AMD/ATI lost another Linux customer

[m4rgin4l]

I was going through some old mails in my account and I realized that I bought my ATI HD4850 over a year ago, and it hit me: In this year I've never had decent support for my card in Linux. Neither the proprietary driver nor the OSS one managed to provide even a tiny fraction of the features I enjoyed when I was an NVIDIA user (or an Intel user).

What drove me to buy an ATI card was the news that they were going to publish the specifications and I thought that maybe for once a manufacturer understood how to interact with the community and that they were actually going to make some positive change in the Linux driver scene. Right then, I earned an appearance on the cover of "Incredibly Naive People" magazine.

After reading a couple of blog posts about how developers and users felt betrayed by the treatment Apple is giving them I realized I was in a similar position: I have a very good piece of hardware, but the company that makes it couldn't care less about me. I belong to a minority of people who actually enjoys having a choice in the matter of operating systems, and that makes me a LOUSY customer.

After this horrible year, I now understand the reasons for ATI/AMD's decision: they can't (won't) support Linux so they'll handle the responsibility to the community. Let them carry the weight of developing and maintaining the drivers. I'm not judging the company's morals on this, it IS a corporation after all, and they can do whatever the hell they want with they're products and their drivers.

NVIDIA won't be releasing their specifications anytime soon, but...guess what? It seems they actually give a crap about Linux users! Back when I was an NVIDIA customer, the longest time I went without support from their drivers was just a month (with the release of Fedora 8 or 9). On top of this, we now have Nouveau, that might not be in good shape yet, but if I was to switch to NVIDIA right now, I would have basically the same features I "enjoy" from the AMD/ATI hydra of drivers (catalyst, ati, radeonhd, etc.).

So, this is me saying "Adios AMD", I won't get suckered again into your bullshit anytime soon.

[Melcar]

AMD *does* care about Linux; that's why they released documentation. After all, isn't that what the Linux world always demands? When it comes to Linux the best driver choice is and open source driver, and AMD has been playing ball by allowing the development of such a driver. Even better, they provide a driver themselves for those customers that don't want to go the open source route.

[m4rgin4l]

AMD *does* care about Linux; that's why they released documentation. After all, isn't that what the Linux world always demands? When it comes to Linux the best driver choice is and open source driver, and AMD has been playing ball by allowing the development of such a driver. Even better, they provide a driver themselves for those customers that don't want to go the open source route.

AMD does not "care" about anything, it is a company, and it has an agenda and a business plan :) Nothing wrong about that, btw.

Anyway, I forgot to divulge that I'm a Fedora user, so there's no choice for me, the proprietary driver doesn't work and the OSS ones only provide the very basics for the latest chipsets (like the one on my HD4850).

With NVIDIA I DO have a choice, their proprietary driver works, and there's the OSS one that basically will give what I have right now with ATI.

I also have an Intel card on my laptop. Although the current driver is somewhat buggy, that's because of the radical changes it suffered. I don't have choice here either (I can't even change the hardware :) but at least the driver is shown constant improvement. From ATI all I got was the occasional "maybe next year, if all planets align, you'll be able to play accelerated video on last year's hardware".

[deanjo]

When it comes to Linux the best driver choice is and open source driver,

That's a very vague comment. In performance, the foss driver is not the best choice, in development time, again not the best choice, feature wise again not the best choice. Really all you can say is "In theory, when it comes to Linux the best driver choice should be the open source driver."

[nanonyme]

That's a very vague comment. In performance, the foss driver is not the best choice, in development time, again not the best choice, feature wise again not the best choice. Really all you can say is "In theory, when it comes to Linux the best driver choice should be the open source driver."GPL opensource drivers stay working easier on Linux than closed drivers though since they can be integrated to the kernel tree which means features they need probably won't get dropped. Just takes a looong time to get them to the working state in the first place. :)

[nanonyme]

I also have an Intel card on my laptop. Although the current driver is somewhat buggy, that's because of the radical changes it suffered. I don't have choice here either (I can't even change the hardware :) but at least the driver is shown constant improvement. From ATI all I got was the occasional "maybe next year, if all planets align, you'll be able to play accelerated video on last year's hardware".Opensource drivers for Intel, ATi, and nVidia chips are planned to have video acceleration over Gallium. Just takes a while to get the infrastructure up, things are very much progressing for most people. (apparently excluding some AGP users regardless of the chip)

[deanjo]

Opensource drivers for Intel, ATi, and nVidia chips are planned to have video acceleration over Gallium. Just takes a while to get the infrastructure up, things are very much progressing for most people. (apparently excluding some AGP users regardless of the chip)

Planning is one thing, executing is quite another.

[energyman]

the hilarious thing is:
in the past people complained (very loudly, constantly and everywhere). 'Give us the doc' they yelled. 'if you release the docs, we will create kickass drivers in no time'. 'Stop fglrx and give the docs' and of course 'as soon as we have the docs, fglrx will be superflous'.

AMD releases docs - and people still complain. Funny. Or sad. Depends on POV.

[m4rgin4l]

the hilarious thing is:
in the past people complained (very loudly, constantly and everywhere). 'Give us the doc' they yelled. 'if you release the docs, we will create kickass drivers in no time'. 'Stop fglrx and give the docs' and of course 'as soon as we have the docs, fglrx will be superflous'.

AMD releases docs - and people still complain. Funny. Or sad. Depends on POV.

Although I wasn't among those yelling, I surely thought that the release of the documentation would have had a positive effect on the free drivers in a timely manner.

It is obvious by know that the release of documentation is only the first step in having oss drivers of acceptable quality. We cannot hope to achive that when only a handful of developers are trying to cope with constant architectural changes on the X side, and hardware changes on AMD/ATI side.

I know that eventually the devs are going to achieve some success, but my guess is that we're still a few months away from it (over a year maybe). This would mean that I would get acceptable drivers for my card TWO years after the purchase date.

I think the worst part of all this is that the card is actually a fantastic piece of hardware and it leaves the comparable NVIDIA products eating dust, both in price and performance.

