FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > ArchLinux > ArchLinux Development

 
 
LinkBack Thread Tools
 
Old 07-21-2008, 10:07 PM
Andreas Radke
 
Default kernel26 2.6.26-1 and related module packages

x86_64: kernel boots well with nvidia modul. give you my signoff.

-Andy
 
Old 07-22-2008, 05:16 AM
Eric Belanger
 
Default kernel26 2.6.26-1 and related module packages

On Mon, 21 Jul 2008, Thomas Bächler wrote:
- Someone should test squashfs, unionfs and aufs and report if they work
(Simo?)


my chroot works fine so unionfs should be OK. Only tested on x86_64

- lirc is broken, maybe an update to 0.8.3 will fix it. Can the maintainer
take care of this? I don't know when I'll get to it.


It's currently orphaned. I could give it a try if no one else wants to
adopt or try to fix it. But I don't use it and might not have time this
week though.




i686 is untested, so I need at least two signoffs. x86_64 works fine.




Signing off x86_64 with nvidia and i686.

Both systems boots and run fine.
--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.
 
Old 07-22-2008, 07:35 AM
Thomas Bächler
 
Default kernel26 2.6.26-1 and related module packages

Thomas Bächler schrieb:

Module packages:
- I patched catalyst with a weird patch to make it work, please anyone
test and report!


Any reports here? Also from users, please.

- intel-536ep and intel-537 don't compile. I didn't look closely yet,
and frankly, I doubt they have much users. Do we really want to maintain
them?
- lirc is broken, maybe an update to 0.8.3 will fix it. Can the
maintainer take care of this? I don't know when I'll get to it.
- madwifi is broken. Do we still need it? AFAIK, all chipsets should now
be supported by ath5k by now. Can a madwifi user please look at this
(Aaron, Dan?)
- martian is broken again. A user sent me patches last time, I have to
look for his email address, maybe he fixes it again (user whose name I
forgot, if you are reading this, please contact me).

- I didn't try openswan-klips, any updates here? (tomk?)
- wlan-ng26 is broken and I'd like to drop it, as I doubt anybody uses it.
- I updated nvidia-71xx and nvidia-96xx to new upstream release for
2.6.26 compatibility.


i686 is untested, so I need at least two signoffs. x86_64 works fine.


Okay, I think I have enough signoffs. Any opinions on dropping
wlan-ng26? According to Dan, we still need madwifi, is there an Atheros
user here who will fix it (apparently, a recent enough snapshot will
compile).
IMO, once madwifi and lirc are fixed and I got at least one positive
report for catalyst for every architecture, we can move. intel-53{6,7ep}
and martian are unmaintained closed-source crap anyway, I won't be held
back by those.
 
Old 07-22-2008, 08:00 AM
Thomas Bächler
 
Default kernel26 2.6.26-1 and related module packages

Thomas Bächler schrieb:
Okay, I think I have enough signoffs. Any opinions on dropping
wlan-ng26? According to Dan, we still need madwifi, is there an Atheros
user here who will fix it (apparently, a recent enough snapshot will
compile).
IMO, once madwifi and lirc are fixed and I got at least one positive
report for catalyst for every architecture, we can move. intel-53{6,7ep}
and martian are unmaintained closed-source crap anyway, I won't be held
back by those.




Ah, and if anyone needs configuration changes or patches/bugfixes
applied: I will rebuild again tonight, because I forgot to enable
uvcvideo in the configuration, which is quite popular. So write to this
list (or to arch-general) if you need anything.
 
Old 07-23-2008, 03:19 AM
Eric Belanger
 
Default kernel26 2.6.26-1 and related module packages

On Tue, 22 Jul 2008, Eric Belanger wrote:


On Mon, 21 Jul 2008, Thomas Bächler wrote:
- Someone should test squashfs, unionfs and aufs and report if they work
(Simo?)


my chroot works fine so unionfs should be OK. Only tested on x86_64

- lirc is broken, maybe an update to 0.8.3 will fix it. Can the maintainer
take care of this? I don't know when I'll get to it.


