Old posts from the morph-sources thread

Message

ahorn · Post by **ahorn** » Sun Jun 26, 2005 8:35 am

2.6.12-morph3 is fine. hope a better ipw2200 patch will be in the next mm-tree, because some features are still not in, but in the ipw2200 package they're. anyway, your tree is stable, so do the ebuild! morph-sources are always stable.

Guinpen · Post by **Guinpen** » Sun Jun 26, 2005 9:57 am

yardbird wrote:
Godji wrote:I downloaded the morph3 patch (a bz2 file) but I'm not sure what I have to do with it. I'm coming from love-sources and there you always had an ebuild
Download the complete tarball of vanilla kernel 2.6.12, extract it, cd in the source tree and apply the morph patch with:
Code: Select all
bzcat /path/to/patch-2.6.12-morph3.bz2|patch -p1
More info here:

http://morph-sources.homelinux.net/medi ... ph-sources

The ebuild will come when the 2.6.12 branch is marked stable.

Thank you!!

rel · Post by **rel** » Sun Jun 26, 2005 10:15 am

So 2.6.12-morph3 already contain the same patches as in vanilla 2.6.12.1 (CAN-2005-1761 & CAN-2005-1913) Sweet.

Maybe a question that is already answered, but alas:
Do you provide split-out patches with a series file? Would be awesome!

Well, going to test 12-morph3 now.
Having an issue though with this kernel and I guess also with the gentoo one, it is about the i8k modules failig to load.
These modules manage my laptop's fans and buttons, especially the fans is important.

First off to find out if this problem is there with vanilla.

-rel

_loki_ · Post by **_loki_** » Sun Jun 26, 2005 9:46 pm

I found the download-section on http://morph-sources.homelinux.net/ empty and the ebuilds from http://sourceforge.net/projects/morph-sources seems not to work (got an error while runnig the digest)..- after all I got a working ebuild from http://sf.gds.tuwien.ac.at/m/mo/morph-sources/ but not for the 2.6.12 series.. Where can I get that or do I have to patch vanilla myself?

yardbird · Post by **yardbird** » Mon Jun 27, 2005 1:03 am

@rel: I'll see what I can do about the split-out patches.

@_loki_: the website had some problems today, it should be working ok now. If you need directions you can take a look here:

http://morph-sources.homelinux.net/medi ... ph-sources

snakie · Post by **snakie** » Tue Jun 28, 2005 1:53 pm

When I try to download overlay package (morph-overlay-0.2.tar.bz2) from sourceforge, http://prdownloads.sourceforge.net/morp ... .2.tar.bz2 there are SYS-APPS/SYSVINIT directories inside with sysvinit ebuild stuff - manifests etc. It seem like wrong package, isnt' it?

yardbird · Post by **yardbird** » Thu Jun 30, 2005 12:31 pm

Hello,

sorry for the site downtime, my flatmate turned down the server -_-

There's an ongoing discussion on the morph website. The issue is whether to base the next morph-sources on -ck or not. More details are available here.

You can post your opinions here too, and I'm eager to hear them

yardbird · Post by **yardbird** » Thu Jun 30, 2005 12:34 pm

snakie wrote:When I try to download overlay package (morph-overlay-0.2.tar.bz2) from sourceforge, http://prdownloads.sourceforge.net/morp ... .2.tar.bz2 there are SYS-APPS/SYSVINIT directories inside with sysvinit ebuild stuff - manifests etc. It seem like wrong package, isnt' it?

You have to download the ebuilds, not the overlay. The overlay was for an older release of morph-sources and at the moment it is of no use. If you need more directions you can look here.

mrfree · Post by **mrfree** » Fri Jul 01, 2005 8:11 am

yardbird wrote:The issue is whether to base the next morph-sources on -ck or not.

Noooooooooooooooooooooooooooooooooooooooooooooooooooooo

I've used ck sources (2.6.11 series) some months ago with evident memory management problems like frequent swap storming and in general bad performance at all (specially in I/O terms). Actual 2.6.11 morph sources series are, in my case, more more more powerful than ck one.

My system is an IBM Thinkpad A31 notebook.

