Old posts from the morph-sources thread

[yardbird]

Hi Yardbird : you are right just not compiling the power management did the trick.
Now the kernel compile nicely and as i am not using power management it's ok.

Nice to hear this. The software-suspend main developer is getting an amd-64 box soon, so expect software-suspend to work also on x86-64 soon.

I don't know what the difference was. I used the same configuration for morph9 that I use for morph8. I got a warning message with morph 9 that said something about bad symbol and missing get_state with reiser4. Morph9 would not boot. It hung at the splash screen and did nothing. I went back to morph8 and booted again with no problem.

This sounds pretty strange, since there are no differences in the reiser4 patches between morph8 and morph9. Let me ask you: are you using a laptop? Are you using software-suspend? Have you compiled reiser4 as module? May you send me the exact error message you are getting?

i'll report in on morph9 tonight. its compiling and i'll test it for a few hours.

Thanks for the help.

[gungholady]

No, I'm using a desktop and do not use software suspend. I wasn't remembering the error correctly. I don't build the reiser4 as a module. It is built in. Here is the error message that occurs after the modules are installed:

WARNING: /lib/modules/2.6.10-morph9/kernel/fs/reiser4/reiser4.ko needs unknown symbol find_get_pages_tag

Whoops, I double checked and somehow my reiser4 got marked as a module instead of being built in like I need it. I'll change that, then see if does it without an error.

Edit:

It works great when I have reiser4 built in like it should have been. Thanks so much for providing an ebuild for a kernel that includes the three things that I use most. Win4Lin, fbsplash and reiser4. Most of the kernels that include Win4Lin, the patches don't get applied properly for win4lin to work. I have to do patching by hand. Not so with yours, it works correctly. The patches that I use all apply correctly.

[darklegion]

Well it turns out that the reason I could not see the scheduler entries was because the /sys dir I had was actually full of files I had from an a hdd backup a while back,so sysfs was not actually mounted...just a copy of it from Novemeber 16.

[feld]

runnin smooth here yardbird. another reliable release from you.


-Feld

[Gentree]

Impressive patch set, nice work!

I grabbed this kernel to try out realtime LSM since I want to set up a decent sound capability.

I've just discovered morph and I love it.

I've had a persistent problem with programs being very slow to start since around the end of October. I thought my system was getting bloated or that I had a library loading issue. Morph just blew that all away.

emerge was taking fully 15 seconds to even start before , now it's straight up.

I just did a quick check with gxine and mplayer whilst doing emerge kdebase since I saw a few posts about jumpiness.

That may have been earlier versions, in any case not a glitch anywhere here.

You've been busy it seems with the number of versions you've pushed out in the last month , so many thanks for you efforts.

A really top kernel patch set. I wave my feathered cap in a large circle and bow low. "Chapeau!"

[rommel]

hey yardbird this kernel is really nice... one suggestion , i see you added the mouse polling patch but the non-configurable one... might possibly be better to add this one instead it will have a kernel configuration entry under hid.

[Evil Dark Archon]

it's not non-configurable, you just have to configure it differently, you have to use either modules.d or a kernel boot paramater.

[rommel]

yes i know Evil.. thanks for pointing out the obvious and not getting the point i was making... using the patch i linked adds a kernel config option to set the polling rate. get this patch and patch -p1 -R then get the one i linked above and apply it. then just run make oldconfig.

[yardbird]

Hi folks,

2.6.10-morph10 is out. A number of updates in there. Added orinoco and hostAP patches (please report!!), updated to alsa-1.0.8rc2 (from 2.6.10-mm2), updated software-suspend version and inotify patch.

Please hammer it

Impressive patch set, nice work!

Thanks, but the real credit goes to all the people that actually develop all these patches. I'm a mere patch-monkey

I just did a quick check with gxine and mplayer whilst doing emerge kdebase since I saw a few posts about jumpiness.

That may have been earlier versions, in any case not a glitch anywhere here.

Yes, this has been resolved since I dropped Con Kolivas' staircase scheduler.

A really top kernel patch set. I wave my feathered cap in a large circle and bow low. "Chapeau!"

Merci!

hey yardbird this kernel is really nice... one suggestion , i see you added the mouse polling patch but the non-configurable one... might possibly be better to add this one instead it will have a kernel configuration entry under hid.

Thanks for trying this out. I think I'll stay with the latest patch though, since it is more flexible and does not require to recompile the kernel to change the polling rate. Sorry.

[cr0w]

Just trying the new 2.6.10-morph10 release - the suspend2 patch seems to be broken as make can't find suspend_dm.c.

[cr0w]

Oh well - on further investigation the suspend2 device mapper module is not required.

[krejler]

yes i know Evil.. thanks for pointing out the obvious and not getting the point i was making... using the patch i linked adds a kernel config option to set the polling rate. get this patch and patch -p1 -R then get the one i linked above and apply it. then just run make oldconfig.

[EDIT... Removed my comment, I misread your post]

Since a lot of people liked the option in Kconfig, I'll just continue with it. It just didn't seem right to put it there.

I'll post a link to the new patch shortly.

[m0sia]

great work! it works great!

btw anybody know how to use swsusp2 with fbsplah?

[krejler]

yes i know Evil.. thanks for pointing out the obvious and not getting the point i was making... using the patch i linked adds a kernel config option to set the polling rate. get this patch and patch -p1 -R then get the one i linked above and apply it. then just run make oldconfig.

[EDIT... Removed my comment, I misread your post]

Since a lot of people liked the option in Kconfig, I'll just continue with it. It just didn't seem right to put it there.

I'll post a link to the new patch shortly.

Original, same as 'add-mousepoll-param....':
http://omfg.linux.dk/pub/configurable-h ... aram.patch.