It's currently orphaned. I could give it a try if no one else wants to adopt
or try to fix it. But I don't use it and might not have time this week
though.


lirc has been updated to 0.8.3 and built against 2.6.26-1 in testing.
I've also fix/implemented 2 bug reports at the same time:
- fixed daemon script to facilitate the use of IR devices that don't
create device (close FS#9767)
- enabled IR transmitters support (close FS#9709)

I don't have any IR device so I could only minimally test it. Feedback
about success or problems would be nice.

--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.
 
Old 07-23-2008, 12:49 PM
Thomas Bächler
 
Default kernel26 2.6.26-1 and related module packages

Eric Belanger schrieb:

lirc has been updated to 0.8.3 and built against 2.6.26-1 in testing.
I've also fix/implemented 2 bug reports at the same time:
- fixed daemon script to facilitate the use of IR devices that don't
create device (close FS#9767)
- enabled IR transmitters support (close FS#9709)

I don't have any IR device so I could only minimally test it. Feedback
about success or problems would be nice.


Thank you very much. Gerhard Brauer sent me patches for madwifi, I will
update it as well.


This leaves the following:

intel-536ep
intel-537
martian

These are i686-only binary modules. I say we leave them there until a) I
get motivation to fix them or b) a user sends us a patch. Maintaining
them has always been a PITA and I wonder if it's worth it.


openswan-klips

Has been broken for quite some time (2.6.24 I think), we just leave it
until Tom fixes it.


wlan-ng26

I say drop it. We used the wlan-ng USB driver for some exotic hardware,
which apart from Tom nobody used (and his is broken now). This hardware
is probably supported by p54usb now anyway.
 
Old 07-23-2008, 08:03 PM
Andreas Radke
 
Default kernel26 2.6.26-1 and related module packages

Am Tue, 22 Jul 2008 10:00:03 +0200
schrieb Thomas Bächler <thomas@archlinux.org>:

> Thomas Bächler schrieb:
> > Okay, I think I have enough signoffs. Any opinions on dropping
> > wlan-ng26? According to Dan, we still need madwifi, is there an
> > Atheros user here who will fix it (apparently, a recent enough
> > snapshot will compile).
> > IMO, once madwifi and lirc are fixed and I got at least one
> > positive report for catalyst for every architecture, we can move.
> > intel-53{6,7ep} and martian are unmaintained closed-source crap
> > anyway, I won't be held back by those.
> >
>
> Ah, and if anyone needs configuration changes or patches/bugfixes
> applied: I will rebuild again tonight, because I forgot to enable
> uvcvideo in the configuration, which is quite popular. So write to
> this list (or to arch-general) if you need anything.
>

report for my notebook:

had to replace ath_pci module from madwifi with ath5k kernel module and
change netcfg2 config for ath0 -> wlan0 change. works well here.

running the catalyst driver works but i get a weird report when the
module gets loaded at bootup: unknown symbol

from dmesg:

fglrx: module license 'Proprietary. (C) 2002 - ATI Technologies, Starnberg, GERMANY' taints kernel.
fglrx: Unknown symbol flush_tlb_page

-Andy
 
Old 07-23-2008, 08:29 PM
Thomas Bächler
 
Default kernel26 2.6.26-1 and related module packages

Andreas Radke schrieb:

had to replace ath_pci module from madwifi with ath5k kernel module and
change netcfg2 config for ath0 -> wlan0 change. works well here.


See my post further down, madwifi will be added again.


running the catalyst driver works but i get a weird report when the
module gets loaded at bootup: unknown symbol

from dmesg:

fglrx: module license 'Proprietary. (C) 2002 - ATI Technologies, Starnberg, GERMANY' taints kernel.
fglrx: Unknown symbol flush_tlb_page


http://projects.archlinux.org/?p=linux-2.6-ARCH.git;a=blob;f=patches/ati-catalyst-2.6.25.patch;h=f90d8b40a7f1c8a0ea9a3ec4229322d4df9 8e7ab;hb=daaa7a2c43699c3b2acffaef1a4fed863194d29f

We needed that patch with catalyst on 2.6.25 (x86_64 only), I thought
they removed the dependence on that symbol. I could re-add the above patch.
 
