**** BEGIN LOGGING AT Sun May 30 02:59:57 2021
May 30 09:12:35 <kanavin> I was wondering why so quiet.
May 30 10:58:08 <RP> kanavin: holiday weekend maybe?
May 30 10:59:28 <kanavin> RP: maybe. Can I ask you, is this the right way to deal with buildhistory races? http://git.yoctoproject.org/cgit.cgi/poky-contrib/commit/?h=akanavin/package-version-updates&id=85da3f88f0d4a20d828ed0e03b40da6ae98c2401
May 30 11:27:25 <RP> kanavin: that looks horrible from a dependency point of view :(
May 30 11:27:34 <RP> kanavin: I saw your email but haven't read into the issue yet
May 30 11:28:17 <RP> kanavin: packagedata should not be poking into package-split :(
May 30 11:39:15 <RP> kanavin: Two possible fixes are to rewrite buildhistory to be writing sstate output or to split buildhistory_list_pkg_files into a separate function hooked on do_package instead of packagedata
May 30 11:39:26 * RP suspects a bigger rewrite might be helpful :/
May 30 11:43:58 <RP> kanavin: to be clear, do_packagedata should be self sufficient, it shouldn't ever be looking at the do_package task output once the task is complete. The code is therefore just buggy and shouldn't be doing what it is doing :(
May 30 13:58:16 <kanavin> RP: if this keep thwarting auh, I might disable buildhistory in it for now
May 30 14:16:23 <RP> kanavin: how often have we seen it?
May 30 14:18:38 <kanavin> RP: I saw it for the first time in the very last auh run
May 30 14:18:48 <kanavin> RP: but it did happen for 4 different recipes in there
May 30 14:19:22 <kanavin> RP: all of do_compile fails except cmake
May 30 14:20:48 <kanavin> RP: ah, it happened in previous auh too
May 30 14:22:57 <RP> kanavin: I'll have to try and fix it, I think there are probably related open bugs too :/
May 30 20:29:27 <manuel1985> kanavin: This channel moved to libera.chat.
**** ENDING LOGGING AT Mon May 31 02:59:56 2021