xx-sources

jeffroman · Posted: Fri Jul 16, 2004 12:32 pm Post subject:

are these sources based on mm or vanilla?
_________________
Thunder k8w, 2 Opteron 244s, 2GB Corsair, Audigy 2 Platinum eX, 6800 GT OC, 3Ware 9500S-4LP

yngwin · Posted: Fri Jul 16, 2004 12:33 pm Post subject:

As far as I understand they are based on mm, but the patch applies to vanilla.
_________________
"Those who deny freedom to others deserve it not for themselves." - Abraham Lincoln
Free Culture | Defective by Design | EFF

eldiablo · Posted: Fri Jul 16, 2004 12:50 pm Post subject:

xiphux, i could try to bug the people on lkml, maybe they might have a clue whats going on, but i dont think they would be to happy with me bugging them because of an experimental patchset

Edit: You don't need to make a broken out dir if you don't wan't to, or don't have time. I do think that it is sufficient enough to make a list of patches that are included, i've gained some experience in kernel hacking now, so i think i'm able to modify dem to apply

_________________
Always running the most experimental

wrc1944 · Posted: Fri Jul 16, 2004 3:21 pm Post subject:

I get the same error. I tried commenting out the last one, as you did, but it didn't work. Original error:

In file included from include/linux/security.h:30,
from init/main.c:35:
include/linux/sysctl.h:141: error: redefinition of `KERN_CPU_SCHED'
include/linux/sysctl.h:136: error: `KERN_CPU_SCHED' previously defined here
make[1]: *** [init/main.o] Error 1
make: *** [init] Error 2
--------------------------------------------------
Then I commented out (2 different ways), and got this:
----------------------------------------------------------------
wrc@mymachine linux-2.6.7 $ make bzImage
make[1]: `arch/i386/kernel/asm-offsets.s' is up to date.
CC init/main.o
In file included from include/linux/security.h:30,
from init/main.c:35:
include/linux/sysctl.h:141:9: invalid preprocessing directive #KERN_CPU_SCHED
make[1]: *** [init/main.o] Error 1
make: *** [init] Error 2
wrc@mymachine linux-2.6.7 $ make bzImage
----------------------------------------------------
Please clue me in as to the correct syntax to comment this out.
(from include/linux/sysctl.h) This is the correct file and lines, isn't it?

--------------------------------------------

*/
KERN_HZ_TIMER=65, /* int: hz timer on or off */
KERN_CPU_SCHED=66, /* CPU scheduler stuff */
KERN_INTERACTIVE=67, /* interactive tasks can have cpu bursts */
KERN_COMPUTE=68, /* adjust timeslices for a compute server */
KERN_KERNEL_PREEMPT=69, /* preemptible kernel scheduling */
KERN_VOLUNTARY_PREEMPT=70, /* voluntary preemption */
KERN_CPU_SCHED=71,
};
-----------------------------------------------

Thanks,
wrc1944
_________________
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-2.40-r5, gcc-14
kernel-6.12.8 USE=experimental python3_12.7-final-0

eldiablo · Posted: Fri Jul 16, 2004 3:30 pm Post subject:

In c/c++ you shouldn't comment out using #. #'s are used for definitions, and includes.
If you want to comment something out use //. It should then say //KERN_CPU_SCHED
It is also postilbe to comment something(often larger areas of text) by using this method

wrc1944 · Posted: Fri Jul 16, 2004 3:39 pm Post subject:

Never mind- instead of commenting, I just deleted it, and bzImage compiled 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-2.40-r5, gcc-14
kernel-6.12.8 USE=experimental python3_12.7-final-0

wrc1944 · Posted: Fri Jul 16, 2004 4:03 pm Post subject:

Thanks much eldiablo,
If I'm going to be doing this stuff, I guess I'd better read and introductory c/c++ tutorial, so I can at least edit files correctly.

Got the new 2.6.8-rc1-xx1 compiled, rebooted, and am using it right now. Looks good so far- guess I'll also try that renicing line in /etc/conf.d/local.start others mentioned.

wrc1944
_________________
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-2.40-r5, gcc-14
kernel-6.12.8 USE=experimental python3_12.7-final-0

Raku · Posted: Fri Jul 16, 2004 10:20 pm Post subject:

Hi,
I've got problems with running xx-sources since 2.6.7-xx6. Right now I'm trying 2.6.8-rc1-xx1. After installing new kernel, it doesn't boot. Screen hangs just after enabling APIC. From my observations, it's the moment, when screen is supposed to change it's resolution from default to 1024x768@85Hz. So, I disabled framebuffer console support (CONFIG_FRAMEBUFFER_CONSOLE) and everything start to work like it should. But now, i've got these ugly big fonts on my console ;-(
I'm using vesafb. Is it a problem with vesafb-tng??? How can I solve this?
_________________
raku

Powered by Archlinux ;-)

eldiablo · Posted: Fri Jul 16, 2004 10:50 pm Post subject:

I'm also suffering from the same problem, but i haven't testet do disable framebuffer, i hope it works

EDIT: Yeah, disabeling FB did work, i'm so glad to have an working experimental kernel again

