| *** gtristan has quit IRC | 04:30 | |
| *** gtristan has joined #baserock | 04:59 | |
| paulsherwood | \/win 48 | 08:16 |
|---|---|---|
| paulsherwood | bah | 08:16 |
| *** ctbruce has joined #baserock | 08:24 | |
| paulsherwood | http://paste.baserock.org/ovepijitow | 08:33 |
| paulsherwood | jjardon, pedroalvarez ^^ ? | 08:33 |
| *** toscalix has joined #baserock | 08:39 | |
| jjardon | paulsherwood: https://gerrit.baserock.org/#/c/1913/ | 08:48 |
| jjardon | paulsherwood: https://gerrit.baserock.org/#/c/1913/ | 08:52 |
| jjardon | btw, ybd doesnt rebuild isodate when I change the build-system | 08:52 |
| *** jjardon_ has joined #baserock | 08:54 | |
| jjardon_ | paulsherwood: https://gerrit.baserock.org/#/c/1913/2 | 09:00 |
| jjardon_ | (also, seems ybd doesnt rebuild isodate after I change the build-system parameter | 09:01 |
| jjardon | Sorry, im havin problem with the internet connection | 09:06 |
| pedroalvarez | paulsherwood: yes, I was planning to do that today | 09:07 |
| pedroalvarez | (to fix that) | 09:07 |
| pedroalvarez | radiofree, jjardon: btw, before I forget, wayland-ivi-extension fails to build with weston 1.10 http://paste.baserock.org/wavetazoce | 09:07 |
| pedroalvarez | so, let's keep weston-genivi for now | 09:08 |
| jjardon_ | pedroalvarez: is upstream aware of that? | 09:08 |
| pedroalvarez | I didn't have time to check | 09:09 |
| jjardon_ | paulsherwood: about GDP: https://gerrit.baserock.org/#/c/1356/ | 09:10 |
| pedroalvarez | but they are versioning w-i-extension with the version number of weston they are compatible with | 09:10 |
| pedroalvarez | so I assume they are aware | 09:10 |
| jjardon_ | rigth, I guess in the meantime we will have to keep the weston-genivi stratum then | 09:11 |
| pedroalvarez | yup | 09:12 |
| *** rdale has joined #baserock | 09:12 | |
| pedroalvarez | jjardon_: could you answer my comment here please? https://gerrit.baserock.org/#/c/1905/ | 09:13 |
| jjardon_ | pedroalvarez: I was doing exactly that :) | 09:13 |
| pedroalvarez | thanks! | 09:13 |
| pedroalvarez | I'm going to give it a go, at the same time I fix gdp-hmi build | 09:16 |
| *** rdale has quit IRC | 09:19 | |
| *** rdale has joined #baserock | 09:21 | |
| *** gtristan has quit IRC | 09:24 | |
| *** rdale has quit IRC | 09:29 | |
| paulsherwood | pedroalvarez: what about settling on (say) ci-minimal and ci-more | 09:41 |
| paulsherwood | i'd be happy for ybd to do both, but this would cut the loop time for mason | 09:41 |
| paulsherwood | (if it just did ci-minimal) | 09:41 |
| jjardon | pedroalvarez: patch updated, thanks for the review! | 09:42 |
| pedroalvarez | that's not a bad idea | 09:43 |
| pedroalvarez | but mason is the only thing people can look at | 09:43 |
| paulsherwood | pedroalvarez: that will change in a small number of weeks at most | 09:44 |
| paulsherwood | c.b.o is a poorly kept secret at best :) | 09:44 |
| pedroalvarez | :) | 09:44 |
| pedroalvarez | looking forward to that | 09:44 |
| * paulsherwood too | 09:45 | |
| paulsherwood | who would be best placed to offer a patch to split out ci into two? i can do it, for example | 09:46 |
| pedroalvarez | i wonder how would we split ti | 09:49 |
| pedroalvarez | s/ti/it/ | 09:49 |
| paulsherwood | i made some comments in https://gerrit.baserock.org/#/c/1356/ | 09:51 |
| *** edcragg has joined #baserock | 09:51 | |
| paulsherwood | how about ci-minimal has build-system + genivi baseline for both arches, maybe minimal | 09:52 |
| paulsherwood | and ci-more has gdp, gnome, openstack etc | 09:52 |
| pedroalvarez | paulsherwood: actually, ci-minimal loooks like release.morph from your example | 09:53 |
| pedroalvarez | maybe we can just use that | 09:53 |
| paulsherwood | oh, fair enough | 09:54 |
| paulsherwood | so strip out the stuff from release.morph in ci.morph? | 09:54 |
| jjardon | talking about release.morph: https://gerrit.baserock.org/#/c/1257/ | 09:54 |
| paulsherwood | and fix mason to use release.morph | 09:54 |
