| *** Prince781 has joined #buildstream | 00:17 | |
| *** Prince781 has quit IRC | 01:04 | |
| *** Prince781 has joined #buildstream | 01:05 | |
| *** Prince781 has quit IRC | 01:56 | |
| *** Prince781 has joined #buildstream | 01:58 | |
| *** bochecha has quit IRC | 02:22 | |
| *** jsgrant has joined #buildstream | 03:24 | |
| *** Prince781 has quit IRC | 04:13 | |
| *** Prince781 has joined #buildstream | 04:14 | |
| *** Prince781 has quit IRC | 05:06 | |
| *** Prince781 has joined #buildstream | 05:07 | |
| *** Prince781 has quit IRC | 05:11 | |
| *** slaf has quit IRC | 05:14 | |
| *** slaf has joined #buildstream | 05:16 | |
| *** slaf has joined #buildstream | 05:16 | |
| *** slaf has joined #buildstream | 05:17 | |
| *** slaf has joined #buildstream | 05:17 | |
| *** slaf has joined #buildstream | 05:17 | |
| *** slaf has joined #buildstream | 05:17 | |
| *** slaf has joined #buildstream | 05:18 | |
| *** slaf has joined #buildstream | 05:18 | |
| *** slaf has joined #buildstream | 05:18 | |
| *** slaf has joined #buildstream | 05:18 | |
| *** slaf has joined #buildstream | 05:19 | |
| *** slaf has joined #buildstream | 05:19 | |
| *** slaf has joined #buildstream | 05:19 | |
| *** slaf has joined #buildstream | 05:19 | |
| *** slaf has joined #buildstream | 05:20 | |
| *** slaf has joined #buildstream | 05:20 | |
| *** Prince781 has joined #buildstream | 05:32 | |
| *** toscalix has joined #buildstream | 06:30 | |
| *** toscalix has quit IRC | 06:31 | |
| *** toscalix has joined #buildstream | 06:31 | |
| *** Phil has joined #buildstream | 07:16 | |
| *** coldtom has joined #buildstream | 07:52 | |
| *** jonathanmaw has joined #buildstream | 08:36 | |
| toscalix | Hi | 08:43 |
|---|---|---|
| paulsherwood | hi | 08:43 |
| *** jennis_ is now known as jennis | 08:51 | |
| *** tiago has joined #buildstream | 08:54 | |
| *** tiago has quit IRC | 08:58 | |
| *** tiago has joined #buildstream | 09:03 | |
| *** finn has quit IRC | 09:15 | |
| *** finn has joined #buildstream | 09:17 | |
| *** bethw has joined #buildstream | 09:21 | |
| finn | Hey | 09:27 |
| tlater | o/ finn | 09:27 |
| finn | https://pastebin.com/h7fvfHyU | 09:28 |
| finn | any idea why I get this? ModuleNotFoundError: No module named 'buildstream._artifactcache.casserver' | 09:28 |
| finn | yet my python can find it: | 09:29 |
| finn | https://pastebin.com/Sdud0cJw | 09:29 |
| tlater | Whenever I run into these kinds of things clearing out my entire ~/.local/lib/python* seems to help. | 09:31 |
| tlater | I believe you probably have an egg file still lying around | 09:31 |
| tlater | Because you tried to uninstall buildstream by removing executables | 09:31 |
| tlater | This in turn means that parts of the old installation are still around | 09:32 |
| tlater | And probably confuse the hell out of python | 09:32 |
| tlater | (Pip is black voodoo magic) | 09:32 |
| finn | ok, time to fry some eggs | 09:33 |
| * tlater recommends using `pip uninstall` instead of trying to manually uninstall things in the python world. | 09:33 | |
| * tlater recommends that for package managers in general | 09:33 | |
| finn | You mean like this: [finn@aguirre buildstream]$ pip3 uninstall buildstream | 09:34 |
| tlater | Yes | 09:34 |
| tlater | Though when using pip install you shouldn't need to do that, really | 09:35 |
| tlater | Also, on a very related note, why do we always end up frying eggs on this channel? | 09:35 |
| finn | ta tlater, just fried my entire ~/.local and it worked | 09:36 |
| tlater | \o/ | 09:37 |
| tlater | One day I will figure out in detail how people get into this mess and avoid the nuclear option | 09:37 |
| skullman | It's one of the reasons I use a container. | 09:40 |
| finn | It's still misbehaving | 09:48 |
