2.6.10-nitro4: The Kernel Formerly Known As Hercules

[EricHsu]

EricHsu: I can't say that I've been experiencing the same problem; did you try running the hdparm script? Because I've noted some issues with that recently which may be related (hdparm takes over the processor, won't die)

No, I didn't use hdparm, since each version of kernels recognized and enable the DMA/UDMA of my maxtor disk quite well, and I have a weak heart, I've read the man page of hdparm, knowing that some of the options are dangerous, I'm afraid to even use it

My system hung yesterday due to the swap problem, then I wrongly pressed the power button for several seconds to force shutdown... then I broke my reiser4 / filesystem... I should have pressed the "reboot" :'(

[amorpheus]

Well, it seems that the ati fix doesnt work with 2.6.10-nitro4 but it works with 2.6.10-nitro2, any thoughts on that?

[justanothergentoofanatic]

Is this kernel now based upon mm-sources instead of ck-sources?

-Mike

[Pink]

Is this kernel now based upon mm-sources instead of ck-sources?

-Mike

Look at the first post, you know, the one by Darkness when he explains all about the patchset.

[Veldrin]

Hi all,

Darckness: Great Work!

I am now back where I was 2 days ago with 2.6.10-nitro2, just that it is nitro4... I do have some strange sound problem with UT2004 and AmericasArmy: the sound has some wierd stuttering, and seems laggy. short: It renders the games unplayable. On the other hand Q3A work flawless, and any other applcation that uses sound ( mplayer, vlc, xmms, bmp, rhythmbox, even arts...). Am I the only one experiencing this problem, or is generally known.

I'd like to do some more testing on other kernels, but I don't know which ones support resier4. I tried mm-sources, but it wouldn't compile my ATI drivers, so it was pointless going on... BTW same happens if I try nitro4-mm...

I switched to resier4 with a 2.6.8 kernel, and there it works... lookily I still have this one...

Greets
V.
Edit: Typo

[bollucks]

Hi all,

Darckness: Great Work!

I am now back where I was 2 days ago with 2.6.10-nitro2, just that it is nitro4... I do have some strange sound problem with UT2004 and AmericasArmy: the sound has some wierd stuttering, and seems laggy. short: It renders the games unplayable. On the other hand Q3A work flawless, and any other applcation that uses sound ( mplayer, vlc, xmms, bmp, rhythmbox, even arts...). Am I the only one experiencing this problem, or is generally known.

I'd like to do some more testing on other kernels, but I don't know which ones support resier4. I tried mm-sources, but it wouldn't compile my ATI drivers, so it was pointless going on... BTW same happens if I try nitro4-mm...

I switched to resier4 with a 2.6.8 kernel, and there it works... lookily I still have this one...

Greets
V.
Edit: Typo

The openAL drivers in UT engine games use OSS by default. This makes them behave badly and recent incarnations of the staircase scheduler make them stutter. Check -ck's website for the faq on audio issues and how to prevent this problem:
http://kernel.kolivas.org

[FirechilD]

someone knows how to turn on the anticipatory sheduler ?

append = "elevator=as" in my lilo.conf wont work :/

[bollucks]

someone knows how to turn on the anticipatory sheduler ?

append = "elevator=as" in my lilo.conf wont work :/

it's not longer as; it's anticipatory
elevator=anticipatory

[FirechilD]

ive tested it with append = "elevator=anticipatory" - its not working too :/

[rmh3093]

I'm experiencing something weird:

If the system is under some kind of heavy load and the swap is in use, the system will hang on "Deactivating swap..." while reboot or shutdown. Then I have to press the "reboot" button (Ctrl + Alt + Del doesn't work...)

Is this only me? I'm using the 2.6.9-nitro4 now...

I know what you are saying, I was having that with nitro4-mm sometime also, plus none of the orinoco drivers would work with mm


and I tried out morph10 for shits and giggles..... alsa 1.08 seems to have better mixing, I wasnt getting any skipping while running Civ3 in cedega.

