Conversation

Vlastimil Babka πŸ‡¨πŸ‡ΏπŸ‡ͺπŸ‡ΊπŸ‡ΊπŸ‡¦

mm patches be like

4
2
2
@vbabka
Oh lol, I'm sure this will be the best kernel meme of the year :)
1
0
2

Vlastimil Babka πŸ‡¨πŸ‡ΏπŸ‡ͺπŸ‡ΊπŸ‡ΊπŸ‡¦

@hyeyoo just a meme, right? No connection to reality, right? RIGHT?

1
0
1

Vlastimil Babka πŸ‡¨πŸ‡ΏπŸ‡ͺπŸ‡ΊπŸ‡ΊπŸ‡¦

oh no :(

1
0
0

@vbabka

--- a/mm/slub.c
+++ b/mm/slop.c
1
1
0

@vbabka What were you saying yesterday about requiring the sender to refute the LLM?

1
0
0

Vlastimil Babka πŸ‡¨πŸ‡ΏπŸ‡ͺπŸ‡ΊπŸ‡ΊπŸ‡¦

@ffmancera yep, it's spreading :(
I don't deny it can be useful and point out bugs, but not like this...

1
0
0

@vbabka FWIW, in netdev it is not required to refute sashiko, although there is a Netdev LLM CI which is reviewed by a maintainer and if they consider the output is useful then you must refute it (if it is wrong).

1
0
0

@vbabka Btw, this netdev LLM is much better than sashiko tbh

1
0
0

Vlastimil Babka πŸ‡¨πŸ‡ΏπŸ‡ͺπŸ‡ΊπŸ‡ΊπŸ‡¦

@ffmancera yeah I heard sashiko points out much more wrong stuff. Could be it's newer and needs more work, or the underlying LLM is worse.

0
0
0

Vlastimil Babka πŸ‡¨πŸ‡ΏπŸ‡ͺπŸ‡ΊπŸ‡ΊπŸ‡¦

0
0
0