| finn | ModuleNotFoundError: No module named 'google.devtools' | 09:48 |
| finn | yet, my buildgrid can find this just fine | 09:48 |
| * tlater recalls both finn and juergbi packaging a second package with the same name, which potentially screws over simultaneous installations of buildstream and buildgrid | 09:49 | |
| juergbi | with an appropriate __init__.py for the google module, it shouldn't be an issue | 09:50 |
| finn | You mean like this: | 09:51 |
| finn | from pkgutil import extend_path | 09:51 |
| finn | __path__ = extend_path(__path__, __name__) | 09:51 |
| juergbi | it might be missing in some distros | 09:51 |
| juergbi | yes | 09:51 |
| tlater | juergbi: Are you sure the two modules aren't placed in the same directory, weirdly overwriting or somehow being put inside each other? | 09:51 |
| finn | jmac, didn't you solve this by removing both google/ folders from buildgrid and buildstream? | 09:52 |
| finn | then placing a google/ folder into ~/.local/lib/python3.6/site-packages/ | 09:52 |
| jmac | Hacked around rather than solved but yes | 09:52 |
| tlater | Yeeeah, that doesn't sound like a great install story | 09:53 |
| finn | didn't work for me either | 09:53 |
| jmac | I manually combined the contents of all the google/ directories into site-packages/google | 09:53 |
| tlater | Very eww | 09:53 |
| finn | I have no site-packages/google folder strangely | 09:53 |
| finn | I wonder if it's trying to pull in a google lib from elsewhere | 09:54 |
| finn | >>> import google | 09:55 |
| finn | >>> import google.devtools | 09:55 |
| finn | Traceback (most recent call last): | 09:55 |
| finn | File "<stdin>", line 1, in <module> | 09:55 |
| finn | ModuleNotFoundError: No module named 'google.devtools' | 09:55 |
| *** bochecha has joined #buildstream | 09:56 | |
| tlater | finn: That sounds like you have a system-installed google package :) | 09:56 |
| finn | does to me too | 09:57 |
| * tlater thinks the solution is not to have google packages in buildstream/buildgrid | 09:57 | |
| gitlab-br-bot | buildstream: merge request (tiagogomes/issue-289->master: Tiagogomes/issue 289) #521 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/521 | 09:57 |
| finn | someone else will run into this | 09:57 |
| paulsherwood | https://felixge.de/2013/03/11/the-pull-request-hack.html | 09:58 |
| finn | I think it's because of: /lib/python3.6/site-packages/ | 09:58 |
| tlater | Yep, if we want anyone to use this we'll need a better way to package this google stuff | 09:58 |
| paulsherwood | https://twitter.com/havocp/status/988592572258975744 | 09:59 |
| skullman | interesting article paulsherwood | 10:01 |
| * paulsherwood thought so, and posted it here because he's a bit worried that maybe buildstream is not being sufficiently open for contributors | 10:05 | |
| skullman | circumstances differ so different projects have different requirements, It's food for thought but I wouldn't recommend just going ahead and doing it, the commentary rightly points out that you still need *some* barrier to entry, but some are just "it compiles, meets the style, isn't just churn, and doesn't change ABI" | 10:08 |
| paulsherwood | ack | 10:09 |
| jmac | 'meets the style' is often a difficult one for newcomers to figure out | 10:13 |
| paulsherwood | true. and then there are the documentation debates | 10:14 |
| finn | so juergbi, I have the cas server set up, I'm getting bst to send bgd commands. I've set it up so that it will send BuildBox everything but a command to execute. I then pass in the input_root_digest. I then leave it running whilst I open a new terminal | 10:31 |
| finn | The fuse directory I'm supposed to be mounting to is empty | 10:31 |
| juergbi | can you try it using only local CAS, without a CAS server? | 10:32 |
