Re: r128 Dual head support (Was Re: Video card register dumps)

[Benjamin Herrenschmidt]

On Thu, 2004-11-25 at 23:38 -0600, David Stanaway wrote:
> On Fri, 2004-11-12 at 19:35, Benjamin Herrenschmidt wrote:
> 
> > Well, if you don't know the chip well and don't have specs, it will be a
> > hard time to get that stuff working I'm afraid. You should look at the
> > initial dual head work for r128 cards that Alex Deucher
> > <[EMAIL PROTECTED]> did for X.org, since you need at least that
> > (dual head) as a first step to enabling TV Out. Then, I suggest you look
> > at the open-beos project (I think it's on SourceForge) for the radeon
> > code that drives the TV encoder.
> 
> Do you have any more information about this? I saw mention back in
> August that he was looking for hardware to work on the dual head
> support.

He got something working afaik, you should ask him.

ben.

Re: r128 Dual head support (Was Re: Video card register dumps)

[David Stanaway]

On Fri, 2004-11-26 at 00:24, Benjamin Herrenschmidt wrote:
> On Thu, 2004-11-25 at 23:38 -0600, David Stanaway wrote:
> > On Fri, 2004-11-12 at 19:35, Benjamin Herrenschmidt wrote:
> > 
> > > Well, if you don't know the chip well and don't have specs, it will be a
> > > hard time to get that stuff working I'm afraid. You should look at the
> > > initial dual head work for r128 cards that Alex Deucher
> > > <[EMAIL PROTECTED]> did for X.org, since you need at least that
> > > (dual head) as a first step to enabling TV Out. Then, I suggest you look
> > > at the open-beos project (I think it's on SourceForge) for the radeon
> > > code that drives the TV encoder.
> > 
> > Do you have any more information about this? I saw mention back in
> > August that he was looking for hardware to work on the dual head
> > support.
> 
> He got something working afaik, you should ask him.

I am looking at r128dh.diff  here

http://www.botchco.com/alex/r128/

I will let people know if this works on ppc.

Re: r128 Dual head support (Was Re: Video card register dumps)

[Benjamin Herrenschmidt]

On Fri, 2004-11-26 at 00:33 -0600, David Stanaway wrote:
> On Fri, 2004-11-26 at 00:24, Benjamin Herrenschmidt wrote:
> > On Thu, 2004-11-25 at 23:38 -0600, David Stanaway wrote:
> > > On Fri, 2004-11-12 at 19:35, Benjamin Herrenschmidt wrote:
> > > 
> > > > Well, if you don't know the chip well and don't have specs, it will be a
> > > > hard time to get that stuff working I'm afraid. You should look at the
> > > > initial dual head work for r128 cards that Alex Deucher
> > > > <[EMAIL PROTECTED]> did for X.org, since you need at least that
> > > > (dual head) as a first step to enabling TV Out. Then, I suggest you look
> > > > at the open-beos project (I think it's on SourceForge) for the radeon
> > > > code that drives the TV encoder.
> > > 
> > > Do you have any more information about this? I saw mention back in
> > > August that he was looking for hardware to work on the dual head
> > > support.
> > 
> > He got something working afaik, you should ask him.
> 
> I am looking at r128dh.diff  here
> 
> http://www.botchco.com/alex/r128/
> 
> I will let people know if this works on ppc.

It doesn't out-of-the box, it needs some PLL probing fixes and other
things of that sort I suppose. Similar to what I did on radeon in xorg
recently

Ben.

Re: [seems SOLVED] Re: the ibook G4 sleep test patch

[Arne Caspari]

Benjamin Herrenschmidt wrote:


On Tue, 2004-11-23 at 21:00 +0100, Pau Rul·lan Ferragut wrote:
 


On Tuesday 23 November 2004 19:06, Colin Leroy wrote:
   


On 23 Nov 2004 at 08h11, Colin Leroy wrote:
 

   


I had a brilliant idea ;-) and thought that my crashes that
I was noticing happened when starting (relatively) heavyweight stuff
like firefox. Then remembered I have powernowd to switch CPU frequency
based on the load, and tried without (with the CPU fixed at 931MHz):
it seems to have disappeared !
 



 


Give a chance to cpudyn
   



The problem is probably not related to the software used to switch
frequencies, but more probably to the kernel support. I've been having
various problems with the PMU-based frequency switching lately, I'm
still investigating.

Ben.
 

My system sometimes freezes sometime after wakeup from suspend. I am 
currently not using a load based cpu frequency switching.


It seems to me that these freezes are related to disk access. When the 
system froze, it was when it either was swapping in large amounts of 
data ( just a moment ago it was thunderbird that was about to be swapped 
in ) or I was trying to open a very full directory with gnome's natilus.


This would also make sense with the previous poster saying it happens 
when he starts firefox: On my system, starting firefox won't caus that 
much load so that the cpu frequency will be increased by cpufreq. But it 
will cause a lot of disc activity.


/Arne

Re: [seems SOLVED] Re: the ibook G4 sleep test patch

[Arne Caspari]

Arne Caspari wrote:

My system sometimes freezes sometime after wakeup from suspend. I am 
currently not using a load based cpu frequency switching.




Whoops! Correction:

I just realized I have load based frequency scaling enabled again :-P 
And I also realized that this came in effect when working with 
thunderbird. So maybe it is the freq scaling.


/Arne

Re: please test new quik with initrd-support (was: Re: getting quik into sarge?

[simon]

Ce jour Fri, 26 Nov 2004, Holger Levsen a dit:

> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
> 
> Hi,
> 
> cc:s added to gain more attention, reply-to: set.
> 
> On Friday 26 November 2004 10:17, [EMAIL PROTECTED] wrote:
> > i'm the current maintainer of quik, the oldworld PPC bootloader. i'm
> > currently sponsored by Jakko Niemi. there have been many code changes
> > made to it, including fixing bug #259359. for instance, initrd support,
> > using chosen/bootargs in OF, and a couple of other things.
> 
> i would like to point out that the default kernel for powerpc is 2.6 and that 
> the debian kernel packages for 2.6 need initrd support - so this is needed to 
> be able to use the default debian kernels on oldworld (with a free 
> bootloader).
> 
> so this quik is desirable for d-i, too.

yes, understood.

