**** BEGIN LOGGING AT Sat Feb 12 02:59:57 2022 Feb 12 17:31:22 RP: have you seen any races in master staging sysroot recently? Feb 12 17:36:13 RP: looks like qemu-native tried linking zlib.h into recipe-sysroot-native before that was available - https://pastebin.com/03Zb01JX Feb 12 17:36:50 just wanted to check if it's known or if I need to open a bugzilla ticket Feb 12 17:41:40 denix: no, not known Feb 12 17:41:55 denix: I'd be interested in the console log for that, see what bits of zlib had run Feb 12 17:59:50 RP: unfortunately, it was a manual/knotty build with -k flag passed Feb 12 18:00:47 denix: doesn't matter, I'd just be interested to see the console log Feb 12 18:01:01 denix: should be saved in log/cooker/MACHINE/xxx Feb 12 18:01:26 denix: the zlib-native bits vs the qemu-native ones Feb 12 18:03:34 got it, one sec Feb 12 18:07:10 RP: https://pastebin.com/TwkusG9J Feb 12 18:50:05 denix: that is very revealing. Why did it run populate_sysroot_setscene after having run populate_sysroot for zlib-native Feb 12 18:50:32 denix: that is the real issue, some kind of runqueue problem I expect :/ Feb 12 18:50:43 denix: how to reproduce and debug though... :/ Feb 12 18:50:54 denix: please do put that in a bug report Feb 12 18:52:00 RP: ok, will do Feb 12 18:55:11 RP: and it ran do_populate_lic 3 times for zlib-native. is it due to "unihash changed to" messages? Feb 12 19:53:29 denix: yes, it is something related to the hash equiv rehashing Feb 13 01:17:30 RP: seems related - https://pastebin.com/QVqbDCtb **** ENDING LOGGING AT Sun Feb 13 02:59:56 2022