[energyman]

well, I am lucky that the fglrx driver is good enough for me. I don't have any use for composite (I usually only turn it on, if I have a desire for eye candy - after a few hours that is satisfied - and the games I play work perfectly fine (ut2004, vegastrike, wesnoth, triplea, xskat, widelands).

[nhaehnle]

Cut AMD some slack. They definitely did the right thing in releasing documentation, which means that those cards will probably end up being supported virtually indefinitely.

Open source driver development is slow, I grant you that, but AMD are even paying developers to work on the open source driver, even though they have no obligation at all to do that (after all, why should they put resources into two different drivers).

To be honest, I have not much experience with fglrx, but it does seem to me like AMD are definitely on the right track - which can't really be said about NVidia. (Of course, Intel is still king of the hill in that respect, even if you account for the desaster that is Poulsbo.)

[m4rgin4l]

Cut AMD some slack. They definitely did the right thing in releasing documentation, which means that those cards will probably end up being supported virtually indefinitely.

Open source driver development is slow, I grant you that, but AMD are even paying developers to work on the open source driver, even though they have no obligation at all to do that (after all, why should they put resources into two different drivers).

To be honest, I have not much experience with fglrx, but it does seem to me like AMD are definitely on the right track - which can't really be said about NVidia. (Of course, Intel is still king of the hill in that respect, even if you account for the desaster that is Poulsbo.)

I'm not implying that what AMD did was in any way wrong. Quite the opposite. I think what they did is great!

What I'm saying is that I took the risk of switch to AMD/ATI after having a pretty good experience with NVIDIA, and that backfired big time.

What I'm also saying is that if the current trend continues, I will never be able to use a latest generation AMD card with the latest of what Linux has to offer. I CAN do that with Intel and NVIDIA (at least for the time being). It would seem that AMD is always a couple of steps behind.

Who knows? Maybe the situation changes in the coming months (I doesn't look like it's going to), and I will be able to switch back. I sincerely hope so. As I said before, I still think that AMD's hardware it pretty good, maybe even the best.

[bridgman]

Just curious, why do you say it doesn't look like things are going to change ? We started the open source graphics work from scratch less than two years ago, working from oldest unsupported chips to newest, and have pretty much caught up already. All shipping GPUs have open source driver support with 2D and Xv acceleration, 5xx/690 and earlier have 3D acceleration, and 3D accel for 6xx/7xx is in a public repo and getting close to being ready for use.

On the fglrx side we started supporting consumer use cases in late 2007, and have been gradually adding consumer-oriented features and distro support since then. "Bleeding edge" distros like Fedora are only being supported with the open source drivers today -- sounds like that is your main complaint ?

Anyways, if you look at where we were two years ago vs where we are today I think you would have to agree the situation is hardly "steady state" or "unlikely to change".

[blindfrog]

Have to agree with one thing. It's taking a way too long! "New" catalyst driver was released almost 2 years a ago and it still can't handle something as "usual" as video playback properly (GL slow and crashy. XV washed out colors and tearing like hell). Also what the hell are those lockups introduced by 9.6?

Free drivers 1st release was too almost 2 years ago and still it hardly supports OpenGL 1.3 with up to r500 and r600+ support second to useless if you consider the amount of money one puts to those cards and it's still going to take at least a year for any proper OpenGL 2 support through gallium.

Quoting billoreilly "******* thing sucks"

[m4rgin4l]

Just curious, why do you say it doesn't look like things are going to change ? We started the open source graphics work from scratch less than two years ago, working from oldest unsupported chips to newest, and have pretty much caught up already. All shipping GPUs have open source driver support with 2D and Xv acceleration, 5xx/690 and earlier have 3D acceleration, and 3D accel for 6xx/7xx is in a public repo and getting close to being ready for use.

On the fglrx side we started supporting consumer use cases in late 2007, and have been gradually adding consumer-oriented features and distro support since then. "Bleeding edge" distros like Fedora are only being supported with the open source drivers today -- sounds like that is your main complaint ?

Anyways, if you look at where we were two years ago vs where we are today I think you would have to agree the situation is hardly "steady state" or "unlikely to change".

I said that it doesn't look like it's going to change because:

a) The software keeps changing (kernel, mesa, X, etc.),
b) New hardware is being released

If you start working from the oldest architecture to the newest, what you do is making the new customers pay for the support of the old ones. Kinda like a Ponzi scheme :)

The support for distros like Fedora is my main concern. If none of the hardware manufacturer provided "support" for it, I would have kept my mouth shut. But this isn't the case. The latest Intel cards are supported, the latest NVIDIA cards are supported. My 1 year old HD4850 isn't. I don't mean official support from the manufacturer. I would settle for decent set of features, even if the driver is provided with an as-is disclaimer.

I thought about where was I two years ago and I realized that I had a working NVIDIA card back then.

[bridgman]

I said that it doesn't look like it's going to change because:

a) The software keeps changing (kernel, mesa, X, etc.),
b) New hardware is being released

That's what I don't understand. We've been dealing with those issues all this time, and in the last ~20 months have pretty much caught up after being out of open source graphics for ~7 years. Open source drivers don't get hit by changes in the underlying software because the developers making those changes can and do update the drivers at the same time - and keeping pace with new hardware introductions is much easier than catching up in the first place.

Progress on the drivers has been pretty much what we said users should expect -- are you saying that you expected the release of specs to result in better, faster drivers "almost immediately" ? If so, I'm sorry but I don't believe that either we or the development community every said anything like that.

If you're saying "I bought an ATI product a year ago because of the announcement about releasing programming specs, but I'm having trouble living with that decision because I really need full-featured 3D support on Fedora" that's a perfectly reasonable statement -- but that isn't how the thread started.

Your original post said "What drove me to buy an ATI card was the news that they were going to publish the specifications and I thought that maybe for once a manufacturer understood how to interact with the community and that they were actually going to make some positive change in the Linux driver scene."

Where do you think we have failed in that regard ?

[crumja]

This thread makes me sad. Believe me though that I've seen equally as many people switch from Nvidia to ATI/AMD (I'm one of them) because Nvidia isn't as great as you make it seem.

A few examples:
The 71xx series of drivers is not receiving support anymore. That means users of everything Geforce2 and older are SOL when it comes to a stable driver. Nouveau crashes all the time and doesn't support any 3D.