| pedroalvarez | I'm not sure about stripping them from ci.morph | 09:55 |
| paulsherwood | ok | 09:56 |
| paulsherwood | your call | 09:56 |
| pedroalvarez | right, I wouldn't mind we add heavy systems to ci.morph anymore | 09:56 |
| pedroalvarez | I'll fix Mason to build release.morph | 09:56 |
| pedroalvarez | although.. i'd prefer to do this change whenever we have something using ci.morph really | 09:57 |
| paulsherwood | c.b.o uses it. and i use it for testing every release of ybd | 09:58 |
| paulsherwood | on x86 always, and on arm when i remember | 09:59 |
| *** jonathanmaw has joined #baserock | 09:59 | |
| paulsherwood | (and i'm aiming for weekly releases of ybd for the foreseeable future) | 10:00 |
| *** vgrade_w has joined #baserock | 10:02 | |
| *** ssam2 has joined #baserock | 10:04 | |
| *** ChanServ sets mode: +v ssam2 | 10:04 | |
| *** jmacs has quit IRC | 10:08 | |
| *** locallycompact has joined #baserock | 10:13 | |
| locallycompact | how do I add comments to gerrit | 10:16 |
| ssam2 | to lines in a patch? | 10:16 |
| ssam2 | or to the patch itself? | 10:16 |
| locallycompact | general comment | 10:17 |
| *** edcragg has quit IRC | 10:17 | |
| ssam2 | click 'reply' in top bar, or press 'a' on your keyboard | 10:17 |
| ssam2 | you can also press ? for a list of all keyboard shortcuts | 10:17 |
| locallycompact | tyvm! | 10:18 |
| *** franred has joined #baserock | 10:18 | |
| locallycompact | I have an odd problem, I can remote update definitions in gerrit just fine, but when I ssh -A into a baserock vm and remote update definitions from gerrit I get a Permission denied (publickey). | 10:37 |
| *** jjardon_ has quit IRC | 10:37 | |
| richard_maw | locallycompact: is your ssh-agent running on your laptop? | 10:38 |
| locallycompact | by running ssh-agent? | 10:39 |
| locallycompact | seems to be | 10:39 |
| locallycompact | running that and doing over doesn't seem to be making a difference | 10:40 |
| richard_maw | does your ssh-agent have your ssh keys in it? I think `ssh-add -L` should show you | 10:41 |
| locallycompact | ah, it does not | 10:42 |
| locallycompact | thanks | 10:42 |
| locallycompact | and what do the branch names correspond to on gerrit? I see for example jjardon's topic wayland_1_10 but no branch called that | 10:46 |
| ssam2 | if you use `git-review` to submit a patch, then the topic name matches whatever branch you had checked out, when you ran git-review | 10:49 |
| ssam2 | unless you pass --topic | 10:49 |
| ssam2 | if you push directly, you set the topic by adding it to the ref you push to | 10:50 |
| ssam2 | the point of it is to help with pretending Gerrit supports patch series | 10:50 |
| *** rdale has joined #baserock | 10:50 | |
| ssam2 | you can submit 10 patches with the same topic and they get grouped together | 10:50 |
| ssam2 | actual branches in definitions.git correspond to actual branches in Gerrit | 10:51 |
| ssam2 | branches in git://git.baserock.org/baserock/baserock/definitions.git, that is | 10:51 |
| locallycompact | I'm still confused. If I look at this page. what on this page tells me what I should check out to review it? https://gerrit.baserock.org/#/c/1906/ | 10:52 |
| locallycompact | or just use the sha? | 10:53 |
| SotK | The download button in the top right gives a bunch of commands that will do it | 10:54 |
| ssam2 | i'd just do `git-review --download 1906` | 10:54 |
| ssam2 | which wraps the necessary git commands needed to download and checkout | 10:54 |
| locallycompact | ok, I see | 10:55 |
| pedroalvarez | anybody knows where morph.log goes if nothing set in morph.conf? | 11:18 |
| *** rdale has quit IRC | 11:19 | |
| *** rdale has joined #baserock | 11:19 | |
| richard_maw | it doesn't | 11:19 |
| pedroalvarez | hah np | 11:20 |
| pedroalvarez | I was trying to figureout how long takes qtwebkit to build | 11:20 |
| pedroalvarez | but it just finished | 11:20 |