> > since it's in Section: base, and Priority: important, and there are many
> > positive changes to it, i thought it should go into
> > sarge-proposed-updates. is there any objection to this?
> 
> no, but as far as i know the initrd code has "only" been tested  on three 
> machines: ANS, pmac 4400 and pmac 7200. If you got access to another oldworld 
> machine please test and report it.

count my power tower pro in there - the one i'm typing from. also this
machine is booted from 2.6.8-powerpc, the debian one, and it booted with
an initrd very well, thank you very much ;). and i bet it will work the
same on my 8500, though that one is currently on the fritz.

> 
> and, better not to forget, the new quik.deb is available at 
> http://www.ulyssis.org/~p2/debjes/

of which i've downloaded the source and going to integrate into the
debian diff. i'm also going to bump the version to 2.1, since it has
these new changes in it. i think it warrants it, IMO.

> 
> regards,
>  Holger
> -BEGIN PGP SIGNATURE-
> Version: GnuPG v1.2.4 (GNU/Linux)
> 
> iD8DBQFBpv5YUHLQNqxYNSARAnr2AJ9uUEnJw33vOqE3OvxrQZx/4tI+9gCfRaxl
> qXY98lBKp9DVE+aI+J7XoAI=
> =8kK4
> -END PGP SIGNATURE-
> 

-- 
FBI CIA NSA IRS ATF BATF DOD WACO RUBY RIDGE OKC OKLAHOMA CITY MILITIA
GUN HANDGUN MILGOV ASSAULT RIFLE TERRORISM BOMB DRUG HORIUCHI KORESH
DAVIDIAN KAHL POSSE COMITATUS RANDY WEAVER VICKIE WEAVER SPECIAL
FORCES LINDA THOMPSON SPECIAL OPERATIONS GROUP SOG SOF DELTA FORCE
CONSTITUTION BILL OF RIGHTS WHITEWATER POM PARK ON METER ARKANSIDE
IRAN CONTRAS OLIVER NORTH VINCE FOSTER PROMIS MOSSAD NASA MI5 ONI CID
AK47 M16 C4 MALCOLM X REVOLUTION CHEROKEE HILLARY BILL CLINTON GORE
GEORGE BUSH WACKENHUT TERRORIST TASK FORCE 160 SPECIAL OPS 12TH GROUP
5TH GROUP SF
You hear those, Echelon?  Any of the words should be enough to "flag" your
communication by Echelon -- let's see how big their storage capacity REALLY
is...*VBEG*  Remember: EVERY day is Jam Echelon day!


signature.asc
Description: Digital signature

please test new quik with initrd-support (was: Re: getting quik into sarge?

[Holger Levsen]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Hi,

cc:s added to gain more attention, reply-to: set.

On Friday 26 November 2004 10:17, [EMAIL PROTECTED] wrote:
> i'm the current maintainer of quik, the oldworld PPC bootloader. i'm
> currently sponsored by Jakko Niemi. there have been many code changes
> made to it, including fixing bug #259359. for instance, initrd support,
> using chosen/bootargs in OF, and a couple of other things.

i would like to point out that the default kernel for powerpc is 2.6 and that 
the debian kernel packages for 2.6 need initrd support - so this is needed to 
be able to use the default debian kernels on oldworld (with a free 
bootloader).

so this quik is desirable for d-i, too.

> since it's in Section: base, and Priority: important, and there are many
> positive changes to it, i thought it should go into
> sarge-proposed-updates. is there any objection to this?

no, but as far as i know the initrd code has "only" been tested  on three 
machines: ANS, pmac 4400 and pmac 7200. If you got access to another oldworld 
machine please test and report it.

> BTW, many thanks to go Peter 'p2' De Schrijver for his tremendous help in
> making it even possible to close that bug, as it was a pretty hefty bug.

yup! :-)

and, better not to forget, the new quik.deb is available at 
http://www.ulyssis.org/~p2/debjes/


regards,
 Holger
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFBpv5YUHLQNqxYNSARAnr2AJ9uUEnJw33vOqE3OvxrQZx/4tI+9gCfRaxl
qXY98lBKp9DVE+aI+J7XoAI=
=8kK4
-END PGP SIGNATURE-

Re: Powerbook produces a lot more heat after kernel upgrade

[Timo Reimerdes]

On Sun, 2004-11-21 at 11:55 +1100, Benjamin Herrenschmidt wrote:
> On Sat, 2004-11-13 at 09:57 +0100, Richard Hult wrote:
> > Hi,
> > 
> > > Today I compiled the 2.6.9 stock kernel + Ben's suspend
> > > patch. However I am experiencing a lot of extra heet production as
> > > opposed to the 2.6.8.1 kernel. I have locked my cpu at the lowest
> > > frequency (it is fast enough), and with the 2.6.8.1 the fan was
> > > almost always off except with excessive loads. But while using
> > > the 2.9.6 the fans have to cool down my albook a lot more.
> > > 
> > > I haven't made any significant change to my kernel config. I have
> > > a PowerBook5,4 running at 1,33 MHz (well 666 MHz actually).
> > 
> > I can confirm this on my PowerBook 5,4 1.5GHz. Although the temperature 
> > rise happened when upgrading from the debian kernel package for 2.6.7 to 
> > 2.6.8. The idle temperature goes up ~4 degrees celcius, just by booting 
> > with 2.6.8 or later instead of 2.6.7. A friend with the same model is 
> > also experiencing this.
> 
> That's weird... I don't have an explanation off-hand.
> 
> Ben.
Hm... I havent researched it but I said to a friend of mine about 15
minutes ago "somehow the temperature management seems to get worse". *g*
In other words - I had the same feeling. (Running Ubuntu 2.6.8.1 now -
preparing to test the sleep patch)

Timo

Re: Debian or Ubantu

[Timo Reimerdes]