The 96xx series took 6 months to get support for Xserver 1.6. It's also likely getting too expensive to support and will be relegated to the waste bin.

In a few years, the same thing will happen to the newer generation of cards.

2D performance is much higher on the ATI/AMD OSS drivers. It's stable, doesn't crash, suspends properly, and offers tear-free Xv. KMS and DRI2 are here already, and Gallium3D will be here by next summer.

Now, those features probably aren't foremost on your needs list, which means that you should've evaluated what the card and driver are capable of before purchasing. I'm as happy as a clam now on my R500 card which I bought one year ago. Tell me, what is it about fglrx that makes it unusable? I was on that driver before I transitioned to the OSS drivers once the support matured.

As bridgman said, the driver team started from scratch on all the cards and have caught up now in all but the 3D for the R6xx/R7xx cards. In the next official release, they will be up to the previous gen level of support. The rate of progress now is fast enough that future generations of cards will be supported much faster than the R6xx/R7xx series is.

[m4rgin4l]

OK, although I still don't understand the premise of the thread. Your original post said "What drove me to buy an ATI card was the news that they were going to publish the specifications and I thought that maybe for once a manufacturer understood how to interact with the community and that they were actually going to make some positive change in the Linux driver scene."

What do you think we have failed to do in terms of interacting with the community etc.. ? Progress on the drivers has been pretty much in line with what we said users should expect -- are you saying that you expected the release of specs to result in better, faster drivers in a shorter period of time ? If so, I don't think any of the developers would have supported that view even 2 years ago.

AMD hasn't failed at all. In terms of interaction with the community AMD did everything right (which is still surprising). The thing is that, at least for me, nothing have changed in the last year, I'm still dealing with a feature-less OSS driver and a non-working proprietary one.

Anyway, I think there's no point in continuing arguing about this. This isn't helpful for anyone. Perhaps the only valuable piece of information that can be extracted from this thread is that Fedora users should stay away from the latest generation AMD cards, a fact that should be obvious for anyone willing to search through the forum.

PS: Please refrain yourselves from suggesting that I should apply any of the patches that are publicly available to the Catalyst driver. Applying those patches is the manufacturer's job, not mine :)

[energyman]

no, patching is the distris job. If your distri can't do it, look for one that does.

gentoo is even more 'moving' than fedora - and there fglrx is not really a problem. The ebuilds include all patches needed. Sure, sometimes you have to get it from bugzilla - or do it yourself in the first days after a release - but it is easily done.

[m4rgin4l]

This thread makes me sad. Believe me though that I've seen equally as many people switch from Nvidia to ATI/AMD (I'm one of them) because Nvidia isn't as great as you make it seem.

A few examples:
The 71xx series of drivers is not receiving support anymore. That means users of everything Geforce2 and older are SOL when it comes to a stable driver. Nouveau crashes all the time and doesn't support any 3D.

The 96xx series took 6 months to get support for Xserver 1.6. It's also likely getting too expensive to support and will be relegated to the waste bin.

In a few years, the same thing will happen to the newer generation of cards.

2D performance is much higher on the ATI/AMD OSS drivers. It's stable, doesn't crash, suspends properly, and offers tear-free Xv. KMS and DRI2 are here already, and Gallium3D will be here by next summer.

Now, those features probably aren't foremost on your needs list, which means that you should've evaluated what the card and driver are capable of before purchasing. I'm as happy as a clam now on my R500 card which I bought one year ago. Tell me, what is it about fglrx that makes it unusable? I was on that driver before I transitioned to the OSS drivers once the support matured.

As bridgman said, the driver team started from scratch on all the cards and have caught up now in all but the 3D for the R6xx/R7xx cards. In the next official release, they will be up to the previous gen level of support. The rate of progress now is fast enough that future generations of cards will be supported much faster than the R6xx/R7xx series is.

CRAP! So I'm screwed either way :)

Anyway, I think you made a very good point. Maybe the next gen cards will be supported faster. In the meantime I have to stick with whatever works NOW.

Thanks for the info.

[m4rgin4l]

no, patching is the distris job. If your distri can't do it, look for one that does.

gentoo is even more 'moving' than fedora - and there fglrx is not really a problem. The ebuilds include all patches needed. Sure, sometimes you have to get it from bugzilla - or do it yourself in the first days after a release - but it is easily done.

Why patching to support the latest kernel (and the rest of the sw stack) is the distro's job? I'm willing to accept that for the OSS drivers, but not for the proprietary drivers.

[doubledr]

Why patching to support the latest kernel (and the rest of the sw stack) is the distro's job? I'm willing to accept that for the OSS drivers, but not for the proprietary drivers.

I think iit is because each distro has its own policy for installing files.

[m4rgin4l]

I think iit is because each distro has its own policy for installing files.

That might be the case for the distro managed packages, but I was referring to the manufacturer provided package.

Also, Fedora does not include proprietary packages on its repos, so they wouldn't do it anyway :)

[energyman]

again, use a distri that puts the user first - like gentoo :P

[Hasenpfote]

Hey m4rgin4l, I'm one of the guys like you! But in my case, I bought a 3870 which doesnt allow me to play Team Fortress 2 in Linux (crashes when loading a map, right now flickering because ATi HDMI is used). I never entered a map. But I bought the 3870 because of AMDs step to release documentation. I could have bought a Geforce 8800GT at that time, but I wanted to support AMD.

In the first step, I justed wanted to tell you:
Are you taking part in OOS driver development? No? Then STFU!
But on the other hand: With buying an AMD card, you already supported them! So, your task is done! Now its their task to give you drivers for your platform.

But the main reason why I'm being so hard to AMD is, because you're right saying "They dont care about costumers!" Need a prove?
The so called "Themen-Woche" (Topic-Week) on planet3dnow: http://www.planet3dnow.de/vbulletin/forumdisplay.php?f=197
It was advertised with users can ask AMD, AMD answers! AMD did NOTHING!!!! Let me repeat: NOTHING!!! They totally blamed one of the biggest sites in Germany! Wanna see how to do it properly? Look at the Intel-Evening: http://www.planet3dnow.de/vbulletin/forumdisplay.php?f=198 They nearly answered everything! Last time, when they couldn't answer, they asked a spezialist at the company and gave the answer later.

