| *** alatiera_ has quit IRC | 00:00 | |
| gitlab-br-bot | sstriker approved MR !946 (jmac/remote_execution_split->master: Split remote execution from artifact cache) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/946 | 00:30 |
|---|---|---|
| *** federico has quit IRC | 00:32 | |
| *** tristan has joined #buildstream | 07:43 | |
| *** mohan43u has quit IRC | 08:01 | |
| *** mohan43u has joined #buildstream | 08:01 | |
| *** tristan has quit IRC | 08:05 | |
| *** phildawson has quit IRC | 08:09 | |
| *** toscalix has joined #buildstream | 09:00 | |
| *** phildawson has joined #buildstream | 09:05 | |
| *** bochecha has joined #buildstream | 09:17 | |
| *** benbrown has quit IRC | 09:21 | |
| *** benbrown has joined #buildstream | 09:21 | |
| *** benbrown has quit IRC | 09:38 | |
| *** benbrown has joined #buildstream | 09:38 | |
| gitlab-br-bot | danielsilverstone-ct closed issue #760 (Plugin.__del__ writes DEBUG messages to stdout) on buildstream https://gitlab.com/BuildStream/buildstream/issues/760 | 09:48 |
| *** raoul has joined #buildstream | 09:49 | |
| jmac | That test_pull_tree issue I mentioned yesterday is happening on CI as well - https://gitlab.com/BuildStream/buildstream/-/jobs/123687142 has been stuck on it for 16 hours | 10:12 |
| tpollard | ouch | 10:14 |
| jmac | I'll investigate it | 10:14 |
| Kinnison | Does anyone know of an issue related to batching requests to the referencestorage service? | 10:14 |
| * Kinnison has heard talk about such a change to the service and wondered | 10:15 | |
| *** jonathanmaw has joined #buildstream | 10:17 | |
| *** tristan has joined #buildstream | 10:23 | |
| *** alatiera_ has joined #buildstream | 10:25 | |
| *** lachlan has joined #buildstream | 10:28 | |
| *** bochecha_ has joined #buildstream | 10:42 | |
| *** bochecha has quit IRC | 10:44 | |
| *** bochecha_ is now known as bochecha | 10:44 | |
| juergbi | raoul: fyi, I've commented on #775 execution environments | 10:55 |
| gitlab-br-bot | Issue #775: Execution Environment Requirements https://gitlab.com/BuildStream/buildstream/issues/775 | 10:55 |
| raoul | ah right, I was just reading up on uname -m and had noticed it does incorporate endianness. Is there reason you've gone for slightly different names compared to uname results for architecture? (x86_64 vs x86-64 for example) | 10:57 |
| *** palimm has joined #buildstream | 10:57 | |
| juergbi | jmac: are you planning to base one of !946 or !952 on top of the other already in the MR, or which one do you prefer to go in first, or are there no merge conflicts? | 10:58 |
| gitlab-br-bot | MR !952: Remote-execution instance configuration support https://gitlab.com/BuildStream/buildstream/merge_requests/952 | 10:58 |
| gitlab-br-bot | valentindavid opened MR !962 (valentindavid/overnight-tests-bump-fdo-sdk-version->master: Bump version of fd.o SDK and bst_external in overnight tests) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/962 | 10:58 |
| jennis | We have an option to checkout a tarball to stdout with `bst checkout --tar foo.bst -`, I can't think why we're supporting this or a use-case for this? Could anyone perhaps enlighten me? | 10:59 |
| valentind | jjardon, I am going to bump the version of fd.o SDK for the overnight test. I think there should be no problem. Here is the MR: https://gitlab.com/BuildStream/buildstream/merge_requests/962 | 10:59 |
| valentind | jjardon, is that OK? | 10:59 |
| jjardon | valentind: sure | 11:01 |
| jjardon | jonathanmaw: do you think it would be possible to have a new release of bst-external? | 11:02 |
| jonathanmaw | jjardon: can do | 11:03 |
| jjardon | jonathanmaw: thanks a lot. valentind maybe better to use the tag when jonathanmaw releases the new one | 11:03 |
| juergbi | raoul: the main reason we can't just use `uname -m` is that it's OS-specific and we want something OS-independent | 11:04 |
| valentind | jjardon, OK | 11:05 |
| juergbi | raoul: another reason is that Linux has multiple values for 32-bit x86: i386 - i686 | 11:05 |
| juergbi | so as we anyway need to define new values, I aimed towards the closest to official naming | 11:06 |
| juergbi | (and vendor-neutral) | 11:06 |
| raoul | cool, cheers for clearing that up | 11:08 |
| valentind | If it is for requirements, note that x86_64 can always build i386. But aarch64 cannot always build armhf. Sometimes it can, sometimes it cannot. uname -m does not give information on that. | 11:08 |