On Mon, 2004-11-22 at 17:35 +1100, Dean Hamstead wrote:
> >>Debian on the other hand may be more complicated than I want. I wasn't [too]
> >>encouraged when X took serious tweaking to get it to boot up.
> >>
> >>
> > 
> >The main advantage for going "straight" Debian is when you need
> >something that's not supported by the Ubuntu distro, and/or you are
> >running a platform not supported by it. Ubuntu is limited to i386,
> >"new-world" PPC and AMD64, whereas it seems like there's a Debian for
> >just about everything.
> >  
> >
> commercial distributions feed the masses, debian is by the
> people for the people. so if your running linux on  more obscure or outdated hardware here> then other like
> minded users (and possibly yourself) are maintaining things
> for the hardware because they/you use it. that or a nice
> package maintainer is cross compiling, which is just fine
> and dandy as well ;)
> 
> you may also find crux usefull, especially if your machine
> will be filling a server roll. but, as ive mentioned many
> times before - im very bsd = servers , linux = workstations
> kind of man.
> 
> Dean
I have just recently installed Ubuntu on my Powerbook G4 and was very
pleased with the fact that everything except the known issues worked
after install:
suspend and DRI are a nono (AlBook G4 - ATI 9600)
after 1 line in fstab USB sticks work smooth
sound's a-ok
xserver runs nicely
fonts are all available
And on top of that - the config files are just like plain, good, old
debian ;). 
So I like it as a starting point.

(ps: there is no development stuff on the computer after install - but
its available on the repositories)

Timo

Re: Powerbook produces a lot more heat after kernel upgrade

[Colin Leroy]

On 26 Nov 2004 at 12h11, Timo Reimerdes wrote:

Hi, 

> Hm... I havent researched it but I said to a friend of mine about 15
> minutes ago "somehow the temperature management seems to get worse". *g*
> In other words - I had the same feeling. (Running Ubuntu 2.6.8.1 now -
> preparing to test the sleep patch)

The fan driver never changed the way it works between its inclusion (around
2.6.2 iirc) and 2.6.10-rc2, so it's not temperature management related.

Since 2.6.10-rc2 however, it now works by incrementing fan speed step 
by step according to the temperature. Try it, maybe works better. 
-- 
Colin

Some trouble on Ibook G4

[Sylvain \"Skarsnik\" Colinet]

Hi everyone.
Today I decide to update the kernel on my ibook G4, with a linux2.6.9
and the last sleep patch (thanks benjamin).
XFree86 and radeon work fine with 3d acceleration. (1100 fps with glxgear).
The sleep work fine too.
But I have lot of trouble with sound,
I can't disable the micro and when I move the screen that produce
sound. The sound is very ugly because of that. Control mixer never
have the same comportement at startup and after I launch kde (arts not
running).
I can't play dvd because the sound seen to slow down xine and have
strange behavor, sound cut off when I go to first tty and return when
I seek in stream.
The last prob is some video don't play at full speed by both xine and
mplayer, but the cpu isn't in big charge.

So if someone can help me, I will be very happy and thankful.
--
Sylvain "Skarsnik" Colinet

Victory was near but the power of the ring couldn't be undone

heat problems with 2.6.9

[Jesus Climent]

hi

Ever since i put 2.6.9 to try the sleep patches my fan fires to max
revolutions and my cpu seems to be hotter than ever. Before it went fired up
when compiling and/or encoding Oggs, but now even scping some files seems to
be too cpu-hungry to keep the cpu cold.

Any similar problems out there?

My machine is a ibook G4 12" 1GHz

-- 
Jesus Climent  info:www.pumuki.org
Unix SysAdm|Linux User #66350|Debian Developer|2.4.27|Helsinki Finland
GPG: 1024D/86946D69 BB64 2339 1CAA 7064 E429  7E18 66FC 1D7F 8694 6D69

And then he ran into my knife... he ran into my knife ten times.
--June (Chicago)

Re: heat problems with 2.6.9

[Colin Leroy]

On 26 Nov 2004 at 14h11, Jesus Climent wrote:

Hi, 

> Ever since i put 2.6.9 to try the sleep patches my fan fires to max
> revolutions and my cpu seems to be hotter than ever. Before it went fired up
> when compiling and/or encoding Oggs, but now even scping some files seems to
> be too cpu-hungry to keep the cpu cold.
> 
> Any similar problems out there?

Not noticed here. What laptop do you have, ibook, powerbook? Precise model can
be found in the machine field of /proc/cpuinfo.

Can you trace the temperatures with another kernel not showing the problem and
this one, see if it's the CPU hotter or the fan kicking in sooner ?

Use that for example:
while true; do cat /sys/devices/temperatures/cpu_temperature; sleep 1; done;

you can also cat /sys/devices/temperatures/cpu_fan_speed to see the fan speed

-- 
Colin

Re: Problems booting kernel on IBM RS/6000 7043-140

[Johannes Martin]

Hi,

On Thu, 25 Nov 2004, Sven Luther wrote:
> Can you try the
> http://ftp.debian.org/debian/dists/unstable/main/installer-powerpc/current/images/powerpc/netboot/minimal/2.4/vmlinuz-prep.initrd
>
> It should not work (the above is oldworld pmac specific), but we will know if
> the problem is with the initrd size.
It didn't work... but I found out what the problem was. I needed to
upgrade the firmware.

Now I can boot the Debian installer image, if I enter
root=/dev/ram console=ttyS0,9600

Unfortunately, the pcnet32 driver doesn't work. It detects the network
card, but it dhclient doesn't get an answer and if I setup an IP manually,
I can't ping the workstation either.

With that old 2.4.15 boot disk & debian installer root, the pcnet32 driver
does work (at least for pings).

I tried the 2.4 vmlinuz-prep.initrd, but it does not seem to support the
serial console. At least I can't see anything after the 'now booting the
kernel' message.