AS a stock owner (yeah, I made that mistake, too), I wrote to AMD. The answer?
I will forward your email to the Director of our
product PR division. We will look into the matter and work to find a
solution to ensure all inquiries are addressed appropriately in the
future.Well, thank you! How about apologizing to planet3dnow and making a REAL topic-week?

Overall: You're right! AMD doesn't take care about costumers, after they bought a card. Well.. yeah.. Wait! They supported a funny overclocking event (http://www.planet3dnow.de/vbulletin/blog.php?cp=8). Thanks AMD! Good Job and cheap advertisement for your superduperuberoverclockable CPUs, that can't compare at stock speed with Intels High-End-CPUs.

Sorry for being offtopic... But I needed to say this!

[L33F3R]

I could have bought a Geforce 8800GT at that time, but I wanted to support AMD.

I find it funny how people can be so generous to a company that doesnt properly support your platform and yet not donate a dime to the various projects that make up that platform.

[bakou]

As a user of a Lenovo T500 with switchable graphics between the Intel 4500X chipset and an ATi 3650, I feel the need to add to this discussion. I can't believe we have an AMD employee here going on about "BLEEDING EDGE" software support. That is completely ridiculous. Let me say I am no open source crusader, I use the NVIDIA closed sourced drivers on other machines, and I wold also use fglrx IF IT WORKED. I respect that you have released documents on open source, but if you really wanted to effect changes quickly in the Linux community you would just release the source for fglrx for Linux, since it is obvious you need the open source community's help developing that buggy piece of shit, rather than forcing them to reverse engineer everything from scratch just to protect some kind of trade secrets, which are no longer really secrets anyway.

Since my machine uses many new components I am required to use these so-called bleeding edge kernels. The oldest kernel I can use to give me wireless support is 2.6.27, 2.6.28 supposedly works with it however it is buggy and I experienced some crashes. Additionally, the oldest kernel I can use to get proper support for my Intel chipset GPU with the new Intel drivers is a heavily patched version of 2.6.29, or preferably 2.6.30 (both of these also work great with the wireless, thank you!). Intel or NVIDIA do not send their employees to troll forums complaining about bleeding edge software, rather they actually work on supporting the current state of Linux software, and they do a fine job of it. Also the NVIDIA closed source drivers add support for these 'bleeding edge' kernels usually before they are even officially out.

Since, as you can see I use a laptop from a company you have obviously made a deal with to officially support, I find it rather insulting that your company has sent you here to make excuses, since you are not supporting it whatsoever in Linux. The fglrx driver serves almost no purpose that I can see, it is too buggy to use for any practical reason 3D drivers are used, even if I was willing to downgrade my kernel and forget about supporting the rest of my perfectly good hardware from companies which make working drivers available in a timely manner. It would be completely impractical for me to downgrade my entire system and compromise all my other hardware anyway, just to use drivers with buggy 3D rendering, buggy compositing and buggy video playback (oh right, all 3 reasons to even use a 3D card in the first place!)

I am sorry for the insulting and ranting tone of this post, but given you appear to be posting on behalf of AMD/ATi you are bringing out the worst of my feelings on this issue, and your stance is illogical and inexcusable.

[bridgman]

I am sorry for the insulting and ranting tone of this post, but given you appear to be posting on behalf of AMD/ATi you are bringing out the worst of my feelings on this issue, and your stance is illogical and inexcusable.

No worries. We're here to listen as well.

As a user of a Lenovo T500 with switchable graphics between the Intel 4500X chipset and an ATi 3650, I feel the need to add to this discussion. I can't believe we have an AMD employee here going on about "BLEEDING EDGE" software support. That is completely ridiculous.

With respect, you are using "bleeding edge" differently from the way I do, and then getting mad about something I did not say. The recent Fedora releases have included code which is 3-6 months *ahead* of the upstream kernels -- that's what I'm calling bleeding edge, not the regular released kernels you need for new hardware support.

For distros with recent kernels from upstream I have been using the term "faster moving" (relative to the enterprise distros which make up most of our customer base) - if you don't think that is fair please let me know what you would consider appropriate.

Let me say I am no open source crusader, I use the NVIDIA closed sourced drivers on other machines, and I wold also use fglrx IF IT WORKED. I respect that you have released documents on open source, but if you really wanted to effect changes quickly in the Linux community you would just release the source for fglrx for Linux, since it is obvious you need the open source community's help developing that buggy piece of shit, rather than forcing them to reverse engineer everything from scratch just to protect some kind of trade secrets, which are no longer really secrets anyway.

Please remember that AMD employees are the ones adding new hardware support to the open source drivers, with full access to the hardware design teams. Not sure where "reverse engineering" comes into it but maybe I'm missing your point.

Since my machine uses many new components I am required to use these so-called bleeding edge kernels.

Again, you are the one using "bleeding edge" to refer to regular upstream kernels. I use it to refer to distros containing code which is *ahead* of the upstream.

The oldest kernel I can use to give me wireless support is 2.6.27, 2.6.28 supposedly works with it however it is buggy and I experienced some crashes. Additionally, the oldest kernel I can use to get proper support for my Intel chipset GPU with the new Intel drivers is a heavily patched version of 2.6.29, or preferably 2.6.30 (both of these also work great with the wireless, thank you!).

We need to shorten the gap between new kernel release and support in fglrx in order to deal with the new hardware enablement issues you mention. Matthew and I have both said this multiple times, however we have also said that this will be a gradual effort, not something that happens overnight.

You don't have to wait for us, however -- the fglrx driver is written to an OS-neutral API, and the install package includes source code for a Kernel Compatibility Layer. This allows distro packagers or any interested third party to adapt the code to newer / different kernels without requiring modifications to the driver itself.

Intel or NVIDIA do not send their employees to troll forums complaining about bleeding edge software, rather they actually work on supporting the current state of Linux software, and they do a fine job of it.

