16:01:29 #startmeeting F32-blocker-review 16:01:29 Meeting started Mon Apr 20 16:01:29 2020 UTC. 16:01:29 This meeting is logged and archived in a public location. 16:01:29 The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01:29 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:01:29 The meeting name has been set to 'f32-blocker-review' 16:01:29 #meetingname F32-blocker-review 16:01:29 #topic Roll Call 16:01:29 The meeting name has been set to 'f32-blocker-review' 16:01:45 Hello all! Good day! :-D 16:01:46 morning folks, who's around for blocekr fun 16:01:47 .hello2 16:01:48 frantisekz: frantisekz 'František Zatloukal' 16:01:50 * kparal is here 16:02:12 .fas lailah 16:02:16 regenbogen: lailah 'Sylvia Sánchez' 16:03:01 .hello2 16:03:02 sgallagh: sgallagh 'Stephen Gallagher' 16:03:20 Why am I not showing up with my actual nick? 16:04:21 regenbogen: i don't know. it may be that when you signed in, lailah was taken (by someone else or just by a dropped connection of yours)? 16:04:31 depends on your client and stuf. 16:04:42 Konversation 16:04:49 It's KDE's default 16:05:06 don't know much about it 16:05:29 Now? 16:05:36 Meh... 16:05:36 .hello2 16:05:37 lruzicka: lruzicka 'Lukáš Růžička' 16:06:19 make that /me 16:06:48 And now? 16:07:01 * pwhalen is here 16:07:03 * satellit_ listening 16:07:06 .hello2 16:07:07 bcotton: bcotton 'Ben Cotton' 16:07:20 .fas lailah 16:07:21 Lailah: lailah 'Sylvia Sánchez' 16:07:24 Lailah: hey, you're back 16:07:25 :P 16:07:28 Yeah! 16:07:38 I had to log out and in again 16:08:07 .hello mohanboddu 16:08:08 mboddu: mohanboddu 'Mohan Boddu' 16:09:12 hi everyone 16:09:18 hi, adamw! 16:09:20 hi adamw 16:09:24 coremodule: are you on secretary duty? 16:09:25 Hiiiii!!! 16:09:32 you got it! 16:09:53 call me terry, and keep it a secret! 16:09:54 #chair bcotton mboddu 16:09:54 Current chairs: adamw bcotton mboddu 16:10:03 shhhhh terry it's supposed to be a SECRET 16:10:10 dammit, sorry 16:10:13 impending boilerplate alert! 16:10:14 #topic Introduction 16:10:14 Why are we here? 16:10:14 #info Our purpose in this meeting is to review proposed blocker and nice-to-have bugs and decide whether to accept them, and to monitor the progress of fixing existing accepted blocker and nice-to-have bugs. 16:10:16 #info We'll be following the process outlined at: 16:10:17 #link https://fedoraproject.org/wiki/QA:SOP_Blocker_Bug_Meeting 16:10:19 #info The bugs up for review today are available at: 16:10:19 #link http://qa.fedoraproject.org/blockerbugs/current 16:10:21 #info The criteria for release blocking bugs can be found at: 16:10:23 #link https://fedoraproject.org/wiki/Basic_Release_Criteria 16:10:25 #link https://fedoraproject.org/wiki/Fedora_32_Beta_Release_Criteria 16:10:27 #link https://fedoraproject.org/wiki/Fedora_32_Final_Release_Criteria 16:10:31 #info for Final, we have: 16:10:33 #info 1 Proposed Blockers 16:10:35 #info 2 Accepted Blockers 16:10:37 #info 5 Proposed Freeze Exceptions 16:10:39 #info 4 Accepted Freeze Exceptions 16:12:07 #info coremodule will secretarialize 16:12:15 #topic Proposed Final blockers 16:12:23 #topic (1825046) NVIDIA Turing GPU "secure boot" is broken and can lead to full system lockups 16:12:23 #link https://bugzilla.redhat.com/show_bug.cgi?id=1825046 16:12:23 #info Proposed Blocker, kernel, ON_QA 16:12:28 (seems like the blockerbugs/current link should be shared more widely) 16:13:02 cyberpear: it's in all the meeting announcement mails, i think 16:13:13 so... I think this should go to F32, but as a FE 16:13:24 and the candidate announcements 16:13:30 it's HW enablement for single GPU generation 16:13:48 yeah, I'm easily FE on this, blocker justification isn't clear. 16:13:53 I don't remember when we last blocked on nouveau 16:14:01 everyone please ignore the shadowy things happening behind curtains here thank you 16:14:07 given its state and nvidia zero help 16:14:10 It is pending stable push now, so the fix is there 16:14:10 nothing to see here, citizens, return to your homes 16:14:16 I haven't tested the change yet, but the patch also touches Pascal-gen code, it looks like 16:14:23 * Lailah cleaning glasses 16:14:29 Eye glasses 16:14:38 * Lailah in mole mode 16:14:40 definitely +1 FE here 16:14:44 yeah, we should definitely let this in 16:14:50 +1 FE as I said in the ticket. 