Johannes

Re: Bug#283104: Installation report on debian-installer

[Sven Luther]

On Fri, Nov 26, 2004 at 02:43:29PM +, Brian Ruth wrote:
> Package: installation-reports
> Debian-installer-version: 26th Nov 04, from URL below
> uname -a: Linux odyssey 2.4.26 #5 Wed Nov 17 13:55:55 GMT 2004 ppc GNU/Linux
> Date: Multiple attempts
> Method: Floppy boot using own kernel and root.img, then net install from
> ftp.uk.linux.org, see below
> Machine: IBM RS/6000 7025-F40
> Processor: PowerPC 604e, 233MHz
> Memory: 512MB
> Root Device: SCSI, /dev/sda
> Root Size/partition table: Probably not relevant
> Output of lspci and lspci -n: Probably not relevant
> Base System Installation Checklist:
> [O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it
> 
> Initial boot worked:[O]
> Configure network HW:   [O]
> Config network: [O]
> Detect CD:  [ ]
> Load installer modules: [ ]
> Detect hard drives: [O]
> Partition hard drives:  [O]
> Create file systems:[O]
> Mount partitions:   [O]
> Install base system:[E]
> Install boot loader:[ ]
> Reboot: [ ]
> 
> Comments/Problems:
> 
> Trying to install debian-testing onto an IBM 7025-F40 (PowerPC PReP) system
> using the debian-installer from:
> 
> http://ftp.debian.org/debian/dists/sarge/main/installer-powerpc/current/images/powerpc/floppy-2.4/
> 
> This machine is rather esoteric and requires a specially patched kernel to
> boot it; I'm therefore trying to install by booting from a floppy containing a
> kernel with all the necessary drivers in it and then using the root.img root
> disk from the above location (this is the method I used to install the machine

floppy install on powerpc on anything except old world powermacs are not
currently supported. If there is a special patch needed for the kernel, please
inform us about it, and let's try to get it in the official kernels. Can you
try booting the plain 2.4.27 and 2.6.8 debian official kernels (the latest
needs some mkvmlinuz invocation before it is usefull to you). That said, there
where a bunch of prep patches which got into the 2.6.8 kernel, so it may be
that it will work already.

Once we have a working kernel, we can go ahead and see if we can produce a
prep floppy set, but first we need a working kernel.

Also, did you try the netboot method with either : 

  
http://people.debian.org/~luther/d-i/images/daily/powerpc/netboot/vmlinuz-prep.initrd

(i guess your box uses the prep boot, and not ppcbug), or :

  
http://people.debian.org/~luther/d-i/images/daily/powerpc/netboot/2.4/vmlinuz-prep.initrd

the initrd size may be too big for your prep OF, and you may want to try :

  
http://people.debian.org/~luther/d-i/images/daily/powerpc/netboot-minimal/2.4/vmlinuz-prep.initrd

Altough i believe that it doesn't contain yet the needed prep modules and
such, since it was hand-crafter for pegasos1, but i know there is also a
initrd size limit on some prep systems, which make the above system a good
guess.

> using the old Debian installation system). I use the following boot 
> parameters:
> 
> console=ttyS0,9600 load_ramdisk=1 ramdisk_size=32768 root=/dev/fd0
> debconf/priority=medium
> 
> This starts up OK and gets me into the installer menu; I can then configure
> country, language and keyboard OK, and opt to load drivers from the
> net-drivers.img disk. This also works, although it generates an error
> complaining about not being able to load kernel modules (because I'm using my
> own kernel); it nevertheless loads the next debian-installer components OK.
> 
> I can then follow through the installation, configuring my hardware and
> specifying to use the ftp.uk.linux.org mirror (I've also tried it manually
> specifying a mirror at www.mirroservice.org), partitioning disks, etc., and
> installing the base system.
> 
> The 'install the base system' part works fine until it reaches the point at
> which it tries to install a kernel. This fails, presumably because I'm using

Can you get us the exact error message (look at console 3 & 4 for details).

> my own kernel. The trouble is, I can't progress the installation any further
> at this point - if I try to select any other option from the main menu, it
> just tries to re-run the 'install base system' part again. There seems to be
> no way out, and I'm left with an unbootable system.

Notice that d-i doesn't know about the prep boot partition yet, so you woukd
have to use manual intervention anyway.

two things are needed, i believe, a partman-prep, based on the partman-alto
.udeb, and a way to install the kernel, which probably means a new
prep-installer .udeb package. You are the best candidate to implement this,
since you have access to the hardware,and an interest ti make it happen. I am
sure we will allbe willing to help you out on this.

> 
> Since I can't use one of the standard kernels on my system anyway, would it be
> possible to have the 'install kernel' menu offer an option to 'not install a
> kernel and skip this step' - I could then finish off the installation and
> ma

Re: please test new quik with initrd-support (was: Re: getting quik into sarge?

[Jaakko Niemi]

On Fri, 26 Nov 2004, Holger Levsen wrote:
> > since it's in Section: base, and Priority: important, and there are many
> > positive changes to it, i thought it should go into
> > sarge-proposed-updates. is there any objection to this?
> 
> no, but as far as i know the initrd code has "only" been tested  on three 
> machines: ANS, pmac 4400 and pmac 7200. If you got access to another oldworld 
> machine please test and report it.

 Also, it works just fine on my 9500. I really need to test
 d-i on it one of these days as I need to change new disk
 in, as the old one makes horrible noise. 

Re: Problems booting kernel on IBM RS/6000 7043-140

[Sven Luther]

On Fri, Nov 26, 2004 at 04:29:02PM +0100, Johannes Martin wrote:
> Hi,
> 
> On Thu, 25 Nov 2004, Sven Luther wrote:
> > Can you try the
> > http://ftp.debian.org/debian/dists/unstable/main/installer-powerpc/current/images/powerpc/netboot/minimal/2.4/vmlinuz-prep.initrd
> >
> > It should not work (the above is oldworld pmac specific), but we will know 
> > if
> > the problem is with the initrd size.
> It didn't work... but I found out what the problem was. I needed to
> upgrade the firmware.

Ok, ...

> Now I can boot the Debian installer image, if I enter
>   root=/dev/ram console=ttyS0,9600
> 
> Unfortunately, the pcnet32 driver doesn't work. It detects the network
> card, but it dhclient doesn't get an answer and if I setup an IP manually,
> I can't ping the workstation either.

Please fill a bug report against kernel-image-2.6.8-powerpc, and let's get
this fixed. What is the self-built kernel that you did use ? 

> With that old 2.4.15 boot disk & debian installer root, the pcnet32 driver
> does work (at least for pings).

Ok. 

> I tried the 2.4 vmlinuz-prep.initrd, but it does not seem to support the
> serial console. At least I can't see anything after the 'now booting the
> kernel' message.