| finn | I've inspected the virtual directory hash that jmac sends me and it indeed contains some kind of directory structure | 10:32 |
| juergbi | to figure out whether it's an issue with fetching the blobs or whether it's a more general issue | 10:32 |
| finn | Would this mean that jmac would need to change his bst client to be sending hashes to some local directory too? | 10:33 |
| jmac | I would just run the worker and buildbox on the artifact server | 10:34 |
| juergbi | or on the machine with the bst client | 10:34 |
| jmac | Yes, that would be even more direct | 10:34 |
| juergbi | you can directly point to ~/.cache/buildstream/artifacts/cas | 10:35 |
| finn | I am running everything on my local machine already | 10:35 |
| finn | Mounted | 10:43 |
| finn | Would you expect to see fuse/bin/sh ? | 10:43 |
| jmac | Yes, if it's the hash you gave me earlier, you should have a bin/sh which is a symlink to busybox. | 10:47 |
| finn | I have a bin/busybox | 10:47 |
| jmac | Are you trying to use dee8bff55be7a56b8f42b2869a0ce47bfa1b65bbd6e00edffc9002f5714e7b3b ? | 10:48 |
| finn | yes | 10:49 |
| jmac | That does contain the symlinks; I can extract them with castool | 10:49 |
| finn | agreed | 10:50 |
| finn | I can see the symlink too your cas-inspector | 10:50 |
| finn | name: "sh" target: "busybox" | 10:51 |
| finn | I can run /bin/busybox through BuildBox | 10:53 |
| finn | And capture the output help page from it | 10:54 |
| finn | Any idea why this wouldn't work? /bin/busybox ls | 10:54 |
| finn | bwrap: execvp /bin/busybox ls: No such file or directory | 10:54 |
| finn | ah got it nvm | 10:55 |
| finn | Interesting how the symlinks aren't working though | 10:56 |
| juergbi | finn: symlink is buildbox code issue, will fix that | 11:05 |
| *** bochecha has quit IRC | 11:05 | |
| *** bochecha has joined #buildstream | 11:06 | |
| finn | ok let me know when that's in and I can continue testing | 11:07 |
| *** bochecha has quit IRC | 11:10 | |
| *** bochecha has joined #buildstream | 11:11 | |
| *** cs_shadow has joined #buildstream | 11:21 | |
| gitlab-br-bot | buildstream: merge request (phil/436-add-ubuntu-install-intructions->master: Phil/436 add ubuntu install intructions) #525 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/525 | 12:58 |
| juergbi | finn: I've pushed symlink support to my wip branch | 13:00 |
| finn | thanks | 13:00 |
| finn | will test rn | 13:00 |
| jmac | I'm going to need to install BuildBox to do more testing | 13:03 |
| jmac | Would you expect it to build correctly on stock Debian 9? | 13:03 |
| finn | New output >> autoreconf: 'configure.ac' or 'configure.in' is required | 13:06 |
| gitlab-br-bot | buildstream: merge request (valentindavid/331_include->master: WIP: Add support for include in project.conf) #471 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/471 | 13:06 |
| finn | also jmac, I added the proper ActionResult to the repo now | 13:06 |
| jmac | Great | 13:07 |
| juergbi | jmac: with grpc 1.12 or 1.13 built from source, yes | 13:08 |
| juergbi | unfortunately, that's only in debian experimental. stable/testing/unstable are many releases behind | 13:09 |
| jmac | Ah. | 13:09 |
| finn | juergbi, https://pastebin.com/HXtEwWgQ | 13:09 |
| juergbi | finn: might this be a working directory issue? | 13:10 |
| finn | oh, append a cd /tmp first or something | 13:11 |
| juergbi | if that's where your configure.ac is | 13:11 |
| juergbi | I'll add a --chdir option to buildbox | 13:11 |
| juergbi | pushed --chdir option | 13:14 |
| finn | thanks, that was quick | 13:14 |
| finn | juergbi, https://pastebin.com/ZmFjeAfD | 13:27 |
| gitlab-br-bot | buildstream: merge request (phil/436-add-ubuntu-install-intructions->master: WIP: Phil/436 add ubuntu install intructions) #525 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/525 | 13:32 |
