Posts
671Following
107Followers
127- Linux Kernel Developer @ Oracle (Linux Kernel MM) (2025.02 ~ Present)
- A slab subsystem co-maintainer and a reviewer for the reverse mapping subsystem
- Former Intern @ NVIDIA, SK Hynix, Panmnesia (Security, MM and CXL)
- B.Sc. in Computer Science & Engineering, Chungnam National University (Class of 2025)
Opinions are my own.
My interests are:
Memory Management,
Computer Architecture,
Circuit Design,
Virtualization
repeated
Looks like exactly the kind of AI abuse I feared could happen in the kernel is happening.
Now you can see why I pushed back so hard on the automated tooling docs to make it clear we should reject this crap out of hand, being attacked in various publications online for saying so.
Wonder if they will talk about this or?
And yes I told you so.
https://lore.kernel.org/linux-mm/cbd0aafa-bd45-4f4d-a2dd-440473657dba@lucifer.local/
5
4
1
Harry (Hyeonggon) Yoo
hyeyoo> While I've been thinking rationally about this, I've a feeling a large part of our peers haven't
... that's truely concerning, yeah. It'll take some time (hopefully until it's too late) for people to realize.
> My take is that the tech is "good enough" (see latest claude etc)
Yeah, it's quite good! the only problem is...
> that people will keep hyping it just enough that it will justify printing more money for it.
...that it doesn't seem like the industry is making a reasonable investment or carefully applying the tech to reality.
> In the long run will we see the harsh reality of this (hard to fix bugs etc)? Sure! Will it be too late ? Pretty sure that's also a yes.
Sadly yes :(
> Anyway sorry my reasoning is a bit all over the place, it's just confusing times
Don't worry. It's confusing time for everybody...
0
1
1
How will it succeed (in an economically viable sense) if the tech isn't great? The bubble will burst at some point anyway if the tech can't back up the expectations.
I feel like people have stopped thinking rationally (cuz I don't see any clue or potential that the machine is going to take over, at least for now) and are just going wild & pouring an insane amount of money into it...
They'll have to face the consequences eventually, but what worries me is that it might affect the people around me (who aren't doing AI stuff).
+1 for questioning my own sanity.... you're not alone! hehe
1
1
1
seems like it's just human nature to feel enormous excitement and worry at the same time whenever a fancy new tech comes out (only to realize later it wasn't as good or bad as expected)
/me stares at his stock account and wondering when to pull out
(but yeah it's not just about the stock market)
1
0
2
repeated
Mathieu
mathieu@fosstodon.orgI can't wait for either the bubble to burst or the machine taking over. This inbetween is such a sad pathetic state.
1
3
1
repeated
Chris Down
cdown@fosstodon.orgCompressed swap is a common lever to improve memory density, but there's a lot of confusion about how to best use it out there, and many people treat zram/zswap as two flavours of the same thing when they are really far more nuanced.
So what works, what doesn't, and why? In this article I go over the tradeoffs, the work we are doing upstream, and a little about what the future looks like. I am, as always, happy to answer questions :-)
https://chrisdown.name/2026/03/24/zswap-vs-zram-when-to-use-what.html
4
15
3
Harry (Hyeonggon) Yoo
hyeyoo
0
0
4
that's awful indeed.
the maintainer should transfer some power and responsibilities (ok, this is already...) to rightful folks and trust/respect them :/
0
0
2
More than anything, I just wish you wouldn't burn yourself out and would take a little rest :'(
I really appreciate and keep being impressed by what you and David have been doing, but at the same time I can't help worrying that you're taking on too much of a burden...
1
0
2
Harry (Hyeonggon) Yoo
hyeyoo
0
0
1
Harry (Hyeonggon) Yoo
hyeyoo(going to say the same thing for b4, I guess)
1
0
3
repeated
repeated
repeated
A blog post I wrote about folios
2
6
2