Shared publicly  - 
 
Linux 3.3.2 kernel is now released.
3.0.28, and 3.2.15, if you like those older releases, are also now released.
I'm announcing the release of the 3.3.2 kernel. All users of the 3.3 kernel series must upgrade. The updated 3.3.y git tree can be found at: git://git.kernel.org/pub/scm/linux/kernel/git/stable/li...
165
31
Miguel Rodriguez's profile photoSebastien Le Lourec's profile photoYan Huang's profile photosilence liu's profile photo
32 comments
 
John & I will build and install immediatly
 
+Ricardo Bánffy they generally don't include btrfs updates in the stable kernels, since it's still under heavy development and rapidly moving. The changes don't really fit into the "stable" release model.
Matt B
+
1
2
1
 
+Ricardo Bánffy clone the linux-btrfs repo

and run a git diff to get a patch against 3.3 - that's what I'm currently running:

git diff c16fa4f2ad19908a47c63d8fa436a1178438c7e7 bc3f116fec194f1d7329b160c266fe16b9266a1e > ../btrfs-3.3+.patch

(that's from bash history, don't know if the commits are still correct ^^)

then manually apply the patch on your kernel directory

or do some more advanced git-magic via cherry-pick, etc.
 
Today I learned the stable kernel is not meant to be stable, but a testing ground for mainline:

http://thread.gmane.org/gmane.linux.kernel/1280458/focus=8575

I would rather use v3.2.x, or even v3.3 rather than v3.3.x, at least until v3.4 is released, because until then the crap coming from the release candidates is going to stay until fixed there.

So, I have a problem with this sentence "All users of the 3.3 kernel series must upgrade". Seriously? Even if there are issues with the patch series, and they were not in v3.3, and you know it, and you still deliberately push the broken patches? I think that's deceiving the users.

+Wiktor Sztwiertnia Seems the ath9k fix is there :)
 
+Felipe Contreras The version of the Linux kernel you use should be determined by your individual needs and demands. There are no guarantees of stability (or security) in any branch. Stable and development are more akin to frozen and in flux rather than bug-free and buggy. Even then, that is subject to change since Linus does as Linus pleases.
 
3.3.1 broke my wifi, let's hope this one fixes it again. :)
 
I also got bitten by the wifi bug. I had recently switched from Wicd to NetworkManager and thought NM was buggy. While researching the problem I saw the bug report and downgraded to 3.2.14 (last available in my pacman cache) -- wifi was again rock-solid. Then I downloaded the 3.3.2 packages from Testing and upgraded; wifi is still rock-solid. So it looks like the bug has been fixed in 3.3.2 (or the bad patch has been removed)...
 
+Felipe Contreras I heard what you said, but in the end it doesn't matter. The stable branch isn't stable because it has fewer bugs. The stable branch is stable because Linus draws a line in the sand and makes it so. Linus can't force you to use 3.x.y. As an admin of your own box, you must make that call. Linus is not always right, but Linux is his baby. If you want proof how hard headed he can be, look up the history of bitkeeper and git.
 
+Earl Matthews I don't think you know how the Linux kernel development works; Linus is not responsible for the stable series, Greg is.

And he is basically following rules, and the rules ensures that as long as v3.4 is in development, v3.3.x will tend to be more unstable. Period.
 
+Zack Buhman +Jason Lee I also got bitten by that.

But v3.2 is not prehistoric, in fact, it's probably the latest stable series. v3.3 will potentially remain more unstable than usual until v3.4 is released, or at least gets closer to the final release.
 
+Felipe Contreras You seem to be stuck thinking that the stable branch means stable, bug free code. This is more wrong than wrong. Stable is literally a label. It could have been called production and test or orange and pear. All these names are equally correct. So to say I don't understand how kernel development works when you don't understand the difference between stable and devel is mind boggling. Seriously, please go back and look at the history of Linux development (especially wifi which has always been troubled and buggy) and try to understand why they do what they do. The point of the process is not to produce bug free releases. It is to produce timely releases with the smallest patches. Wifi has been broken before, it wil be fixed and broken again. Such is Linux.
 
Simple reality when developing is no matter how good, you do miss things. Work hard to find most problems during pre-release as wider testing phases, so fewer problems to fix when released.

Matters little what products considered, problems usually found when production is flat out supplying mass markets.

Risk-sensitive businesses either wait a while before updating in hope fewer bugs left in need of a fix, or they select higher commercial grade support (eg Novell) specialists who will quickly fix any serious problems they chance upon.

Such is life.

The best measure is each release achieve a lower count for serious issues arising in production stage.

When cosmetic issues come to be regarded as serious :-)


My thanks to all the Technical types, who work at it doing a great job :-)
 
+Paul Parker I agree. Bugs also get triaged. Bugs in WiFi will always take a back seat to memory management or security. A fix might even be complete, but unapproved due to not having enough people to check the patch for issues. It happens.
 