Neither do we. You are getting mad about things which we have not said or done. I might do it in the future though, so there's no harm in having a good rant just in case :D

The point I am trying to make (without success, apparently ;)) is that we are continuing to improve the open source and the proprietary drivers in parallel, however right now if you want to use the newest kernel versions then the open source drivers are your best bet.

Also the NVIDIA closed source drivers add support for these 'bleeding edge' kernels usually before they are even officially out.

Our open source drivers also track the latest kernels and X servers, and they are used by developers *making* changes to the underlying code. The proprietary drivers do not, although I expect they will close the gap over time.

Since, as you can see I use a laptop from a company you have obviously made a deal with to officially support, I find it rather insulting that your company has sent you here to make excuses, since you are not supporting it whatsoever in Linux. The fglrx driver serves almost no purpose that I can see, it is too buggy to use for any practical reason 3D drivers are used, even if I was willing to downgrade my kernel and forget about supporting the rest of my perfectly good hardware from companies which make working drivers available in a timely manner. It would be completely impractical for me to downgrade my entire system and compromise all my other hardware anyway, just to use drivers with buggy 3D rendering, buggy compositing and buggy video playback (oh right, all 3 reasons to even use a 3D card in the first place!)

What happens when you run the open source drivers with your kernel of choice ? I realize you won't have 3D HW acceleration yet, although we're getting pretty close, but display, 2D acceleration and video playback should be solid today. If they are not please let us know.

[bakou]

Well you may be using bleeding edge to describe that, but the fact of the matter is fglrx (as far as I'm aware) doesn't support any kernels newer than .28, Which I believe was released around 8 months ago. I am a Gentoo user, I could downgrade but as I explained in my post it just would not make any sense in my situation, as my Intel card works great (I would LIKE to use the more powerful 3D capabilities of my ATi GPU!).

I have no desire to test the open source drivers until they have reliable 3D support for the same reason as above. My idea of good video playback is using mplayer OpenGL with direct rendering and shader based scaling (that is the only technology I am aware of on Linux that can compete with WinXP/Vista VMR, for example).

I didn't know it is the AMD engineers who are adding hardware support to the open source drivers, or working on them at all. In that case, wouldn't it be more a matter of cutting and pasting code from the fglrx driver then re engineering it to be more compatible with X protocols?

I have heard that the R600 drivers are just now starting to render triangles which is very good, but I don't think they are available for much public testing short of checking out drivers on git yourself (which can lead to some very messy situations...).

Why then do you continue to allocate resources to fglrx if you could just allocate them all to the open source effort which is what most people on Linux want anyway, and will probably result in a much better driver in the long run? I see that progress is being made, I just don't understand the logic behind the management of resources, especially in light of your response. Of course with collaboration between the OS community supporting new software and kernels will never be an issue, that is the whole point! That is why it will be better! Why pay your employees to constantly track kernel changes like nVidia when people are willing to do it for you? It makes a lot of sense to go totally open on Linux when there are an army of programmers willing to help. That is exactly why I am so frustrated with this situation where I am in limbo with two drivers where neither one is really useful, I'm sure they are both useful to people who have 2 year old PCs, but this is not how the world works.

Anyway Thank you for your response, I feel somewhat better about the situation now, my main concern is then why aren't more engineers and programmers from AMD working on the R600 open source effort rather than beating the dead horse that is fglrx on Linux (I know it works fine on Windows, that is probably why I'm logged in on Windows now lol)

No worries. We're here to listen as well :D



With respect, you are using "bleeding edge" differently from the way I do, and getting mad about things I did not say. The recent Fedora releases have had code that is 3-6 months *ahead* of the upstream kernels -- that's what I'm calling bleeding edge, not the regular released kernels you need for new hardware support.

For distros with recent kernels from upstream I have been using the term "faster moving" (relative to the enterprise distros which make up most of our customer base) - if you don't think that is fair please let me know what you would consider appropriate.



Please remember that AMD employees are the ones adding new hardware support to the open source drivers. Not sure where "reverse engineering" comes into it but maybe I'm missing your point.






We need to shorten the gap between new kernel / X server release and support in fglrx. Matthew and I have both said this multiple times. We have also said that this will be an incremental effort.



Neither do we. You are ssying things which we did not.



Our open source drivers also track the latest kernels and X servers, and they are used by developers *making* changes to the underlying code. The proprietary drivers do not, although I expect they will close the gap over time.



What happens when you run the open source drivers with your kernel of choice ? I realize you won't have 3D HW acceleration yet, although we're getting pretty close, but display, 2D acceleration and video playback should be solid today. If they are not please let us know.

[bridgman]

I have no desire to test the open source drivers until they have reliable 3D support for the same reason as above. My idea of good video playback is using mplayer OpenGL with direct rendering and shader based scaling (that is the only technology I am aware of on Linux that can compete with WinXP/Vista VMR, for example).

You might be surprised how well Xv works for you on the open source drivers. Same shader based scaling, plus lower CPU utilization and some additional filtering & colour space options. Not saying you should rip out your current setup on my say-so, but it might be worth checking what other users are saying.

I didn't know it is the AMD engineers who are adding hardware support to the open source drivers, or working on them at all. In that case, wouldn't it be more a matter of cutting and pasting code from the fglrx driver then re engineering it to be more compatible with X protocols?

Just to keep us on the same page for names, fglrx is the "engineering" name for the Linux Catalyst driver, which in turn is basically an X/DRI driver that uses common code shared across multiple OSes.

The hardware layers of the Catalyst drivers (including fglrx) are perhaps 20x the size of the corresponding open source code, at least on the 3D side, and use a different underlying architecture. The extra size is what it takes for the last bit of performance and functionality. We were originally hoping to be able to leverage the hardware layers from the proprietary drivers but it turned out not to be practical because of the size and the architectural differences.

On the display/modesetting side, the open source drivers use AtomBIOS, which lets us run the same code used by the proprietary drivers. For functinality not covered by AtomBIOS we generally obtain pseudocode from the Catalyst driver teams and use that as an implementation guide.

