hehe thank you Peter !!! I'll try that later in the afternoon. :thumbsup:


EDIT : i made an upgrade from Dapper and using your boot image it seems like there's a problem with the framebuffer, the ubuntu splash screen during boot is corrupted, but it boots though.

Beryl and 3D effects work nice too with my Radeon 8500 here :)

This worked, as did Peter's install and boot kernels. Yay :)

I still wonder, this X.org bug has been there since the summer and I know everyone knows the real problem* behind it.. why isn't it being fixed in the X.org mainline or on any reasonable distro?

because this bug only affects the Pegasos ? :roll:

no problems on a mac ppc here.

thanks again Peter. It's the first time that compiz and beryl actually work without a problem on the Pegasos. Nice one. I wrote a tutorial about Ubuntu Feisty installation on Pegasos, i'll send it to Geoffrey for the next Pegasos Book.

It affects ANY system that has multiple domains and the VGA adapter is not in domain 0.

https://bugzilla.novell.com/attachment.cgi?id=98419

Here's a good example of the bus fudging in action; Note at the bottom (where addresses are listed) that the domain 0 devices are all on bus 0. This is correct.



Note that the domain 1 devices are all on bus 1. This is due to the bus fudging enabled by the Linux kernel variable 'pci_assign_all_buses' which purposefully increments bus numbers globally. This is an old hack which enabled /proc filesystem entries to be on multiple domains but still be accessed through ancient PCI accessors which did not properly support domains; by looking at the bus number (which is unique globally with that flag) it could derive the domain the bus was sitting in.



The problem here is that when X.org uses sysfs to scan the buses, it the starts to put these values VERBATIM through domain-aware PCI accessors, which do not use the fudging table. Therefore, a device which is truly on "0001:00:08.0" (Pegasos AGP slot graphics adapter) is passed through as pci_config_read_something(1, 1, 8, 0, *value) - this adapter does NOT exist.

There could be many fixes. You could turn off pci_assign_all_buses in the PCI code, which would return all bus numbers to normal. Sysfs would then report the 'correct' verbatim values to pass through to domain-aware pci accessors.

A hack might be to stop using domain-aware pci accessors, and use the Linux bus fudging non-domain-aware functionality instead (this is what the /proc interface does as the /proc interface does not list PCI domain numbering).

A hack might be to detect that buses are numbered globally (sysfs/sysctl variable?) and restore those bus numbers before passing in values.

Another hack might be to fudge the domain-aware pci accessors so that they also access the fudging table, but this might break a thousand things.

All in all, Linux and X.org are both doing this wrong and both of them need to be fixed in tandem.. unfortunately it seems the likelihood of this bug being fixed correctly to support PCI domains in the clean, simple manner it could be, is obfuscated by the code process of throw things at the mainline and hope it doesn't break, and not "how should this be done and what components do we need to overhaul to do it".

I don't think the work is truly that difficult to spec and design :(



I would much prefer, in real life, that we had

* proper Yaboot support (I don't see why this is so difficult to be honest.. there are patches to make it work, even if our firmware 'does the wrong thing')

* This X.org bug fixed, or the pci_assign_all_buses stuff removed from the Linux kernels on PowerPC, or something done to alleviate the trouble?

I'd like to propose a test; can someone install Ubuntu and recompile their kernel so that pci_assign_all_buses (it's in powerpc/platforms/chrp/pci.c or something similar) is turned off and see if X.org works? I don't have a Pegasos here to truly test this and my Efika has no PCI devices.

hey i can only agree with that, a tutorial isn't satisfying but at least some unexperienced users can give it a try from now.




i can try that next weekend if no one does it before.
I'll keep u tuned.

I would say just as well, that just because you CAN get it to work does not mean this should be the final solution to the problem.