16:14:50 i was +1 FE in the BZ and remain so 16:14:54 I'm also -1 blocker 16:15:14 +1FE 16:15:22 At this point, with the fix in pending stable, does it matter? 16:15:26 -1 blocker, too, because it's not clear there's a criterion we could stretch 16:15:30 jforbes: it does if the fix doesn't work 16:15:42 or if the kernel update borks other stuff 16:15:56 if the bug is only an FE we can then drop the kernel to an older version and ship 16:15:57 jforbes: only if the fix fails to work properly 16:16:06 if it's a blocker we're committed to fixing it 16:16:12 fair 16:16:32 -1 blocker from me as well 16:16:38 definitely +1FE as well 16:16:57 if the kernel changes other nvidia gen codepaths it'd be good to have confirmation it still works on them... 16:16:57 -1FE 16:17:03 i don't immediately see any such feedback in the bug 16:17:12 (decathorpe tested with proprietary driver) 16:17:13 +1FE, sorry about the typo 16:17:33 * Lailah has sight again :-) 16:17:46 -1B/+1FE 16:17:50 proposed #agreed 1825046 - RejectedBlocker (Final) AcceptedFreezeException (Final) - we think the impact here is too narrow to qualify as blocker (it only affects a very new NVIDIA hardware generation), but certainly significant enough to accept as an FE 16:17:56 ack 16:17:59 ack 16:18:00 ack 16:18:01 ack 16:18:04 ack 16:18:06 ack 16:18:11 lruzicka, can you try your P51 with that kernel? I think it's even older than Pascal, but it'd be nice to be sure 16:18:12 ack 16:18:20 ack 16:18:23 also, coremodule, your P53 testing would be nice 16:18:31 ack 16:18:49 frantisekz, can do 16:18:53 frantisekz, I will try 16:18:55 thanks! :) 16:18:55 #agreed 1825046 - RejectedBlocker (Final) AcceptedFreezeException (Final) - we think the impact here is too narrow to qualify as blocker (it only affects a very new NVIDIA hardware generation), but certainly significant enough to accept as an FE 16:20:10 #topic Proposed Final freeze exceptions 16:20:14 #info moving onto proposed FEs 16:20:34 skipping the backgrounds one as it's an accepted blocker anyhow 16:20:34 #topic (1824075) File Roller missing in the default Silverblue 32 installation 16:20:34 #link https://bugzilla.redhat.com/show_bug.cgi?id=1824075 16:20:34 #info Proposed Freeze Exceptions, file-roller, POST 16:20:56 +1 FE, could be fixed by an update, but it'd nice to have on GA release 16:21:03 +1 FE 16:21:15 +1FE 16:21:15 sure, change is very restricted to SB so it can't break anything blocking 16:21:17 +1 FE 16:21:18 frantisekz: It could not be fixed by an update 16:21:20 +1 FE 16:21:23 frantisekz: it can't be...right 16:21:28 +1 FE 16:21:30 still +1 FE 16:21:35 ehm... okay, didn't know that 16:21:49 +1 FE 16:22:04 proposed #agreed 1824075 - AcceptedFreezeException (Final) - this is a desired change to the default SB app loadout and can't be fixed with a post-release update, change is very limited and cannot break anything else 16:22:08 s/limited/targeted/ 16:22:14 why can't this be fixed with an update, btw? 16:22:16 ack 16:22:27 ack 16:22:29 ack 16:22:31 ack 16:22:33 ack 16:22:39 ack 16:22:55 kparal: we can't really add apps to the 'default set' post-release 16:22:57 kparal, because it needs to be on the image so that anaconda could install it? just gessing 16:23:10 kparal: https://bugzilla.redhat.com/show_bug.cgi?id=1824075#c6 16:23:31 ok 16:23:40 #agreed 1824075 - AcceptedFreezeException (Final) - this is a desired change to the default SB app loadout and can't be fixed with a post-release update, change is very targeted and cannot break anything else 16:23:52 #topic (1819843) libgit2-1.0.0 is available 16:23:53 #link https://bugzilla.redhat.com/show_bug.cgi?id=1819843 16:23:53 #info Proposed Freeze Exceptions, libgit2, MODIFIED 16:24:03 still -1 FE 16:24:04 -1, as voted in bug 16:24:08 -1 FE 16:24:09 #link https://pagure.io/fesco/issue/2377 16:24:33 #info igor is asking FESCo for approval to do a soname bump post-release 16:24:55 I don't think this is worth a FE 16:25:02 -1 FE 16:25:03 -1FE 16:25:19 0 16:25:19 well, but it seems to work ok 16:25:34 would not that be a good thing to have it? 16:25:38 we don't just pull things through freeze because they "seem to work ok" :P 16:25:46 it has to have some actual benefit 16:25:52 that is worth breaking a freeze for 16:26:04 adamw, isn't a newer version of software a benefit? 