| juergbi | valentind: strictly speaking I don't think this is correct. x86-32 compat support is optional in the kernel | 11:09 |
| juergbi | in practice it's unlikely to be disabled on a build server | 11:09 |
| juergbi | as all x86-64 hardware supports x86-32, afaik | 11:09 |
| valentind | right. | 11:09 |
| juergbi | although there are environments that disable x86-32 for certain processes using seccomp to reduce attack surface | 11:10 |
| jonathanmaw | jjardon, valentind: tags pushed for 0.7.0 https://gitlab.com/BuildStream/bst-external/tree/0.7.0 | 11:10 |
| jjardon | jonathanmaw: \o/ thanks | 11:10 |
| jmac | juergbi: I was planning to merge both into one MR after !946 had passed review | 11:11 |
| gitlab-br-bot | MR !946: Split remote execution from artifact cache https://gitlab.com/BuildStream/buildstream/merge_requests/946 | 11:11 |
| jmac | But I can do them separately, if that makes more sense | 11:11 |
| juergbi | doesn't matter to me. i'll just make sure to let you do the actual merging | 11:12 |
| juergbi | !946 looks good to me overall, just had the two comments | 11:12 |
| jmac | Cool, thanks for the review | 11:12 |
| jmac | I want to figure out why test_pull_tree is hanging first | 11:12 |
| juergbi | sure | 11:13 |
| juergbi | we're still missing user configuration for remote-execution. should cover this soon as well | 11:13 |
| valentind | Talking about "uname -m". When is it going to be removed from the cache key? | 11:14 |
| juergbi | raoul will work on this as part of resolving #775 | 11:14 |
| gitlab-br-bot | Issue #775: Execution Environment Requirements https://gitlab.com/BuildStream/buildstream/issues/775 | 11:14 |
| juergbi | is a blocker for 1.4 | 11:14 |
| raoul | it's why I've been talking about it valentind :P | 11:15 |
| valentind | raoul, ah great. | 11:17 |
| *** ctolentino has joined #buildstream | 11:17 | |
| benschubert | When running Buildstream, I get errors like: "Exception ignored when trying to write to the signal wakeup df: BlockingIOError: [Errno 11] Resource temporarily unavailable. | 11:18 |
| benschubert | Is that expected? Seeing it printed and burried in the output is scary... | 11:18 |
| juergbi | certainly not something that you should get, no | 11:19 |
| valentind | raoul, so you are fixing https://gitlab.com/BuildStream/buildstream/issues/523 ? | 11:19 |
| benschubert | juergbi: it happens all around with a project or around 3000 elements | 11:20 |
| benschubert | Seems like all my fetch jobs are stalled now | 11:21 |
| benschubert | I only have build jobs after that (Fetched: 204 3120 0) | 11:21 |
| phildawson | <jennis> We have an option to checkout a tarball to stdout with `bst checkout --tar foo.bst -`, I can't think why we're supporting this or a use-case for this? Could anyone perhaps enlighten me? | 11:23 |
| benschubert | I'm on the latest master | 11:23 |
| phildawson | perhaps tiagogomes_ can answer this? | 11:23 |
| phildawson | jennis' question that is | 11:23 |
| tiagogomes_ | bst checkout --tar foo.bst - | docker import | 11:24 |
| gitlab-br-bot | jjardon approved MR !962 (valentindavid/overnight-tests-bump-fdo-sdk-version->master: Bump version of fd.o SDK and bst_external in overnight tests) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/962 | 11:25 |
| phildawson | thanks tiagogomes_ | 11:25 |
| tiagogomes_ | bst checkout --tar foo.bst - | gzip | 11:25 |
| tiagogomes_ | bst checkout --tar foo.bst - | dd of=/dev/sda | 11:27 |
| raoul | valentind, yeah hopefully fixing #775 should fix #523 too. | 11:28 |
| gitlab-br-bot | Issue #775: Execution Environment Requirements https://gitlab.com/BuildStream/buildstream/issues/775 | 11:28 |
| gitlab-br-bot | Issue #523: BuildStream doesn't hit the cache when building foreign arches https://gitlab.com/BuildStream/buildstream/issues/523 | 11:28 |
| valentind | Awesome. | 11:29 |
| gitlab-br-bot | jjardon merged MR !962 (valentindavid/overnight-tests-bump-fdo-sdk-version->master: Bump version of fd.o SDK and bst_external in overnight tests) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/962 | 11:41 |
| jennis | tiagogomes_, ahh, thanks | 11:46 |
| jjardon | valentind: I'm hitting https://gitlab.com/BuildStream/buildstream/issues/645 ; is the patch pushed there in the bst-1.2 branch as well? | 11:48 |
| valentind | jjardon, on Freedesktop SDK or another project? | 11:48 |
