Linux Kernel 2.6.29 released!

skottish · 2009-03-26 21:45:28

I just installed the third Arch release of this kernel and the disk errors haven't reappeared.

fukawi2 · 2009-03-26 22:18:57

skottish wrote:

It was mentioned on the mailing list and I'll bring it up here because it happened to me. All four of my ext4 partitions threw the "primary superblock features different from backup" error and ran the disk check. Three out of four were on the second boot after installing the new kernel, that last on the third. I searched around and most of the hits were related to ext4-dev (or whatever it was called). Is anyone else here seeing this?

Just to clarify -- does this cause data loss, or just failure to mount?

skottish · 2009-03-26 22:48:42

fukawi2 wrote:

skottish wrote:
It was mentioned on the mailing list and I'll bring it up here because it happened to me. All four of my ext4 partitions threw the "primary superblock features different from backup" error and ran the disk check. Three out of four were on the second boot after installing the new kernel, that last on the third. I searched around and most of the hits were related to ext4-dev (or whatever it was called). Is anyone else here seeing this?
Just to clarify -- does this cause data loss, or just failure to mount?

Fails to mount (error)--> restarts computer automatically--> runs disk check--> finds no errors (in my case)-- > boots normally.

I haven't had a chance to look at what the default settings are for ext4 now; I'm mounting with 'defaults'. Maybe something changed upstream?

Ranguvar · 2009-03-27 00:25:41

I haven't looked either, but you could grep though this for ext4.

skottish · 2009-03-27 03:50:54

Ranguvar wrote:

I haven't looked either, but you could grep though this for ext4.

It's a good idea, but there's far too many hits for a decent search. So, I'll either have to read the whole document or do it the scientific way: ignore it and hope that it goes away. I like the latter.

dr/owned · 2009-03-27 05:33:51

Y'all know what i'll be doing this weekend...tweaking and breaking! (the kernel)....and the install in general

Last edited by dr/owned (2009-03-27 05:34:15)

atomopawn · 2009-03-30 23:03:59

After upgrading, I have some very odd X server issues. But they may be related to upgrades of other packages. I'm using the intel driver so we'll see if updates to intel-dri and xf86-video-intel down the road can fix this.

skottish · 2009-03-30 23:13:32

atomopawn wrote:

After upgrading, I have some very odd X server issues. But they may be related to upgrades of other packages. I'm using the intel driver so we'll see if updates to intel-dri and xf86-video-intel down the road can fix this.

It probably is the intel drivers. It seems like everyone is having issues ranging from quirky to disastrous.

agapito · 2009-03-30 23:22:54

Why kms is not enabled by default on ARCH testing Kernel?

wonder · 2009-03-30 23:38:23

agapito wrote:

Why kms is not enabled by default on ARCH testing Kernel?

from http://kernelnewbies.org/Linux_2_6_29

In the X.org side, it's even more difficult. Because when the kernel modesetting support is enabled, it is /!\ REQUIRED /!\ that the X.org driver also has modesetting support. If you enable kernel modesetting and you don't have a modesetting-enabled driver, X.org will NOT be able to work in any way and it even may crash your machine. There's no way to workaround this, except disabling kernel modesetting (running a modesetting enabled X.org driver in a modesetting disabled kernel is allowed). Right now, only the Intel driver seems to have a stable release with modesetting support, alpha support is being worked on for other drivers.

imho kms should be disabled until majority of drivers can work or be enought stable.

Last edited by wonder (2009-03-30 23:39:40)

big_gie · 2009-03-30 23:40:43

Enablling kms in kernel requires kms-enabled userland. And its quite buggy. A kms enabled kernel will probably crash if userland does not support kms.

I suspect they will enable kms in intel driver/libdrm/... first. So people could try building their kms enabled kernel and test it. THEN a kms-enabled kernel will be provided.

KMS requires way to much things updated. Its too big of a risk.

sa · 2009-03-30 23:52:25

is btrfs stable?

burgulgoth · 2009-03-30 23:53:58

anrxc wrote:

Careful, networking seems to be broken http://thread.gmane.org/gmane.linux.ker … cus=811417

Yea, it seems to be like that. I'm getting very poor performances with my wireless card (b43). Transfer goes up to 200KB/s top. Wired network seems to be ok.

fukawi2 · 2009-03-31 00:15:07

sa wrote:

is btrfs stable?

