| *** OhYash1 is now known as ohyash | 12:01 | |
| lbt | Thaodan_: I need to review PR15 when I get back off holiday - feel free to remind me :) | 15:51 |
|---|---|---|
| Thaodan_ | f | 15:52 |
| Thaodan_ | sure :) | 15:52 |
| Thaodan_ | when is that ca? | 15:52 |
| Thaodan_ | have a nice vacation | 15:53 |
| lbt | back on 11/march - off on a ski & salsa holiday \o/ | 15:53 |
| Thaodan_ | interesting combination cold and hot | 16:01 |
| T4 | <dashinfantry> 查看 @thefxtec 的推文：https://twitter.com/thefxtec/status/1101530864305999872?s=09 | 17:31 |
| T4 | <dashinfantry> [Edit] @thefxtec ：https://twitter.com/thefxtec/status/1101530864305999872?s=09 | 17:31 |
| heroic_1 | sledges: any way to get repos for kagura/tone in mer-hybris? my port is buildable now, sources are at https://github.com/marina-bay/ | 18:27 |
| heroic_1 | I am relying on some patched repos, but I've PR'd all the necessary changes to d-h-d, bionic and hybris-boot, it's just a matter of the patched repos getting merged | 18:28 |
| mal | heroic_1: were you the one making those PRs to dhd? | 18:28 |
| heroic_1 | yes, ix5 on gh | 18:29 |
| heroic_1 | mal: you'll be glad to know I am no longer relying on droid-hal-img-boot-sony-tone but using the community-adaptation way with hybris-boot instead | 18:30 |
| heroic_1 | mal: which android version is your fp2 port based on? I am running into this because of a change in Oreo's bootable/recovery https://github.com/mer-hybris/hybris-boot/issues/162 | 18:31 |
| mal | heroic_1: aosp6, soon aosp7 I hope (waiting for sources) | 18:33 |
| heroic_1 | ok, so that means no community ports are using oreo atm? | 18:33 |
| mal | there are some | 18:33 |
| heroic_1 | what about the latte one? which android version is that on | 18:33 |
| mal | latte is non-android based | 18:34 |
| mal | i.e. native port | 18:34 |
| mal | there were some hybris-15.1 ports | 18:34 |
| heroic_1 | non-android means there isn't even a dependency on mediaserver and all the android hals? very interesting | 18:35 |
| mal | heroic_1: hybris-15.1 is for lineage ports so this PR in not relevant https://github.com/mer-hybris/android_bionic/pull/10 | 18:35 |
| heroic_1 | mal: which version of bionic am I supposed to use on oreo? | 18:35 |
| mal | heroic_1: depends on device | 18:35 |
| heroic_1 | the nile ports use https://github.com/mer-hybris/android_bionic/tree/hybris-sony-aosp-8.1.0_r52_20190206 | 18:36 |
| mal | heroic_1: for sony device you use the sony hybris branch | 18:36 |
| heroic_1 | ok then I will stick to the sony branches | 18:37 |
| mal | heroic_1: everything is native in latter, it's x86 device and uses native graphics and stuff | 18:37 |
| mal | *latte | 18:37 |
| heroic_1 | very cool | 18:37 |
| heroic_1 | where are the sony branches for bionic cut out from? like, what do they use as a base | 18:39 |
| heroic_1 | btw this is also needed for the sony branches: https://github.com/mer-hybris/android_bionic/pull/9 | 18:39 |
| heroic_1 | I mean the sony branches seem to be snapshots of something, so what upstream are they from | 18:39 |
| mal | heroic_1: those are taken from the manifest mentioned in sony open devices build instructions | 18:41 |
| mal | heroic_1: what build error do you get without that? | 18:42 |
| heroic_1 | that manifest is just the default android 8.1 one | 18:42 |
| heroic_1 | but the hybris-sony branches have lots of patches on top | 18:42 |
| mal | mostly yes, plus the local manifest from sony | 18:42 |
| heroic_1 | I am only talking about bionic now | 18:42 |
| mal | like what commits? | 18:43 |
| heroic_1 | sony's local_manifests don't touch bionic | 18:43 |
| heroic_1 | we use upstream aosp bionic | 18:43 |
| heroic_1 | mal: the build error was that __get_tls was undefined | 18:43 |
| mal | heroic_1: I find that odd, considering nile devices build fine | 18:43 |
| heroic_1 | I don't have an android 8 tree synced any more as of yesterday, but I can go verify it for you on the weekend | 18:44 |
| mal | heroic_1: why did you remove it? | 18:46 |
| heroic_1 | because my main focus isn't sailfish | 18:46 |
| heroic_1 | we have pressing issues on pie that need fixing | 18:47 |
| mal | so you are not planning on maintaining the port? | 18:47 |
| heroic_1 | sure I want to maintain it, it's just I can't do everything in a single day | 18:47 |
| sledges | PSA: XA2 build instructions available for early testing: https://sailfishos.org/wiki/DRAFT-Sailfish_X_Xperia_XA2_Build_and_Flash | 18:48 |
| sledges | heroic_1: thanks for contribs, please recheck your port against ^ | 18:48 |
| sledges | is your port not only buildable but deployable and bootable? ;) | 18:48 |
