**** BEGIN LOGGING AT Thu Aug 22 02:59:56 2019 Aug 22 05:01:37 Morning! Aug 22 05:24:05 Tofe: I'll have a look if I can quickly mock a Hammerhead 7.1 build together, shouldn't be rocket science hopefully seeing there are plenty of 7.1 ports already out in the wild. So Halium side should be easy. Then just some minor adjustmnets in our layer. FIngers crossed it would boot. **** BEGIN LOGGING AT Thu Aug 22 06:45:35 2019 Aug 22 07:08:40 Morning! Aug 22 07:09:13 Herrie: yes, good idea; I see there's even an official LineageOS 14.1 port https://forum.xda-developers.com/google-nexus-5/orig-development/rom-cm14-1-nexus-5-hammerhead-t3510548 Aug 22 07:29:18 Tofe: Yeah, it shouldn't be rocket science. Probably need to get the GCC patches in there, but well Aug 22 08:07:02 yep Aug 22 09:04:48 Halium 5.1 kernel didn't work, so now patching the cm 14.1 kernel with GCC fixes, PID namespace etc ;) Aug 22 09:07:16 Herrie: pong Aug 22 09:07:23 novaldex: Hi :) Aug 22 09:07:33 how's things? Aug 22 09:07:58 novaldex: Good in general, busy as always ;) Just noticed some performance issues on the various webservers, so thought to give you a nudge ;) Aug 22 09:08:22 webos-ports.org is slow, but even the IRC logs are slow like hell and those are extremely small ;) http://logs.nslu2-linux.org/livelogs/webos-ports.txt Aug 22 09:08:40 is that just today? i've been doing some patching this morning & noticed a slow down too Aug 22 09:09:41 bad CVE on Jenkins has been fixed Aug 22 09:11:59 novaldex: Seems last couple of days Aug 22 09:12:55 My ping is 140-145ms to webos-ports.org Aug 22 09:13:06 Sometimes peaking around 225ms Aug 22 09:14:46 that's a bit much, will have a look Aug 22 09:15:17 Both from work connection and home fiber Aug 22 09:18:44 i'd hate to think what my latency would be like over here, home or wor Aug 22 09:19:09 (in fact my home @ 6Mbps is probably better than this client's overloaded connection!) Aug 22 09:37:31 Damn, I can't get my rosy to boot with the new images Aug 22 09:37:40 even a sumo image doesn't boot Aug 22 09:42:45 Tofe: Hmmz weird :S Aug 22 09:46:30 yup Aug 22 09:52:55 Tofe: Got Halium bits to build. Kernel patches integrated as well, now testing kernel build inside ours ;) Aug 22 09:56:28 :) Aug 22 10:37:55 Herrie: I don't suppose things look any better now? Aug 22 11:29:01 novaldex: Not really Aug 22 11:29:03 Still 140ms Aug 22 11:29:16 okay, thanks Aug 22 11:41:50 Tofe: Kernel builds, memnotify didn't but well.. Just removed it now for Hammerhead Aug 22 11:42:21 Finishing up some modules now and should have a 7.1 image Aug 22 11:43:08 \o/ Aug 22 11:56:44 Tofe: What am I doing wrong? https://bpaste.net/show/zoTy Aug 22 11:57:08 I build it as module now as per https://github.com/Herrie82/android_kernel_lge_hammerhead-1/commit/ff3e68f5b85fb22a17c8b55d7386feebd6ef86ec Aug 22 12:03:04 I can simply remove it from MACHINE_EXTRA_RDEPENDS in hammerhead.conf and it would build the image, but that's not what we want I guess? Aug 22 12:03:15 Though for testing oFono, not having WiFi is not end of the world I guess Aug 22 12:06:25 for now you can exclude it from rdepends yes; somehow the module isn't build, maybe an issue/inconsistency in the kernel config file Aug 22 12:13:48 Yeah that's what I've done. It's on the FTP in case you want to give it a stab too Aug 22 12:32:37 Hmmz stuck @ Google logo it seems. Log: https://bpaste.net/show/FQ9F Aug 22 12:33:53 Seems some mounts go wrong Aug 22 12:34:17 At least kernel not pannicking ;) Aug 22 12:38:43 Tofe: Pushed meta-smartphone to: https://github.com/Herrie82/meta-smartphone/tree/herrie/warrior-hammerhead-7.1 Aug 22 12:38:59 Let me also PR some stuff to Halium. Just have to do some git-fu with branches etc Aug 22 13:17:33 Herrie: looks more like "systemd-modules-load.service: Main process exited, code=exited, status=1/FAILURE" is the culprit somehow Aug 22 13:25:39 Herrie: "User namespace: missing" Aug 22 13:25:48 "Cgroup cpuset: missing" Aug 22 13:27:02 it boots ok-ish, but the android container doesn't start Aug 22 13:27:06 Ah ok Aug 22 13:27:37 logs in /var/log/lxc/android.log Aug 22 13:30:47 Let me check some defconfig options. Aug 22 13:30:57 Seems I missed the CONFIG_CPUSETS=y