Mmm, it should : 

  CONFIG_SERIAL=y
  CONFIG_SERIAL_CONSOLE=y

But if we can get the 2.6 pcnet32 driver fixed, it would be nicer, as i don't
like doing lot of work with the 2.4 kernel, which will hopefully be dropped
quickly after the sarge release.

Friendly,

Sven Luther

Re: please test new quik with initrd-support (was: Re: getting quik into sarge?

[Sven Luther]

On Fri, Nov 26, 2004 at 10:58:48AM +0100, Holger Levsen wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
> 
> Hi,
> 
> cc:s added to gain more attention, reply-to: set.
> 
> On Friday 26 November 2004 10:17, [EMAIL PROTECTED] wrote:
> > i'm the current maintainer of quik, the oldworld PPC bootloader. i'm
> > currently sponsored by Jakko Niemi. there have been many code changes
> > made to it, including fixing bug #259359. for instance, initrd support,
> > using chosen/bootargs in OF, and a couple of other things.
> 
> i would like to point out that the default kernel for powerpc is 2.6 and that 
> the debian kernel packages for 2.6 need initrd support - so this is needed to 
> be able to use the default debian kernels on oldworld (with a free 
> bootloader).
> 
> so this quik is desirable for d-i, too.
> 
> > since it's in Section: base, and Priority: important, and there are many
> > positive changes to it, i thought it should go into
> > sarge-proposed-updates. is there any objection to this?
> 
> no, but as far as i know the initrd code has "only" been tested  on three 
> machines: ANS, pmac 4400 and pmac 7200. If you got access to another oldworld 
> machine please test and report it.

Mmm, i could test it on the umax apus 2000 and on the pmac 8200/120 i have
here, need to test that this weekend. quik didn't work yet on the umax, and i
have never really used it, so some kind of help would be welcome in order to
set it up. 

Mmm, wondering if there are some kind of ps2 or usb -> adb adapters, so i
could try using the KVM switch for the apple boxes too.

> > BTW, many thanks to go Peter 'p2' De Schrijver for his tremendous help in
> > making it even possible to close that bug, as it was a pretty hefty bug.
> 
> yup! :-)
> 
> and, better not to forget, the new quik.deb is available at 
> http://www.ulyssis.org/~p2/debjes/

Thanks,

Friendly,

Sven Luther

Re: alsa mixer controls

[Joerg Sommer]

Michel Dänzer <[EMAIL PROTECTED]> wrote:
> On Thu, 2004-11-25 at 00:32 +0100, Florian Reitmeir wrote:
>> 
>> i just wonderd about alle the mixer controls of my iBook G4.
>> 
>> there is:
>> 
>> Headphon
>> Bass
>> Trebele
>> PCM
>> PCM 1
>> PC Speak
>> Auto Mute
>> Beep
>> DRC
>> DRC Rang
>> Monitor
>> 
>> some are clear, others not. i just searched with google for a while and
>> controllers like 'DRC'/'DRC Rang'/'Monitor' didn't show up with some useful
>> explanation.
>
> DRC is Dynamic Range Compression. The range mixer defines the range it
> compresses to, so if DRC is enabled with an empty range you mostly hear
> nothing.
>
> Monitor is the level at which input is forwarded to output directly
> (watch out for the feedback loop via the built-in microphone!).
>
> I've wondered myself what 'PCM 1' is, anyone?

AFAIK a second PCM channel.

-- 
Richard is only to blame for his inability to spell /dev/disk correctly.
Albert Cahalan, <[EMAIL PROTECTED]>, lkml

Re: please test new quik with initrd-support (was: Re: getting quik into sarge?

[Clive Menzies]

On (26/11/04 18:04), Sven Luther wrote:
> On Fri, Nov 26, 2004 at 10:58:48AM +0100, Holger Levsen wrote:
> > -BEGIN PGP SIGNED MESSAGE-
> > Hash: SHA1
> > 
> > Hi,
> > 
> > cc:s added to gain more attention, reply-to: set.
> > 
> > On Friday 26 November 2004 10:17, [EMAIL PROTECTED] wrote:
> > > i'm the current maintainer of quik, the oldworld PPC bootloader. i'm
> > > currently sponsored by Jakko Niemi. there have been many code changes
> > > made to it, including fixing bug #259359. for instance, initrd support,
> > > using chosen/bootargs in OF, and a couple of other things.
> > 
> > i would like to point out that the default kernel for powerpc is 2.6 and 
> > that 
> > the debian kernel packages for 2.6 need initrd support - so this is needed 
> > to 
> > be able to use the default debian kernels on oldworld (with a free 
> > bootloader).
> > 
> > so this quik is desirable for d-i, too.
> > 
> > > since it's in Section: base, and Priority: important, and there are many
> > > positive changes to it, i thought it should go into
> > > sarge-proposed-updates. is there any objection to this?
> > 
> > no, but as far as i know the initrd code has "only" been tested  on three 
> > machines: ANS, pmac 4400 and pmac 7200. If you got access to another 
> > oldworld 
> > machine please test and report it.
> 
> Mmm, i could test it on the umax apus 2000 and on the pmac 8200/120 i have
> here, need to test that this weekend. quik didn't work yet on the umax, and i
> have never really used it, so some kind of help would be welcome in order to
> set it up. 
> 
> Mmm, wondering if there are some kind of ps2 or usb -> adb adapters, so i
> could try using the KVM switch for the apple boxes too.
Hi Sven

This might help:

http://www.novatech.co.uk/novatech/specpage.html?F1DS104T

Regards

Clive

> 
> > > BTW, many thanks to go Peter 'p2' De Schrijver for his tremendous help in
> > > making it even possible to close that bug, as it was a pretty hefty bug.
> > 
> > yup! :-)
> > 
> > and, better not to forget, the new quik.deb is available at 
> > http://www.ulyssis.org/~p2/debjes/
> 
> Thanks,
> 
> Friendly,
> 
> Sven Luther
> 
> 
> -- 
> To UNSUBSCRIBE, email to [EMAIL PROTECTED]
> with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
> 

-- 
www.clivemenzies.co.uk ...
...strategies for business

Configuring X on a Powerbook G4 17" with Mobility Radeon 9600 (128 MB)

[Jens-Erik Weber]

Hello,

I described my problem in
.