No.

Ranguvar · 2009-03-31 02:33:51

sa wrote:

is btrfs stable?

The disk format isn't even settled yet! Easy, tiger (Well, it's done most likely, but they're still waiting before setting it in stone to make sure it has no huge flaws).

Remember, ext4 was merged into the main kernel two years ago... 2.6.28 was just when it was marked as stable. Btrfs is like ext4 from two years ago. Worse, actually. ext4 was still based off ext2/3 - not nearly as big a departure from previous filesystems as Btrfs - and Btrfs has far more features being worked on.

Btrfs aims to be the ZFS of Linux, but even if it was stable now (and nothing else), it's slower than molasses compared to most other filesystems now. ZFS is slower than most common FSes too, even on Solaris or FreeBSD, but not to the extent that Btrfs is. Don't worry, I'm sure they'll work on optimizing it before it's done, that's just not their main goal

Cypher · 2009-03-31 02:37:43

big_gie wrote:

Enablling kms in kernel requires kms-enabled userland. And its quite buggy. A kms enabled kernel will probably crash if userland does not support kms.
I suspect they will enable kms in intel driver/libdrm/... first. So people could try building their kms enabled kernel and test it. THEN a kms-enabled kernel will be provided.
KMS requires way to much things updated. Its too big of a risk.

if wanna try KMS, where can find the option to enable KMS?

big_gie · 2009-03-31 02:51:07

Cypher wrote:

if wanna try KMS, where can find the option to enable KMS?

For the userland tools I really don't know. Intel's driver had a --enable-kms back in the 2.4 or 2.5 days, but now it does not exist anymore. So I don't know what needs to be done. Same things for the rest. I think you need dri2 and GEM...

As for the kernel, its easy:

  │ │  Device Drivers  --->
  │ │    Graphics support  ---> 
  │ │      <*> Direct Rendering Manager (XFree86 4.1.0 and higher DRI support)  ---> 
  │ │        <*>   Intel 830M, 845G, 852GM, 855GM, 865G (i915 driver)  --->                              │ │
  │ │           i915 driver                                                                         │ │
  │ │             [*]       Enable modesetting on intel by default

skottish · 2009-03-31 03:25:36

For everyone asking about KMS, check out this thread. There may be something helpful there:

http://bbs.archlinux.org/viewtopic.php?id=67891

CheesyBeef · 2009-03-31 15:23:53

Ranguvar wrote:

sa wrote:
is btrfs stable?
The disk format isn't even settled yet! Easy, tiger (Well, it's done most likely, but they're still waiting before setting it in stone to make sure it has no huge flaws).
Remember, ext4 was merged into the main kernel two years ago... 2.6.28 was just when it was marked as stable. Btrfs is like ext4 from two years ago. Worse, actually. ext4 was still based off ext2/3 - not nearly as big a departure from previous filesystems as Btrfs - and Btrfs has far more features being worked on.
Btrfs aims to be the ZFS of Linux, but even if it was stable now (and nothing else), it's slower than molasses compared to most other filesystems now. ZFS is slower than most common FSes too, even on Solaris or FreeBSD, but not to the extent that Btrfs is. Don't worry, I'm sure they'll work on optimizing it before it's done, that's just not their main goal

Hmm, I did not know that... thanks for sharing!

Cypher · 2009-03-31 16:17:25

big_gie wrote:

Cypher wrote:
if wanna try KMS, where can find the option to enable KMS?
For the userland tools I really don't know. Intel's driver had a --enable-kms back in the 2.4 or 2.5 days, but now it does not exist anymore. So I don't know what needs to be done. Same things for the rest. I think you need dri2 and GEM...
As for the kernel, its easy:
  │ │  Device Drivers  --->
  │ │    Graphics support  ---> 
  │ │      <*> Direct Rendering Manager (XFree86 4.1.0 and higher DRI support)  ---> 
  │ │        <*>   Intel 830M, 845G, 852GM, 855GM, 865G (i915 driver)  --->                              │ │
  │ │           i915 driver                                                                         │ │
  │ │             [*]       Enable modesetting on intel by default

tnkz, now im recompiling 2.6.29 to test KMS.

in intel i trust

big_gie · 2009-03-31 17:05:17

Cypher wrote:

in intel i trust

I lost faith not long ago abotu kms...