| sledges | anyhow, i'm afk for the weekend now:) | 18:48 |
| mal | heroic_1: well I find it odd to remove sources unless you are running out of disk space | 18:49 |
| heroic_1 | mal: that is exactly the issue if you only have 500gb and also some work-related stuff | 18:49 |
| heroic_1 | throw in a vm and you have to be very conservative with your space | 18:50 |
| mal | heroic_1: I have several 100 GB of hybris sources :) | 18:50 |
| heroic_1 | sledges: great, that looks like you need very few patches... cool | 18:50 |
| heroic_1 | my port is bootable and the basics are there, but I don't know what you'd consider deployable | 18:51 |
| mal | heroic_1: I still find it odd that why you even touched hybris-15.1 branches, those are not relevant for you at all | 18:51 |
| sledges | flashable | 18:51 |
| heroic_1 | mal: I thought that the hybris-15.1 were the upstream of the hybris-sony branches, that's why I PR'd against them | 18:52 |
| heroic_1 | sledges: no, not flashable, because of https://github.com/mer-hybris/hybris-boot/issues/162 | 18:52 |
| mal | heroic_1: nope, those are not related, those share similar things but not the same always | 18:52 |
| mal | sledges: he reverted to community build way | 18:52 |
| sledges | heroic_1: if your port can be built against the wiki guide just by replacing the env vars like HABUILD_DEVICE, FAMILY, etc, then it's a good candidate to mer-hybris upstream, will talk next week! | 18:52 |
| sledges | (that should clear some doubts you had of how to use the FAMILY env var too that you raised in some issue/PR, etc) | 18:54 |
| heroic_1 | ye, thanks | 18:54 |
| heroic_1 | one thing that you might not have noticed though(cuz you're building official images): https://github.com/mer-hybris/droid-hal-device/pull/208 | 18:54 |
| mal | heroic_1: that should not be needed if following the new XA2 build instructions | 18:55 |
| heroic_1 | it means that it will look for "droid-config-kagura-flashing", while the d-h-c only provides "droid-config-f8331-flashing" | 18:55 |
| heroic_1 | sledges: I see that your instructions are for the "official" way with pre-packaged droid-system and droid-system-vendor | 18:56 |
| heroic_1 | but let's talk next week, don't want to steal your weekend :) | 18:56 |
| mal | heroic_1: it's simple to skip pre-packaged droid-system and droid-system-vendor | 18:56 |
| mal | just tiny changes to spec | 18:57 |
| heroic_1 | nope | 18:57 |
| heroic_1 | I mean yeah, it's simple to skip them, but the resulting structure is very different | 18:57 |
| mal | not really different | 18:58 |
| heroic_1 | yes it is | 18:58 |
| heroic_1 | git clone --recursive https://github.com/mer-hybris/droid-hal-img-boot-sony-$FAMILY hybris/mw/droid-hal-img-boot-sony-$FAMILY | 18:58 |
| heroic_1 | this means the lvm-based structure will be in place | 18:58 |
| heroic_1 | while I changed my port to not use droid-hal-img-boot-sony-* | 18:58 |
| mal | you just skip a few lines | 18:59 |
| mal | lvm based build is not really related to droid-system | 18:59 |
| heroic_1 | yes mal, that is true | 19:00 |
| mal | those are different things | 19:00 |
| heroic_1 | But they correlate with a different build system using droid-hal-img-boot-sony-* | 19:00 |
| mal | I really can't understand your problem | 19:01 |
| mal | heroic_1: I think one of the problems you could be that you used the fXXXX codenames here whereas nile used the actual codenames | 19:03 |
| mal | which messed up the instructions | 19:03 |
| heroic_1 | no | 19:03 |
| mal | yes | 19:03 |
| heroic_1 | I am using the exact same scheme: DEVICE=f8331, HABUILD_DEVICE=kagura, FAMILY=tone | 19:04 |
| heroic_1 | The reason why sledges doesn't run into the issue with droid-config-*-flashing is because droid-hal-img-boot-sony-* provides that package | 19:05 |
| heroic_1 | meaning: | 19:05 |
| heroic_1 | wait I have it backwards | 19:07 |
| mal | heroic_1: anyway you have useless specs in droid-hal repo because you didn't use the same way nile did | 19:08 |
| mal | it might work but still not the proper way | 19:08 |
| heroic_1 | see this: if NOT using droid-hal-img-boot-sony-* and using the community-build way, you have d-h-d that wants "droid-config-kagura-flashing" | 19:09 |
| heroic_1 | https://github.com/mer-hybris/droid-hal-device/pull/208/files | 19:09 |
| heroic_1 | but if you ARE using the custom boot image in the official sailfish x way, it only wants the generic droid-config-flashing https://github.com/mer-hybris/hybris-initrd/blob/7b52ab04ad19f0c46cccbb2fe3b53c63a04b0c5f/droid-hal-device-img-boot.inc#L80 | 19:10 |
| heroic_1 | hence my PR | 19:10 |