The first round of new hardware support was done entirely by community developers while we were building an internal team, but the model that we are using today is that AMD developers focus on adding new hardware support which leaves community developers free to add features and functionality to the existing drivers. The lines are pretty fuzzy though -- community developers are already helping to troubleshoot and fix bugs in the 6xx/7xx 3D driver, and agd5f has been helping with the KMS/GEM/TTM implementation effort.

I have heard that the R600 drivers are just now starting to render triangles which is very good, but I don't think they are available for much public testing short of checking out drivers on git yourself (which can lead to some very messy situations...).

First triangle was 9 months ago. Now we have most of the major functions working and are trying to hunt down a nagging problem with vertex buffers so that we can expand the testing. The back-to-front copy used in GLSwapBuffers is not yet HW accelerated but agd5f is working on that now. We are approaching the point where distro packagers should think about picking up the code but we're not there yet.

Why then do you continue to allocate resources to fglrx if you could just allocate them all to the open source effort which is what most people on Linux want anyway, and will probably result in a much better driver in the long run? I see that progress is being made, I just don't understand the logic being the management of resources, especially in light of your response. Of course with collaboration between the OS community supporting new software and kernels will never be an issue, that is the whole point! That is exactly why I am so frustrated with this situation where I am in limbo with two drivers where neither one is really useful, I'm sure they are both useful to people who have 2 year old PCs, but this is not how the world works.

Simple. The fglrx driver is what we need for the professional workstation market. Over the last 18 months or so we have also started using it to deliver high end features to consumer users. I expect that most consumer users will be perfectly happy with the open source drivers, however.

We could have done all the development in secret and only announced when we were finished, but I think that would have taken longer and annoyed the community in the process. We wanted to hire developers who were familiar with the open source stack, and it seemed to make the most sense to have them keep working closely with the rest of the community. That meant you had to watch and suffer while we development took place rather than just being given the end product, but I still think this is the right way to proceed.

Anyway Thank you for your response, I feel somewhat better about the situation now, my main concern is then why aren't more engineers and programmers from AMD working on the R600 open source effort rather than beating the dead horse that is fglrx on Linux (I know it works fine on Windows, that is probably why I'm logged in on Windows now lol)

The critical path items for the open source drivers have been (a) writing and releasing the programming docs, and (b) coming up to speed on the latest driver stack. I don't think adding more developers would have really made that much difference.

Our largest Linux market is still the professional workstation market, which is almost entirely based on enterprise distros and relatively stable hardware platforms. That's where fglrx comes from, and where it continues to be essential. As long as we are making a proprietary driver, we are also trying to use it as a vehicle to deliver neat new features to consumer users, such as MultiView and Crossfire.

I'm logged in on Windows too, but for different reasons -- after 2 years and a big box of scrapped hardware I still haven't found a Linux analog modem driver or standalone modem that can handle my crappy rural phone line ;(

[energyman]

Hey m4rgin4l, I'm one of the guys like you! But in my case, I bought a 3870 which doesnt allow me to play Team Fortress 2 in Linux (crashes when loading a map, right now flickering because ATi HDMI is used). I never entered a map. But I bought the 3870 because of AMDs step to release documentation. I could have bought a Geforce 8800GT at that time, but I wanted to support AMD.

In the first step, I justed wanted to tell you:
Are you taking part in OOS driver development? No? Then STFU!
But on the other hand: With buying an AMD card, you already supported them! So, your task is done! Now its their task to give you drivers for your platform.

But the main reason why I'm being so hard to AMD is, because you're right saying "They dont care about costumers!" Need a prove?
The so called "Themen-Woche" (Topic-Week) on planet3dnow: http://www.planet3dnow.de/vbulletin/forumdisplay.php?f=197
It was advertised with users can ask AMD, AMD answers! AMD did NOTHING!!!! Let me repeat: NOTHING!!! They totally blamed one of the biggest sites in Germany! Wanna see how to do it properly? Look at the Intel-Evening: http://www.planet3dnow.de/vbulletin/forumdisplay.php?f=198 They nearly answered everything! Last time, when they couldn't answer, they asked a spezialist at the company and gave the answer later.

AS a stock owner (yeah, I made that mistake, too), I wrote to AMD. The answer?
Well, thank you! How about apologizing to planet3dnow and making a REAL topic-week?

Overall: You're right! AMD doesn't take care about costumers, after they bought a card. Well.. yeah.. Wait! They supported a funny overclocking event (http://www.planet3dnow.de/vbulletin/blog.php?cp=8). Thanks AMD! Good Job and cheap advertisement for your superduperuberoverclockable CPUs, that can't compare at stock speed with Intels High-End-CPUs.

Sorry for being offtopic... But I needed to say this!

and cost less then a third.

[energyman]

I find it funny how people can be so generous to a company that doesnt properly support your platform and yet not donate a dime to the various projects that make up that platform.

'generous'?

what would you suggest? buying from nvidia who gives a shit?

[bakou]

First triangle was 9 months ago. Now we have most of the major functions working and are trying to hunt down a nagging problem with vertex buffers so that we can expand the testing. The back-to-front copy used in GLSwapBuffers is not yet HW accelerated but agd5f is working on that now.


My mistake, I was confused. After re-reading the news it seems the latest was that radeon-rewrite R600/700 can now run most simple OpenGL apps like glxgears etc. And with double buffering and shader support I assume we'll have full 3D support soon, good!


Simple. The fglrx driver is what we need for the professional workstation market. Over the last 18 months or so we have also started using it to deliver high end features to consumer users. I expect that most consumer users will be perfectly happy with the open source drivers, however.

The critical path items for the open source drivers have been (a) writing and releasing the programming docs, and (b) coming up to speed on the latest driver stack. I don't think adding more developers would have really made that much difference.

Our largest Linux market is still the professional workstation market, which is almost entirely based on enterprise distros and relatively stable hardware platforms. That's where fglrx comes from, and where it continues to be essential. As long as we are making a proprietary driver, we are also trying to use it as a vehicle to deliver neat new features to consumer users, such as MultiView and Crossfire.