[chunderbunny]

I'm getting a weird error on compiling:

Code: Select all

 ...
  CC      fs/reiser4/plugin/dir/hashed_dir.o
  CC      fs/reiser4/plugin/dir/pseudo_dir.o
  CC      fs/reiser4/plugin/dir/dir.o
fs/reiser4/plugin/dir/dir.c:1538: error: stray '`' in program
fs/reiser4/plugin/dir/dir.c:1538: error: stray '\131' in program
fs/reiser4/plugin/dir/dir.c:1538: error: parse error before '%' token
fs/reiser4/plugin/dir/dir.c:1538: error: stray '\211' in program
fs/reiser4/plugin/dir/dir.c:1539: warning: function declaration isn't a prototype
fs/reiser4/plugin/dir/dir.c: In function `detach_fsdata':
fs/reiser4/plugin/dir/dir.c:1542: error: `f' undeclared (first use in this function)
fs/reiser4/plugin/dir/dir.c:1542: error: (Each undeclared identifier is reported only once
fs/reiser4/plugin/dir/dir.c:1542: error: for each function it appears in.)
make[2]: *** [fs/reiser4/plugin/dir/dir.o] Error 1
make[1]: *** [fs/reiser4] Error 2
make: *** [fs] Error 2

This is using a .config from my old 2.6.9-nitro3 kernel. Any ideas?

[Veldrin]

The openAL drivers in UT engine games use OSS by default. This makes them behave badly and recent incarnations of the staircase scheduler make them stutter. Check -ck's website for the faq on audio issues and how to prevent this problem:
http://kernel.kolivas.org

Yep, that did the trick...
I had to create a file ~/.openalrc and add

Code: Select all

(define speaker-num 2)
(define devices '(alsa))
(define alsa-out-device default)

It working fine....

Greets
V.
And THX again

[DaDead]

yup ati fix dont work

[jannis]

Hey I found some kernel patches on:

http://kerneltrap.org/node/4493

Would this be something to check out when you include it to nitro-sources? If it's crap, throw it away but there have been some positive reactions.

[Guinpen]

The kernel sounds like a fine piece of work! I can't wait to try it out. Thanks for your work!

[bollucks]

ive tested it with append = "elevator=anticipatory" - its not working too :/

Check you haven't stripped your config down and removed support for the anticipatory scheduler in the process. Note you can also set it once the kernel is already running with

echo anticipatory > /sys/block/hda/queue/scheduler

And you check it with

cat /sys/block/hda/queue/scheduler

it looks like this:
noop deadline [cfq] anticipatory
where [] means cfq is running

[discomfitor]

jannis: I was looking into them, but I can't get the as patch for some reason. If somebody else has it and could link me to it, I would appreciate it.

[Teh Penguin D00d]

Got it working. Awesome.

For those curious...hotplug depends on General Setup -> sysctl. That's what creates the directories in /proc.

This is neither covered in the Gentoo UDEV documentation, nor the guide I used to originally setup UDEV, http://webpages.charter.net/decibelshel ... rimer.html

Maybe that'll save some poor sap endless hours of frustration.

[borkdox]

Nvidia kernel fails to emerge in this kernel?

I am running AMD64 and when trying to emerge nvidia-kernel I get:

Code: Select all

/var/tmp/portage/nvidia-kernel-1.0.6629-r1/work/NVIDIA-Linux-x86_64-1.0-6629-pkg2/usr/src/nv/nv.c: In function `_get_phys_address':
/var/tmp/portage/nvidia-kernel-1.0.6629-r1/work/NVIDIA-Linux-x86_64-1.0-6629-pkg2/usr/src/nv/nv.c:2529: error: structure has no member named `pud'
/var/tmp/portage/nvidia-kernel-1.0.6629-r1/work/NVIDIA-Linux-x86_64-1.0-6629-pkg2/usr/src/nv/nv.c: In function `nv_agp_init':
/var/tmp/portage/nvidia-kernel-1.0.6629-r1/work/NVIDIA-Linux-x86_64-1.0-6629-pkg2/usr/src/nv/nv.c:3010: warning: `inter_module_put' is deprecated (declared at include/linux/module.h:577)
make[3]: *** [/var/tmp/portage/nvidia-kernel-1.0.6629-r1/work/NVIDIA-Linux-x86_64-1.0-6629-pkg2/usr/src/nv/nv.o] Error 1
make[2]: *** [_module_/var/tmp/portage/nvidia-kernel-1.0.6629-r1/work/NVIDIA-Linux-x86_64-1.0-6629-pkg2/usr/src/nv] Error 2
NVIDIA: left KBUILD.
nvidia.ko failed to build!
make[1]: *** [module] Error 1
make: *** [module] Error 2