Cypher · 2009-04-01 00:41:23

big_gie wrote:

Cypher wrote:
in intel i trust
I lost faith not long ago abotu kms...

works but have some problems: kdm4 shows garbage so you have problems to read things like "user" and "password", cant splashy works so have to disable and somethimes when you change from vt1 to vt7 kdm restart. But the text in vt1(init3) looks really nice

big_gie · 2009-04-01 00:42:32

My laptop screen is 15 inches but 1920x1200. So a 10pt font looks waaaaay to small. Last time I tried kms on it I could not see anything...

Alphalutra1 · 2009-04-01 00:55:43

Ranguvar wrote:

ZFS is slower than most common FSes too, even on Solaris or FreeBSD, but not to the extent that Btrfs is. Don't worry, I'm sure they'll work on optimizing it before it's done, that's just not their main goal

Any benchmarks to back this up?

Ranguvar · 2009-04-01 02:28:25

Alphalutra1 wrote:

Ranguvar wrote:
ZFS is slower than most common FSes too, even on Solaris or FreeBSD, but not to the extent that Btrfs is. Don't worry, I'm sure they'll work on optimizing it before it's done, that's just not their main goal
Any benchmarks to back this up?

Google is all I have Keep in mind that if the benchmarks of ZFS are on Linux, you're looking at ZFS hacked into FUSE, and not even optimized for FUSE, so it's hard to compare. If you really want to see ZFS's performance, you have to look at in on OpenSolaris or FreeBSD. It won't make a huge difference. but it's there.

http://kerneltrap.org/mailarchive/linux … 524/thread (FUSE)

I don't know a ton about high-end filesystem benchmarking, mail traiffc and such, but it seems those are situations in which ZFS can dominate. No clue.

Wait... looking at more stuff, ZFS might actually be _fast_. I have no clue anymore.
http://www.phoronix.com/scan.php?page=a … 2008&num=1
http://www.phoronix.com/scan.php?page=a … 2008&num=6

Arch Linux

#51 2009-03-26 21:45:28

Re: Linux Kernel 2.6.29 released!

#52 2009-03-26 22:18:57

Re: Linux Kernel 2.6.29 released!

#53 2009-03-26 22:48:42

Re: Linux Kernel 2.6.29 released!

#54 2009-03-27 00:25:41

Re: Linux Kernel 2.6.29 released!

#55 2009-03-27 03:50:54

Re: Linux Kernel 2.6.29 released!

#56 2009-03-27 05:33:51

Re: Linux Kernel 2.6.29 released!

#57 2009-03-30 23:03:59

Re: Linux Kernel 2.6.29 released!

#58 2009-03-30 23:13:32

Re: Linux Kernel 2.6.29 released!

#59 2009-03-30 23:22:54

Re: Linux Kernel 2.6.29 released!

#60 2009-03-30 23:38:23

Re: Linux Kernel 2.6.29 released!

#61 2009-03-30 23:40:43

Re: Linux Kernel 2.6.29 released!

#62 2009-03-30 23:52:25

Re: Linux Kernel 2.6.29 released!

#63 2009-03-30 23:53:58

Re: Linux Kernel 2.6.29 released!

#64 2009-03-31 00:15:07

Re: Linux Kernel 2.6.29 released!

#65 2009-03-31 02:33:51

Re: Linux Kernel 2.6.29 released!

#66 2009-03-31 02:37:43

Re: Linux Kernel 2.6.29 released!

#67 2009-03-31 02:51:07

Re: Linux Kernel 2.6.29 released!

#68 2009-03-31 03:25:36

Re: Linux Kernel 2.6.29 released!

#69 2009-03-31 15:23:53

Re: Linux Kernel 2.6.29 released!

#70 2009-03-31 16:17:25

Re: Linux Kernel 2.6.29 released!

#71 2009-03-31 17:05:17

Re: Linux Kernel 2.6.29 released!

#72 2009-04-01 00:41:23

Re: Linux Kernel 2.6.29 released!

#73 2009-04-01 00:42:32

Re: Linux Kernel 2.6.29 released!

#74 2009-04-01 00:55:43

Re: Linux Kernel 2.6.29 released!

#75 2009-04-01 02:28:25

Re: Linux Kernel 2.6.29 released!

Board footer