Whats wrong with xrandr? In my limited experience with fglrx (Yes I tried it a while ago on Ubuntu 8.10 but it didn't last long on my system) multiview didn't work very well at all, especially with compiz which is what makes dual screens on Linux these days much more usable (windows actually maximize to the right size properly when using two screens of different sizes across a large virtual desktop etc, which I often do at work). xrandr under xorg 1.6.2, intel 2.7 and compiz 8.2 and the latest 2.6.30 kernel patch is very solid for me, Also according to most benchmarks I have seen crossfire performance is not great compared to Windows. You don't plan to implement crossfire in the open source driver? Sorry but I have a hard time believing fglrx is so perfect that people actually want to use the most powerful ATi cards on Linux over other options for enterprise rendering workstations, even though you surely know better than me.



We could have done all the development in secret and only announced when we were finished, but I think that would have taken longer and annoyed the community in the process. We wanted to hire developers who were familiar with the open source stack, and it seemed to make the most sense to have them keep working closely with the rest of the community. That meant you had to watch and suffer while we did the development rather than just being given the end product, but I still think this is the right way to proceed.


I am glad you didn't keep it secret, that would probably make everyone even more irritated and hopeless if you did :o




I'm logged in on Windows too, but for different reasons -- after 2 years and a big box of scrapped hardware I still haven't found a Linux analog modem driver or standalone modem that can handle my crappy rural phone line ;(

Lame. I once had a rural phone line random disc. issue (in Windows with a softmodem) and we complained so much to the phone company that they came and did some fixes on the line nearby, worked fine after! But yes it sounds like you understand where we are coming from when people on this forum complain about the state of Linux drivers.

[AdrenalineJunky]

As a user of a Lenovo T500 with switchable graphics between the Intel 4500X chipset and an ATi 3650, I feel the need to add to this discussion. I can't believe we have an AMD employee here going on about "BLEEDING EDGE" software support. That is completely ridiculous. Let me say I am no open source crusader, I use the NVIDIA closed sourced drivers on other machines, and I wold also use fglrx IF IT WORKED. I respect that you have released documents on open source, but if you really wanted to effect changes quickly in the Linux community you would just release the source for fglrx for Linux, since it is obvious you need the open source community's help developing that buggy piece of shit, rather than forcing them to reverse engineer everything from scratch just to protect some kind of trade secrets, which are no longer really secrets anyway.

Since my machine uses many new components I am required to use these so-called bleeding edge kernels. The oldest kernel I can use to give me wireless support is 2.6.27, 2.6.28 supposedly works with it however it is buggy and I experienced some crashes. Additionally, the oldest kernel I can use to get proper support for my Intel chipset GPU with the new Intel drivers is a heavily patched version of 2.6.29, or preferably 2.6.30 (both of these also work great with the wireless, thank you!). Intel or NVIDIA do not send their employees to troll forums complaining about bleeding edge software, rather they actually work on supporting the current state of Linux software, and they do a fine job of it. Also the NVIDIA closed source drivers add support for these 'bleeding edge' kernels usually before they are even officially out.

Since, as you can see I use a laptop from a company you have obviously made a deal with to officially support, I find it rather insulting that your company has sent you here to make excuses, since you are not supporting it whatsoever in Linux. The fglrx driver serves almost no purpose that I can see, it is too buggy to use for any practical reason 3D drivers are used, even if I was willing to downgrade my kernel and forget about supporting the rest of my perfectly good hardware from companies which make working drivers available in a timely manner. It would be completely impractical for me to downgrade my entire system and compromise all my other hardware anyway, just to use drivers with buggy 3D rendering, buggy compositing and buggy video playback (oh right, all 3 reasons to even use a 3D card in the first place!)

I am sorry for the insulting and ranting tone of this post, but given you appear to be posting on behalf of AMD/ATi you are bringing out the worst of my feelings on this issue, and your stance is illogical and inexcusable.

if you wanna complain about FGLRX - its at least understandable, but at the same time i don't think you understand the full situation AMD inherited from ATI, or just how far it has come.

but calling the AMD employees who spend their time here trolls is rediculous, as they spend alot of time providing very interesting and usefull information, as well as insight into what exactly is going on at AMD.

furthermore, despite the dmesg spamming (which can be stopped) with patches FGLRX actually works quite well with 2.6.29/2.6.30, i use arch linux and have never had to wait for more then a day or two after a new kernel or driver is released to install it, yes this is some measure of extra work for the community, but bridgeman has explained before that due the the huge codebases and the amount of work to be done AMD decided to try to support kernels in a time frame that would work for the majority of people allowing them to still maximize effort on adding new features and fixing bugs, in an order that once again benefits the majority of people - which is hardly unreasonable considering 2 years ago the drivers were so far beyond horrible it was a joke.

[bakou]

if you wanna complain about FGLRX - its at least understandable, but at the same time i don't think you understand the full situation AMD inherited from ATI, or just how far it has come.


but calling the AMD employees who spend their time here trolls is rediculous, as they spend alot of time providing very interesting and usefull information, as well as insight into what exactly is going on at AMD.


I understand its a bad situation, providing information is good, making excuses as to why you can't compete with your two major competitors is not. You have to understand how angry that makes me as a long time supporter of Linux and finally deciding to become an ATi customer for the first time since the Radeon 9600 pro era, largely because I didn't want an nVidia card that would die after a week in my laptop (despite knowing they have much more practical Linux support at the time, and having better performance at the same price point). I am not an AMD apologist or fanboy, I go with the company who I think will give me a better value for my dollar.


furthermore, despite the dmesg spamming (which can be stopped) with patches FGLRX actually works quite well with 2.6.29/2.6.30, i use arch linux and have never had to wait for more then a day or two after a new kernel or driver is released to install it, yes this is some measure of extra work for the community, but bridgeman has explained before that due the the huge codebases and the amount of work to be done AMD decided to try to support kernels in a time frame that would work for the majority of people allowing them to still maximize effort on adding new features and fixing bugs, in an order that once again benefits the majority of people - which is hardly unreasonable considering 2 years ago the drivers were so far beyond horrible it was a joke.

If the patch works so well why is it not integrated into the main driver releases, even after several releases? I have heard many people complaining about crashing after using the patches so I decided not to try it.

[Hasenpfote]

and cost less then a third.
No, the cost only slightly less than samespeed Intel-CPUs.
What I meant was, why do they show how great their CPUs are overclockable (with stuff everybody has at home like liquid nitrogen or even liquid helium), and can't compare with Intel at stock speeds?

