19:00:15 #startmeeting 19:00:15 Meeting started Wed Aug 14 19:00:15 2024 UTC. The chair is carnil. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:00:15 Useful Commands: #action #agreed #help #info #idea #link #topic. 19:00:31 hello, anybody around for the meeting? 19:00:49 i am 19:01:16 \o/ 19:01:54 I have not much more than the usual going trough the open bug list, so let's start (maybe bwh is joining, some will concern him, e.g. initramfs-tools) 19:02:26 sounds good 19:02:27 #chair carnil waldi 19:02:27 Current chairs: carnil waldi 19:03:01 #topic initramfs-tools autopkgtests failures 19:03:35 I wanted to ask on this actually bwh if he knows if there was some progress of Benjamin needs to be approached who has written those autopkgtests TTBOMK 19:03:46 Hi, sorry, got carried away reading a bug report 19:03:46 if bwh is not around we can skip this 19:03:51 ah good hello 19:04:34 I have not got anything done this week, sorry 19:04:50 but we do have a bug report highlighting what went wrong in the initramfs-tools autopkgtest 19:05:01 so I think I know how to fix it 19:05:04 bwh: I guess you are fine if we define again an action to have a look assigned to you? 19:05:08 yes 19:05:40 #action bwh will have a look at the initramfs-tools autopkgtest failures preventing testing migration 19:06:19 there are two longer standing bugs to quickly look at: 19:06:34 #topic #1072004: linux: regression in the 9p protocol in 6.8 breaks autopkgtest qemu jobs (affecting debci). 19:07:05 I just updated that report - the fix is now upstream 19:08:01 ok pefect, so that will likely flow automatically in the 6.11-rcX upload to experimental and into 6.10.y stable series as well 19:08:17 Yes 19:08:34 #topic #1050256: AppArmor breaks locking non-fs Unix sockets. 19:09:06 for this just a status update: I pinged as well offlist again John. He mentioned "patches will be submitted for stable this week" 19:09:22 so looking forward that the solution will flow in the next 6.1.106 update. 19:09:50 I'm not very in favour of picking one of his proposed approaches and not waiting for this stable submission. 19:10:00 OK 19:10:19 Given the report is now almost one year old, I assume we continue to wait and then have it at some point. Or someone prefers to still try to push that more? 19:10:50 No, let's wait for it to go through stable 19:11:20 I will continue monitoring it and then close the bug once it hits 6.1.y for real and included in an update 19:11:36 ok now to the longer list of bugs which are current 19:12:08 #topic #1063754 fat-modules: SD corruption upon opening file on Linux desktop 19:12:23 I'm still composing a reply to this one 19:13:06 so, leave it with me 19:13:35 ok, I was asking bcause the last reply from the reporter seems he got upset 19:14:01 #action bwh is taking care of writing to the followup of the reporter of #1063754 19:14:29 #topic #1076372 linux-image-6.5.0-0.deb12.4-amd64: ext4 file corruption with newer kernels 19:15:00 Just looking at the last message, we now have a test case but we also know that it's hardware-dependent 19:15:29 this was paritally with me last week. I followed up to the bug as we discussed last time (but I was late). Stefan provided more information on what it already did. He still wants to test other HW combinations. 19:15:35 bwh: correct 19:16:01 so here I would say we have no further actions for now and wait another round for Stefans. if nothing happens at the next meeting we might just do a re-ping. 19:16:10 I agree 19:17:09 #topic #1078168 initramfs-tools: failing autopkgtest prevents testing migration 19:17:17 this is the one discussed above, so 19:17:43 #action bwh takes care of handling #1078168: initramfs-tools: failing autopkgtest prevents testing migration 19:18:03 I skip the next one as we discussed it now already 19:18:15 #topic #1078702 Please re-add the AHCI driver in the cloud kernel 19:18:34 waldi: I think this is best in your hands. I tried to get an idea why zigo says "re-add" 19:18:43 but it was never added in the cloud images AFAIU 19:18:52 but maybe I'm wrong. 19:18:59 it was never 19:19:07 Is it correct that this is out of scope for the images and can we reply so? 19:19:37 yes. i'll respond 19:20:04 #action waldi will respond to the #1078702 bug requesting addition of AHCI driver in the cloud kernel 19:20:07 thanks waldi 19:20:53 topic #1062421 linux-image-6.5.0-5-686-pae: [iwlegacy] kernel oops 19:20:59 #topic #1062421 linux-image-6.5.0-5-686-pae: [iwlegacy] kernel oops 19:21:31 I said I would investigate this and I haven't yet, so give me that as an action again 19:21:38 Nothing new here in the upstream report, just Thorsten Leemhuis responded with hints how to report the regression 19:22:04 #action bwh will investigate the #1062421 issue 19:22:28 #topic #1078030 lpfc: lost all san paths 19:22:44 waldi: we had last time as action that you might be a good one to ask the correct questions 19:23:10 do you have time for it? Or should we try to redirect the reporter somewhere more specific? Any other ideas? 19:23:14 yes, i did not manage to do that 19:24:33 keep the action, or would you feel more to put it on hold? 19:24:49 i'll keep the action 19:25:39 #action waldi might figure out the questions to ask for tackling #1078030 and takes care to repond to the bug followup 19:26:25 #topic #1013211 System freeze followed by kernel panic or blank screen 19:27:31 an update on this one: this was on the list of bugs I did close as housekeeping, the reporter quickly responded back that the problem persist. I suggested to report the problem upstream. It seems reproducible under other distro kernels as well. 19:27:56 so propose no action, but as a general note it seems worth to try to close older bugs and see to which reporters still respond :) 19:28:26 Right. This seems like some weird hardware-specific problem that we can't debug 19:29:02 #topic #1070159 i915: CPU usage spikes with monitor powered down or unplugged 19:29:18 same story here: