2.6.16-rc5-no2 "Rock like Ninja"|(All Of -mm)(Suspend2*)

cheater1034 · Veteran Joined: 09 Sep 2004 Posts: 1558

Looks to me as if it is a problem with the cpu scheduler (plugsched)

It is by default set to ingosched, I would try something different, such as nicksched, and see if you still get the problem.

Just specify the boot param, cpusched="", in the "" put nicksched, etc
_________________
IRC!: #zen-sources on irc.rizon.net
zen-kernel.org
--
Lost in android development land.

wrc1944 · Posted: Sat Mar 11, 2006 6:33 pm Post subject:

Same as KAMIKAZE_ & CoronaLover.

Checking if this processor honours the WP bit even supervisor mode... Ok.
divide error: 0000 [#1]
PREEMPT
last sys file:
Modules linked in:
CPU: 0
EIP: 0060:[<c02cff27>] Not tained VLI
EFLAGS: 00010246 (2.6.16-rc5-no2 #3)
...
<0> Kernel panic - not syncing: Attempted to kill idle task!

I was using FB, unlike KAMIKAZE_.

EDIT:
Just specify the boot param, cpusched="", in the "" put nicksched-

Doesn't Work. Will try others, but this might be something else, probably beyond my knowledge.
_________________
Main box- AsRock x370 Gaming K4
Ryzen 7 3700x, 3.6GHz, 16GB GSkill Flare DDR4 3200mhz
Samsung SATA 1000GB, Radeon HD R7 350 2GB DDR5
OpenRC Gentoo ~amd64 plasma, glibc-2.39-r7, gcc-14
kernel-6.9.3 USE=experimental python3_12.3-final

mbar · Veteran Joined: 19 Jan 2005 Posts: 1990 Location: Poland

wrc1944 · Posted: Sat Mar 11, 2006 6:57 pm Post subject:

When I use cpusched="nicksched" or staircase, ingo_11, etc in grub.conf kernel line, I'm getting" error 27 unrecognized command"???

Shouldn't that command be OK?
_________________
Main box- AsRock x370 Gaming K4
Ryzen 7 3700x, 3.6GHz, 16GB GSkill Flare DDR4 3200mhz
Samsung SATA 1000GB, Radeon HD R7 350 2GB DDR5
OpenRC Gentoo ~amd64 plasma, glibc-2.39-r7, gcc-14
kernel-6.9.3 USE=experimental python3_12.3-final

cheater1034 · Veteran Joined: 09 Sep 2004 Posts: 1558

The problem at my knowledge looks to be the scheduler.

It may be the sched-* -mm patches conflicting or something, with plugsched.

I am at my dad's house until sunday so looking at this problem further is impossible for me.

It looks like a cpu scheduler problem more than anything however.
---
Since the version of plugsched included is for -mm2, a -mm3 is not released yet. Possibly there are problems there somewhere.

I will look at this tomorrow, I will see if using -mm2 and plugsched fixes the issue.
_________________
IRC!: #zen-sources on irc.rizon.net
zen-kernel.org
--
Lost in android development land.

cheater1034 · Veteran Joined: 09 Sep 2004 Posts: 1558

Here is a final solutin to your kernel panic problem(s)

I need someone to go to #no-sources or #conrad tonight (on irc @ Freenode), and find joecool or jasonf. If you find joecool talk to him and show him your problem, if you find jasonf tell him to tell joecool to go to FGO in the no-sources release thread and look at the problem.

I can't do anything on this PC,and I have no laptop or anything. So I am very limited. Joecool may have a solution, or may be able to find one.

EDIT: Just thought I'd add, when I first booted this kernel up at my house on my desktop, both nicksched and ingo_ll (low latency) worked, I would try one of those 2 on boot.
_________________
IRC!: #zen-sources on irc.rizon.net
zen-kernel.org
--
Lost in android development land.

Tiger683 · Posted: Sat Mar 11, 2006 11:00 pm Post subject:

These are the -mm woes :/ I remember having a patch from -mm in nitro
wich made the generic IDE driver cause a kernel panic... So, all in all, the work we as patchsets
do ourselves ain't THAT bad after all...

If you don't mind i might look into this issue tomorrow too

PS: it would be nice for cheater if someone actually posted __USEFULL AND VERBOSE__ debug output,
please activate ALL debug options in kernel's debugging menu and post what it spits out then....
_________________
Retired gentoo user

cheater1034 · Veteran Joined: 09 Sep 2004 Posts: 1558

mbar · Veteran Joined: 19 Jan 2005 Posts: 1990 Location: Poland

Now I went back to rc4-no1 because rt2500 module (my wlan card) from portage does not compile with 2.6.16_rc5-no2

This and Suspend2 not working means rc5-no2 is no go on my laptop.

Phlogiston · Posted: Sun Mar 12, 2006 12:42 pm Post subject:

omg, there's a big mess with kernels at the moment. With no, suspend does not work , with latest nitrowar network does not work, and I with older ones I can't use fglrx.

