Re: [Regression] Bug 216961 - Severe IO scheduling starvation issues with btrfs

[Linux regression tracking (Thorsten Leemhuis)]

On 28.02.23 20:40, David Sterba wrote:
> On Fri, Feb 17, 2023 at 06:17:58AM +0100, Linux regression tracking (Thorsten Leemhuis) wrote:
>> On 16.02.23 23:39, Josef Bacik wrote:
>>> On Thu, Feb 16, 2023 at 12:43 PM Thorsten Leemhuis
>>> <regressions@xxxxxxxxxxxxx> wrote:
>>>>
>>>> I might be missing something, but it looks like the report was ignored.
>>>> Is there a reason for that?
>>> Mostly just being blind.
>> Happens.
>>
>>>  I've responded to the BZ, I'll investigate
>>> through the BZ.
> According to the bug it's a problem in BFQ.

Yeah, I noticed yesterday, after I looked into the issue again, as it
looked stalled. But as I already wrote in a comment in the ticket
yesterday:

Did anyone tell the bfq developers about this?

Doesn't look like it. Adding them to the list of recipients them now,
even if they might know about it. If not:

Paolo, Jens, there seems to be a regression in BFQ likely introduced
between 5.19 and 6.0 that people apparently notice often with Btrfs. For
details see:
https://bugzilla.kernel.org/show_bug.cgi?id=216961

Josef apparently has seen report about this as well where switching to
another io scheduler helped:
https://bugzilla.kernel.org/show_bug.cgi?id=216961#c8

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat)
--
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
If I did something stupid, please tell me, as explained on that page.

#regzbot title: bfq: severe IO scheduling starvation issues with btrfs