January 2025 - openSUSE Kernel Bugs - openSUSE Mailing Lists

[Bug 1235020] Hard freeze workqueue lockup mediatek bluetooth
by bugzilla_noreply＠suse.com 03 Jan '25

03 Jan '25

https://bugzilla.suse.com/show_bug.cgi?id=1235020 https://bugzilla.suse.com/show_bug.cgi?id=1235020#c8 --- Comment #8 from Andrea Ippolito <andrea.ippo(a)gmail.com> --- (In reply to Takashi Iwai from comment #7) > Could you check 6.12.8 kernel in OBS Kernel:stable repo? > https://download.opensuse.org/repositories/Kernel:/stable/standard/ > > It contains a few Mediatek BT patches and those might address your problem. Thanks, I'll try and report back. Please be aware that the issue isn't too frequent, so it may take a while :) -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1235020] Hard freeze workqueue lockup mediatek bluetooth
by bugzilla_noreply＠suse.com 03 Jan '25

03 Jan '25

https://bugzilla.suse.com/show_bug.cgi?id=1235020 https://bugzilla.suse.com/show_bug.cgi?id=1235020#c7 --- Comment #7 from Takashi Iwai <tiwai(a)suse.com> --- Could you check 6.12.8 kernel in OBS Kernel:stable repo? https://download.opensuse.org/repositories/Kernel:/stable/standard/ It contains a few Mediatek BT patches and those might address your problem. -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1235020] Hard freeze workqueue lockup mediatek bluetooth
by bugzilla_noreply＠suse.com 03 Jan '25

03 Jan '25

https://bugzilla.suse.com/show_bug.cgi?id=1235020 https://bugzilla.suse.com/show_bug.cgi?id=1235020#c6 --- Comment #6 from Andrea Ippolito <andrea.ippo(a)gmail.com> --- (In reply to Takashi Iwai from comment #5) > Thanks! So it's triggered at btusb_disconnect() -> hci_unregister_dev(), > and likely the same issue as bsc#1231599. > > The Mediatek driver in the upstream is still in a bad shape, unfortunately. > We applied a revert but this doesn't seem covering the all issues. No problem, thanks to you for looking into it! Is there anything else I can help with? I know this is missing steps to reproduce but TBH I don't think I was doing anything special at the time, actually I'm not even sure my only BT device (a pair of earbuds) was out of the charging case and connected at the time this happened! In any case I will try and pay more attention next time this happens and update here, although TBH I hope you already have a clue about what could be the problem :-) Every time this happens it completely kills productivity (and potentially causes loss of work, too :( ) Have a nice day :) -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1235020] Hard freeze workqueue lockup mediatek bluetooth
by bugzilla_noreply＠suse.com 03 Jan '25

03 Jan '25

https://bugzilla.suse.com/show_bug.cgi?id=1235020 https://bugzilla.suse.com/show_bug.cgi?id=1235020#c5 Takashi Iwai <tiwai(a)suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(andrea.ippo@gmail | |.com) | --- Comment #5 from Takashi Iwai <tiwai(a)suse.com> --- Thanks! So it's triggered at btusb_disconnect() -> hci_unregister_dev(), and likely the same issue as bsc#1231599. The Mediatek driver in the upstream is still in a bad shape, unfortunately. We applied a revert but this doesn't seem covering the all issues. -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1235020] Hard freeze workqueue lockup mediatek bluetooth
by bugzilla_noreply＠suse.com 03 Jan '25

03 Jan '25

https://bugzilla.suse.com/show_bug.cgi?id=1235020 https://bugzilla.suse.com/show_bug.cgi?id=1235020#c4 --- Comment #4 from Andrea Ippolito <andrea.ippo(a)gmail.com> --- (In reply to Takashi Iwai from comment #3) > (In reply to Andrea Ippolito from comment #1) > > There are also entries like: > > > > Jan 02 18:19:35 andromeda kernel: Oops: general protection fault, probably > > for non-canonical address 0xdead000000000108: 0000 [#1] PREEMPT SMP NOPTI > > That's crucial info, please give the full output including this stack trace. Hi Takashi, sorry for not explaining this well. That line was an excerpt from a pastebin I linked in my initial post: https://pastebin.com/n3m1S8B7 That paste however was a stripped down version of the journal, where I removed previous events. In case the full boot is needed, you will find it here: https://pastebin.com/zy6fWF0V relevant timestamps are in the title (command: sudo journalctl -b -5 > 2025-01-02-workqueue-lockup-timestamps-18:19:35-18:20:33.txt) Thanks a lot in advance! -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1234784] nvme nvme0: I/O tag 579 (a243) opcode 0x1 (I/O Cmd) QID 1 timeout, aborting req_op:WRITE(1) size:524288
by bugzilla_noreply＠suse.com 03 Jan '25

03 Jan '25

https://bugzilla.suse.com/show_bug.cgi?id=1234784 https://bugzilla.suse.com/show_bug.cgi?id=1234784#c3 Daniel Wagner <daniel.wagner(a)suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |daniel.wagner(a)suse.com Status|NEW |IN_PROGRESS --- Comment #3 from Daniel Wagner <daniel.wagner(a)suse.com> --- It doesn't look like there is a problem in the nvme driver itself. The device is just not returning the IO requests in time. I've checked if there are any known quirks for this particular device but it doesn't seem to be the case. A bit of searching on the IRQ setup I found some discussions on the VMD MSI-X domain remapping and some performance impact. Anyway, not sure if this is could be a problem here. Also a good candidate for troubles is PCI bus power management. I suggest you try some some of the kernel knobs if they address the problem, e.g. msi_remap=off pci=noacpi and see if it changes anything. -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1235020] Hard freeze workqueue lockup mediatek bluetooth
by bugzilla_noreply＠suse.com 03 Jan '25

