Linus Torvalds Comments On Bcachefs Prospects For Linux 6.6

[avis]

Well, all this makes me worry about the future of kernel. I don't believe Kent is some dumb student who do not understand the basics, at least his previous work, bcache is within kernel and somehow it had been merged. So either all these procedures are new or not so obligatory. All this "you think you are special" grunt reminds of Linus' age. He believes his old trusted lieutenants, and his lieutenant cares only about proper bureaucratic procedure and usage of iomap. Same lieutenant by the way, who've merged into 6.6 a code, sole purpose of which is to break nvidia driver. You can break userspace in case you are one of maintainers, and maybe you can do so even without going through linux-next branch.

Every new major Linux kernel feature (nope, not just that, even new device drivers and filesystems) has gone through Linux next and this practice is neither "new", nor "optional".

His lieutenants guarantee the kernel which, when works, is reliable as it can humanly be.

I don't know why you tried to make this refusal "personal" as there was nothing personal about it.

Two 100% technicalities: 1) not pulled into Linux next first 2) has compilation errors. The kernel has strived to have zero compiler warnings, let alone errors.

[timofonic]

I think he should learn to be more humble and polite, despite he may be right in some points. Sometimes it's better to have a good attitude than being right. That's part of having a mature and psychologically balanced behavior.

It's no excuse to become an ashole if others become harsh, sectarian and other negative behaviors against you.

I hope he apologizes for his attitude, get mental health assistance and even make compromises. Maybe even merging into linux-next first.

He's an extremely talented developer, that's the hard part. He just has to polish his behavior.

[anarki2]

So at this stage it's still not clear yet if Bcachefs will be merged this week for the Linux 6.6 cycle.

Oh it's pretty clear alright. It won't be.

[Quackdoc]

I think he should learn to be more humble and polite, despite he may be right in some points. Sometimes it's better to have a good attitude than being right. That's part of having a mature and psychologically balanced behavior.

While I dont disagree with the dev trying to become more polite (it's that case for all of them) they are kernel devs, they should NEVER put a good attitude over being right, When kernel devs are dancing around issues that could cause problems, they SHOULD be called out, same goes for when devs shut down discussion by disagreeing and not discussing it, just like it goes for when you post a massive patch to linux-mainline and not linux-next.

Bcachefs patch is kind amazing, since it has brought out bad points from both "sides of the table" despite being a table that should be round. both many people had shameful behavior and bickering.

[oiaohm]

Well, all this makes me worry about the future of kernel. I don't believe Kent is some dumb student who do not understand the basics, at least his previous work, bcache is within kernel and somehow it had been merged. So either all these procedures are new or not so obligatory..

Its not that the procedures are new or obligatory. The Linux-Next before mainline only starts in 2002. Bcache got mainline in 2013 while all the current rules were in effect. He obeyed them back in 2013 to get bcache merged. It is now 10 years latter and Kent Overstreet is either trying to ignore the rules or forgot key steps.

Part of why Linus Torvalds did not notice that bcachefs was not in linux-next tree is that Kent Overstreet back in 2012-2013 before talking to Linus had bcache in the linux-next tree before talking to Linus Torvalds. Linus basically incorrectly assumed Kent would be following Linux kernel procedures correctly.

Khrundel there is many ways to be a human to be dumb. There is a form of dumb you see it with complex machines where people believe they still remember how to use the machine when they have not used it for decade then use it and kill themselves because they forget some critical step to use the machine correctly. Basically Kent Overstreet just done a mistake with software that with real machines at times wins you a Darwin award if its not intentional.

Yes if you look at the linux kernel mailing list history that is kept on line in archives so you can go back to 2012/2013 and see back then with bcache Kent Overstreet does the process absolutely correctly including having bcache in linux-next before talking to Linus Torvalds. Something gone wrong with Kent Overstreet in the decade since the merge in bcache 2013 personally I would say human memory error leading to current mess bcachefs.

I would say it only human to screw up this way when you have not done something for a decade and don't read the manual carefully how to do it. As I pointed out in another post the Linux kernel manual on posts are not as clear on you must do linux-next branch is it really should be.

Human forgetfulness of procedures(Kent Overstreet)+documentation issue+Linus Torvalds assume resulting in not noticing is the most likely cause of current bcachefs mess. We are in this mess now with the only valid way out really is correct the mistakes and go forwards.

Yes a decade is a long time between drinks for Kent Overstreet.