Like I said there, I can get X running with ati, radeon and fbdev
drivers, but there's still a horizontal flickering.

Can I somehow combine the kernel parameters video=ofonly (btw: what
does it mean?) for eliminating the horizontal flickering and
video=radeonfb to get X running without displaying a wild moving
pattern?

I tried everything I found in the WWW except compiling a kernel by
myself. Is there an easy solution?

Jens




___
Gesendet von Yahoo! Mail - Jetzt mit 100MB Speicher kostenlos - Hier anmelden: 
http://mail.yahoo.de

Installing K2.6.9 with sleep patch #5

[Erik Chakravarty]

After patching, compiling and installing the new kernel, I get a kernel
panic and init halts because the modules.dep file cannot be found, even
though I installed the modules, and the new modules directory can be
found under /lib/modules.

In /boot, I have the System.map for the new kernel but not the
initrd.img file which I cannot find in the tree. Is there a separate
"make" required to build this image?

(I didn't have this issue on i386 but this is the first ppc kernel I
have tried to build - sorry for the ppc newbie question :-)  )

Cheers

-- 
_

Erik Chakravarty
(+44) 7720 850521

Re: Installing K2.6.9 with sleep patch #5

[Eugen Dedu]

mkinitrd -o outfile version

HTH,
Eugen
--
Eugen Dedu
Assistant Professor / Maître de conférences
http://lifc.univ-fcomte.fr/~dedu

Re: TEST: Sleep patch #5

[Vincent Hanquez]

On Wed, Nov 24, 2004 at 06:09:05PM +1100, Benjamin Herrenschmidt wrote:
> Ok, here's the 5th version of the sleep patch for ATI based albooks &
> iBook G4. Other machine users, please test too as it may cause
> regressions (or improvements) as well.
> 
> Not many changes since last version, I don't know why sometimes, the
> panel doesn't come back properly (though a new sleep cycle tends to fix
> it) and I have no fix for the USB issues. However, I added some hacks to
> the cache flush code based on what Apple does in Darwin that may help
> make the thing more robust, and I added a hook mecanism that allow the
> video to be restored _very_ early on wakeup, pretty much before
> everything else, thus making it easier to spot & debug crashes on
> wakeup.
> 
> There are still pending issues, like cpufreq on some machines will
> "think" it's running at full speed on wakeup while it's in fact running
> at slow speed (thankfully not the opposite), I will try to fix those in
> the next iteration.

Hi Ben,

It works nicely, no more green garbage after resume on Xv.

I track the module that make me crash, it was 'radeon' (drm/AGP code)

when the module usage count is not null, the screen is black after
resume, and I can only power off (button hold for 5s)
when the module is loaded and not use (i.e. the X server didn't detect
it at start time, module load after X startup, the resume works well)

Do you know how can I debug (and hopefully fix) that further ? with the
black screen, I don't really know what I can do.

for now I'll live without that, I do not use 3d often (actually once a
year .. ;))

Thanks you for your work (and thanks others which contributes too)

-- 
Tab

Re: TEST: Sleep patch #5

[Benjamin Herrenschmidt]

On Fri, 2004-11-26 at 22:57 +0100, Vincent Hanquez wrote:
> On Wed, Nov 24, 2004 at 06:09:05PM +1100, Benjamin Herrenschmidt wrote:
> > Ok, here's the 5th version of the sleep patch for ATI based albooks &
> > iBook G4. Other machine users, please test too as it may cause
> > regressions (or improvements) as well.
> > 

> Hi Ben,
> 
> It works nicely, no more green garbage after resume on Xv.
> 
> I track the module that make me crash, it was 'radeon' (drm/AGP code)
> 
> when the module usage count is not null, the screen is black after
> resume, and I can only power off (button hold for 5s)
> when the module is loaded and not use (i.e. the X server didn't detect
> it at start time, module load after X startup, the resume works well)
> 
> Do you know how can I debug (and hopefully fix) that further ? with the
> black screen, I don't really know what I can do.
> 
> for now I'll live without that, I do not use 3d often (actually once a
> year .. ;))

There seem to be a problem with sleep & radeon when using DRI indeed,
unless you force the AGPMode to 4 in your XF86Config file.

Ben.

Re: Some trouble on Ibook G4

[Benjamin Herrenschmidt]

On Fri, 2004-11-26 at 14:01 +0100, Sylvain "Skarsnik" Colinet wrote:
> Hi everyone.
> Today I decide to update the kernel on my ibook G4, with a linux2.6.9
> and the last sleep patch (thanks benjamin).
> XFree86 and radeon work fine with 3d acceleration. (1100 fps with glxgear).
> The sleep work fine too.
> But I have lot of trouble with sound,
> I can't disable the micro and when I move the screen that produce
> sound. The sound is very ugly because of that. Control mixer never
> have the same comportement at startup and after I launch kde (arts not
> running).
> I can't play dvd because the sound seen to slow down xine and have
> strange behavor, sound cut off when I go to first tty and return when
> I seek in stream.
> The last prob is some video don't play at full speed by both xine and
> mplayer, but the cpu isn't in big charge.

get an alsa mixer and disable DRC (or increase it's range)


Ben.

Re: please test new quik with initrd-support (was: Re: getting quik into sarge?

[Romain Francoise]

Holger Levsen <[EMAIL PROTECTED]> writes:

> no, but as far as i know the initrd code has "only" been tested on
> three machines: ANS, pmac 4400 and pmac 7200. If you got access to
> another oldworld machine please test and report it.

Tested on a Power Mac 6500/250 installed about two hours ago using d-i
daily floppies: booting the 2.6.8 kernel from sarge with its initrd
works fine.  The machine has 32MB of physical RAM.

-- 
  ,''`.
 : :' :Romain Francoise <[EMAIL PROTECTED]>
 `. `' http://people.debian.org/~rfrancoise/
   `-


pgpKiWYRZv3So.pgp
Description: PGP signature

Re: please test new quik with initrd-support (was: Re: getting quik into sarge?