03 Jan '25

https://bugzilla.suse.com/show_bug.cgi?id=1235020 https://bugzilla.suse.com/show_bug.cgi?id=1235020#c3 Takashi Iwai <tiwai(a)suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |andrea.ippo(a)gmail.com, | |tiwai(a)suse.com Flags| |needinfo?(andrea.ippo@gmail | |.com) --- Comment #3 from Takashi Iwai <tiwai(a)suse.com> --- (In reply to Andrea Ippolito from comment #1) > There are also entries like: > > Jan 02 18:19:35 andromeda kernel: Oops: general protection fault, probably > for non-canonical address 0xdead000000000108: 0000 [#1] PREEMPT SMP NOPTI That's crucial info, please give the full output including this stack trace. -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1229285] Xorg does not retrieve EDID properly, resulting in incorrect screen size (kernel-default 6.7 and above).
by bugzilla_noreply＠suse.com 02 Jan '25

02 Jan '25

https://bugzilla.suse.com/show_bug.cgi?id=1229285 https://bugzilla.suse.com/show_bug.cgi?id=1229285#c10 Donnie Menig <djmenig(a)outlook.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |CONFIRMED Flags|needinfo?(djmenig(a)outlook.c | |om) | --- Comment #10 from Donnie Menig <djmenig(a)outlook.com> --- (In reply to Jiri Slaby from comment #9) > By a chance, has this been fixed in the meantime? > > I see there is not progress in the upstream tracker :/. Unfortunately, no; however, openSUSE provides the kernel-longerm package, and that can be used as a work-around. Since the regression was introduced with kernel version 6.7, anything before that version should work. Hopefully they fix the regression before kernel-longterm gets to version 6.7! :) https://software.opensuse.org/package/kernel-longterm -- You are receiving this mail because: You are the assignee for the bug.

1 0

[Bug 1235020] New: Hard freeze workqueue lockup mediatek bluetooth
by bugzilla_noreply＠suse.com 02 Jan '25

02 Jan '25

https://bugzilla.suse.com/show_bug.cgi?id=1235020 Bug ID: 1235020 Summary: Hard freeze workqueue lockup mediatek bluetooth Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: Other OS: openSUSE Tumbleweed Status: NEW Severity: Normal Priority: P5 - None Component: Kernel Assignee: kernel-bugs(a)opensuse.org Reporter: andrea.ippo(a)gmail.com QA Contact: qa-bugs(a)suse.de Target Milestone: --- Found By: --- Blocker: --- Hi, I've been experiencing random lockups on TW for a while now (more than a month): https://community.frame.work/t/workqueue-lockup-opensuse-tw/60430 (I'm the OP here) https://forums.opensuse.org/t/workqueue-lockup-should-i-file-a-bug/180387 I had to switch to Fedora for the past couple of weeks, and surprisingly I never experienced this issue there, which makes me think it might be related to the opensuse kernel (or maybe, just bad luck on TW). The problem seems to be related to bluetooth, as we can see by checking at the logs at the timestamp preceding the BUG - lockup entry by X seconds as specified in the error: Examples: https://pastebin.com/79dj1PeX https://pastebin.com/n3m1S8B7 In both cases if we go back by 58s we get around this part: Jan 02 18:19:33 andromeda kernel: Bluetooth: hci0: Execution of wmt command timed out Jan 02 18:19:33 andromeda kernel: Bluetooth: hci0: Failed to send wmt patch dwnld (-110) Jan 02 18:19:33 andromeda kernel: Bluetooth: hci0: Failed to set up firmware (-110) Jan 02 18:19:33 andromeda kernel: Bluetooth: hci0: HCI Enhanced Setup Synchronous Connection command is advertised, but not supported. Jan 02 18:19:35 andromeda kernel: Bluetooth: hci0: Failed to read MSFT supported features (-110) Jan 02 18:19:35 andromeda kernel: Bluetooth: hci0: command 0xfd30 tx timeout Jan 02 18:19:35 andromeda kernel: Bluetooth: hci0: AOSP get vendor capabilities (-110) Jan 02 18:19:35 andromeda bluetoothd[1147]: Failed to set mode: Authentication Failed (0x05) When this happens, my entire DE becomes unresponsive, even systemctl reboot won't work and I'm forced to hard power off. I'm running 6.12.6-1-default Thanks -- You are receiving this mail because: You are the assignee for the bug.

1 2

[Bug 1234782] Kernel 6.12.6 with amdgpu breaks brightness control after resume
by bugzilla_noreply＠suse.com 02 Jan '25

02 Jan '25

https://bugzilla.suse.com/show_bug.cgi?id=1234782 https://bugzilla.suse.com/show_bug.cgi?id=1234782#c9 --- Comment #9 from Frank Krüger <fkrueger(a)mailbox.org> --- (In reply to Jiri Slaby from comment #7) > Pushed the fix from: > https://lore.kernel.org/all/20241227073700.3102801-1-alexander.deucher@amd. > com/ Thank you! Version 6.12.8 works just fine. -- You are receiving this mail because: You are the assignee for the bug.

1 0