| paulsherwood | pedroalvarez: git clone https://github.com/devcurmudgeon/build-logs ; cd build-logs ; git grep qtwebkit | grep Elapsed | 11:23 |
| paulsherwood | ybd-build-to-qt.1x60.log:15-07-18 20:57:30 [135/151/282] [qtwebkit] Elapsed time for build 01:28:05 | 11:25 |
| paulsherwood | that was with max-jobs set to 60 on a 40 core machine | 11:26 |
| paulsherwood | also http://wiki.baserock.org/build-times/ | 11:27 |
| pedroalvarez | [qtwebkit] Elapsed time 00:41:13 on moonshot | 11:29 |
| paulsherwood | really? | 11:30 |
| pedroalvarez | for real :P | 11:30 |
| paulsherwood | that's pretty *good* actually :) | 11:30 |
| * paulsherwood has lost his moonshot ip | 11:31 | |
| paulsherwood | s/lost/forgotten/ | 11:31 |
| pedroalvarez | paulsherwood: I can remind it to you | 11:31 |
| paulsherwood | yes please :) | 11:31 |
| pedroalvarez | but i think they are being powered of today | 11:31 |
| pedroalvarez | off* | 11:31 |
| paulsherwood | ack | 11:33 |
| paulsherwood | pedroalvarez: is that ARMv7 build, or aarch64 btw? | 11:34 |
| pedroalvarez | armv7 | 11:35 |
| pedroalvarez | I've never done anything with aarch64 | 11:35 |
| paulsherwood | ack | 11:38 |
| *** jjardon_ has joined #baserock | 11:43 | |
| *** ctbruce has quit IRC | 11:45 | |
| *** ctbruce has joined #baserock | 11:49 | |
| paulsherwood | looks like they've been turned off already? | 12:04 |
| locallycompact | ssam2, so how do I interpret it when for example there are five patches in the same topic (wayland_1_10), only two of which I see commit history for the git-review --download for the most recent commit (Upgrade wayland to 1.10.0, I see that and Upgrade libinput), and one of which claims 'Merge Conflict' | 12:05 |
| locallycompact | s/see commit/see in the commit/ | 12:05 |
| pedroalvarez | paulsherwood: it looks like that | 12:06 |
| paulsherwood | k | 12:06 |
| pedroalvarez | :( | 12:08 |
| paulsherwood | we'll sort something soon :) | 12:08 |
| pedroalvarez | yay | 12:08 |
| ssam2 | locallycompact: first, ignore 'Merge Conflict' -- that means it depends on something which isn't merged | 12:11 |
| ssam2 | locallycompact: but it's perfectly OK if one patch depends on another and neither are merged yet | 12:11 |
| ssam2 | locallycompact: probably the 5 patches don't all depend on each other though | 12:11 |
| ssam2 | locallycompact: if there's not an obvious 'top' commit that depends on the other 4, then you should treat them as seperate commits that can be merged in any order | 12:12 |
| locallycompact | ok | 12:12 |
| locallycompact | In some sense I'd prefer to expect that I could review all those simultaneously, but that's just not the case because that's not how they were submitted? | 12:13 |
| ssam2 | git-review doesn't automate that for you in this case | 12:14 |
| ssam2 | you could create your own branch and merge all 5 commits into it, then test that | 12:14 |
| locallycompact | I see | 12:15 |
| locallycompact | Are all the things that are on gerrit things in active review? i.e do they go away | 12:17 |
| locallycompact | oh Isee | 12:17 |
| locallycompact | nvm | 12:17 |
| locallycompact | defaults to open | 12:17 |
| *** rdale has quit IRC | 12:22 | |
| pedroalvarez | alsa-lib and alsa-utils have moved some of their tags | 12:57 |
| pedroalvarez | **again** | 12:57 |
| pedroalvarez | and this time, one of the tags we are using | 12:59 |
| * pedroalvarez will anchor those tags in baserock/* and allow lorry to forcepush things once | 13:00 | |
| paulsherwood | when we notice this kind of tag-moving, could we email upstream, cc baserock-dev noting what's happened? eventually i expect we'll encounter 'no, we didn't.... oh, sh**, we've been hacked) | 13:05 |
| pedroalvarez | I remember doing that once with alsa | 13:10 |
| pedroalvarez | long time ago | 13:10 |
| pedroalvarez | answer was "no, we didn't.." | 13:10 |
| * pedroalvarez will do | 13:10 | |
| pedroalvarez | GDP with new wayland seems to work. although I hardly can test it due | 13:11 |
