**** BEGIN LOGGING AT Fri Nov 17 03:00:03 2017 Nov 17 08:22:26 Any tips on how to slim down an SDK? Mine is 4.2GB extracted now, and I see that /usr/lib/.debug is 1.2GiB Nov 17 08:22:31 Can I skip that one? Nov 17 08:23:45 I'd like to skip /usr/src and the debug packages, as long as I can still build stuff. Nov 17 08:27:06 tasslehoff: have you checked how that stuff gets pulled in? maybe http://www.yoctoproject.org/docs/2.1/ref-manual/ref-manual.html#var-SDKIMAGE_FEATURES Nov 17 08:28:10 LetoThe2nd: currently checking. I have a meta-toolchain recipe that requires meta-toolchain-qt5-bb Nov 17 08:28:30 tasslehoff: ah qt stuff? i'm out then, totally no experience there. Nov 17 08:31:30 LetoThe2nd: hehe. not a Qt-question I think. I seem to have .debug variants of all my libs Nov 17 08:31:54 and src packages for everything as well Nov 17 08:33:33 LetoThe2nd: but you nailed it. -eee shows SDKIMAGE_FEATURES="dev-pkgs dbg-pkgs " Nov 17 08:37:11 * LetoThe2nd celebrates! https://www.youtube.com/watch?v=3GwjfUFyY6M Nov 17 08:37:49 deserved :) Nov 17 08:41:38 so I can likely set SDKIMAGE_FEATURES just to dev-pkgs, but that won't get rid of the src packages? Nov 17 08:44:07 no idea :-P Nov 17 10:30:29 tasslehoff: src files are from dbg-pkgs Nov 17 10:30:36 (why would you need upstream source for dev?) Nov 17 10:30:55 src files are for eg source listing in gdb Nov 17 10:51:25 rburton: thanks. I just figured maybe something else dragged in src. all good then. Nov 17 17:34:41 what's missing in my image if I get a device node for a usb drive if it is plugged in at boot but not if I plug it in if my system is already running Nov 17 18:18:14 but if I execute lsusb -v the device is detected Nov 17 22:22:02 armpit: will you be updating patchwork to reflect the patch status as well Nov 17 22:22:27 armpit: I see that the patches have landed in master but they are not updated state on pw Nov 17 22:41:25 for reference usb autosuspend was at fault. disabling it with echo "on" > /sys/bus/usb/devices/1-1/power/control fixes things... Nov 17 22:46:40 khem, hi there. About the -klibc failures for arm64, I'll try to update mtd-utils but for kexec-tools the arm64 support is very recent, I'd say backport is not worth Nov 17 22:47:27 COMPATIBLE_HOST_aarch64 = "null" would fix that Nov 17 22:50:39 ant_home: yes that would be fine I think Nov 17 22:50:55 ...or both recipes...in fact afaik these recipes are only used by initramfs-kexcboot-klibc-image which is almost surely used only by me ;) Nov 17 22:51:15 me = armv4 / armv5 Nov 17 22:51:50 its fine either way, I am sure its useful on armv8 but probably you are right about users Nov 17 22:52:15 now and then I test the x86 version ;) Nov 17 22:53:00 I would have put all in meta-handheld but apparently someone out there is using klibc Nov 17 22:53:57 once I sent the proposal I got immediately protests Nov 17 22:54:04 years ago... Nov 17 22:55:47 khem, but if klibc goes there it exits from your radar :) Nov 17 22:55:52 so no no Nov 17 22:57:58 khem, btw, would it make sense to add a musl-static optimized class to try to replace klibc? Nov 17 23:00:14 ant_home: yes using musl static would be great Nov 17 23:07:48 ok, I've almost finished my kernel journey, back to userland soon ;) Nov 17 23:27:10 gn **** ENDING LOGGING AT Sat Nov 18 03:00:01 2017