| *** paulw has joined #baserock | 06:19 | |
| richard_maw | tiagogomes_: it would definitely make more sense than in the system, they are defined in the systems for legacy reasons | 07:38 |
|---|---|---|
| richard_maw | tiagogomes_: though I'd prefer if they were defined in the chunks and aggregated as well, since if you include a component, you probably want to configure it | 07:39 |
| tiagogomes_ | on the chunks, maybe it would make more sense to inline them in a `deployment-commands` section, instead of referring to a list of configuration extensions to run | 07:56 |
| *** bashrc_ has joined #baserock | 08:01 | |
| *** jonathanmaw has joined #baserock | 08:05 | |
| richard_maw | well, yes and no | 08:08 |
| richard_maw | for simple stuff deployment commands would be the right level | 08:08 |
| richard_maw | but if it gets interesting, you want a way to share the code | 08:08 |
| richard_maw | and it would also imply the need for them all to be shell scripts | 08:09 |
| richard_maw | unless we decide to write them out to a file and exec them | 08:09 |
| tiagogomes_ | if they are really chunk-only deployment commands, no sharing will be needed | 08:16 |
| richard_maw | ah I misunderstood, I thought you wanted `deployment-commands` in the clusters | 08:19 |
| tiagogomes_ | I am talking about the case where some chunk needs to have information from the cluster env variables to be able to configure itself. For example, a web service that needs to know which port to listen to | 08:19 |
| richard_maw | tiagogomes_: do you anticipate it needing these values at build time, or just deployment time? | 08:19 |
| tiagogomes_ | at deploy time, because the user may want to tweak them | 08:20 |
| richard_maw | good, I'd be nervy of the desire to specify build-time variables | 08:20 |
| *** mariaderidder has joined #baserock | 08:48 | |
| paulsherwood | are there any gerrit admin wizards here? | 08:53 |
| paulsherwood | i'm wondering if the baserock contrib documentation is generally applicable | 08:54 |
| paulsherwood | in particular the commit-msg hook thing - is that always required with gerrit etc? | 08:55 |
| *** edcragg has joined #baserock | 08:55 | |
| paulsherwood | (i appreciate this is slightly off-topic, i'm trying to help AGL) | 08:55 |
| paulsherwood | pedroalvarez: ^^ ? | 08:55 |
| SotK | paulsherwood: it is always required I believe | 08:56 |
| paulsherwood | ok thanks | 08:57 |
| richard_maw | it's not strictly necessary, as there is a way to push without it… but you don't want to do that, as when you need to send a follow-up patch, you need to specify all the replacement commits manually | 08:58 |
| edcragg | i thought the one used in baserock was a fairly standard hook | 09:09 |
| edcragg | you can add commit IDs manually, surely | 09:09 |
| edcragg | s/commit/change/ | 09:09 |
| SotK | I'd guess so | 09:09 |
| richard_maw | that is also an option | 09:09 |
| edcragg | when i first looked at the contributing page i was confused by the section on submitting to definitions (system branches etc), but it seems that this isn't really necessary | 09:19 |
| edcragg | unless i'm missing something | 09:21 |
| *** locallycompact has quit IRC | 10:38 | |
| *** Jude has joined #baserock | 12:24 | |
| Jude | Hi all, I have a question that i need help with. " Is there a specific baserock definition for building i.MX6 boards' kernel files into SDKs or can one use the same definitions used to build on other boards like Rcar H2 ? " | 12:28 |
| *** zoli__ has joined #baserock | 12:39 | |
| richard_maw | I guess that question's for me, since I wrote the SDK stuff | 12:40 |
| richard_maw | let me reload some context | 12:41 |
| Jude | okay, thanks richard_maw | 12:42 |
| *** zoli__ has quit IRC | 12:43 | |
| richard_maw | ok, unless you're messing with the userland significantly (which isn't usually kernel work), you can re-use the existing sdk stuff by replacing the "systems/devel-system-armv7lhf-highbank.morph" in clusters/sdk-example-cluster.morph with the system definition for the board you want | 12:44 |
| richard_maw | Jude: however, our SDK work was more focused on handling userland applications development, so I'm not sure whether ""SDK"" as defined by your question asker matches up with our definition | 12:45 |
| richard_maw | there's a *lot* of assumed definition of SDK in that question, which doesn't necessarily hold for other contexts | 12:46 |
| Jude | thanks for your answer richard_maw. I will get more info from the asker. | 13:01 |
| *** edcragg has quit IRC | 13:15 | |
| *** edcragg has joined #baserock | 13:16 | |
| *** zoli__ has joined #baserock | 13:27 | |
| *** mariaderidder has quit IRC | 13:36 | |
| *** mariaderidder has joined #baserock | 13:51 | |
| *** inara has quit IRC | 13:52 | |
| *** inara has joined #baserock | 13:53 | |
| paulsherwood | richard_maw: highbank? wouldn't wandboard be better? | 14:13 |
| *** paulw has quit IRC | 14:18 | |
| richard_maw | paulsherwood: probably, or a Jetson one, we've just not had cause to change it since situations changed. | 14:19 |
| richard_maw | s/situations/circumstances/ | 14:20 |
| paulsherwood | wandboard is imx6, is all :) | 14:25 |
| Jude | richard_maw: in that case, does this still stand "replacing the 'systems/devel-system-armv7lhf-highbank.morph" in clusters/sdk-example-cluster.morph' or do i need to change highbank to wandboard? | 14:29 |
| paulsherwood | Jude: if you are actually doing this work, it would perhaps be better to create new definitions (specifically bsp, system and cluster) named to suit the board you are targeting | 14:39 |
| paulsherwood | (and base those on the information from the existing examples as a start point) | 14:39 |
| Jude | paulsherwood: We are not doing this work. The asker wanted me to find out for him, how to go about this. Thanks for your clarification, i will pass this one to him. :) | 14:41 |
| *** petefoth has quit IRC | 14:41 | |
| *** jonathanmaw has quit IRC | 16:01 | |
| *** tiagogomes_ has quit IRC | 16:25 | |
| *** lachlanmackenzie has joined #baserock | 16:38 | |
| *** bashrc_ has quit IRC | 16:52 | |
| *** mariaderidder has quit IRC | 16:55 | |
| *** zoli__ has quit IRC | 17:15 | |
| *** zoli__ has joined #baserock | 17:20 | |
| *** Jude has quit IRC | 17:20 | |
| *** Jude has joined #baserock | 17:51 | |
| *** edcragg has quit IRC | 17:58 | |
| *** lachlanmackenzie has quit IRC | 18:20 | |
| *** zoli__ has quit IRC | 18:59 | |
| *** zoli__ has joined #baserock | 18:59 | |
| *** zoli__ has joined #baserock | 19:00 | |
| *** zoli__ has quit IRC | 19:06 | |
| *** Jude has quit IRC | 21:25 | |
| *** edcragg has joined #baserock | 22:36 | |
| *** edcragg has quit IRC | 23:02 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!