| pedroalvarez | "[ 321.478822] nouveau 57000000.gpu: DRM: skipped size 0" errors | 13:11 |
| pedroalvarez | and with wayland 1.9 it just works.. :/ | 13:14 |
| paulsherwood | 13:14 <+pedroalvarez> and with wayland 1.9 it just works.. :/ | 13:15 |
| paulsherwood | is that a bad thing? | 13:15 |
| pedroalvarez | the bad thing is that it fails with wayland 1.10 | 13:15 |
| pedroalvarez | given this problem.. I don't know what to vote in wayland 1.10 patches :/ | 13:23 |
| paulsherwood | ack | 13:23 |
| paulsherwood | vote -1, ask jjardon/radiofree to investigate? | 13:23 |
| pedroalvarez | I remember radiofree suggesting swiching back to nouveau-drm | 13:26 |
| pedroalvarez | paulsherwood: btw, these patches should fix the gdp-hmi build https://gerrit.baserock.org/#/q/status:open+project:baserock/baserock/definitions+branch:master+topic:fix-gdp-hmi | 13:36 |
| radiofree | pedroalvarez: yes, use the out of tree module | 13:49 |
| radiofree | i've already pushed a working branch for that http://git.baserock.org/cgit/delta/nouveau.git/log/?h=baserock/tegra/4.4 | 13:50 |
| pedroalvarez | I'll give it a go | 13:56 |
| pedroalvarez | thanks! | 13:56 |
| pedroalvarez | Patch to update alsa unpetrify-refs: https://gerrit.baserock.org/1917 | 13:57 |
| locallycompact | jjardon, radiofree ok so from a clean deployed system weston --log=/weston.log, does appear something is consuming the foreground but all I see is the terminal still | 14:08 |
| jjardon | pedroalvarez: Id try upgrade to linux 4.5 first | 14:08 |
| locallycompact | is there a way to exit that | 14:08 |
| locallycompact | without rebooting | 14:08 |
| * locallycompact reboots | 14:08 | |
| locallycompact | jjardon, https://paste.fedoraproject.org/329860/95806145/ | 14:10 |
| pedroalvarez | jjardon: do you have time for that now? I wanted to test this to unblock your wayland/weston series | 14:14 |
| jjardon | pedroalvarez: nope, sorry; I sugested that as there are fixes in the nouveau drm in the new linux release | 14:15 |
| jjardon | pedroalvarez: this happens in exactly the same system but only upgrade wayland to 1.10? | 14:16 |
| pedroalvarez | it also happens with wayland 1.9, but only sometimes.. with wayland 1.10 it happens all the time | 14:17 |
| pedroalvarez | I understand your suggestion now, makes sense | 14:18 |
| pedroalvarez | rc5 though.. | 14:19 |
| radiofree | locallycompact: it's a drm error, try with --backend=fbdev-backend.so | 14:20 |
| radiofree | i think there's some configuration you have to do when launching qemu | 14:20 |
| radiofree | -vga foo | 14:20 |
| radiofree | jjardon: ^ | 14:20 |
| jjardon | locallycompact: what is your version of qemu? | 14:20 |
| jjardon | 2.5 I guess? | 14:21 |
| locallycompact | how to tell? | 14:21 |
| locallycompact | whatever's in arch | 14:21 |
| jjardon | qemu-system-x86_64 --version | 14:21 |
| locallycompact | 2.4.1 | 14:21 |
| jjardon | ok, upgrade and try with -vga virtio, to check if its a bug in the bochs drm driver | 14:22 |
| jjardon | (the default in 2.4) | 14:22 |
| locallycompact | What's -vga virtio I use virt-manager to launch all this | 14:23 |
| radiofree | locallycompact: what does lsmod give you | 14:23 |
| locallycompact | in the vm? | 14:23 |
| radiofree | in the vm yes | 14:23 |
| locallycompact | Module Size Used by | 14:24 |
| locallycompact | cirrus 18000 1 | 14:24 |
| locallycompact | ttm 68619 1 cirrus | 14:24 |
| radiofree | ok you're using cirrus, don't use that | 14:24 |
| radiofree | use bochs | 14:24 |
| jjardon | locallycompact: oh | 14:24 |
| locallycompact | Don't got that | 14:24 |
| radiofree | you need to configure it in virt-manager | 14:24 |
| radiofree | graphics card settings or something? | 14:25 |
| locallycompact | yeah no I don't have that option | 14:25 |
| * radiofree has never used virt-manager so doesn't know | 14:25 | |
| locallycompact | got Cirrus, QXL, VGA, VMVGA and Xen | 14:25 |
| jjardon | locallycompact: you need the qlx driver, I think; let me check if we have that in our bsp | 14:25 |