_________________
Always running the most experimental

Raku · Posted: Fri Jul 16, 2004 11:57 pm Post subject:

trevorj · n00b Joined: 13 May 2003 Posts: 68

damn fb!

vesafb-tng obviously has some issues then. plus it doesn't switch from X to the fb res right sometimes. it makes it look like I'm having tunnel vision.

maybe it's our card? nvidia here... ill go try a couple things...
_________________
\\ trevorj

Raku · Posted: Sat Jul 17, 2004 9:24 am Post subject:

ok then - we know, that vesafb-tng sucks :-)

but i still wanna have a hi-res on my console. Is rivafb working fine? I tried to use it some time ago, but screen always locked when I was trying to get back from X to console. And I don't remember if it had high resolution support :-/
And don't wanna spend whole day trying it out :-/
_________________
raku

Powered by Archlinux ;-)

fallow · Posted: Sat Jul 17, 2004 9:24 am Post subject:

I had some problems with tng , so I`m using previous vesa-rrc version.
I don`t need any of tng features , so rrc is 100% efficient for me

_________________
"Time is a companion that goes with us on a journey. It reminds us to cherish each moment, because it will never come again. What we leave behind is not as important as how we have lived" J-L. Picard

Raku · Posted: Sat Jul 17, 2004 9:27 am Post subject:

well, maybe it's a good idea to bring again the possibility of choice beetween vesa-tng and vesa-rrc like it was in 2.6.7-xx4? (or sencond brunch of kernel tree with vesa-rrc only support?
_________________
raku

Powered by Archlinux ;-)

trevorj · n00b Joined: 13 May 2003 Posts: 68

meh. cvs is only up to 2.6.7-xx7. :evil:

_________________
\\ trevorj

eldiablo · Posted: Sat Jul 17, 2004 12:41 pm Post subject:

The CVS build is based upon 2.6.8-rc1-mm1, it is also called 2.6.8-rc1-xx1.
I don't know if there are any differences int the CVS version, and the main version.
_________________
Always running the most experimental

4nykey · Apprentice Joined: 11 Feb 2004 Posts: 180

BTW, there's a new tng patch for 2.6.8-rc1, seems to work here.

trevorj · n00b Joined: 13 May 2003 Posts: 68

problem with swsusp2:

xiphux · Posted: Sun Jul 18, 2004 12:03 am Post subject:

Hey guys, sorry for the delays. My time and network bandwidth have been a little monopolized lately.

I have an updated tree with kexec, 2.6.8-rc1-np1, compile fixes, etc, I'm just waiting for a chance to commit it...

Eldiablo, honestly, I've lost track of which patches I've applied. But you would probably be able to take advantage of cvs's log, by using cvs2cl.pl to see what was committed. If you really wanted, you could try checking out cvs snapshots of different times, to at least get an idea of where it started to break.
One thing that could help, though: have you tried enabling the magic sysrq key (in kernel debugging) and doing a register dump with Alt-SysRq-P? You don't have to post all the hex junk, but the human-readable parts would be helpful.

If you wanted, I could try the whole selectable vesafb tng/rrc thing again. But the option in xx4 was all sorts of busted, and trying again could quite likely end up the same way. So the people who are happily using vesafb now would have to be willing to take that risk...

trevorj, at line 137 of that file, change this code:

Raku · Posted: Sun Jul 18, 2004 12:37 am Post subject:

Someone has written here about new vesa-tng patch for 2.6.8-rc1. xiphux, maybe you can apply this patch instead of the one currently used (if currnet is older than that)?
here is link to it: http://dev.gentoo.org/~spock/projects/vesafb-tng/vesafb-tng-0.9-rc3-2.6.8-rc1.patch.bz2
_________________
raku

Powered by Archlinux ;-)

4nykey · Apprentice Joined: 11 Feb 2004 Posts: 180

I tried to revert some older patch then apply new one, there were couple of rejects and it ended up with this diff here: http://4nykey.nm.ru/tmp/vesafb-up.patch.bz2

xiphux · Posted: Sun Jul 18, 2004 3:02 am Post subject:

Vesafb-tng 0.9-rc3 is in cvs, as well as the other stuff I mentioned.

xiphux · Posted: Sun Jul 18, 2004 8:27 am Post subject:

I made the vesafb rrc/tng configure option again, from scratch. It's cleaner this time around, and hopefully a little more stable. I committed it into cvs - I'll wait a little and leave it as the latest commit, so it'll be easy to revert if it goes horribly wrong. And since I don't use vesafb, you guys'll have to be the guinea pigs...

eldiablo · Posted: Sun Jul 18, 2004 8:46 am Post subject:

I'll try this new kernel now, and see if it hangs at boot as well.
It's also nice that you included kexec, it's fine consept, but it's need a lot of work.

EDIT: I have a little problem cheking out the cvs resporitory, i only get this message after som while:

xiphux · Posted: Sun Jul 18, 2004 9:10 am Post subject:

That's weird, I'm not doing anything with the repository. You might have to wait for a bit until the server "figures out" that I'm not using it anymore.