but in this case the reporter did already reported the issue upstream 19:29:26 That got some response upstream but no fix proposed yet 19:30:11 correct, reference: https://gitlab.freedesktop.org/drm/i915/kernel/-/issues/11919#note_2526367 19:30:36 -> no action for now 19:30:59 right 19:31:14 #topic #1077827 even with [[featureset]] name = 'rt' enable = false in config.local/defines.toml, linux-headers-*-common-rt is still getting built 19:31:20 I think we can close this? 19:31:26 yes 19:31:35 seems i forgot to close this 19:31:41 waldi: Late to the party and catching up, but for 1078030 (Emulex LightPulse) could be an HBA firmware version; there is a known workaround for frame drop issues similar to that shown in the log 19:32:07 #action close the bug as this not a Debian bug per se and modifications are pushed to the packaging 19:32:24 #action close the bug #1077827 as this not a Debian bug per se and modifications are pushed to the packaging 19:32:46 #topic #1078109 linux-image-6.10.3-amd64: No WiFi When Booting with linux-image-6.10.3-amd64 19:33:02 turns out that it was the wl module not building here because of missed installation of the kernel-headers 19:33:39 before closing the bug I wanted to ask about if someone has idea on what happened here behind the scene 19:34:02 it might have been the usaal race between the signed packages not yet available, and maybe the user had installed manually the unsigned kernel images? 19:35:03 i would assume the user does not have the headers meta package installed 19:35:03 We could ask for a dpkg.log from that time if it's still present 19:35:42 Those issues are often due to someone wanting to 'clean up' / 'tidy up' packages they think are not needed 19:35:53 Actually, no the apt log would be more meaningful 19:36:50 ok we might ask for this additional information. 19:36:55 If someone wants to spend time investigating this they can, but I would ignore it unless we see multiple reports where linux-headers got uninstalled 19:38:10 Let's close the bug then with a note that if logs form that time to further investigate the issue are still prepsent it is welcome to attach it to the bug, but other than that we think the bug can be closed 19:39:20 #action carnil does a final followup to #1078109 asking if apt logs form that time are still present and close the bug 19:39:35 #topic #980555 Missing ec_sys module 19:39:43 I need help with this one 19:40:01 I closed it beeing an old bug, and with our rationale that it looks more as a debugging and development help 19:40:11 bakc from the time as question from waldi 19:40:19 I think the upstream links are quite clear; the functionality should be integrated into the appropriate driver 19:40:20 one of the contributors disagrees 19:40:45 but I think we stay on our position and would not enable it. So tag it wontfix, add again this rationale and close it again? 19:40:55 (suspect it will be reopened though) 19:41:28 I read through all the threads earlier in the week and it the module can poke the hardware directly from userspace and is said to potentially be able to open up security holes 19:41:55 Sure, but this is in debugfs which becomes read-only if Secure Boot is enabled 19:42:17 and is root-only otherwise 19:42:26 unless this driver does something very wrong 19:42:37 There was also some good upstream sub-system maintainer support for integrating functionality this can be used to affect into the appropriate driver 19:43:46 That's obviously what should happen long-term. I think the question is whether it is useful to let people experiment using this before doing so 19:45:20 This comment is interesting: https://bugzilla.redhat.com/show_bug.cgi?id=1943318#c12 19:46:49 bwh: personally I'd enable it but that is because I always want to tinker! Also, Ubuntu has it enabled and I don't recall seeing any complaints about it being abused there 19:47:56 bwh: he follows up in comment #19 though, when told it is in the kernel-debug package 🙂 19:48:13 does someone wants to have a closer look if it would be safe to enable for our Debian builds? 19:48:16 right, so seems like Fedora does have it as an option 19:48:25 carnil: I will try to do that 19:48:35 Thank you Ben 19:48:44 (I just had a look and it is really tiny...) 19:49:11 #action bwh will try to have a closer look for #980555 and if it's safe to enable the module build in Debian. 19:49:47 so we are almost running out of time, let's skip the next two bugs and just mention the IPU6 support one 19:49:58 #topic #1078170 linux-image-6.10.3-amd64: Please enable CONFIG_VIDEO_INTEL_IPU6 for Intel IPU6 based webcams 19:50:10 there is as well a proposed merge request for this. 19:50:36 I asked the merge request author if he could test the combination of enabled config options to verify support works 19:50:50 but other than that and the missing descriptions of the changes it should be fine 19:51:33 Merge request: https://salsa.debian.org/kernel-team/linux/-/merge_requests/1157 19:51:39 I will look at the MR at some point, but I don't promise it will be this week 19:51:55 ok thanks 19:52:12 along with the 6.11-rcX upload for experimental I guess? 19:52:21 or can the later happen earlier already? 19:52:58 I need to fix the blhc complaint for 6.11. If I have time I will go through relevant MRs before uploading. 19:53:04 https://salsa.debian.org/kernel-team/linux/-/merge_requests/1153 there are tough failures from blhc test, but not looked exactly 19:53:14 bwh: ok thank you 19:53:49 #action bwh will try to go along those MRs (but no promise) before uploading 6.11 if time permits 19:53:56 not perfectly formulated, sorry 19:54:03 #topic AOB 19:54:13 anything else or some other comments? 19:54:14 nothing from me 19:54:37 Nothing from me 19:54:40 waldi: are you taking the chair for next week (meeting on Jitsi)? 19:55:14 i will do that 19:55:18 thank you 19:55:28 so thanks all for contributing to todays meeting 19:55:46 Thank you for chairing 19:56:02 #endmeeting