Re: Thorsten is tracking regression again and could need a little help (was: Re: Linux 5.15)

[Thorsten Leemhuis]

On 01.11.21 14:03, Greg KH wrote:
> On Mon, Nov 01, 2021 at 01:44:01PM +0100, Thorsten Leemhuis wrote:
>> On 01.11.21 13:33, Greg KH wrote:
>>> On Mon, Nov 01, 2021 at 05:49:40AM +0100, Thorsten Leemhuis wrote:
>>>> example by simply forwarding the mail to regressions@xxxxxxxxxxxxx or
>>>> CCing that address on a reply. I'll handle everything else then and tell
>>>> regzbot about it. But if you feel adventurous, you can also skip me as
>>>> the man-in-the-middle and tell the bot directly. To do that, just send a
>>>> reply to the report to the regressions mailing list
>>>> (regressions@xxxxxxxxxxxxxxx) either directly or by CCing it on a reply
>>>> you would have written anyway; when doing so, place something like
>>>> '#regzbot ^introduced v5.15..' (separated by blank lines) somewhere in
>>>> the text, as outlined in regzbot's 'getting started guide' or its
>>>> reference documentation:
>>>> https://gitlab.com/knurd42/regzbot/-/blob/main/docs/getting_started.md
>>>> https://gitlab.com/knurd42/regzbot/-/blob/main/docs/reference.md
>>>>
>>>> That's it, regzbot then on its next run will add the report to the list
>>>> of tracked regression. I'll keep an eye on things and try to fix any
>>>> problems I notice, as there likely will be a few. But then doesn't need
>>>> to bother you.
>>>>
>>>> There is one thing that would really help: if one or two subsystem
>>>> maintainers could give regzbot a shot for all the regression reports
>>>> they get, even for easy fixes, as the bot really needs something to chew
>>>> on. Any volunteers?
>>>
>>> I'll try it for the USB subsystem this merge cycle. 
>>
>> That will be a great help, many thx.
>>
>>> Do you want a bug
>>> report email redirected to that address or will a simple forward work
>>> well enough?
>>
>> Redirecting will make it a little easier for me, but a simple forward is
>> fine, too.
> 
> Ok, I did that now for a USB bug report, hopefully that worked.  If not,
> I can forward it on.

Got it, but I could need some advice on it if you have a minute.

Does that report really look like a regression from your point of view?
The part "The code has been this way in the kernel for a very long time,
which suggests that it has been working, [...]" sounds like it is, but
OTOH it's quite vague.

I'm asking, because with my regression tracking work and regzbot I focus
on regressions and ignore things that were always broken, as I (at least
for now) don't want it to become yet another bug tracker (and I guess I
would quickly drown in bugs as well).

But if you think this case looks like a regression, I'll add it. Guess I
just need to be creative then how to tell regzbot when it got
introduced. Guess I'll settle on v2.6.13..v5.15", that should indicate
there is something strange here. It's a use-case I hadn't expected. But
well, that's why I wanted to something for testing. :-D

Ciao, Thorsten