We've been doing this kernel dance for 2 years now. X.org has been busted for 6 months. I would like to be able to tell a user to just download a certain file and boot their Efika with it; not entertain 100 little quirks and a 200k instructional PDF on "how to install Linux".

some pegasos users warned you and Genesi several times about this issue (on this forum), i thought you had priviledged contact with the ubuntu devs ? but then nothing happened since "Dapper". We still have no kernel image to boot and no functional xorg for Feisty.

Sorry but as i'm not qualified to fix that bug myself, at least i tried to make something useful for those who want to try the newest ubuntu. Now thanks for you encouraging words Matt :-(

The tutorial will remain in french on a french site, why should i bother tranalating it.

Who told you that?

We are an Ubuntu partner; we contributed systems and we can put Ubuntu preinstalled on systems. We do have the email addresses and a couple phone numbers of developers. But Open Source Software is not a special secret club where you need to be a corporate affiliate to get things done.

I'm in the PowerPC team on Launchpad, but that's a namedrop, all I do is read the bugreports and wonder if it will ever be fixed. Most of the Pegasos errors I have no ability to fix; my Pegasos has no CPU card and it's 4000 miles away anyway.

We have no more privileged access to Ubuntu developers, Novell developers or Fedora/RedHat developers than you do.



Things hardly get fixed if you do not report bugs; Peter did a great job running into problems with Fiesty while it was in release candidate, and we refreshed all the bugs relevant. However, nobody else tried the system..



I wasn't trying to be discouraging, nor criticising you personally - we have 1500 people registered to the forum and 500 of them most certainly have received free hardware from Genesi and a lot of support.

I find it hard to believe that 500 developers out there do not give a shit, and expect things to be done for them by Genesi, Ubuntu and whoever else they expect to 'provide' for them.

Last time I checked, this is not how Open Source Software is developed or organised; If it's broken, if you have the code, you can fix it. If you cannot fix it yourself, but know someone who can..

Nobody else wants to take it on though? I find it hard to believe that NOBODY managed to work out why the bug happens (see above, I cheated and asked Gerald what he thought it was).

They are seemingly happy to have 3 consecutive Ubuntu releases be broken, 6 months of X.org bugs, receive a patch (which is hideously broken) and rely on a post-release hack and some test kernels?

FIVE HUNDRED FREE SYSTEMS and the result is we lag behind for the past 3 years. This is not what I envisioned..

Soundsquare

Keep in mind that Genesi doesnt have a staff of software developers that work for them and their resources are limited in this respect. As Matt said, this is the point of the Free ODW and Efika systems so that those who benefit from free HW can contribute to the cause...

I'm interested in what happened with Pegasos getting dropped from official support by Ubuntu a while back? Their was some weird deal with this iirc

magnetic

i think a part of the answer should be in Matt's signature. ;-)

Neko:
I think this is what the PowerPC community get after Ubuntu has
dropped the support. Sure, they use fine words and say that it's
still community supported.. I am just an outsider when it comes
to Ubuntu - I've never used it myself. But from what I understand
it's very much a commercial distro up to the point that they don't
charge you for it, unless you want special support.

What do I mean? Well, Canonical is the company behind Ubuntu.
it's not a normal developer community. It's a company. And they
have decided to drop PowerPC support. I'm pretty sure they don't
have many developers who does anything the lead x86 developers
haven't told them to do.

I've told you before, drop Ubuntu. I know it's a popular distro but
they have dumped PowerPC support and you really should just
accept the fact and move on. You can't possibly think you can
market your hardware with Ubuntu when there is no support for
it from the distribution itself!

You have a good OpenSUSE guy on your team. Use him, focus
on OpenSUSE for desktop, Debian for servers and Gentoo for
developers. And that's it. There is no use to try to support every
distro out there. Genesi is a very small company and this com-
munityisn't the largest there is. It's better to focus on the thing
s we can do good instead of wasting resources on something
that just won't happen. Ubuntu for PowerPC is dead. Just accept it.

ironfist

Good and interesting advice. However, its always been part of the Genesi business plan to support as many distros/os as possible to make their HW versatile. Its a shame about Ubuntu as IMHO its the best and most user friendly distro.

magnetic

We are in touch with the Ubuntu folks. They had the ODW and they have the EFIKA. The issue is one of the desktop itself. There has not been a contender from PowerPC that had any relevance to the opportunity/market until now - after Apple that is and even then PowerPC was/is considered small or insignificant. We are working on a couple of things that can change that.