| *** flatmush has joined #buildstream | 13:47 | |
| gitlab-br-bot | buildstream: merge request (phil/436-add-ubuntu-install-intructions->master: WIP: Phil/436 add ubuntu install intructions) #525 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/525 | 13:49 |
| gitlab-br-bot | buildstream: merge request (phil/436-add-ubuntu-install-intructions->master: Phil/436 add ubuntu install intructions) #525 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/525 | 13:50 |
| flatmush | I need to build a minimal Linux distribution using BuildStream, likely just Linux+BusyBox type thing, can anyone point me in the right direction? | 13:52 |
| coldtom | flatmush: there's a link to an example of pretty much Linux+BusyBox in the documentation, here's the repo: https://gitlab.com/BuildStream/buildstream-examples/tree/master/build-x86image | 13:56 |
| flatmush | cool, thanks | 13:57 |
| finn | juergbi, is this bug likely in BuildBox or something we're not setting up correctly? | 14:03 |
| *** bethw has quit IRC | 14:07 | |
| jmac | juergbi: I'm now building in a chroot with grpc 1.13 and I'm getting this: https://pastebin.com/RnvwFVsH - looks like google/protobuf should exist but doesn't. Any ideas? | 14:07 |
| finn | jmac, that's unusual | 14:08 |
| finn | After installing grpc 1.13, that lib exists on my machine... | 14:09 |
| finn | if it's just the proto it wants, could you copy it over for now? | 14:09 |
| jmac | grpc now no longer compiles (it did before) | 14:13 |
| finn | if you make a pastebin of your setup, I can take a look | 14:14 |
| jmac | Not sure what you mean by 'my setup' | 14:20 |
| jmac | But if it's looking for google/protobuf/descriptor.proto, that file does not exist in any protobuf library I have | 14:22 |
| finn | I just meant the steps you were taking so I could repeat | 14:23 |
| jmac | It's quite a long set of steps from the creation of a chroot | 14:24 |
| finn | https://github.com/google/protobuf/blob/master/src/google/protobuf/descriptor.proto | 14:24 |
| finn | can't you just copy that over? | 14:24 |
| jmac | I could try | 14:24 |
| finn | I think originally I installed grpc from source using this repo on my last machine | 14:25 |
| gitlab-br-bot | buildstream: merge request (valentindavid/331_include->master: WIP: Add support for include in project.conf) #471 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/471 | 14:26 |
| jmac | You installed grpc from https://github.com/google/protobuf? | 14:27 |
| finn | Sorry, I meant I installed the protobuf library using this | 14:28 |
| gitlab-br-bot | buildstream: issue #452 ("Mistyped urls can cause a stack trace") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/452 | 14:28 |
| jmac | Complains about having been built with a different version of protoc | 14:29 |
| finn | I installed grpc from source too: https://github.com/grpc/grpc | 14:29 |
| jmac | Yes, I've got that | 14:29 |
| finn | and you tried this? https://github.com/google/protobuf/blob/master/src/README.md | 14:30 |
| *** tristan has joined #buildstream | 14:32 | |
| jmac | I'm trying that now. Previously I was using the debian-supplied protocol buffer libraries, I think | 14:32 |
| jmac | That builds and installs fine but meson/ninja is still looking in the local directory for google/protobuf/descriptor.proto | 14:38 |
| gitlab-br-bot | buildstream: merge request (phil/436-add-ubuntu-install-intructions->master: Phil/436 add ubuntu install intructions) #525 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/525 | 14:43 |
| gitlab-br-bot | buildstream: merge request (tiagogomes/issue-289->master: Tiagogomes/issue 289) #521 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/521 | 14:45 |