| locallycompact | qxl? | 14:25 |
| jjardon | one second | 14:26 |
| radiofree | -vga std would probably work (if you use the command line) | 14:27 |
| *** jjardon_ has quit IRC | 14:27 | |
| locallycompact | let me try all the ui options | 14:27 |
| locallycompact | qxl gives | 14:28 |
| locallycompact | [14:26:08.189] logind: session control granted | 14:28 |
| locallycompact | [14:26:08.189] no drm device found | 14:28 |
| locallycompact | [14:26:08.189] fatal: failed to create compositor backend | 14:28 |
| jjardon | https://www.irccloud.com/pastebin/4t45mPPo/ | 14:28 |
| jjardon | locallycompact: you need the patch above | 14:28 |
| locallycompact | alright | 14:29 |
| jjardon | locallycompact: if you use qemu it should work though | 14:29 |
| locallycompact | what, exactly? | 14:30 |
| locallycompact | it should have already worked? | 14:30 |
| jjardon | locallycompact: if you use virt-manager, you need the qxl driver, if you use qemu directly, bochs is used, wich is the one I normally test with | 14:32 |
| locallycompact | ok yeah it worked with the VGA menu option | 14:33 |
| locallycompact | I'll make that default | 14:33 |
| locallycompact | tyvm | 14:33 |
| *** faybrocklebank has quit IRC | 14:36 | |
| jjardon | pedroalvarez: so, to be clear here: are you saying gdp doesnt work when upgrading wayland or upgrading weston? | 14:43 |
| *** rdale has joined #baserock | 14:49 | |
| *** faybrocklebank has joined #baserock | 14:50 | |
| pedroalvarez | there are 2 problems I don't want to mis: | 15:05 |
| pedroalvarez | mix* | 15:05 |
| pedroalvarez | 1) wayland-ivi-extension doesn't build with latest weston 1.10, but that is irrelevant for me given that we are not upgrading weston-genivi | 15:06 |
| pedroalvarez | 2) Jetson GDP is much more unstable with wayland 1.10. It's not sure that the wayland upgrade triggers the issue, but I couldn't test the upgrade because of that issue | 15:08 |
| pedroalvarez | Conclusion: I'd like to fix that stability problem to make sure the upgrade to wayland 1.10 doesn't break things | 15:08 |
| radiofree | Using the oot module should fix that | 15:37 |
| *** vgrade_w has quit IRC | 15:53 | |
| pedroalvarez | indeed, same problem with linux 4.5-rc5 | 15:53 |
| *** locallycompact has quit IRC | 16:21 | |
| *** jjardon_ has joined #baserock | 16:28 | |
| radiofree | pedroalvarez: did you try with the nouveau out of tree module | 16:32 |
| pedroalvarez | radiofree: I'm currently building and deploying | 16:33 |
| *** locallycompact has joined #baserock | 16:38 | |
| *** tiagogomes___ has quit IRC | 17:00 | |
| *** tiagogomes has joined #baserock | 17:00 | |
| *** ctbruce has quit IRC | 17:00 | |
| *** franred has quit IRC | 17:02 | |
| *** ratmice_ has joined #baserock | 17:14 | |
| *** bfletcher has quit IRC | 17:20 | |
| *** cyndis has quit IRC | 17:20 | |
| *** ratmice has quit IRC | 17:20 | |
| *** jjardon_ has quit IRC | 17:22 | |
| *** jjardon_ has joined #baserock | 17:23 | |
| *** bfletcher has joined #baserock | 17:23 | |
| *** toscalix has quit IRC | 17:26 | |
| *** cyndis has joined #baserock | 17:27 | |
| *** jjardon_ has quit IRC | 17:29 | |
| pedroalvarez | radiofree: same problem with the oot module | 17:37 |
| pedroalvarez | this makes me sad :( | 17:39 |
| pedroalvarez | I won't have time to debug what's going on | 17:39 |
| radiofree | Bugger | 17:51 |
| *** locallycompact has quit IRC | 17:54 | |
| *** jonathanmaw has quit IRC | 17:55 | |
| pedroalvarez | btw, there is some discussion here: https://devtalk.nvidia.com/default/topic/916337/-nouveau-e-drm-failed-to-create-ce-channel-22/?offset=10 | 17:57 |
| *** bashrc_ has quit IRC | 18:01 | |
| *** ssam2 has quit IRC | 18:09 | |
| *** tiagogomes has quit IRC | 19:01 | |
| *** Stanto has left #baserock | 20:22 | |
| *** zoli_ has quit IRC | 20:22 | |
| *** zoli_ has joined #baserock | 20:30 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!