**** BEGIN LOGGING AT Fri Aug 16 03:05:29 2019 Aug 16 03:25:19 Crofton|work, i can change the virtual to be : PREFERRED_PROVIDER_virtual/gpu ?= "${@bb.utils.contains("MACHINE_FEATURES", "mali", "mali-t62x", "mesa", d)}" Aug 16 03:25:40 in the odroid-xu-default.inc Aug 16 03:26:21 then if you remove 'mali' from MACHINE_FEATURES then it should do the right thing Aug 16 13:19:06 armpit, | WARNING: /home/balister/opensdr/test-builds/odroid-xu4/build/tmp-glibc/work/odroid_c2-oe-linux/kernel-module-mali-utgard/r9p0-01rel0-meson-gx-r0/temp/run.do_compile.20042:1 exit 1 from 'exit 1' Aug 16 13:19:09 with c2 Aug 16 15:29:25 * armpit so behind for training Aug 16 16:06:59 armpit, what kind of training? Aug 16 16:08:37 YP devday Aug 16 16:10:45 AH Aug 16 16:10:51 Thanks for helping out Aug 16 18:06:38 Crofton|work, I think its call 'Sucker' ; ) Aug 16 18:25:44 armpit, did not give away our secrets Aug 16 20:34:37 armpit: can http://lists.openembedded.org/pipermail/openembedded-core/2019-June/283876.html be pack-ported to warrior please? Aug 16 20:52:45 tlwoerner, so are you seeing this issue on warrior? Aug 16 20:54:11 armpit: yes, if i set my kernel to linux-dummy, the build errors out complaining that make-mod-scripts depends on a non-existent job Aug 16 20:56:13 but i don't think this patch fully fixes the problem. eventually the build fails anyway because make-mod-scripts, when it does run, looks for a kernel source directory, of which there isn't one Aug 16 21:14:47 i'm trying to figure out why make-mod-scripts runs at all. if i don't have a kernel, obviously i don't have kernel modules either Aug 17 02:31:56 * armpit hmm, 5.4 looks to be the next lts Aug 17 02:32:45 * armpit kernel **** ENDING LOGGING AT Sat Aug 17 03:00:04 2019