**** BEGIN LOGGING AT Wed Oct 16 02:59:57 2019 Oct 16 03:47:18 @NotKit do i neeed fix-mountpoints for halium-boot? Oct 16 03:47:41 @Akatsu . [i would just change second one from oom_score_adj = p->signal->oom_adj to oom_ …], ok Oct 16 03:47:45 will try Oct 16 05:00:27 @Luka [@NotKit do i neeed fix-mountpoints for halium-boot?], no Oct 16 06:35:07 @NotKit [no], thanks Oct 16 06:35:21 Akatsu helped but new error Oct 16 06:35:32 in https://github.com/LinerSRT/n370b_n1/blob/f2063c3db04b2341e46c21bb9a2cfcbedee1e89c/mm/vmscan.c#L2116 Oct 16 06:35:38 error: 'lowmem_minfree' undeclared (first use in this function) Oct 16 06:36:49 hmm itis in drivers/staging/android/lowmemorykiller.c Oct 16 06:37:35 may be it must be like extern int lowmem_minfree[2]; Oct 16 06:40:44 with that i haev ../mm/vmscan.c:2116:31: error: array subscript is above array bounds [-Werror=array-bounds] Oct 16 06:43:40 now ../fs/ext4/inode.c:4736:2: error: #error POSIX_ACL not supported in 3.18 backport … #error POSIX_ACL not supported in 3.18 backport Oct 16 06:48:22 Akatsu idea? Oct 16 06:51:39 @Luka [may be it must be like extern int lowmem_minfree[2];], not 2 but 9 Oct 16 06:56:06 again Oct 16 07:11:28 damn cant finde that at all Oct 16 07:14:38 I was able to narrow down the issue with glamor, it appears something in/called by glamor_prepare_access_picture_box extremely slows down the X server for me. Oct 16 07:15:20 Commenting out the following lines in the if-statement makes the UI work fluid again: Oct 16 07:15:25 https://github.com/NotKit/glamor-hybris/blob/master/src/glamor_render.c#L1728 Oct 16 07:15:42 Any ideas, @NotKit ? Oct 16 07:16:36 With those lines the X server goes to 100% cpu especially when opening GTK programs like eg caja... Oct 16 07:18:00 (without it’s ok, but also rather slow as the code falls back to fbcomposite) Oct 16 07:20:29 Luka maybe try with oreo kernel instead? Oct 16 07:22:08 Akatsu don't, MediaTek uses different BSP for Oreo, will break blobs Oct 16 07:23:21 Luka your config seems to have diverged pretty much, maybe it's easier to enable options one by one? Oct 16 07:24:40 @Luka [/home/hello/doogeek/out1/../mm/vmscan.c:2577: undefined reference to `lowmem_min …], you probably have disabled Android OOM killer? Oct 16 07:26:30 @NotKit [Akatsu don't, MediaTek uses different BSP for Oreo, will break blobs], So kernel makes big diff for mediatek? I always thought newer kernel is backward compatible Oct 16 07:27:03 So like run newer kernel with old vendor should pass Oct 16 07:28:01 it depends on particular case. If BSP kernel was not updated for the device, it normally will be backwards compatible, as it is same old kernel with patches on top Oct 16 07:29:42 If it was updated, then interfaces exposed for blobs may be totally incompatible, as vendor obviously doesn't test running new kernel with old userspace Oct 16 07:30:36 @Daniel [(without it’s ok, but also rather slow as the code falls back to fbcomposite)], does it even work without those ifs? Oct 16 07:33:14 so must fix this kernel Oct 16 07:33:14 ? Oct 16 07:43:14 @NotKit [does it even work without those ifs?], Surprisingly yes. But I have to leave the fbcomposite statement (otherwise there is garbage on the screen) Oct 16 07:45:18 I don't know much how glamor works, unfortunately. This was forked from Rockchip's xserver and modifed to work with hybris buffers. They now have a newer version which tries to do memory mapping when possible instead of downloading/uploading textures through GL, but I didn't try to adapt it Oct 16 08:34:08 @NotKit [I don't know much how glamor works, unfortunately. This was forked from Rockchip …], Thanks anyways. I’ll do a little further testing about the impact and performance. All in all it’s definitely usable now. Oct 16 09:02:56 @NotKit [I don't know much how glamor works, unfortunately. This was forked from Rockchip …], Ah, sorry, one more basic question: I have seen a lot of debug messages inside the code, how and where can I make them show up? Oct 16 15:39:38 yay got kernel compiled Oct 16 15:41:16 @NotKit where must i get device/common Oct 16 15:43:57 @Daniel [Ah, sorry, one more basic question: I have seen a lot of debug messages inside t …], if there are no compile-time disabled macros, probably you need to enable verbose logging for Xorg by argument and check /var/log/Xorg.0.log Oct 16 15:44:33 @Luka [@NotKit where must i get