**** BEGIN LOGGING AT Mon Feb 15 02:59:58 2021 Feb 15 07:19:25 Morning! Feb 15 07:31:49 Good morning! I'm lurking here to get an overview what's happening here as I'm interested in LuneOS. Information on the web pages is a bit sparse, so I had to become an IRC user again (after 25 years or so). I'm impressed by what you achieved to keep the webOS heritage alive. Thank you very much for your work and your endurance. Feb 15 07:38:08 Morning! Feb 15 07:38:14 As for "a bit sparse": three out of five links in this channel's topic are dysfunctional. Feb 15 07:40:00 FreeRadical: for me, the issues link is down, the donations might not be the best one anymore, but the rest seems to work Feb 15 07:41:13 FreeRadical: (and thanks for the encouragement :) ) Feb 15 07:43:21 FreeRadical: currently, the elephant in the room is the lack of apps, obviously. We'll need to improve this status quo, but the plan isn't clear yet Feb 15 07:47:19 @Tofe Incidentally, the only mobile app I ever wrote was a set top box remote for the rare beast we had at home back then. It was a webOS app and had exactly two users, me and my wife. Meanwhile, I became a father of three kids and don't have the time any more for this, although I miss it. Feb 15 07:48:02 Tofe: the donations link is only dysfunctional if you really want to donate :-) Feb 15 10:01:50 Tofe: Now flashing the Tissot image to see if wifi/bt work again Feb 15 10:02:09 Seeing we'll be on nice new 4.x kernels, Anbox should be a lot easier for the app side of things ;) Feb 15 10:11:12 Tofe: bluebinder is now there, but seems it needs some tweak (similar to the one from Sagit I pasted? Sep 20 12:17:41 tissot bluebinder[3266]: Failed to open /dev/vhci deviceUnable to open virtual device Feb 15 10:11:51 Seems I don't have a /dev/vhci Feb 15 10:17:10 I'm building kernel with SRCREV = "28cfffc8ab361e49654bb376d385f425e017e357" Feb 15 10:17:12 So that should be OK Feb 15 10:18:19 Hmmz WLAN still same issue :S Sep 20 12:17:45 tissot modprobe[3452]: modprobe: ERROR: could not insert 'wlan': Exec format error Feb 15 10:26:32 logcat for BT looks pretty OK, only see this: 09-20 10:18:42.024 40 1141 E vendor.qti.bluetooth@1.0-async_fd_watcher: ThreadRoutine unable to set SCHED_FIFO for pid 40, tid 1141, error Operation not permitted Feb 15 10:26:43 But do seem to get MAC assigned etc Feb 15 10:27:16 LEt me reboot, maybe it's a first boot issue Feb 15 10:33:03 Hmmz still no /dev/vhci Feb 15 10:35:32 My logcat: https://paste.ubuntu.com/p/xZFZtFVrc2/ Feb 15 10:36:54 Herrie: for the "Exec format error", there should be more in the kernel ring buffer. I suspect the module to not have been compiled with a reasonably similar toolchain. Feb 15 10:38:03 Well it's all built in the same place ;) Feb 15 10:38:17 Weird thing is that it works for Tofe, but not for me and we're building exactly the same Feb 15 10:38:21 Or should be at least ;) Feb 15 11:41:21 FreeRadical: You have any spare device laying around where you'd like to see LuneOS running on? Feb 15 12:09:51 Tofe: I got also pointed to this commit wrt BT: https://github.com/SailfishOS-sagit/android_kernel_xiaomi_msm8998/commit/de845066bc95f7d1ffea6d8e02cc32aa7c360990 Feb 15 12:16:43 Herrie: Actually, I own a Pinephone and do try to run LuneOS on it ... but it's hard to get started, as even setting up accounts like IMAP or CalDAV don't work out of the box. I'm not scared by workarounds using the shell (coming from SXMO), I just didn't invest much time finding out where the problem is (which is always the limiting factor here: Feb 15 12:16:44 time ...). Luna seems to restart about once every 10min, too, so it's a bit rough at the moment, but I immediately felt so familiar with all of it and was really grateful for what you already achieved. Feb 15 12:16:44 I have a sibon (Fairphone2) at the moment, running SailfishOS as a daily driver, and I probably receive one or two more sibons over the year to experiment with. I don't expect you to do anything about it; I hope to learn a bit by lurking here such that I probably can try to do something myself. As I wrote above, it was quite difficult to actually Feb 15 12:16:45 find something recent about LuneOS development, so I'm happy to have ended here finally! Feb 15 12:27:44 Herrie: bluebinder seems a bit fragile, for instance it you deactivate BT, then bluebinder is all lost Feb 15 12:28:18 FreeRadical: I'm doing the pinephone port, I can help there Feb 15 12:29:47 FreeRadical: what edition of the Pinephone do you have ? Feb 15 12:33:45 Tofe: I have a v1.2a, tested the February 13 image and the one before that. Feb 15 12:35:49 ok good, I'll check; lately I've worked more on getting Halium-9.0 to work, I didn't touch the pinephone image Feb 15 12:40:19 Tofe: Please see that I'm not here to beg you to do something! One day, I'd like to be able to do something myself, like (on the pinephone port) upgrading to a later kernel or adapting