Re: [PATCH] mm/damon: introduce DAMON-based NUMA memory tiering module

[Krzysztof Kozlowski]

On 26/03/2026 16:14, Josh Law wrote:
> 
> 
> On 26 March 2026 12:50:40 GMT, "Lorenzo Stoakes (Oracle)" <ljs@xxxxxxxxxx> wrote:
>> On Thu, Mar 26, 2026 at 01:40:19PM +0100, Krzysztof Kozlowski wrote:
>>> On 26/03/2026 13:29, Lorenzo Stoakes (Oracle) wrote:
>>>>>
>>>>> Answering to reviewers with whatever confirmation they look for is also
>>>>> a warning sign of non-trustworthy behavior. Or rather behavior trying to
>>>>> get trust.
>>>>>
>>>>> [1]
>>>>> https://lore.kernel.org/all/D47F8215-FD08-45ED-AB01-0A5C48CD41DD@xxxxxxxxxxxxx/
>>>>>
>>>>> [2]
>>>>> https://lore.kernel.org/all/2F84DD09-2880-45E0-AA98-204F10848F85@xxxxxxxxxxxxx/
>>>>>
>>>>> Best regards,
>>>>> Krzysztof
>>>>
>>>> Yeah this is really adding up to somebody abusing the kernel process. I think
>>>> 'Josh Law' or any other pseudonym that can be traced back to whoever's behind
>>>> this should be banned from the mailing list altogether.
>>>>
>>>> Also, as per Christian ([3]), it seems this guy was using the pseudonym
>>>> "techyguyperplexable".
>>>
>>> Yes, repos on that Github carry commits like:
>>>
>>> commit c04af501bacbef54cf97cdfb904ddf295f8327c1
>>> Author:     techyguyperplexable <objecting@xxxxxxxxxxxxx>
>>> AuthorDate: Sun Jan 11 04:21:53 2026 +0530
>>> Commit:     techyguyperplexable <objecting@xxxxxxxxxxxxx>
>>> CommitDate: Sun Jan 11 04:21:53 2026 +0530
>>>
>>>     ANDROID: sched/rt: reduce lock hold time in update_curr_rt
>>>
>>>     Move resched_curr and do_start_rt_bandwidth outside the spinlock
>>>     critical section to reduce lock contention
>>>
>>>     Change-Id: I8dfb4e33460ae4eeb0bb854cc2cb084ab5edf7e4
>>>     Signed-off-by: techyguyperplexable <objecting@xxxxxxxxxxxxx>
>>>
>>
>> Thanks, it looks pretty clear cut then that 'Josh Law' is an openclaw bot.
>>
>> To think he was scheduled to be added as a reviewer for literally all of lib/
>> recently (5) until various of us pushed back on it :/
>>
>> [5]: https://lore.kernel.org/all/20260308202425.C9EE4C116C6@xxxxxxxxxxxxxxx/
>>
>> Is it really that easy to game the system?...
>>
>> I hate the asymmetry of this - we have to jump on these every time and they can
>> send WAY more mail than we can reasonably reply to.
>>
>> I think we need a clear 'yeeting' mechanism to block these people on everything
>> once discovered.
>>
>> Sadly the kernel has been historically very poor on getting rid of problem
>> people (The CoC is essentially a swear monitor at this stage).
>>
>> In the face of this new kind of threat we probably need to do a lot better.
>>
>>>>
>>>> There's a github with the same name ([4]), and that same user made a comment on an
>>>> openclaw issue relating to linking it to gemini ([5]).
>>>>
>>>> So it seems likely he/she is likely using openclaw to fully automate this 'Josh
>>>> Law' user and we're speaking to a bot here.
>>>>
>>>> (Screenshots have been taken for the (possibly inevitable) deleting that will
>>>> happen later.)
>>>>
>>>> [3]:https://lore.kernel.org/all/20260313-halskette-annahme-94e782eb4ae4@brauner/
>>>> [4]:https://github.com/techyguyperplexable
>>>> [5]:https://github.com/openclaw/openclaw/issues/44134#issuecomment-4106247302
>>>
>>>
>>>
>>> Best regards,
>>> Krzysztof
>>
>> Thanks, Lorenzo
> 
> 
> Hello! Wow i came back to a lot.
> 
> 
> I can promisw you now i am a human, hand on the bible, and i will improve! 
> 
> 
> So i will add assisted by to any commit descriptions written by AI, or anything even touching AI, i understand your apprehension and i apoligize very deeply.
>

If 20 times you were asked for something already on the list - to wrap
replies - and you still ignore it, then I don't trust you are meaning this.

Anyway, you did not address the actual problems here.

Best regards,
Krzysztof