device/common], same place as where device tree for LOS is from? Oct 16 16:31:14 @NotKit not there, imeant something like https://android.googlesource.com/device/common/+/refs/heads/nougat-cts-release Oct 16 16:31:39 what is it for? Oct 16 16:32:13 Android parts should come from Lineage. If that's custom repo, then you need to find it from the same author Oct 16 17:06:41 @NotKit https://github.com/LineageOS/android_device_common Oct 16 17:06:52 but no cm14 branch Oct 16 17:08:51 found Oct 16 17:08:51 https://github.com/LineageOS/android_device_common/tree/staging/cm-14.0 Oct 16 17:15:57 @NotKit i compiled halium-boot Oct 16 17:15:58 but Oct 16 17:16:05 cant compile systemimage Oct 16 17:16:06 ninja: error: '/home/hello/halium7d/out/target/product/n370b/obj/SHARED_LIBRARIES/libcurl_intermediates/export_includes', needed by '/home/hello/halium7d/out/target/product/n370b/obj/SHARED_LIBRARIES/gps.mt6737m_intermediates/import_includes', missing and no known rule to make it Oct 16 17:16:21 still same device? Oct 16 17:16:37 yep Oct 16 17:16:57 but no problems with halium boot as kernel is prebuilt Oct 16 17:18:29 Luka you need to readd libcurl to Halium Oct 16 17:19:43 it was deleted? Oct 16 17:20:23 yes, during Android build minification process Oct 16 17:20:29 https://github.com/Halium/android_build/blob/halium-7.1/core/main.mk#L527 - it needs to be also listed in here Oct 16 17:31:45 @NotKit damny device Oct 16 17:31:47 ninja: error: '/home/hello/halium7d/out/target/product/n370b/obj/SHARED_LIBRARIES/libandroid_runtime_intermediates/export_includes', needed by '/home/hello/halium7d/out/target/product/n370b/obj/SHARED_LIBRARIES/gps.mt6737m_intermediates/import_includes', missing and no known rule to make it Oct 16 17:32:47 you can probably skip gps for the moment Oct 16 17:32:56 just comment out whatever pulls it in Oct 16 17:33:17 (Photo, 744x139) https://irc.ubports.com/Xk7mrx6g.png xd Oct 16 17:33:27 lol Oct 16 17:33:43 stop Oct 16 17:33:51 isnt this triada big deal back in time/ Oct 16 17:33:51 it is related to my device Oct 16 17:33:59 look for that xD Oct 16 17:34:06 was in stock romof device Oct 16 17:36:53 https://www.maketecheasier.com/triada-malware-preinstalled-low-cost-android-phones/ Oct 16 17:37:02 seems like x5 max pro was affected Oct 16 17:37:54 [Edit] seems like x5/ max pro was affected Oct 16 17:48:16 wtf. go github>type libandroid_runtime.so in search> get first repo>click on it > find some random files one called "Latest Free Pron Videos on GotPron" Oct 16 17:48:20 (Sticker, 512x368) https://irc.ubports.com/V5dVqKY7.webp Oct 16 17:49:11 uh ok Oct 16 18:07:49 @Akatsu . [seems like x5/ max pro was affected], i know Oct 16 18:08:05 have you fixed this error already? Oct 16 18:08:28 no working Oct 16 18:08:46 maybe thekit is right Oct 16 18:08:52 just giveup for now on gps Oct 16 18:09:11 ok Oct 16 18:09:17 https://github.com/LinerSRT/android_device_n370b/blob/5d0413c0acc17efbb041ad76d7f0dd6d384035aa/gps/hal/Android.mk Oct 16 18:09:39 you can comment libandroid_runtime too tho under local shared libraries Oct 16 18:14:41 where to coment that Oct 16 18:15:06 link above Oct 16 18:15:10 #60 Oct 16 18:18:15 now i feel broken af Oct 16 18:18:54 i read tizen slides and it says i really need to bring modem up to get aDSP however when i want to do it i get errors from rmt_storage Oct 16 18:27:52 thanks Oct 16 18:27:55 building Oct 16 18:34:59 тщ Oct 16 18:35:01 no* Oct 16 18:35:03 libsepol.avtab_write_item: policy version 29 does not support ioctl extendedpermissions rules and one was specified Oct 16 18:35:45 here is big part https://bpaste.net/show/RISJ Oct 16 18:36:04 @Akatsu . [link above], thanks helped Oct 16 18:37:20 revert_sepolicy_to29.patch Oct 16 18:37:27 let my try to inde Oct 16 18:39:00 Cant finde for my platform Oct 16 18:39:35 have you used use_host_lex? Oct 16 18:39:47 export USE_HOST_LEX=yes Oct 16 18:40:08 inside of working terminal Oct 16 18:42:27 not helped Oct 16 18:42:51 may be https://github.com/LinerSRT/android_device_n370b/blob/LS14.1/patches/0001-sepolicy-v29-patch.patch Oct 16 18:42:53 nanu a was added by: nanu a Oct 16 18:42:54 problem Oct 16 18:44:07 its applied? Oct 16 18:44:15 if not try do it Oct 16 18:45:31 inside device tree ./patches/apply-patches.sh Oct 16 19:31:04 Jace N was added by: Jace N **** BEGIN LOGGING AT Thu Oct 17 02:29:00 2019 **** ENDING LOGGING AT Thu Oct 17 02:59:57 2019