16:26:12 doesn't have to be lruzicka 16:26:15 no 16:26:22 i mean, follow the logic there 16:26:23 You can get it with an update lruzicka 16:26:34 Lailah, I alredy got it, so I am fine :D 16:26:38 a freeze is specifically about *not* just pulling in "newer shinier" everything 16:26:52 if "it's newer and shinier" is a legitimate reason for a freeze exception, what's the point of a freeze at all? 16:27:21 Bureaucracy for its own sake? 16:27:31 proposed #agreed 1819843 - RejectedFreezeException (Final) - there is no significant benefit to be gained by a freeze exception for this. it would be safer to do it as a post-release update 16:27:37 ack 16:27:37 sgallagh: as much as you know i love bureaucracy...:P 16:27:43 adamw: I thought "newer and shinier" was the very reason of Fedora... :-P 16:27:46 ack 16:27:47 ack 16:27:51 acl 16:27:55 ack 16:28:00 ack (because of the majority) 16:28:00 #agreed 1819843 - RejectedFreezeException (Final) - there is no significant benefit to be gained by a freeze exception for this. it would be safer to do it as a post-release update 16:28:16 #topic (1823678) Tiled (usually 5k) monitors doesn't work properly 16:28:16 #link https://bugzilla.redhat.com/show_bug.cgi?id=1823678 16:28:16 #info Proposed Freeze Exceptions, mutter, POST 16:28:51 +1 FE, but we'll need them to submit an update soon 16:28:52 +1FE, so that the shiny new 5k monitors shine 16:28:53 i'll poke the team 16:28:53 i was -1 in the BZ becuase it looked like the fix was still in the process. that's apparently not the case 16:29:16 +1 FE 16:29:33 I'm not really in favour of this being a FE 16:29:34 i'm not sure there are a lot of people using live images on a 5k monitor, but i guess i don't see the harm 16:29:40 -1FE from me 16:29:41 0 FE for me 16:29:46 why not, Lailah? 16:30:01 +1 FE 16:30:11 How many people use live with tiling on 5k ? 16:30:24 bcotton, what if they want to install Fedora Workstation? 16:31:02 Lailah, all those who have such monitor and want to install Workstation? 16:31:05 right 16:31:08 Lailah, how would you do it? 16:31:18 Lailah... more every day :) 16:31:41 I wouldn't think of trying tiling TBH 16:32:15 Lailah, the monitor is tiling, but it can display a huge picture on all those tiles, I believe 16:32:16 When I try a live ISO, I'm more concerned about drivers and main software working properly, not having system crashes. 16:32:31 Lailah, and therefore this FE goes in there :D 16:32:48 Lailah, because you do not want your system to behave badly on such monitors 16:32:49 Well, I'm not saying it's wrong, and y'all already voted. 16:34:19 Lailah: i think you may be misunderstanding 'tiled' here 16:34:32 this isn't about the window manager feature 'tiling' where it puts all your windows into a tile arrangement 16:34:52 And what is it then? 16:35:00 it's a hardware thing: 5K displays are such huge resolutions that the display manager has to do some fancy trickery to actually render to them 16:35:19 (this used to be the case for 4K as well, before HDMI and Displayport standards caught up) 16:35:39 uhm... okay... 16:36:03 you have to connect the monitor to the system with two cables 16:36:14 and the system has to send output as if it were two monitors side-by-side 16:36:42 https://www.unixtutorial.org/project-connect-lg-5k-display-to-pc explains it in a bit more detail 16:37:09 so the practical result of this bug is just: if you're using one of these displays, you're missing half the screen and you can't really use it. 16:37:35 * adamw counts votes 16:37:39 Oh 16:37:44 Okay, now I understand 16:37:53 I changed my mind adamw 16:37:58 +1 FE 16:38:04 roger! 