IRQ error on boot with 6.4.1

[dzyndzla]

Hi there, I got the below error together with 6.4.1 kernel update. Unfortunately, 6.4.1-arch2-1 didn’t solve this. Google didn’t help much. How should I deal with it? Any hints appreciated.

inxi - http://ix.io/4zNT

PC is LENOVO T490

[    4.695194] irq 31: nobody cared (try booting with the "irqpoll" option)
[    4.695206] CPU: 3 PID: 0 Comm: swapper/3 Not tainted 6.4.1-arch2-1 #1 cf34d70ffed66439727ee92a8197bd2b3e0b11de
[    4.695208] Hardware name: LENOVO <redacted>/<redacted>, BIOS N2IETA1W (1.79 ) 01/16/2023
[    4.695210] Call Trace:
[    4.695211]  <IRQ>
[    4.695213]  dump_stack_lvl+0x47/0x60
[    4.695218]  __report_bad_irq+0x2b/0xc0
[    4.695221]  note_interrupt+0x2a6/0x2f0
[    4.695223]  handle_irq_event+0x75/0x80
[    4.695226]  handle_fasteoi_irq+0x8e/0x1f0
[    4.695228]  __common_interrupt+0x43/0xc0
[    4.695231]  common_interrupt+0x81/0xa0
[    4.695234]  </IRQ>
[    4.695234]  <TASK>
[    4.695235]  asm_common_interrupt+0x26/0x40
[    4.695238] RIP: 0010:cpuidle_enter_state+0xcc/0x440
[    4.695240] Code: 2a 3f 3c ff e8 c5 f3 ff ff 8b 53 04 49 89 c5 0f 1f 44 00 00 31 ff e8 63 41 3b ff 45 84 ff 0f 85 56 02 00 00 fb 0f 1f 44 00 00 <45> 85 f6 0f 88 85 01 00 00 49 63 d6 48 8d 04 52 48 8d 04 82 49 8d
[    4.695242] RSP: 0018:ffffa53c80137e90 EFLAGS: 00000246
[    4.695243] RAX: ffff8e6e2e6f3f40 RBX: ffff8e6e2e6ffa00 RCX: 0000000000000000
[    4.695245] RDX: 0000000000000003 RSI: fffffffd2a185f92 RDI: 0000000000000000
[    4.695245] RBP: 0000000000000001 R08: 0000000000000002 R09: 00000000471c7c22
[    4.695246] R10: ffff8e6e2e6f2944 R11: 0000000000000008 R12: ffffffffb6545920
[    4.695247] R13: 0000000117dac32f R14: 0000000000000001 R15: 0000000000000000
[    4.695249]  cpuidle_enter+0x2d/0x40
[    4.695252]  do_idle+0x1d8/0x230
[    4.695254]  cpu_startup_entry+0x1d/0x20
[    4.695255]  start_secondary+0x12b/0x150
[    4.695258]  secondary_startup_64_no_verify+0x10b/0x10b
[    4.695261]  </TASK>
[    4.695261] handlers:
[    4.695266] [<000000002834e634>] irq_default_primary_handler threaded [<000000002fa0492f>] tis_int_handler

EDIT: booting with irqpoll kernel parameter resulted in kernel panic and chrooting system

[thefrog]

If the new kernel doesn’t solve the issue then downgrade to 6.3.9 see if that fixes it.

[dzyndzla]

This is what I did but it’s just a workaround rather than a permanent solution. That’s why I’m asking for a hand.

[ricklinux]

Disable TPM in UEFI and you may have to rid yourself of secure boot and it should work.

Edit: Think I already mentioned to do this?

https://bugzilla.redhat.com/show_bug.cgi?id=1770021

[dzyndzla]

I think disabling TPM decreases security I’d like to keep TPM enabled, especially in case of mobile devices. Also, I tried turning it off, it didn’t solve that IRQ error.

[mihalycsaba]

I don’t think TPM does much anyway if your drives aren’t encrypred. It’s used for some random number generation if I remember correctly.

I had stutters a few months ago with amd fTPM enabled, disabling it solved my problem. The other suggestion was to update your bios, I didn’t update because the bios was still in beta.

[dzyndzla]

yea, I have all drives encrypted, some using bitlocker, so I need TPM. In my case bios is also up to date.

[mihalycsaba]

Lot of people had similar issues last week on AMD, downgrade I’m sure it will be fixed soon.

[ricklinux]

I don’t use TPM or secure boot and i always remove the security keys on the initial install.

[dzyndzla]

yea, that’s what I hope for; I downgraded and wait for resolution

EDIT: I found a root cause and a solution for that. It looks like this is very Lenovo-specific issue and is caused by the way Lenovo implements / handles TPM module. So, @ricklinux you had a good feeling about that. Following this: https://support.lenovo.com/us/en/solutions/ht512219-the-message-firmware-bug-tpm-interrupt-not-working-polling-instead-is-reported-when-executing-dmesg-on-whitley-based-systems-lenovo-thinksystem and adding tpm_tis.interrupts=0 as a kernel parameter solves that IRQ error.