somehow Aug 22 13:33:11 and CONFIG_MODULES=y as well Aug 22 13:33:21 That explains why it didn't work for WLAN I guess Aug 22 13:33:31 Though the Mer script didn't flag these... Aug 22 13:34:50 yes, it explains it all :) Aug 22 13:34:58 let's hope it'll be sufficient Aug 22 13:35:15 didn't I implement - partially - user namespaces? Aug 22 13:36:02 ah no it was pid ns Aug 22 13:36:38 Yeah and pid ns as well Aug 22 13:36:41 I added those commits ;) Aug 22 13:37:19 ok :) Aug 22 13:37:39 Without them couldn't build kernel Aug 22 13:37:50 I'll double check MEr's script, maybe these need adding Aug 22 13:38:01 Then I'll just need to redownload boot.img I think, in order to test a bit -- if it boots Aug 22 13:38:17 Yeah Aug 22 13:38:21 Building now Aug 22 13:54:03 Doesn't like the modules somehow Aug 22 13:54:40 https://bpaste.net/show/HON2 Aug 22 13:56:04 Anyway trying again without modules now Aug 22 14:02:23 For backlight you can always have it in the kernel Aug 22 14:02:31 as for wifi, it's weird errors here Aug 22 14:02:51 IPV6 is set to "yes", so... mystery Aug 22 14:03:26 Without modules it works :S Aug 22 14:03:31 Which file you need? zImage-dtb-hammerhead.fastboot ? Aug 22 14:05:58 yes Aug 22 14:06:14 Should be there :) Aug 22 14:07:30 booting... Aug 22 14:09:02 Here too Aug 22 14:09:51 still the same "cgfsng.c:cgfsng_create:1385 - Invalid argument - Failed to create /sys/fs/cgroup/net_cls//lxc/android: Invalid argument" Aug 22 14:11:42 looks like this might be fixed via https://github.com/lxc/lxc/pull/3014/commits/36f7018103cd66cc16128b04200df15320472a54 ? I don't see the relation with halium 7.1 here Aug 22 14:12:09 maybe it's not fatal Aug 22 14:12:11 I guess I need https://github.com/Herrie82/android_kernel_lge_hammerhead/commit/533a628f03d4d0be2f1c4e969aa1c6d7a7e564bb Aug 22 14:13:39 ah, yes, good idea Aug 22 14:14:46 Another try Aug 22 14:18:37 New file there Aug 22 14:18:45 download it already Aug 22 14:20:29 much better logs Aug 22 14:21:11 but crash in kernel in proc_ns_follow_link ... Aug 22 14:21:23 (Process lxc-start) Aug 22 14:21:59 NULL pointer it seems... yay... Aug 22 14:22:10 That could be https://github.com/Herrie82/android_kernel_lge_hammerhead/commit/bee43040a491ed49b2abf4924f957d7d51349129 somehow? Aug 22 14:22:33 not sure; I don't think so Aug 22 14:23:21 Ah seems I might have missed this one as well: https://github.com/Halium/android_kernel_lge_hammerhead/commit/25437b2a54dd619a96e268ecaf303b089aa785e4 Aug 22 14:24:01 it could be related Aug 22 14:29:59 mmh this method is a new one, introduce in the cm-14.1 kernel Aug 22 14:32:47 we could also try applying the last commit here https://github.com/Herrie82/android_kernel_lge_hammerhead-1/commits/lineage-16.0/fs/proc/namespaces.c or the 3 last commits Aug 22 14:34:13 Onyx has this commit: https://github.com/LineageOS/android_kernel_oneplus_onyx/commit/ceebc12a0ad78884dee035c6874d028fe259f4f6#diff-3b35875fb9a81612b098b440bddcf5a8 Aug 22 14:35:24 I'm tempted to just try last commit first ;) Aug 22 14:35:30 yes, please do Aug 22 14:35:36 it's the most convincing Aug 22 14:38:36 OK building Aug 22 14:51:53 Failed & patching Aug 22 14:52:06 Need to add nd_jump_link Aug 22 14:55:12 New fastboot on FTP Aug 22 14:57:45 Still issues on the container it seems Aug 22 14:58:35 https://bpaste.net/show/4Xxg Aug 22 15:02:43 "lxc-wait: monitor.c: lxc_monitor_open: 237 Failed to connect to monitor socket' Aug 22 15:03:24 it's still crashing Aug 22 15:03:34 (look in dmesg) Aug 22 15:05:47 what if we just reverted these problematic commits that create the crash in the first place ? Aug 22 15:06:41 https://github.com/Herrie82/android_kernel_lge_hammerhead-1/commits/cm-14.1/fs/proc/namespaces.c the last 3 there... Aug 22 15:11:33 Let me try that Aug 22 15:35:27 Tofe: Well the history for our branch is a bit different: https://github.com/Herrie82/android_kernel_lge_hammerhead/commits/halium-7.1/fs/namespace.c Aug 22 15:35:36 The 5.1 one: https://github.com/Herrie82/android_kernel_lge_hammerhead/commits/halium-5.1/fs/namespace.c Aug 22 16:51:43 oh **** BEGIN LOGGING AT Thu Aug 22 19:11:52 2019 **** BEGIN LOGGING AT Thu Aug 22 19:36:37 2019 **** ENDING LOGGING AT Fri Aug 23 03:00:40 2019