16:38:44 proposed #agreed 1823678 - AcceptedFreezeException (Final) - this is a significant problem for affected displays and can't be fully fixed with a post-release update as the lives will still have the bug 16:38:49 ack 16:38:52 ack 16:38:53 ack 16:38:58 #agreed 1823678 - AcceptedFreezeException (Final) - this is a significant problem for affected displays and can't be fully fixed with a post-release update as the lives will still have the bug 16:39:35 ack 16:39:39 ack 16:39:40 #topic (1824947) MOTD missing from zezere-ignition 16:39:40 #link https://bugzilla.redhat.com/show_bug.cgi?id=1824947 16:39:40 #info Proposed Freeze Exceptions, zezere, MODIFIED 16:39:53 +1fe 16:39:59 +1 FE 16:40:00 +1 FE 16:40:38 No idea what is zezere 16:40:43 Sorry! 16:40:56 so what's the impact here? can you not provision the device without the info in the MOTD? 16:40:57 it's a thing that runs on the IoT images when they boot 16:41:10 i think it's a convenience thing, but it could be fairly darn inconvenient if you can't discover the info any other way i guess 16:41:45 When an IoT system is booted, it shows the url and mac address to use with the provisioning server. It will only affect IoT 16:41:52 bcotton, you can, but you have to figure out what the IP is by either plugging in a monitor to the IoT device or using a tool to scan your network for devices (ie nmap) 16:42:04 okay, seems annoying enough 16:42:06 +1 fe 16:42:13 for a lot of IoT devices, the monitor thing doesn't make a lot of sense 16:42:33 monitor/serial console 16:43:57 proposed #agreed 1824947 - AcceptedFreezeException (Final) - this is a significant quality-of-life improvement for IoT that cannot be done post-release and is very targeted as nothing else uses zezere 16:44:05 ack 16:44:07 ack 16:44:09 ack 16:44:09 ack 16:44:11 ack 16:44:19 #agreed 1824947 - AcceptedFreezeException (Final) - this is a significant quality-of-life improvement for IoT that cannot be done post-release and is very targeted as nothing else uses zezere 16:44:20 ack 16:44:32 .fire lruzicka cause late ack 16:44:32 adamw fires lruzicka cause late ack 16:44:46 #topic Accepted blockers 16:44:51 #info moving onto review of accepted blockers 16:44:59 #topic (1824216) final f32-backgrounds version isn't in stable repo 16:44:59 #link https://bugzilla.redhat.com/show_bug.cgi?id=1824216 16:44:59 #info Accepted Blocker, f32-backgrounds, MODIFIED 16:45:04 adamw, thats not fair, pwhalen and Lailah werent fired last time 16:45:12 whoever said life was fair 16:45:41 so we have a lot of chaos in the ticket here, but bottom line, we have images and whenever this gets to be the last blocker we can throw together and ship SOMETHING 16:45:42 heh :) 16:45:44 +1FE, when we slipped partly for this 16:45:49 this isn't FE review any more 16:45:53 it's accepted blocker review :) 16:46:13 i am trying to find out from mizmo whether to expect updated images or not 16:46:14 ho ho ho ho, santa claus is coming to town 16:47:02 well that explains why i'm cold 16:47:30 lruzicka: are we having a mini-winter in April? 16:47:57 Lailah, I longed for that lovely piece of laughter 16:48:17 #info there's a lot of activity in the bug here and uncertainty about whether and how to switch to the animated background by default, but bottom line all the pieces are there and we can definitely ship *something* even if it means provenpackager Executive Decisions 16:48:41 #info mizmo may be working on tweaked images to address some of the concerns expressed on devel@ and in the ticket, adamw will clarify that before requesting candidates 16:50:15 any other notes here? 16:50:32 I don't think so... 16:50:44 alrighty 16:50:53 But then what happens to this last bug? 16:51:36 well, the big one is this one 16:51:37 #topic (1824418) Rescue Mode doesn't recognize LVM -Partitions 16:51:37 #link https://bugzilla.redhat.com/show_bug.cgi?id=1824418 16:51:38 #info Accepted Blocker, python-blivet, NEW 16:51:43 this is the thing we're really blocked on 16:51:47 oh... 16:52:12 so, i was hoping we'd get more movement on it over the weekend / monday daytime European time, but, it's still broken :P 16:52:28 vtrefny and I have been poking it for the last couple hours 16:52:32 over in #anaconda 16:52:59 i am currently running a build of a netinst image with my tweak to invalidate the LVsInfo cache, to see if we can test that on the HW reproducers 16:53:26 who had hw reproducers, again? 