On Mon, Jan 13, 2020 at 8:49 AM pmkellly(a)frontier.com
<pmkellly(a)frontier.com> wrote:
On 1/11/20 5:58 PM, pmkellly(a)frontier.com wrote:
>
>
> On 1/11/20 4:27 PM, Chris Murphy wrote:
>> On Sat, Jan 11, 2020 at 1:26 PM pmkellly(a)frontier.com
>> <pmkellly(a)frontier.com> wrote:
>>>
>>>
>>> On 1/11/20 2:18 PM, Chris Murphy wrote:
>>>> On Sat, Jan 11, 2020 at 9:00 AM pmkellly(a)frontier.com
>>>> <pmkellly(a)frontier.com> wrote:
>>>>>
>>>>> This is in regard to machines running Fully updated F31 and getting
>>>>> Kernel tainted errors and tpm2-abrmd errors on a recurring basis.
>>>>
>>>> Is this a 5.4 kernel? And does the Computer have Secure Boot enabled?
>>>> Could be this:
>>>>
https://bugzilla.redhat.com/show_bug.cgi?id=1777072#c2
>>>>
>>>> Reported to be fixed in 5.4.7. If you're using that you'll need
to be
>>>> more specific about what's causing the taint.
>>>> sudo journalctl -b | grep -i taint
>>>>
>>>> And tpm stuff is a known bug
>>>>
https://bugzilla.redhat.com/show_bug.cgi?id=1769215
>>>>
>>>>
>>>>
>>>
>>> I'm not running secure boot. I'm running kernel 5.4.8 since first
thing
>>> this morning. Here's the grep result:
>>>
>>> grep -i taint Taint-TMP-Journal.txt
>>> Jan 10 12:54:38 localhost.localdomain kernel: CPU: 0 PID: 9713 Comm:
>>> tpm2-abrmd Not tainted 5.4.8-200.fc31.x86_64 #1
>>>
>>> I was getting the kernel tainted error right along with 5.4.7. I've
been
>>> using that PC today just to see what happens. I just checked and I
>>> haven't gotten any more of the kernel tainted errors, but I'm still
>>> getting the tpm2 errors. I just checked the screen shots I made of the
>>> Details pages in Problem Reporter and it shows I was running
>>> 5.4.8-200.fc31_x86-64. The taint error is "E - Unsigned module has
been
>>> loaded".
>>
>> Anything out of tree or unsigned taints the kernel. This might have a
>> better chance of finding out what's causing it:
>>
>> journalctl -b | grep -i -B4 taint
>>
>> Usually the message right before the message "tainting kernel" will
>> say what module and what the issue is, or maybe it's on the same line.
>>
>>
>
> NEWS!!!:
>
> I think this might have been a leftover from 5.4.7.
>
> This morning after I did the upgrade that got me to 5.4.8 I did a
> restart and when I looked, I found the kernel tainted error, but as I
> have been using that PC all day that error has not reoccurred. I just
> did another restart and then I ran:
>
> journalctl -b | grep -i -B4 taint
>
> and got no response at all. Could it be that during the restart I did
> this morning right after the upgrade that there was at some point enough
> of 5.4.7 around to cause the error, but the id had been changed to 5.4.8?
>
>
The tpm2-abrmd errors are continuing at a frantic pace. there is
constant hard disk activity. I'm guessing from the errors being logged
about every 3 seconds. I verified that secure boot is not enabled in the
UEIF/BIOS.
Is there a way to turn off tpm2-abrmd or tpm2 in general or will things
I need break? I've been thinking that I should revert the users PCs to
F30. Any advice or thoughts are welcome.
Did you read the bug I cited?
https://bugzilla.redhat.com/show_bug.cgi?id=1769215
It's supposed to be fixed now. If you're seeing something different
than what's reported in that bug, then it needs a new bug filed. If
it's the same thing, you need to let them know what version you're
using and that it's not fixed.
That bug also has a couple of workarounds depending on the hardware -
it seems if you enable TPM in the firmware setup it goes away. In my
case, it was already enabled but there are other kernel bugs related
to the TPM I have, so I removed that particular package.
--
Chris Murphy