 22:09 | |
| *** juanalday has quit IRC | 22:20 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!