Looking at the bigger market, it is not surprising that this trend continues. The money is being made on the server side and the margins are getting tighter. At the same time, the PC market is growing, but fragmenting to higher end processors and games and lower cost machines that don't need the processor power to run a Windows desktop. In other words a Ubuntu Desktop is fine and there are plenty of non-Microsoft applications that work well. Back to the server vendors... why would they do anything to create a problem for Intel? They need Intel! Encouraging a low cost SoC that gets the job done on the low end is a challenge for them. Sun has the thought-leadership here, HP is in a distant second...

Ubuntu is about good intentions and Canonical. The link is very close, maybe too close. Without the support of Mark, there would still be Debian, and Gentoo, and Crux and....

When we have created a compelling opportunity for Canonical we will have Ubuntu involved in the EFIKA and Open Client. All good things in time!

R&B :)

We are still here:

http://www.ubuntu.com/partners/findapartner

R&B :)

Ubuntu for PowerPC is not dead; nor is Pegasos support "dropped". It's simply gone the way (as bbrv said) of all PowerPC desktop systems, which is the sidelines and doldrums.

Without some extreme major desktop support like Apple there's nothing even the nicest, best developer relations guy can do about it. A couple million dollars would speak louder. All we can do is throw hardware at people and hope, because we don't HAVE a couple million dollars to throw at it (and I'd hate to waste it)

The moment we can support running Yaboot on the firmware properly, Ubuntu support will just appear, and work. Then again it's hardly worth building an Ubuntu CD for the Efika as it has no real potential as a system which has a CD drive; the netboot kernel, I guess we can rebuild.. if anyone wants to take that on right now, I am all for it. We gave out nearly 400!

*quietly dashes that hope against the rocks*

As for supporting as many distros as possible, this relies on this free hardware being used. I don't really see that if there is no incentive like having your name flash up on the screen of 100,000 Mac users or being a high-profile (nee bullshit hype) project like OLPC. Too many free systems just get put in closets. We gave away a lot of ODWs only to find that the next month, developers got a Mac Mini out of their own pocket and put the ODW in the closet, only to be touched when we nagged and nagged and nagged.

Well, I happen to think that hardware support in an operating system is a virtue; however it is done. One obstacle is that we provide a certain firmware solution and a certain way of doing things, and Linux developers may well disagree. Unfortunately they are unwilling to accept a small patch which fixes things on our system as it may be 'unclean' to them. This results in the system NOT being supported, even after we gave them this hardware.

I don't see the problem in patching yaboot or the device tree in the OS to fix firmware errors. This happens on x86 all the time with ACPI tables and weird laptops and so on. I don't think PowerPC needs to be this super-clean architecture with no quirks on it to succeed. Certainly it will never go anywhere if developers refuse to entertain any device with even the slightest, minor buggette or semi-proprietary firmware solution..

There is a very, VERY obvious fix for it which is to choose a better distribution or operating system which isn't on a high horse. Maybe that's Gentoo or maybe it's something built on Ubuntu/Debian or Crux or SuSE, with a team of developers creating our very own Genesi Linux (or bplan Linux :) however we have only as many developers as I have fingers on my hands internally, a disconcerting lack of resources (we would all love a PowerPC compile farm but do you realise how much those IBM blades cost?!) and certainly not the ability to start creating operating system solutions from scratch.

Would any of the Linux distribution guys here (who don't work for us already) actually entertain building a new distribution from scratch or modifying their current distribution for the purposes of a highly integrated, specialised support for Pegasos and Efika? The ULTIMATE board support package - it's very own Linux optimized for power management and performance, and application stability?!

There's always MorphOS, Solaris and QNX though.. stay tuned :D