| heroic_1 | mal: you mean those are useless? https://github.com/marina-bay/droid-hal-sony-tone | 19:10 |
| heroic_1 | they are completely the same as nile https://github.com/mer-hybris/droid-hal-sony-nile/blob/master/droid-hal-discovery.spec | 19:11 |
| heroic_1 | https://github.com/marina-bay/droid-hal-sony-tone/blob/sfos-3.0.1.11-android-8.1/droid-hal-f8131.spec | 19:11 |
| heroic_1 | device=kagura and rpm_device=kagura | 19:11 |
| heroic_1 | same as nile | 19:11 |
| mal | heroic_1: half of those are useless | 19:15 |
| mal | fXXX1 and fXXX2 are just variant of the same device so you could use for example dora as the common spec | 19:15 |
| mal | like here https://github.com/mer-hybris/droid-hal-sony-nile | 19:16 |
| heroic_1 | lunch device is different | 19:16 |
| heroic_1 | important for getting the right system.prop | 19:16 |
| mal | so pioneer is for both h4113 and h3113 | 19:16 |
| heroic_1 | I did this for a reason mal | 19:16 |
| mal | there was also a reason why it wasn't done for nile devices | 19:17 |
| heroic_1 | else you won't get the dualsim props that are needed by rild | 19:17 |
| mal | to minimize useles builds | 19:17 |
| mal | nope | 19:17 |
| heroic_1 | well either way, it doesn't really hurt | 19:18 |
| mal | well for nile devices single and dual sim variants use same system image eventhough in android those have separate builds | 19:19 |
| mal | do the dual-sim properties matter in single sim device, since we only use one slot | 19:19 |
| heroic_1 | I guess it works via the pre-packaged build.prop in droid-system-$DEVICE | 19:20 |
| mal | because we delete the dual-sim ofono config for single-sim image | 19:20 |
| heroic_1 | we start the second rild based on that prop, so I was very careful to keep it in for dualsim https://github.com/sonyxperiadev/device-sony-common/blob/master/rootdir/vendor/etc/init/rild2.rc | 19:21 |
| heroic_1 | but anyway, can we please stop debating this minute detail | 19:21 |
| heroic_1 | I somehow lost bluetooth when switching to community builds | 19:21 |
| mal | well it doesn't seem to cause any issues in nile device sailfish builds | 19:21 |
| heroic_1 | and still haven't gotten usb to work | 19:22 |
| heroic_1 | that's what I'll try to focus on next week | 19:22 |
| mal | heroic_1: I know what the reason is, I just don't know how to fix it | 19:22 |
| heroic_1 | for what exactly? | 19:22 |
| mal | heroic_1: the reason is that the configfs init in .rc is not actually doing anything for some reason, in logs you can see it runs the .rc file but the configfs configs cannot be seen in actual path | 19:24 |
| mal | so there is only the stuff from usb configs from early init but not from .rc | 19:24 |
| heroic_1 | oh yeah, I also noticed some locations it tries to write to don't exist | 19:25 |
| heroic_1 | but I can better help you when I pull logs from the device, either this weekend or next week | 19:25 |
| mal | that is the thing, the path that should have been created by .rc file are not created | 19:25 |
| mal | heroic_1: I have all the logs I need from another device | 19:25 |
| heroic_1 | ok great | 19:30 |
| heroic_1 | I will try some stuff on my side as well | 19:31 |
| heroic_1 | have a nice evening everyone! | 19:31 |
| *** Gizmokid2005 is now known as Fri-Day | 21:13 | |
| *** Fri-Day is now known as Gizmokid2005 | 21:13 | |
| ctroy | How to run anbox on sailfishOS? | 22:56 |
| mal | depends on the device you have | 22:57 |
| *** doniks_ is now known as doniks | 22:57 | |
| ctroy | moto g3 | 22:58 |
| mal | what kernel version does it have? | 22:58 |
| ctroy | I'm using sailfishos 2.2 | 22:59 |
| ctroy | The kernel is 3.10.49-g55f9ac8-dirty | 23:02 |
| ctroy | it works with my kernel? | 23:05 |
| vknecht | ctroy, if it has overlayfs and lxc patches, it might work | 23:06 |
| ctroy | Where do I find these patches? | 23:20 |
| vknecht | iirc the last two here : https://github.com/mlehtima/android_kernel_sony_msm/commits/anbox | 23:20 |
| vknecht | but there was an etherpad, with better instructions... | 23:21 |
| mal | https://public.etherpad-mozilla.org/p/anbox-sailfishos | 23:22 |
| vknecht | that's it, thanks mal :) | 23:23 |
| mal | had to grep logs because I forgot the link | 23:23 |
| r0kk3rz | i did link it in the faq mal :P | 23:24 |
| mal | ok | 23:25 |
| ctroy | huge thanks :) | 23:26 |
| vknecht | my anbox notes : https://pastebin.com/sefNgMMe | 23:26 |
| mal | vknecht: I think it's better to use the other repo linked in etherpad, my repo is not very up to date anymore | 23:28 |
| ctroy | the lxc patch is editing a file? | 23:40 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!