| valentind | Because it should not happen to freedesktop sdk. | 11:49 |
| jjardon | valentind: other project | 11:49 |
| valentind | jjardon, So if we put it in 1.2 we change cache keys. | 11:49 |
| valentind | jjardon, what I recommend is you take the strip-command from freedesktop sdk. | 11:49 |
| valentind | The strip-commands from Buildstream are not that good. But we found there is no way to make something good and portable. | 11:50 |
| valentind | So in general, just always redefine them. | 11:50 |
| jjardon | valentind: yeah, but we need a good solution for bst-1.2 users; is changing the cache keys not allow during 1.2.x ? | 11:51 |
| valentind | jjardon, we discussed to replace "strip-commands" to be just "true". That is it does not do anything. And then users can define it. | 11:52 |
| jjardon | valentind: sure, but I guess that is not a solution for bst-1.2 users? | 11:52 |
| juergbi | I don't think we should change this for 1.2. but we should really fix this for 1.4 | 11:53 |
| jjardon | juergbi: can we apply the master fix in 1.2, even if we change the cache key? | 11:54 |
| juergbi | not sure. we don't change the cache key algorithm, so it wouldn't affect projects that already override strip-commands, right? | 11:54 |
| valentind | juergbi, for reference the changes are https://gitlab.com/BuildStream/buildstream/merge_requests/918 | 11:54 |
| juergbi | ta | 11:54 |
| valentind | juergbi, exactly. It would only affect those who have not defined strip-commands. | 11:55 |
| juergbi | so it could make sense to apply this fix to 1.2 (but of course not the complete move away from default strip commands) | 11:55 |
| valentind | OK then. | 11:55 |
| juergbi | I'd like to get input from tristan on this, though | 11:56 |
| valentind | I will make the MR on 1.2 | 11:56 |
| juergbi | it's a corner case whether it's acceptable on 1.2 | 11:56 |
| jjardon | thanks guys | 11:57 |
| gitlab-br-bot | valentindavid opened MR !963 (valentindavid/fix-recursive-strip-debug-1.2->bst-1.2: [backport 1.2] Fix infinite recursion in default strip debug command) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/963 | 12:03 |
| valentind | jjardon, ^ | 12:03 |
| gitlab-br-bot | lchlan closed MR !917 (lachlanmackenzie/trigger-benchmark-tests->master: Add trigger for benchmarking) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/917 | 12:09 |
| *** tristan has quit IRC | 12:17 | |
| *** alatiera_ has quit IRC | 12:35 | |
| *** alatiera_ has joined #buildstream | 12:37 | |
| benschubert | Starting from an empty cache, could someone help me understand this output: "(Fetched:204 3120 0)_ (Built:217 0 3)" ...? | 12:47 |
| benschubert | Oh, elements with no sources, that might be it | 12:49 |
| benschubert | Just don't understand why I haven't fetched more than 204 elements | 12:49 |
| *** tristan has joined #buildstream | 12:53 | |
| juergbi | benschubert: maybe they were already fetched in a previous session? (cached in ~/.cache/buildstream/sources) | 13:27 |
| *** alatiera_ has quit IRC | 13:27 | |
| *** alatiera_ has joined #buildstream | 13:30 | |
| *** alatiera_ has quit IRC | 13:33 | |
| *** cheapie has joined #buildstream | 13:36 | |
| *** alatiera_ has joined #buildstream | 13:36 | |
| gitlab-br-bot | jmacarthur opened MR !964 (jmac/fix-test-hangs-2->master: Avoid hanging artifact cache tests) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/964 | 13:52 |
| WSalmon | is tristan about? | 13:54 |
| Kinnison | He mentioned elsewhere earlier that he is a bit feverish, so perhaps not | 13:55 |
| WSalmon | indeed, i thought i would ping to see, but it can wait. glad to hear he is making progress, i hate being ill, get well soon :) | 13:57 |
| benschubert | juergbi: I had cleaned up everything just before | 14:02 |
| benschubert | And they weren't skipped as would be the case right? | 14:02 |
| juergbi | benschubert: second number is skipped | 14:04 |
| benschubert | oups just realized that yup | 14:04 |
| benschubert | seems weird what just happened... oh well, I'll see if/when it finishes what's the state of it | 14:04 |
| benschubert | thanks! | 14:04 |
| *** toscalix has quit IRC | 14:23 | |
| *** toscalix has joined #buildstream | 14:24 | |
| gitlab-br-bot | chiaratolentino opened issue #776 (Add support for running tests in BuildStream) on buildstream https://gitlab.com/BuildStream/buildstream/issues/776 | 14:26 |