| *** bethw has joined #buildstream | 14:49 | |
| gitlab-br-bot | buildstream: issue #453 ("Testing locally fails due to wrong import order") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/453 | 14:53 |
| *** flatmush has quit IRC | 15:03 | |
| gitlab-br-bot | buildstream: merge request (dp0/453/import-order->master: Reorder app.py imports) #526 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/526 | 15:03 |
| gitlab-br-bot | buildstream: issue #289 ("Need better errors when running something on nothing") changed state ("closed") https://gitlab.com/BuildStream/buildstream/issues/289 | 15:15 |
| gitlab-br-bot | buildstream: merge request (tiagogomes/issue-289->master: Tiagogomes/issue 289) #521 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/521 | 15:15 |
| laurence | jmac, when you have 2 mins, please could you update this ticket re protobuf https://gitlab.com/BuildStream/buildstream/issues/431 | 15:18 |
| laurence | wondering if this still impacts you atm | 15:18 |
| jmac | Yes it does and I don't have anything to update it with | 15:18 |
| finn | I can get a bit further with BuildBox now, though it seems to just hang: | 15:34 |
| finn | https://pastebin.com/PMSkCbrg | 15:34 |
| finn | It gets to that last part and just stops, not returning me a digest | 15:34 |
| jmac | and previously, when you were just running 'ls' or something would you get a digest from that? | 15:37 |
| finn | exactly | 15:37 |
| jmac | Hmm, odd | 15:37 |
| jmac | In unrelated good news I have a buildbox binary | 15:37 |
| finn | ha, now you get to test the chain | 15:38 |
| finn | wondering actually if buildbox has gotten into a weird state, I'm getting this now: buildbox: ../buildbox.cc:989: void flush_to_cas(inode*): Assertion `fd >= 0' failed. | 15:39 |
| finn | buildbox: ../buildbox.cc:989: void flush_to_cas(inode*): Assertion `fd >= 0' failed. | 15:39 |
| finn | buildbox: ../buildbox.cc:989: void flush_to_cas(inode*): Assertion `fd >= 0' failed. | 15:39 |
| finn | buildbox: ../buildbox.cc:989: void flush_to_cas(inode*): Assertion `fd >= 0' failed. | 15:39 |
| finn | sorry, that meant to be a pastebin | 15:39 |
| *** finn has quit IRC | 15:40 | |
| * tlater kindly points finn to https://hastebin.com/, as pastebin's page layout is awful. | 15:40 | |
| *** finn has joined #buildstream | 15:41 | |
| *** Phil has quit IRC | 16:51 | |
| *** Phil has joined #buildstream | 16:51 | |
| *** toscalix has quit IRC | 16:51 | |
| coldtom | can someone review this MR please? https://gitlab.com/BuildStream/buildstream/merge_requests/518 | 16:53 |
| *** phildawson has joined #buildstream | 16:54 | |
| *** Phil has quit IRC | 16:55 | |
| *** phildawson has quit IRC | 16:57 | |
| *** bethw has quit IRC | 17:00 | |
| *** coldtom has quit IRC | 17:02 | |
| jmac | Oh, he's gone | 17:03 |
| jmac | I left a comment anyway | 17:03 |
| *** tristan has quit IRC | 17:07 | |
| *** jonathanmaw has quit IRC | 17:15 | |
| *** Prince781 has quit IRC | 17:18 | |
| *** Prince781 has joined #buildstream | 17:20 | |
| *** jennis has quit IRC | 17:33 | |
| *** jsgrant has quit IRC | 17:59 | |
| *** jennis has joined #buildstream | 18:32 | |
| *** Prince781 has quit IRC | 18:37 | |
| *** Prince781_ has joined #buildstream | 18:38 | |
| *** Prince781_ has quit IRC | 19:00 | |
| *** Prince781 has joined #buildstream | 19:01 | |
| *** Prince781 has quit IRC | 19:37 | |
| *** Prince781 has joined #buildstream | 19:38 | |
| *** Prince781 has quit IRC | 21:59 | |
| *** Prince781 has joined #buildstream | 21:59 | |
| *** bochecha has quit IRC | 22:36 | |
| *** bochecha has joined #buildstream | 22:36 | |
| *** bochecha has quit IRC | 22:40 | |
| *** bochecha has joined #buildstream | 22:41 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!