This is my 2 euro-cents

Ciao

ahorn · Post by **ahorn** » Sat Jul 02, 2005 5:34 am

i want it mm based

dhalsiim · Post by **dhalsiim** » Sun Jul 03, 2005 9:54 pm

Bleh .. I sync'ed my tree yesterday and it pulled 2.6.13-r1 .. I don't suppose the patchset would work with it. I'm going to give it a shot.. if it doesn't work I'm reverting back to 2.6.12.1

lamekain · Post by **lamekain** » Mon Jul 04, 2005 8:10 pm

Software suspend 2 doesn't work. To be precise: suspending works like a charm, but resuming hangs just after copying the old kernel back (after the good old textmode progressbar finishes). The same problem occured with nitro-2.6.12.. probably the swsusp version 2.1.9.5 is to blame.

baeksu · Post by **baeksu** » Mon Jul 04, 2005 11:02 pm

lamekain wrote:Software suspend 2 doesn't work. To be precise: suspending works like a charm, but resuming hangs just after copying the old kernel back (after the good old textmode progressbar finishes). The same problem occured with nitro-2.6.12.. probably the swsusp version 2.1.9.5 is to blame.

Works fine here. Maybe some problem with the settings (or nvidia?)?

m0sia · Post by **m0sia** » Tue Jul 05, 2005 5:30 am

[quote="baeksu"][quote="lamekain"][b]Software suspend 2[/b] doesn't work. To be precise: suspending works like a charm, but [u]resuming[/u] hangs just after copying the old kernel back (after the good old textmode progressbar finishes). The same problem occured with nitro-2.6.12.. probably the swsusp version 2.1.9.5 is to blame.[/quote]
Works fine here. Maybe some problem with the settings (or nvidia?)?[/quote]

what problems can cause nvidia?

lamekain · Post by **lamekain** » Tue Jul 05, 2005 6:01 am

baeksu wrote:
lamekain wrote:Software suspend 2 doesn't work. To be precise: suspending works like a charm, but resuming hangs just after copying the old kernel back (after the good old textmode progressbar finishes). The same problem occured with nitro-2.6.12.. probably the swsusp version 2.1.9.5 is to blame.
Works fine here. Maybe some problem with the settings (or nvidia?)?

Hmm.. The settings should be fine (I have encryption disabled and compression is lzf in the config-file). And I don't have nvidia. Suspending worked fine with earlier morphs ( if I remember right it was 2.6.12-rc6 ).

yardbird · Post by **yardbird** » Tue Jul 05, 2005 11:22 am

Hello,

the next morph release should come soon, once a remaining issue is solved (more details on the website).

@lamekain: suspend2 still has some quirks and its success rate may well be heavily dependent upon its version. In the next release there will be and updated suspend2 patch, I hope it solves the problem for you.

@m0sia: nvidia's binary driver has caused problems in suspend2 in the past (most likely because of non standard PCI suspend/resume code), but it should work ok now.

mayday147 · Post by **mayday147** » Wed Jul 06, 2005 11:35 am

Why isn't the mppe-mppc patch present anymore in your patchset? Do I have to apply it myself?

nightfrost · Post by **nightfrost** » Wed Jul 06, 2005 9:22 pm

I see there's been some talk about 2.6.12 and ipw2200. What's the deal really? I can't compile ipw2200 against the latest morph - is that expected behaviour?

mlybarger · Post by **mlybarger** » Fri Jul 08, 2005 10:25 am

i've read that nitro and morph sources contain the suspend2 patchset. does anyone have these recent sources working w/ ndiswrapper? are there other sources containing the suspend2 patchset (outside the suspend2-sources)?

those are my two _main_ needs in a kernel, suspend2 and ndiswrapper (which seems to be a separate ebuild?). i have no idea why anyone would use the resume in the vanilla kernel outside of in an apache serrver.

i can't build ndiswrapper w the morph-sources-2.6.11-r12

nightfrost · Post by **nightfrost** » Fri Jul 08, 2005 10:34 am

take a look at the "known issues" section of the wiki. Here's a link: http://morph-sources.homelinux.net/medi ... own_Issues. It might be what you need.