What I need would be a good old nitro from Tiger

BTW: rc6 is out
_________________
Workstation: 5.1 SurroundSound, LIRC remote control; Laptop [IBM-T43]: patched sources, s2disk/ram, fingerprint sensor

wrc1944 · Posted: Sun Mar 12, 2006 7:05 pm Post subject:

For what it's worth, I just compiled a 2.6.16-rc6-mm1 on my nesl247 conrad "Evolution" box.

No problems compiling, and it runs fine so far. I'm using deadline scheduler.

EDIT: Regarding the mm kernels new ext3 multiple block allocation patches problem I mentioned on Page 1 of this thread:

I set all my ext3 partitions with tune2fs -c 1 /dev/hdxx just to make sure this didn't ocurr with this new mm kernel, and rebooted once and no problems so far. I'll leave this check setting on for a few more reboots so I'm convinced it is indeed not going to recurr. I don't really know if these patches were dropped, or fixed in the mm kernels. From what I originally had read about them, they seemed very promising.

EDIT 2: At ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.16-rc6/2.6.16-rc6-mm1/patch-list I found the following, so I guess they are still in mm kernels, and fixed, as I've rebooted again with check enabled, and no inode problems on any partition.

ext3-get-blocks-maping-multiple-blocks-at-a-once.patch
ext3-get-blocks-maping-multiple-blocks-at-a-once-vs-ext3_readdir-use-generic-readahead.patch
ext3-get-blocks-maping-multiple-blocks-at-a-once-get-block-chain-confliction-fix.patch
ext3-get-blocks-multiple-block-allocation.patch
ext3-get-blocks-support-multiple-blocks-allocation-in.patch
ext3-get-blocks-adjust-accounting-info-in.patch
ext3-get-blocks-adjust-accounting-info-in-fix.patch
ext3-get-blocks-adjust-reservation-window-size-for.patch
ext3-multi-block-get_block.patch
ext3-add-o-bh-option.patch
ext3-nobh-writeback-support-for-filesystems-blocksize.patch
_________________
Main box- AsRock x370 Gaming K4
Ryzen 7 3700x, 3.6GHz, 16GB GSkill Flare DDR4 3200mhz
Samsung SATA 1000GB, Radeon HD R7 350 2GB DDR5
OpenRC Gentoo ~amd64 plasma, glibc-2.39-r7, gcc-14
kernel-6.9.3 USE=experimental python3_12.3-final

SchAmane · Tux's lil' helper Joined: 11 Sep 2004 Posts: 107

cheater1034 · Veteran Joined: 09 Sep 2004 Posts: 1558

Ok, I am back, which means time to get to work.

Once plugsched-for-2.6.16-rc6-mm1 is released, expect a new unstable no-sources release.
---

Although I am slightly disappointed with this release, I won't give up

Suspend2 patches will not come until stable 2.6.16 is released, until then you need to settle for the suspend* from vanilla/mm.

2.6.16-rc6-no1 is still going to come, hopefully the day the new plugsched comes.

This release works great for a few people I heard, but the people who posted here have other thoughts

Here is a plan for future -no released.

2.6.16-rc6-no1 | UNSTABLE
-----
1) based on 2.6.16_rc6 and 2.6.16-rc6-mm1
2) Plugsched as cpu scheduler.
3) NO Suspend2 until stable release
4) Usual patches (tty config, fbsplash, vesafb-tng, filesystems, etc)
5) More -CK (except staircase/sched patches, no defaultcfq, and probally no dynticks, <It does not patch to -mm close to clean>)
6) BadRAM Patch being added to the mix

2.6.16-no1 | STABLE
1) based on 2.6.16
2) Plugsched as cpu scheduler.
3) Brand new *modified suspend2 patch
4) Usual Patches (tty config, fbsplash, vesafb-tng, filesystems, etc)
5) All of -CK except select few patches, such as staircase, which is in plugsched
6) BadRAM Patch from 2.6.16_rc6-no1 for 2.6.16
............

Again, this is the plan for the new releases, the scheduler issues are going to be fixed when the new plugsched is released. I emailed the BadRAM developer and got a 2.6.16_rc5 patch I am going to play with. Should be fun

EDIT:
Hotfix release coming tomorrow (2.6.16-rc5-no3) will be based on -mm2 instead of -mm3 for plugsched purposes, plugsched has been stripped and patched clean, should work to everyone's standards.
_________________
IRC!: #zen-sources on irc.rizon.net
zen-kernel.org
--
Lost in android development land.

g-d · n00b Joined: 13 Mar 2006 Posts: 3

squashfs don't work with 2.6.16-rc5-no2 and worked with 2.6.16_rc4-no1 and the same kernel config. Somebody knows the same problem ?

uncleben · n00b Joined: 21 May 2004 Posts: 23

I get the same kernel panic using nickshed

cheater1034 · Veteran Joined: 09 Sep 2004 Posts: 1558

2.6.16-rc5-no3 coming soon, fixing the issues on a lot of things, hopefully the squashfs issue (stripped and taken care of), the cpu scheduler issue should be non-existent, and badram is a nice feature for x86 and x86_64.
_________________
IRC!: #zen-sources on irc.rizon.net
zen-kernel.org
--
Lost in android development land.