!!! ERROR: media-video/nvidia-kernel-1.0.6629-r1 failed.
!!! Function linux-mod_src_compile, Line 284, Exitcode 2
!!! Unable to make IGNORE_CC_MISMATCH=yes V=1 SYSSRC=/usr/src/linux SYSOUT=/usr/src/linux clean module.
!!! If you need support, post the topmost build error, NOT this status message.

[Guinpen]

I may be a retard, but upon startup I get a message:

Coldplugging pci devices...
... can't load module shpchp
missing kernel or user mode driver shpchp
... can't load module shpchp
missing kernel or user mode driver shpchp

I try modprobe shpchp and get:

FATAL: Error inserting shpchp (some path I don't remember): Operation not permitted.

I don't know what this module is for, and everything seems to be working. But it's really annoying, so anyone have any ideas?

Otherwise, awesome kernel! Waay better than gentoo-dev-sources 2.6.9-r13!!

[thubble]

Just curious: what bad things happen with CONFIG_IOPRIO_WRITE=y ?

well i think fs corruption ? ... booting nitro3 and running it for a bit toasted one install using reiser4 fs. possibly this option was the cause... havent tried it on any others.

I noticed that the latest version of cfq-ts (050108 in -ckdev) no longer has this warning at the top, and Con has included a patch (cfq_queue_pid_fix.diff) which appears to be an attempt to fix the problem. Does this mean it's now safe to enable this option? Also, what's the worst that can happen, is it just a crash/hang or could there be actual data loss or corruption?

[bollucks]

Just curious: what bad things happen with CONFIG_IOPRIO_WRITE=y ?

well i think fs corruption ? ... booting nitro3 and running it for a bit toasted one install using reiser4 fs. possibly this option was the cause... havent tried it on any others.

I noticed that the latest version of cfq-ts (050108 in -ckdev) no longer has this warning at the top, and Con has included a patch (cfq_queue_pid_fix.diff) which appears to be an attempt to fix the problem. Does this mean it's now safe to enable this option? Also, what's the worst that can happen, is it just a crash/hang or could there be actual data loss or corruption?

CONFIG_IOPRIO_WRITE just doesn't build so there is no danger in it at all. The newer ckdev snapshots have support for it so it builds and is safe. No filesystem corruption is likely; the worst possible is a hang (according to Jens Axboe)

[discomfitor]

As bollucks says, there's no possibility of real damage, but it isn't marked as safe yet so I'm just making sure people know not to use it. Otherwise they'll post here to tell me it doesn't work . I'll remove the warning when I can run it at full stability.

On the next nitro release:
Unless people start giving me tons of patches that they need implemented very soon, or exposing huge bugs in this version, I'm gonna let this one go for a few days. I've been irritating a lot of people on my end with my constant kernel patching, and I think it's time to give them a little break. Of course, if there's need for another release soon I'll be all over it.

[bartmasz]

Hi!

I’ve been struggling with nitro2 and nitro4 (without mm) to get my soundcard work (onboard SiS 7012). I tied various ways (build-in | module, with and without OSS). There were no errors, soundcard was detected. Aplay, mplayer and xmms played every thing except the fact that I couldn’t hear the sound!
Finally I emerged gentoo-dev-sources-2.6.9-r13 and without changing anything in the alsa config sound works fine.
That’s all

[Teh Penguin D00d]

I had problems with ALSA as well. I had to open alsa mixer, and change some of the settings. 2.6.9 and 2.6.10 seem to have different ALSA mixer settings. Try loading alsamixer and checking to see if everything is unmuted/turned up.