[timofonic]

Its not that the procedures are new or obligatory. The Linux-Next before mainline only starts in 2002. Bcache got mainline in 2013 while all the current rules were in effect. He obeyed them back in 2013 to get bcache merged. It is now 10 years latter and Kent Overstreet is either trying to ignore the rules or forgot key steps.

Part of why Linus Torvalds did not notice that bcachefs was not in linux-next tree is that Kent Overstreet back in 2012-2013 before talking to Linus had bcache in the linux-next tree before talking to Linus Torvalds. Linus basically incorrectly assumed Kent would be following Linux kernel procedures correctly.

Khrundel there is many ways to be a human to be dumb. There is a form of dumb you see it with complex machines where people believe they still remember how to use the machine when they have not used it for decade then use it and kill themselves because they forget some critical step to use the machine correctly. Basically Kent Overstreet just done a mistake with software that with real machines at times wins you a Darwin award if its not intentional.

Yes if you look at the linux kernel mailing list history that is kept on line in archives so you can go back to 2012/2013 and see back then with bcache Kent Overstreet does the process absolutely correctly including having bcache in linux-next before talking to Linus Torvalds. Something gone wrong with Kent Overstreet in the decade since the merge in bcache 2013 personally I would say human memory error leading to current mess bcachefs.

I would say it only human to screw up this way when you have not done something for a decade and don't read the manual carefully how to do it. As I pointed out in another post the Linux kernel manual on posts are not as clear on you must do linux-next branch is it really should be.

Human forgetfulness of procedures(Kent Overstreet)+documentation issue+Linus Torvalds assume resulting in not noticing is the most likely cause of current bcachefs mess. We are in this mess now with the only valid way out really is correct the mistakes and go forwards.

Yes a decade is a long time between drinks for Kent Overstreet.

While I agree, I'm worried that the damage may be too much to fix it.

[timofonic]

This seems a quite interesting point too...

It looks like this series hasn't been in -next at all? That seems like a pretty important step.

Also, when I look at the PR, it seems to be a branch history going back _years_. For this kind of a feature, I'd expect a short series of "here's the code" in incremental additions (e.g. look at the x86 shstk series), not the development history from it being out of tree -- this could easily lead to ugly bisection problems, etc.

There's other coding issues found that may motivate to make a general kernel cleanup too.

[L_A_G]

Sadly this kind of blasé attitude is quite common in the tech sector and probably isn't helped by Linux kernel development not being done with much face-to-face interaction between developers. Purely text-based communication leaves out a lot of the nuances you get in face-to-face interaction and it becomes easy for your imagination to start filling in the gaps such that what's actually a 100% professional refusal becomes hostile or even a mocking one. To him, this is clearly personal.

So instead of dealing with maintainers that he thinks are hostile, mocking or both he tried to bypass them and go straight to Linus. Which obviously isn't how you upstream things.

If Linus refuses him, as he probably should, he's probably going to take it personally, again, and he's only going to get worse.

[Khrundel]

Its not that the procedures are new or obligatory. The Linux-Next before mainline only starts in 2002. Bcache got mainline in 2013 while all the current rules were in effect. He obeyed them back in 2013 to get bcache merged. It is now 10 years latter and Kent Overstreet is either trying to ignore the rules or forgot key steps.

Well, in that case, and assuming this is not some overcomplicated plan to make an excuse to drop bcachefs from Kent's side, all this may be just misunderstanding.

Kent is not trying to merge bcachefs, he tries to merge prerequisite, a new sync primitive to kernel. He may not realize this is a feature which should come through linux-next, more like bugfix. I assume he've merged some patches to bcache with some kind of shortcut process.

From what I've read during previous attempts, it was like: "Please, review my code... Anybody?". I don't know, maybe it was done via some private channel. Up to last month I was sure, nobody in kernel team wants to touch bcachefs, when suddenly Linus himself had mentioned it within notes to one of RCs. Then he himself reviewed code, all this looked like Linux himself is eager to merge this asap. And now he turns 180 degrees and say like "go away and never turn back. Nobody likes you here anyway".

[Nuc!eoN]

What do you people think about this?

Would Kent attitude be more tamed to be suffice? Will he be able to change that in his personal behavior? Linus achieved it.

What about not submitting it to linux-next before?

What about that compile error?

Is Bcachefs doomed after this?

Kent is doing the right thing. He is strategically puting a stick into the ant nest. Letting them bleed out one by one. In the end his efforts will work out. The biggest talents will always get the most backlash, by those who feel threatened. BcacheFS will be a new era and thus some controversy is needed to cement its position.