Old 07-24-2008, 07:58 PM
Thomas Bächler
 
Default kernel26 2.6.26-1 and related module packages

Thomas Bächler schrieb:
So this is it, 2.6.26 is going to testing. I adjusted the configuration
according to http://bugs.archlinux.org/task/10512 and changed the
maximum number of CPUs to 16. I hope I didn't miss any configuration
bugs, if so, please tell me. Except some driver updates, there is
nothing too exciting here. Just a few remarks:


Kernel stuff:
- uvesafb needs the new v86d from testing now (old one segfaults)
- bcm43xx has been removed, all users must switch to b43 or b43legacy
and use b43-fwcutter for firmware extraction. I am removing
bcm43xx-fwcutter from core, as it is obsolete.
- Someone should test squashfs, unionfs and aufs and report if they work
(Simo?)


If there is more, please tell me

Module packages:
- I patched catalyst with a weird patch to make it work, please anyone
test and report!
- intel-536ep and intel-537 don't compile. I didn't look closely yet,
and frankly, I doubt they have much users. Do we really want to maintain
them?
- lirc is broken, maybe an update to 0.8.3 will fix it. Can the
maintainer take care of this? I don't know when I'll get to it.
- madwifi is broken. Do we still need it? AFAIK, all chipsets should now
be supported by ath5k by now. Can a madwifi user please look at this
(Aaron, Dan?)
- martian is broken again. A user sent me patches last time, I have to
look for his email address, maybe he fixes it again (user whose name I
forgot, if you are reading this, please contact me).

- I didn't try openswan-klips, any updates here? (tomk?)
- wlan-ng26 is broken and I'd like to drop it, as I doubt anybody uses it.
- I updated nvidia-71xx and nvidia-96xx to new upstream release for
2.6.26 compatibility.


i686 is untested, so I need at least two signoffs. x86_64 works fine.



I'm building 2.6.26-2 now. Only change is that I am adding uvcvideo and
some DVB driver (hope there wasn't something else). As I will be gone
tomorrow, here is what needs to be done before this is moved to core:


- Build madwifi. Gerhard Brauer sent me a fixed PKGBUILD, I attached it
to this mail.
- Andy said there is a new version for catalyst, which should hopefully
fix the problem he reported. Maybe we need to reapply
http://projects.archlinux.org/?p=linux-2.6-ARCH.git;a=blob;f=patches/ati-catalyst-2.6.25.patch;h=f90d8b40a7f1c8a0ea9a3ec4229322d4df9 8e7ab;hb=daaa7a2c43699c3b2acffaef1a4fed863194d29f
- Drop wlan-ng26. It's not worth maintaining it, as it's probably
unneeded by now.
- Ignore intel-536ep, intel-537 and martian until somebody fixes it. If
nobody sends a patch, move to core anyway, or just drop the packages. I
really don't care.


Have a great week.
 
Old 07-25-2008, 02:29 PM
Andreas Radke
 
Default kernel26 2.6.26-1 and related module packages

Am Wed, 23 Jul 2008 22:29:05 +0200
schrieb Thomas Bächler <thomas@archlinux.org>:

> Andreas Radke schrieb:
> > had to replace ath_pci module from madwifi with ath5k kernel module
> > and change netcfg2 config for ath0 -> wlan0 change. works well here.
>
> See my post further down, madwifi will be added again.
>
> > running the catalyst driver works but i get a weird report when the
> > module gets loaded at bootup: unknown symbol
> >
> > from dmesg:
> >
> > fglrx: module license 'Proprietary. (C) 2002 - ATI Technologies,
> > Starnberg, GERMANY' taints kernel. fglrx: Unknown symbol
> > flush_tlb_page
>
> http://projects.archlinux.org/?p=linux-2.6-ARCH.git;a=blob;f=patches/ati-catalyst-2.6.25.patch;h=f90d8b40a7f1c8a0ea9a3ec4229322d4df9 8e7ab;hb=daaa7a2c43699c3b2acffaef1a4fed863194d29f
>
> We needed that patch with catalyst on 2.6.25 (x86_64 only), I thought
> they removed the dependence on that symbol. I could re-add the above
> patch.
>

Just uploading catalyst 8.7-1 pkg. It still depends on that symbol.
Please add that patch for next release again.

Xorg and glxgears start fine without. So it shouldn't be a showstopper.

-Andy
 

Thread Tools




All times are GMT. The time now is 03:36 AM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright ©2007 - 2008, www.linux-archive.org