16:53:36 kparal, you did right? 16:53:42 I did on some of the machines. 16:53:58 I have a bare metal reproducer, yes 16:53:58 cool 16:54:02 i have one too 16:54:10 i'll put that image up when it's done 16:55:44 adamw: should be soon-ish, right? 16:55:47 yeah 16:55:53 ok, will try to test it asap 16:55:53 say 15-20 mins 16:56:06 (gotta wait for the openqa job to run then i have to move the image elsewhere because Reasons) 16:56:33 .fire Reasons, no reason 16:56:33 adamw fires Reasons, no reason 16:56:39 :PP 16:56:45 LOL 16:57:08 that's how you deal with Reasons 16:59:46 #info we are still investigating this. we have a hack that may possibly be sufficient, but there are still a lot of things we don't entirely know yet 16:59:53 kparal: unreasonably? 17:00:37 adamw, where will be the link for that new image? 17:01:58 lruzicka: probably happyassassin 17:02:04 oo, it's done 17:02:58 I can do an emergency Anaconda build if changes in Anaconda are needed 17:03:18 mkolman: thanks 17:03:52 adamw: is there any place some gentle pressure would help, or is it best if i stay out of people's way so they can figure things out 17:04:11 adamw, so send it my way, when it's done 17:04:18 thanks 17:05:37 adamw: Can I go to make a cup of tea meantime? 17:08:17 bcotton: i think 'stay out of the way' right now 17:08:19 Lailah: yes 17:08:30 sorry, i am talking to three people about the lvm bug and one about backgrounds, it got a bit busy :) 17:08:35 contrary to popular opinion, i can do that :-) 17:08:57 #info adamw is talking to various folks about this bug, we'll try to get a fix ASAP 17:09:25 kparal: lruzicka: frantisekz: https://www.happyassassin.net/temp/00822949-43567473-netinst-x86_64.iso 17:09:28 bcotton: so popular belief has it that you're always in someone's way? 17:09:30 that should be a netinst with my hacked blivet in it 17:09:39 Lailah: pretty much ;-) 17:10:15 adamw, thanks 17:10:55 adamw, do I install from this first or can I just use to rescue the installed system? 17:11:11 lruzicka: no need to install, just try the reproducer 17:11:17 lruzicka: I'd say you can try rescue from this 17:11:19 and post the syslog either way 17:11:47 thanks 17:14:41 bcotton: Glad to know I shouldn't believe popular belief :-P 17:16:01 it's slow to download :/ 17:16:22 30-50min to download it 17:16:39 I guess we shouldn't stall the meeting and wait 17:16:43 How big is it? 17:16:49 700M 17:17:01 Uhm... a bit slow. 17:17:02 kparal: i can maybe shuffle it somewhere else on fedoraproject.org if that would be faster... 17:17:20 #topic Open floor 17:17:24 oh, wait 17:17:27 #undo 17:17:27 Removing item from minutes: 17:17:43 #topic (1824216) final f32-backgrounds version isn't in stable repo 17:17:43 #link https://bugzilla.redhat.com/show_bug.cgi?id=1824216 17:17:43 #info Accepted Blocker, f32-backgrounds, MODIFIED 17:18:04 #info updated info on this one: mizmo has decided not to change the backgrounds again, so we will ship the existing 'final' versions 17:18:18 we just need to thrash out the packaging questions and get an update done. 17:18:28 #topic Open floor 17:19:21 Is there anything for Open Floor? 17:19:34 adamw: ^^ Will be requesting the update stable or just include it in the compose? 17:19:48 * mboddu planning to run the big push now 17:21:06 mboddu: i think it may be wise to just pull in a lot of the late-breaking stuff to the compose 17:21:11 so we can back it out again easy if we change our minds 17:21:36 adamw: Okay, it makes sense and helps me as well :) 17:21:39 Thanks 17:29:30 adamw: it works! 17:30:07 Yay! 17:31:10 don't celebrate yet, it's only a hack :D 17:31:16 really for data 17:31:19 Ah... 17:31:29 what do you need? apart from journal? 17:31:36 okay, i'm gonna close the meeting out. we're working on hte lvm bug in #anaconda 17:31:41 frantisekz: just the journal output 17:31:48 so please join over there to follow that one 17:31:52 thanks for coming, everyone 17:32:00 coremodule: if you can get the secretarialization done soon that'd be great 17:32:03 #endmeeting