[simon]

Ce jour Fri, 26 Nov 2004, Sven Luther a dit:

> Mmm, i could test it on the umax apus 2000 and on the pmac 8200/120 i have
> here, need to test that this weekend. quik didn't work yet on the umax, and i
> have never really used it, so some kind of help would be welcome in order to
> set it up. 

can start off here:

http://simonraven.nuit.ca/macppc/models.shtml
http://wiki.kisikew.org/bin/view/Know/MacModels

as to the umaxen, i've had relatively easy success with a umax 3000 -
desktop case, though the IDE/ATA on them is pretty fucked. it seems to
take only one device at a time. the SCSI is a generic MESH chipset IIRC. 

> Mmm, wondering if there are some kind of ps2 or usb -> adb adapters, so i
> could try using the KVM switch for the apple boxes too.

i think so. dunno how common they are, or are still sold (i don't hang
around computer/electronics stores, i get most of my stuff from dumpsters,
etc.)

> > and, better not to forget, the new quik.deb is available at 
> > http://www.ulyssis.org/~p2/debjes/
> 
> Thanks,
> 
> Friendly,
> 
> Sven Luther

eric

-- 
Software Patents are patently wrong:
http://swpat.ffii.org/papiere/eubsa-swpat0202/ustr0309/index.en.html


signature.asc
Description: Digital signature

Re: gdesklets uses 100 % of my cpu

[Mick Weiss]

At what intervals does it check for mail? Is there an open bug on it on
the gnome site?

I haven't tried it out, but that is what I'd look for.

Best Regards,

- Mick

(o> Web / software developer
( ) UNIX Systems Admin
---   ~ www.mickweiss.com ~



Jaonary Rabarisoa wrote:

On Tue, Nov 23, 2004 at 09:32:28PM +0100, Domingo Fiesta Segura wrote:


Are you (both) using the mail checker desklet? If yes remove it; I've seen
that desklet eating up all the CPU when is not properly configured.




In fact there's a problem with the mail checker desklet. I removed it
and it works better now. It sill uses about 15 %  of the cpu even when I
do nothing. 

Re: gdesklets uses 100 % of my cpu

[simon]

Ce jour Sat, 27 Nov 2004, Mick Weiss a dit:

> At what intervals does it check for mail? Is there an open bug on it on
> the gnome site?
> 
> I haven't tried it out, but that is what I'd look for.
> 
> Best Regards,
> 
> - Mick
> 
> (o> Web / software developer
> ( ) UNIX Systems Admin
> ---   ~ www.mickweiss.com ~
> 

the subject is why i don't use it or recommend it. sitting still and
doing nothing it sucks up more resources than gnome-panel and nautilus.

eric

> -- 
> To UNSUBSCRIBE, email to [EMAIL PROTECTED]
> with a subject of "unsubscribe". Trouble? Contact 
> [EMAIL PROTECTED]
> 

-- 
Cold pizza and cold coffee, second best thing to cold pizza and warm beer.


signature.asc
Description: Digital signature

bootsplash

[Martin Lohmeier]

Hi,

have anyone bootsplash working on his ppc? I got the source from
bootsplash.de, compiled it with no problem but dpkg can't install it.


dpkg: error processing bootsplash (--install):
 subprocess post-installation script returned error exit status 10

Anyone out there who got it working?

by, Martin
-- 
powered by Debian GNU / Linux

Re: TEST: Sleep patch #5

[Jesus Climent]

On Sat, Nov 27, 2004 at 09:15:59AM +1100, Benjamin Herrenschmidt wrote:
> 
> There seem to be a problem with sleep & radeon when using DRI indeed,
> unless you force the AGPMode to 4 in your XF86Config file.

Options "AGPMode" "4" ??

-- 
Jesus Climent  info:www.pumuki.org
Unix SysAdm|Linux User #66350|Debian Developer|2.4.27|Helsinki Finland
GPG: 1024D/86946D69 BB64 2339 1CAA 7064 E429  7E18 66FC 1D7F 8694 6D69

Sticking feathers up your butt does not make you a chicken.
--Tyles Durden (Fight club)

jdk for ppc

[Jose A. Ortega Ruiz]

hi. i've been browsing and googling and haven't been able to find a
jdk 1.4.x for debian ppc (or gnu/linux ppc for that
matter)... blackdown, for one, seems to have abandoned support for
powerpc since the 1.2 days. 

anyone knows of a 1.4.x implementation for this platform (i'm running
on a new ibook G4, btw)?

thanks,
jao
-- 
An eye for an eye leaves everyone blind.
-Mohandas Karamchand Gandhi (1869-1948)

Re: jdk for ppc

[Barry Hawkins]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Jose A. Ortega Ruiz wrote:
|
| hi. i've been browsing and googling and haven't been able to find a
| jdk 1.4.x for debian ppc (or gnu/linux ppc for that
| matter)... blackdown, for one, seems to have abandoned support for
| powerpc since the 1.2 days.
|
| anyone knows of a 1.4.x implementation for this platform (i'm running
| on a new ibook G4, btw)?
|
| thanks,
| jao
You will need to get the IBM 1.4.2 JDK for pSeries Linux (32-bit
PowerPC).  If you are looking for a fully-functional JDK for your
platform, that's pretty much it.

- --
Barry Hawkins
All Things Computed
site: www.alltc.com
weblog: www.yepthatsme.com

Registered Linux User #368650
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFBp+gp7bZ6kUftWZwRAnSeAJ9EqEb1RycQ4PIMFodnsPU2lGF6SACg0a2K
Yw0rH4LoZQaLF0fxoLImLPM=
=l097
-END PGP SIGNATURE-

Re: jdk for ppc

[John Steele Scott]

"Jose A. Ortega Ruiz" <[EMAIL PROTECTED]> writes:

> hi. i've been browsing and googling and haven't been able to find a
> jdk 1.4.x for debian ppc (or gnu/linux ppc for that
> matter)... blackdown, for one, seems to have abandoned support for
> powerpc since the 1.2 days. 
>
> anyone knows of a 1.4.x implementation for this platform (i'm running
> on a new ibook G4, btw)?

