**** BEGIN LOGGING AT Wed Jun 22 02:59:58 2016 Jun 22 04:26:29 kergoth: yes, tons of them Jun 22 07:52:35 morning **** BEGIN LOGGING AT Wed Jun 22 23:20:46 2016 **** BEGIN LOGGING AT Thu Jun 23 00:23:33 2016 Jun 23 00:56:02 anyone else occasionally seeing bitbake re-running a ton of setscenes it really shouldn't have run, when nothing changed for those tasks? i can verify the checksums didn't change, it just re-ran the setscenes Jun 23 01:39:07 kergoth: havent seen but dont have enough builds going lately either Jun 23 01:39:34 kergoth: I do see the task hash mismatch errors a lot though especially on rpi2 Jun 23 01:41:12 i've seen some of those too. i ran into it with do_populate_lic in gcc-runtime-external in meta-sourcery for no apparent reason. changing it to remove GPLv3 made the mismatches go away. not a clue. Jun 23 01:43:56 I wonder if its due to use of TIME Jun 23 01:43:59 or something like that Jun 23 01:44:13 which changes between stamp and compare operations Jun 23 01:54:40 khem: i think that usually results in a basehash mismatch rather than a taskhash mismatch. i *think* a taskhash mismatch might be due to a task dependency on a task with a basehash mismatch, but i'm not seeing any basehash mismatches Jun 23 01:54:43 * kergoth shrugs Jun 23 01:54:50 also not seeing any uses of date or time for the affected tasks Jun 23 01:55:59 it will annoy me enough one day and I Will chase it out, so far so good Jun 23 02:00:15 kergoth: errors I See are like this http://hastebin.com/josusilopu.avrasm Jun 23 02:01:28 * kergoth nods Jun 23 02:08:17 it's like I said to Richard... these errors aren't very helpful Jun 23 02:09:05 if we can't figure out what they mean, that's a bit of a problem Jun 23 02:09:28 yeah, without the details, it doesn't really tell us anything. we need the sigdata, or at least better messages if nothing else Jun 23 02:09:28 I understand that it's not trivial to determine the cause but it's not impossible Jun 23 02:10:11 if even those of us with something of a clue about what's going on can't figure it out, the average user is just screwed :) Jun 23 02:13:46 indeed Jun 23 02:19:02 what disturbs me as well is that it seems like it only sometimes happens, if you run the build again it's fine Jun 23 02:22:02 i suspect that's why it's hard to diagnose, because the taskhash mismatchi s due to a basehash mismatch that was only shown once and vanished thereafter, but that's just a wild guess Jun 23 02:22:14 * kergoth shrugs Jun 23 02:22:24 guess we should open a bug in bugzilla if one isn't already open Jun 23 02:25:37 i should revert my meta-sourcery fix pull req and do a --no-setscene build and see if anything more useful shows up Jun 23 02:29:36 there it goes again, i cleaned and rebuilt systemtap and it's re-running setscenes for python3, gcc-runtime-external, etc **** ENDING LOGGING AT Thu Jun 23 02:59:57 2016