+Zack Buhman Thanks! I didn't know about the Rollback Machine; I'll check it out next time I need something older. I did quickly go ahead and upgrade to 3.3.2 from Testing; I really wanted my Virtualbox back. :)

I'd also like to give props to all of the devs; I think they do a fantastic job on the kernel. Regressions like this don't happen very often, or if they have they haven't affected my day-to-day work. :)
 
+Earl Matthews You seem to be stuck thinking that the stable branch means stable, bug free code.

It's easy to win an argument by using straw man and argue against something your opponent has never said. No, I haven't said 'stable' means bug free code, there's no such thing as bug free code, that's nonsensical. OK?

What I'm saying is that if you find a patch that was introduced in v3.3.1 and breaks things, it makes sense to revert it for v3.3.2. That would make the system more stable.

That's not how 'linux-stable' works.

Stable is literally a label ... you don't understand the difference between stable and devel is mind boggling.

So you are saying 'linux-stable' is not meant to be stable?

All right.
 
+Felipe Contreras I get the feeling that no matter how many times I explain that linux-stable name does not guarantee stability and that stable is a label for the developers to stop adding code to the tree if they can, you'll just be angry that WiFi doesn't work in 3.3.x. The interesting part is that the argument is moot. The development model for Linux won't change. The patch is probably already in the works (if not done by now) and given enough time the issue will be resolved. This is how it has always been. The system is working. Please give people time to do their jobs.

Or just stay angry, I guess that works too.
 
+Earl Matthews You are not listening to what I am saying at all. The problem is not that ath9k got broken in 3.3.1, that's not it at all.
 
+Felipe Contreras Then we are at an impasse. I don't find a word you say convincing and you provide no hard data. If you can convince someone to modify the Linux kernel to meet your standard of stability in your time frame, for free, then I concede. Good luck.
 
IMHO this forum is NOT where convincing anyone shall modify Linux kernel to meet any desires.

Such convincing needs takes place in bug reporting, resolving and development groups which are elsewhere - and require a lot more effort.
 
+Felipe Contreras How does making this personal fix your problem? In fact, how has anything you've posted fixed your problem? Paul has a point. If you have a problem with an aspect of Linux you have to isolate it, collect data, and present it to the right people. Venting on Google+ is as useless as screaming at the sky.

As for my "straw man", I can only shrug. There is a difference between creating and presenting constructive arguments and bellyaching / complaining. All you need is solid data to make your point. Take your time. Google+ isn't going anywhere.
 
Was support for the ACPI driver for the Toshiba Tecra M7 dropped in the 3.3.x kernel series?
 
+Earl Matthews How does making this personal fix your problem? In fact, how has anything you've posted fixed your problem?

How is anything you've posted going to reduce global warming?

Whoa! Wait a second, you never said you had a problem with problem with global warming right?

So, what if I tell you to do X, Y and Z to fight against global warming? Well, that might be helpful, but it has nothing to do with what you said.

That's called a straw man, it's nothing personal, that's what it is.

So stop telling me how to fix global warming, or the economy, or any number of issues out there, that has nothing to do with what I said. You don't understand what I am saying, I do not have a problem that needs fixing.

I already told you but I will repeat it for you:

You are not listening to what I am saying at all. The problem is not that ath9k got broken in 3.3.1, that's not it at all.

As for my "straw man", I can only shrug. There is a difference between creating and presenting constructive arguments...

You may be presenting constructive and brilliant arguments to fight against global warming, they are still against something I never said, and thus straw man arguments.

I believe the issue (not mine, not related to ath9k) is beyond your capabilities to grasp. But if you want to propose solutions to me, I would suggest you try to understand what the issue really is, and ask first 'is this really the problem', and wait for a 'yes, that's the problem'. In the mean time you just don't get, and any 'solutions' you propose to whatever issues are afflicting the world out there are irrelevant, and a distraction.
 
+Felipe Contreras Hey, I remember this thread being about Linux 3.3.2 coming out. You know, ten days ago. Now Linux 3.3.3 is out. That's whats important. If you have no problems with 3.3.2 or 3.3.3 there isn't anything to discuss in a Linux thread.

Also, in the future, try to ease up with the insults and accusations, please. I've tried my best to be civil and I have a thick skin, but you have no idea what my capabilities are and zero data with which to estimate them. Slinging mud isn't going to convince me that you are right, and to be honest no one else cares.
 
+Earl Matthews If you have no problems with 3.3.2 or 3.3.3 there isn't anything to discuss in a Linux thread.

You still have no idea what the problem is about; it didn't happen in 3.3.2, but it might happen later. We'll see.

Also, in the future, try to ease up with the insults and accusations, please.

I have never insulted, or accused anybody.

If you think saying "you don't get it" is an insult, and "this is straw man" an accusation. Well, that's your problem.
Add a comment...