Jose,

I have this one:

bash-2.05b$ java -version
java version "1.4.1"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.1)
Classic VM (build 1.4.1, J2RE 1.4.1 IBM build cxppc321411-20040301 (JIT 
enabled:jitc))

cheers,

John

<#secure method=pgpmime mode=sign>

Re: jdk for ppc

[Jose A. Ortega Ruiz]

Barry Hawkins <[EMAIL PROTECTED]> writes:

> You will need to get the IBM 1.4.2 JDK for pSeries Linux (32-bit
> PowerPC).  If you are looking for a fully-functional JDK for your
> platform, that's pretty much it.
>

thanks for the quick reply. i've actually installed ibm's jdk:

java version "1.4.2"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.2)
Classic VM (build 1.4.2, J2RE 1.4.2 IBM build cxppc32142-20040917 (JIT enabled: 
jitc))

but, whenever i try to run any java program, i get at some point an
'illegal hardware instruction' error (plus a java core file). for
instance, if i run policytool (which comes with the jdk):

JVMDG217: Dump Handler is Processing Signal 4 - Please Wait.
JVMDG303: JVM Requesting Java core file
JVMDG304: Java core file written to /tmp/javacore.20041127.042722.8729.txt
JVMDG215: Dump Handler has Processed Exception Signal 4.
Illegal instruction

(i'm running 2.6.9, in case that matters). any idea about what's going
on?

thanks,
jao
-- 
It is often forgotten that (dictionaries) are artificial repositories, put
together well after the languages they define. The roots of language are
irrational and of a magical nature. -Jorge Luis Borges, writer (1899-1986)

Re: TEST: Sleep patch #5

[Benjamin Herrenschmidt]

On Sat, 2004-11-27 at 02:15 +0100, Jesus Climent wrote:
> On Sat, Nov 27, 2004 at 09:15:59AM +1100, Benjamin Herrenschmidt wrote:
> > 
> > There seem to be a problem with sleep & radeon when using DRI indeed,
> > unless you force the AGPMode to 4 in your XF86Config file.
> 
> Options "AGPMode" "4" ??

 Option "AGPMode" "4" in your "Device" section, yes.

> 
> -- 
> Jesus Climent  info:www.pumuki.org
> Unix SysAdm|Linux User #66350|Debian Developer|2.4.27|Helsinki Finland
> GPG: 1024D/86946D69 BB64 2339 1CAA 7064 E429  7E18 66FC 1D7F 8694 6D69
> 
> Sticking feathers up your butt does not make you a chicken.
>   --Tyles Durden (Fight club)
> 
> 
-- 
Benjamin Herrenschmidt <[EMAIL PROTECTED]>

Re: jdk for ppc

[John Steele Scott]

"Jose A. Ortega Ruiz" <[EMAIL PROTECTED]> writes:
> but, whenever i try to run any java program, i get at some point an
> 'illegal hardware instruction' error (plus a java core file). for
> instance, if i run policytool (which comes with the jdk):

I remember reading something about how the IBM jdk uses some POWER ISA instead
of PowerPC. The Gentoo ebuild has this code:

if use ppc; then
dosed s:/proc/cpuinfo:/etc//cpuinfo:g 
/opt/${P}/jre/bin/libjitc.so
dosed s:/proc/cpuinfo:/etc//cpuinfo:g 
/opt/${P}/jre/bin/libjitc_g.so
insinto /etc
doins ${FILESDIR}/cpuinfo
fi

Which hacks the .so files to look at /etc/cpuinfo to get the cpu type, then
/etc/cpuinfo contains simply the line:

cpu : 604e

I imagine that the two forward-slashes in the sed expression is so that the
replacement string is the same length as the original.

Hope this helps.

cheers,

John


pgpqFsugK88TX.pgp
Description: PGP signature

Re: jdk for ppc

[Gregory Seidman]

On Sat, Nov 27, 2004 at 04:28:42AM +0100, Jose A. Ortega Ruiz wrote:
[...]
} but, whenever i try to run any java program, i get at some point an
} 'illegal hardware instruction' error (plus a java core file). for
} instance, if i run policytool (which comes with the jdk):
} 
} JVMDG217: Dump Handler is Processing Signal 4 - Please Wait.
} JVMDG303: JVM Requesting Java core file
} JVMDG304: Java core file written to /tmp/javacore.20041127.042722.8729.txt
} JVMDG215: Dump Handler has Processed Exception Signal 4.
} Illegal instruction
[...]

Digging around in the messages I've saved from debian-powerpc over the
years, I found the following:

If you want to use the JIT compiler under IBM's jdk please put the
following in your .bashrc or .profile

export JITC_PROCESSOR_TYPE=6

More information can be found at:

http://lists.terrasoftsolutions.com/pipermail/yellowdog-devel/2003-June/000209.html
http://lists.terrasoftsolutions.com/pipermail/yellowdog-devel/2003-July/000251.html

} thanks,
} jao
--Greg

Re: VGA out and Albook 15"

[Federico Gamio]

On Wed, 2004-11-24 at 18:43 +0100, Daniele Menozzi wrote:
> Hi all, I'm trying to set up the vga out over an Albook 15", but I've
> different problems ; I've successfully cloned the screen, but I cannot choose 
> 2
> different resolutions for my laptop screen and the external monitor, it
> works only applying the same resolution to both. I've installed dri-trunk
> Xfree, but the problem remains.
> 
> Have you a working configuration for an ati based albook?
> Thanks to all, I attach mine.

I am unable to get anything on the DVI port of my PB G4 (ati9600).
Can you briefly explain how do you clone the display (programs,
packages, patches to kernel, etc).

Thanks
Federico

> 
> Bye
>   Menoz
> 
> 
-- 
My software never has bugs.   | ASCII Ribbon Campaign /"\
It just develops random features. | For Standards-Complaint Email \ /
Public GnuPG key available at: http://www.keyserver.netX
1024D/203E154B 2003-10-03 Federico Gamio <[EMAIL PROTECTED]> / \
Key fingerprint = 0B9E 3A19 C88B EBAC 5422  C05F 76B5 B922 203E 154B
sub  2048g/32D2F465 2003-10-03


signature.asc
Description: This is a digitally signed message part