to new interfaces. The pinephone is a rather special beast because it's not got all those Android interfaces, but at least it doesn't have the Android limitations Feb 15 12:40:19 ;-) Feb 15 12:41:25 FreeRadical: no worries ! but the pinephone port is currently our only mainline device, so I'd like to have it in good shape :) Feb 15 12:42:22 (and I see you'd like that too :-p ) Feb 15 12:43:08 Herrie: I see they advice to just disable rfkill for bluetooth, which matches the bugs I've seen Feb 15 12:44:48 Tofe: Well I guess we could check HADK Feb 15 12:45:02 See what they say there, just got this pointed out from the Sagit user ;) Feb 15 12:45:12 Who also did a SFOS port and UBTouch port Feb 15 12:48:10 for bluebinder I'm not too worried, but for wlan I still don't have a clear picture as to what happens precisely Feb 15 12:50:55 Doesn't seem anything specific in HADK-FAQ (which does have a lot of other useful things we can use probably): https://github.com/mer-hybris/hadk-faq#bluetooth Feb 15 12:53:37 HADK part of BT is mainly TODO ;) Feb 15 12:53:43 https://sailfishos.org/content/uploads/2021/02/SailfishOS-HardwareAdaptationDevelopmentKit-4.0.1.2.pdf Feb 15 12:53:46 So not very helpful Feb 15 12:53:51 I guess we could try this patch Feb 15 12:54:04 Seeing it's also Xiaomi and similar-ish chipset it might help Feb 15 13:45:59 Tofe: Also came across this for A2: https://github.com/SailfishOS-MI-A2/droid-config-jasmine_sprout/commit/857ab725b1640f60b6824b90491edcca7460b2b6 Feb 15 13:56:40 Herrie: interesting. But let's not take this commit as-is, I see plenty of mistakes in that one Feb 15 13:57:20 Much easier to just add a "PreStart" command in bluebinder itself that would start rfkill unblock hci0 Feb 15 13:57:49 bluebinder service itself, I mean Feb 15 14:11:57 Tofe: Just pasted some references I came across so we can look into them and see what's most suitable for our situation Feb 15 15:52:13 Tofe: My mounts: https://paste.ubuntu.com/p/cDVJgxSYhk/ Feb 15 15:59:02 Herrie: lgtm Feb 15 15:59:57 Yeha thought that might be a first clue as to what might be wrong, but looked OK to me too Feb 15 19:45:35 Tofe: I've killed my tmp-glibc and sstate-cache and rebuilding from scratch see if it makes any difference ;) Feb 15 19:46:24 Herrie: for wlan you mean? Feb 15 19:46:35 Yeah Feb 15 19:46:40 Who knows Feb 15 19:46:53 ok, that's radical, but for you it's not that long Feb 15 19:46:55 I doubt it'll make a difference but can afford a full rebuild Feb 15 21:12:11 Herrie: https://github.com/Tofee/meta-smartphone-1/commit/f4ff251bba2572ebcffc73b7257b889c4c4ce9b5 this variant gives me good results so far. I just fear it is more fragile than a well-known property... Feb 15 21:22:27 12:28 <+Herrie> I did a cleanall for the linux-xiaomi-mido which should delete any artifacts also of the wlan module Feb 15 21:23:16 "cleanall" is the worst, "cleansstate" removes everything except the sources from downloads directory (which _very_ rarely the issue) Feb 15 21:43:16 Tofe: nope still no go: https://paste.ubuntu.com/p/2FgQV5t7pP/ Feb 15 21:43:25 Sep 20 06:17:41 tissot kernel: wlan: disagrees about version of symbol module_layout Feb 15 21:43:31 I'm a bit lost now :S Feb 15 21:43:52 JaMa: I know, we just have a weird issue with kernel wlan module Feb 15 21:43:59 It seems to work for Tofe but not for me Feb 15 21:44:05 Herrie: mismatching kernel/modules ? Feb 15 21:44:22 Tofe: And logcat: https://paste.ubuntu.com/p/gR3PrMHHNd/ Feb 15 21:44:54 BT still the same too :S Feb 15 21:45:27 damn... Feb 15 21:45:50 CAn you paste your kernel recipe, just in case Feb 15 21:45:57 And the .conf? Feb 15 21:46:49 sure Feb 15 21:47:26 https://bin.disroot.org/?8fd356e4b3bbc1b9#73az8Wq3xEyJQngGoKBtB3V7ZHG46yKqAqwaWPPjEz5z kernel recipe (same as on GH) Feb 15 21:47:34 .conf of ? Feb 15 21:49:05 From tissot, the one where the kernel modules are mentioned Feb 15 21:49:22 Kernel-module-wlan and the other ones Feb 15 21:50:20 https://bin.disroot.org/?5815e76072bb50f4#7dHoCsFVigD4kZhCDVGNhG4bdCaoBoTCtL2bfh969o92 tissot.conf Feb 15 21:50:49 in general, my meta-smartphone is (right now at least) in sync with GH Feb 15 21:51:27 OK some stuff for tomorrow early morning then Feb 15 21:51:41 I'm pretty sure I picked everything as well Feb 15 21:51:55 Not sure what can cause this mismatch Feb 15 21:52:07 Also still no /dev/vhci Feb 15 21:52:44 Udev rules? Feb 15 21:52:58 I'm really a bit lost Feb 15 21:53:11 Journal and logcat look pretty good for both Feb 15 21:55:39 re-flash radio partitions ? I use this one (doesn't overwrite userdata nor boot_a) https://bin.disroot.org/?ce93f8c4084d6319#DE5KSq9LfB3237TjT4TFuRnwcdNwGdtExqmm4ECnGynd Feb 15 21:56:04 for me it did reset the bluetooth address for instance Feb 15 21:59:20 we'll continue the investigation tomorrow, but we do have the same device, so we'll manage to get identical results :p **** ENDING LOGGING AT Tue Feb 16 03:00:33 2021