The Kconfig entry:
http://omfg.linux.dk/pub/configurable-h ... nfig.patch.

Make the Kconfig entry actually work:
http://omfg.linux.dk/pub/configurable-h ... iver.patch.

Instead of the three ones above, the full one, which includes all three, is available:
http://omfg.linux.dk/pub/configurable-h ... FULL.patch.

This should, hopefully, make everyone happy!

Farewell!

[krejler]

I just noticed your comment on this, yardbird, sorry for not noticing it earlier.

Thanks for trying this out. I think I'll stay with the latest patch though, since it is more flexible and does not require to recompile the kernel to change the polling rate. Sorry.

Actually, it doesn't.

The old one incorporated a module parameter too, although, it didn't have a sysfs entry for it (which, for instance, the patches above provides). It would use value set in Kconfig as the default, and use the value provided by the parameter, if a parameter was actually passed.

The patches I posted above should work like I just described. It only uses the value fron the config as a last resort.

Anyway, it's up to you.

[Robin79]

The alsa still dosent show any sign of the azx driver in the kernel config and that sucks

[mayday147]

I patched the vanilla sources with morph10 , but I cannot see any reiser4 in the .config. The patch has applied flawlessly, though.

[yoyo]

I patched the vanilla sources with morph10 , but I cannot see any reiser4 in the .config. The patch has applied flawlessly, though.

You have to unselect the "4k stack" (or stuff like this) in another part of menuconfig to see reiser4 options.

Enjoy !

[mayday147]

I patched the vanilla sources with morph10 , but I cannot see any reiser4 in the .config. The patch has applied flawlessly, though.

You have to unselect the "4k stack" (or stuff like this) in another part of menuconfig to see reiser4 options.

Enjoy !

You're right. It did the trick. Thanks.

But isn't 4k stacks supposed to be a good thing? What happened?

[yardbird]

The old one incorporated a module parameter too, although, it didn't have a sysfs entry for it (which, for instance, the patches above provides). It would use value set in Kconfig as the default, and use the value provided by the parameter, if a parameter was actually passed.

The patches I posted above should work like I just described. It only uses the value fron the config as a last resort.

Anyway, it's up to you.

Ok, thanks for the head-up. I'll include your updated "total" patch in the next release. Oh, and thanks also for your work on it

great work! it works great!

btw anybody know how to use swsusp2 with fbsplah?

You mean fbsplash turning on and showing progress-bar when hibernating? It is in the TODO list of the devs, can't say when it will be implemented though.

The alsa still dosent show any sign of the azx driver in the kernel config and that sucks

It seems that the azx driver is not ready for prime time yet. I'm sorry to say that you'll have to wait

You're right. It did the trick. Thanks.

But isn't 4k stacks supposed to be a good thing? What happened?

Reiser4 used to have problems with 4k stacks on. Now these problems seems to be resolved, and I'm running ok with reiser4+4stacks at the moment. I may remove this limitation in the next morph release.

I'm working on -morph11. The highlights are:
- the cfq-ioprio patches by Jens Axboe, which now promise full io priority support also for writes (check the -ck mailing list archives for details). This means yet more system responsivity under high io-loads, yay!
- the update to -ac6, with many bugfixes with respect to the vanilla kernel;
- an up-to-date reiser4 patch from 2.6.10-mm2;
- more sofware-suspend fixes.

Thanks to all who are testing morph and helping in ironing out the issues they find. I really appreciate it

Cheers!

[mayday147]

If you appreciate it so much , then I let you know that I compiled your patch with no errors and running smoothly, yet . Good job.

[m0sia]

great work! it works great!

btw anybody know how to use swsusp2 with fbsplah?

You mean fbsplash turning on and showing progress-bar when hibernating? It is in the TODO list of the devs, can't say when it will be implemented though.

Thanks to all who are testing morph and helping in ironing out the issues they find. I really appreciate it :D

Cheers!

thanks for your answer.

ps Maybe you know why vesafb-tng works with nvidia cards only 800x600,when tv is connected? i think it is problem with vesa_find_vbe_mode(), but i have no time to explore this bug. I'll try to find it out next week.

pps I have a little patch to kernel here: http://bugme.osdl.org/show_bug.cgi?id=3931
but it is still marked "new". maybe i must post it in lkml? I really want to add new unusial usb device in drivers/usb/storage/unusual_devs.h

sorry for bad english.

[feld]

looks good yardbird, but nothing in this new release that affects my setup at all. i'll try -11 since it will have new reiser4 stuff and the new io stuff.


btw, are u gonna keep the same thread for all your kernels? if so i'll just bookmark it...



-Feld

[krejler]

yes i know Evil.. thanks for pointing out the obvious and not getting the point i was making... using the patch i linked adds a kernel config option to set the polling rate. get this patch and patch -p1 -R then get the one i linked above and apply it. then just run make oldconfig.

[EDIT... Removed my comment, I misread your post]

Since a lot of people liked the option in Kconfig, I'll just continue with it. It just didn't seem right to put it there.

I'll post a link to the new patch shortly.

Original, same as 'add-mousepoll-param....':
http://omfg.linux.dk/pub/configurable-h ... aram.patch.

The Kconfig entry:
http://omfg.linux.dk/pub/configurable-h ... nfig.patch.

Make the Kconfig entry actually work:
http://omfg.linux.dk/pub/configurable-h ... iver.patch.

Instead of the three ones above, the full one, which includes all three, is available:
http://omfg.linux.dk/pub/configurable-h ... FULL.patch.

This should, hopefully, make everyone happy!

Farewell!

Oooops.

The patches above were broken. They've been updated.
Sorry for any inconvenience it may have caused.

[babykart]

Hi yardbird
I've just discovered morph and I love it...
very great work!