rel · Post by **rel** » Fri Jul 08, 2005 6:09 pm

latest ndiswrapper (1.2) ebuild should work wihout any problems on morph sources.
The suspend2 patch is now provided by gentoo.

yarbird,

Don't make ck-patchset the base of morph. There is for me no speed gain or any other advantages.
Strange thing yet, vanilla __feels__ more responsive. And if ck was that good or provided somthing good/better than
the vanilla tree does, then why isn't ck not in mainline by now? Con is trying to get it accepted for some time AFAIHR?

yardbird · Post by **yardbird** » Fri Jul 08, 2005 11:17 pm

Hi,

2.6.12-morph5 is out now. It's marked "stable", so there's the ebuild too. Lots of updates in it, let me know how it goes. As usual more details are available on the website.

@mayday147: mppe-mppc is back in this release.

@nightfrost: could you be more specific with the problem you have in compiling ipw2x00?

@rel: yes, I had some problems with -ck too. This latest release is based on vanilla.

Cheers

rel · Post by **rel** » Sat Jul 09, 2005 12:19 am

Oops, while compiling 2.6.12-morph5:
(USE="ext3-latency-fixes suspend2" emerge morph-sources)

CC drivers/video/fbsplash.o
CC drivers/video/cfbsplash.o
CC drivers/video/fbmem.o
CC drivers/video/fbmon.o
CC drivers/video/fbcmap.o
CC drivers/video/fbsysfs.o
CC drivers/video/modedb.o
LD drivers/video/fb.o
CC drivers/video/cfbfillrect.o
CC drivers/video/cfbcopyarea.o
CC drivers/video/cfbimgblt.o
CC drivers/video/softcursor.o
CC drivers/video/vesafb-tng.o
CC drivers/video/vesafb-thread.o
drivers/video/vesafb-thread.c: In function `vesafb_thread':
drivers/video/vesafb-thread.c:542: error: `PF_FREEZE' undeclared (first use in this function)
drivers/video/vesafb-thread.c:542: error: (Each undeclared identifier is reported only once
drivers/video/vesafb-thread.c:542: error: for each function it appears in.)
drivers/video/vesafb-thread.c:543: error: too many arguments to function `refrigerator'
make[2]: *** [drivers/video/vesafb-thread.o] Error 1
make[1]: *** [drivers/video] Error 2
make: *** [drivers] Error 2

yardbird · Post by **yardbird** » Sat Jul 09, 2005 12:32 am

rel wrote:Oops, while compiling 2.6.12-morph5:
(USE="ext3-latency-fixes suspend2" emerge morph-sources)

CC drivers/video/fbsplash.o
CC drivers/video/cfbsplash.o
CC drivers/video/fbmem.o
CC drivers/video/fbmon.o
CC drivers/video/fbcmap.o
CC drivers/video/fbsysfs.o
CC drivers/video/modedb.o
LD drivers/video/fb.o
CC drivers/video/cfbfillrect.o
CC drivers/video/cfbcopyarea.o
CC drivers/video/cfbimgblt.o
CC drivers/video/softcursor.o
CC drivers/video/vesafb-tng.o
CC drivers/video/vesafb-thread.o
drivers/video/vesafb-thread.c: In function `vesafb_thread':
drivers/video/vesafb-thread.c:542: error: `PF_FREEZE' undeclared (first use in this function)
drivers/video/vesafb-thread.c:542: error: (Each undeclared identifier is reported only once
drivers/video/vesafb-thread.c:542: error: for each function it appears in.)
drivers/video/vesafb-thread.c:543: error: too many arguments to function `refrigerator'
make[2]: *** [drivers/video/vesafb-thread.o] Error 1
make[1]: *** [drivers/video] Error 2
make: *** [drivers] Error 2

Ooops... did not get that. Fix:

http://morph-sources.homelinux.net/~yar ... -fix.patch

Apply with patchlevel -p0. Thanks for reporting this

ahorn · Post by **ahorn** » Sat Jul 09, 2005 5:47 am

yeah, finally, no more swsusp2 crap with the new ebuild

what about the sata, staircase and sk98 use flags?

Old posts from the morph-sources thread

ebulids?