| gitlab-br-bot | chiaratolentino opened issue #777 (Add abstract dependencies) on buildstream https://gitlab.com/BuildStream/buildstream/issues/777 | 14:27 |
| benschubert | Again another question: Is it possible to set the message-format from the command line? I'd like to avoid having to setup a project.conf :D | 14:29 |
| benschubert | not a project, a user config | 14:30 |
| *** raoul_ has joined #buildstream | 14:31 | |
| tpollard | benschubert: anything helpful in the bst main options? bst --help | 14:31 |
| *** raoul has quit IRC | 14:32 | |
| benschubert | tpollard: already looked, nothing there for this, almost all other keys for the logging section are overridable but not this one | 14:32 |
| juergbi | I think that requires buildstream.conf | 14:32 |
| benschubert | ok! thanks | 14:33 |
| *** alatiera_ has quit IRC | 14:38 | |
| gitlab-br-bot | willsalmon opened issue #778 (bst example html in the doc's is out of date and is hard to maintain) on buildstream https://gitlab.com/BuildStream/buildstream/issues/778 | 14:39 |
| *** federico has joined #buildstream | 14:40 | |
| *** alatiera_ has joined #buildstream | 14:41 | |
| gitlab-br-bot | BenjaminSchubert opened issue #779 (It should be possible to disable error and message lines completely) on buildstream https://gitlab.com/BuildStream/buildstream/issues/779 | 14:46 |
| gitlab-br-bot | willsalmon opened MR !966 (willsalmon/sessionsBug->master: Keep the Bst2html output up to date.) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/966 | 14:48 |
| *** raoul_ has quit IRC | 14:51 | |
| skullman | hmm, had a non-deterministic test failure on tests/artifactcache/expiry.py::test_never_delete_required | 15:04 |
| * skullman reruns tests from the start to see if it happens again | 15:04 | |
| skullman | weird, the only way I can see that happening is if the artifacts generated from the contents of /dev/urandom were deduplicated by CAS | 15:07 |
| *** raoul has joined #buildstream | 15:09 | |
| *** tiagogomes_ has quit IRC | 16:15 | |
| *** tiagogomes has joined #buildstream | 16:16 | |
| *** abderrahim has joined #buildstream | 16:20 | |
| *** bochecha_ has joined #buildstream | 16:29 | |
| *** bochecha has quit IRC | 16:31 | |
| *** bochecha_ is now known as bochecha | 16:31 | |
| *** alatiera_ has quit IRC | 16:53 | |
| *** alatiera_ has joined #buildstream | 16:53 | |
| *** lachlan has quit IRC | 16:55 | |
| *** lachlan has joined #buildstream | 16:58 | |
| *** alatiera_ has quit IRC | 17:19 | |
| *** alatiera_ has joined #buildstream | 17:19 | |
| *** alatiera_ has quit IRC | 17:20 | |
| *** alatiera_ has joined #buildstream | 17:20 | |
| *** alatiera_ has quit IRC | 17:22 | |
| *** alatiera_ has joined #buildstream | 17:23 | |
| *** bochecha has quit IRC | 17:35 | |
| *** federico has quit IRC | 17:45 | |
| gitlab-br-bot | juergbi opened issue #780 (Remote execution: TLS support for execution service) on buildstream https://gitlab.com/BuildStream/buildstream/issues/780 | 17:57 |
| *** raoul has quit IRC | 18:12 | |
| gitlab-br-bot | jmacarthur opened issue #781 (Remove unused functions from ArtifactCache) on buildstream https://gitlab.com/BuildStream/buildstream/issues/781 | 18:19 |
| *** toscalix has quit IRC | 18:19 | |
| *** lachlan has quit IRC | 18:36 | |
| *** jonathanmaw has quit IRC | 18:47 | |
| *** tristan has quit IRC | 18:50 | |
| *** bochecha has joined #buildstream | 19:41 | |
| gitlab-br-bot | cs-shadow closed issue #362 (Allow workspace commands to work on multiple elements) on buildstream https://gitlab.com/BuildStream/buildstream/issues/362 | 19:41 |
| *** bochecha has quit IRC | 19:47 | |
| *** ctolentino has quit IRC | 20:13 | |
| *** benschubert has quit IRC | 20:52 | |
| *** toscalix has joined #buildstream | 20:57 | |
| *** toscalix has quit IRC | 21:01 | |
| *** tristan has joined #buildstream | 21:15 | |
| *** Dr-G has joined #buildstream | 21:53 | |
| *** Guest99472 has joined #buildstream | 22:32 | |
| *** tristan has quit IRC | 22:37 | |
| *** alatiera_ has quit IRC | 22:41 | |
| *** alatiera_ has joined #buildstream | 23:23 | |
| *** alatiera_ has quit IRC | 23:28 | |
| *** mohan43u has quit IRC | 23:45 | |
| *** mohan43u has joined #buildstream | 23:49 | |
| *** mohan43u has quit IRC | 23:52 | |
| *** mohan43u has joined #buildstream | 23:55 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!