Well, it's not so bad, that AMD is only delivering mid-range CPUs. But then they should give users other advantages, like real good customer support and good drivers (780G on Windows XP anyone?, GTA IV after release on HD 48xx, anyone?). Where is the ATI PowerXpress for Desktops? Why do I need to have a card that uses 60W while being idle? btw: @bridgeman: it is planned to implement PowerXpress on Linux?

[AdrenalineJunky]

No, the cost only slightly less than samespeed Intel-CPUs.
What I meant was, why do they show how great their CPUs are overclockable (with stuff everybody has at home like liquid nitrogen or even liquid helium), and can't compare with Intel at stock speeds?

Well, it's not so bad, that AMD is only delivering mid-range CPUs. But then they should give users other advantages, like real good customer support and good drivers (780G on Windows XP anyone?, GTA IV after release on HD 48xx, anyone?). Where is the ATI PowerXpress for Desktops? Why do I need to have a card that uses 60W while being idle? btw: @bridgeman: it is planned to implement PowerXpress on Linux?

depends on the price point -it goes back and forth a bit. also platform prices make a huge difference as well. for instance, the AMD 955 gives similar performance to the i7 920, costs $65 less, but the cheapest motherboard supporting the i7 920 on newegg is $170, where as motherboards for the 955 can be found much, much cheaper.

as far as knocking ATI for issues - its not like nvidia has a perfectly clean slate there either - drivers for vista right after launch? display lockups on linux a while back? mobile cards falling dead left and right?

[bakou]

depends on the price point -it goes back and forth a bit. also platform prices make a huge difference as well. for instance, the AMD 955 gives similar performance to the i7 920, costs $65 less, but the cheapest motherboard supporting the i7 920 on newegg is $170, where as motherboards for the 955 can be found much, much cheaper.

as far as knocking ATI for issues - its not like nvidia has a perfectly clean slate there either - drivers for vista right after launch? display lockups on linux a while back? mobile cards falling dead left and right?

Yeah as I mentioned earlier the only reason I have an ATi card and am posting here is because I was scared of the mobile GPUs dying. Again, it's not really an excuse. There's a reason why this thread has the title it has. Looks like OP is willing to take the risk of going back to NVIDIA, doesn't it? I've had a small issue with one nVidia card out of like 5 that I've owned but I did overclock it too, and any issue I had with it in Linux relating to drivers was fixed very promptly. They may not be willing to open source much but at least they take the closed source drivers seriously. Seriously enough to not fuck over anyone who buys a new PC for a year.

[L33F3R]

'generous'?

what would you suggest? buying from nvidia who gives a shit?

at least it works. Same for intel, at least to a degree it works. Dont get me wrong i loved the athlon x2 and my old 9600 but you wont see me buy a phenom or a new ati card because for linux (and in the case of phenoms) they suck, period. Im not a dumb consumer, I buy shit that works. Maybe in the future the tides will change but as it stands you buy an ati card for the open source drivers, and thats it. What you propose is rewarding companies for bad behavior.

[energyman]

oh? phenoms suck with linux?

I posted on lkml because of a usb bug in sb700 - and amd devs were very quick in responding.

I once posted on lkml because of an ahci bug in nvidias 520 chipset. It also was fixed quickly - by the ahci maintainer.

Amd's amd64 architecture was even developed with linux devs together:
http://marc.info/?l=linux-kernel&m=107763851825114&w=2

and what have the cpus to do with the graphics? Nothing that is. AMD has still the best prize/performance ratio.

But hey, buy intel and tell us how much 3d sucks - or how poulsbo graphics are working out for you.

[bridgman]

Whats wrong with xrandr? In my limited experience with fglrx (Yes I tried it a while ago on Ubuntu 8.10 but it didn't last long on my system) multiview didn't work very well at all, especially with compiz which is what makes dual screens on Linux these days much more usable (windows actually maximize to the right size properly when using two screens of different sizes across a large virtual desktop etc, which I often do at work).

The RandR screen model doesn't map particularly well onto all of the options and modes supported by the proprietary drivers, so until RandR makes some more advances mapping RandR operations onto fglrx will be a bit of a crapshoot. The open drivers have a much easier time because they generally don't implement anything which RandR doesn't support.

Corbin (MostAwesomeDude) is working on a project over the summer called "shatter", which will allow 2D acceleration to operate across multiple GPUs. This is an essential pre-requisite to adding multi-GPU support to RandR. Once that happens, the RandR vs proprietary driver situation should improve quite a bit.

Multiview was a FireGL-only feature until very recently, so unless you tried it in the last month or two it's likely you were just seeing the work-in-process code.

xrandr under xorg 1.6.2, intel 2.7 and compiz 8.2 and the latest 2.6.30 kernel patch is very solid for me, Also according to most benchmarks I have seen crossfire performance is not great compared to Windows.

My understanding was that Crossfire performance was pretty similar on Linux and Windows; what differed was the number of apps which had profiles included with the driver (since most of the apps are different). Will check.

You don't plan to implement crossfire in the open source driver?

Crossfire is not just "something you turn on", it involves some fairly big changes to all levels of the graphics stack. Most of the programming information required is available already, but so far I haven't run into any devs who think spending time on something like Crossfire in the open drivers is a particularly good idea.

Sorry but I have a hard time believing fglrx is so perfect that people actually want to use the most powerful ATi cards on Linux over other options for enterprise rendering workstations, even though you surely know better than me.

If you have time, it might be worth reading some of the reviews. Most of the hate for fglrx involves using it in environments for which it had not originally been designed or tested. We have started ramping up consumer support but that is relatively recent (<2 years).

[Zhick]

Crossfire is not just "something you turn on", it involves some fairly big changes to all levels of the graphics stack. Most of the programming information required is available already, but so far I haven't run into any devs who think spending time on something like Crossfire in the open drivers is a particularly good idea.
Wouldn't Gallium3d make things like multi-gpu (maybe even across vendors since it's all behind a uniform api) "relatively" straightforward? Probably not, since then somebody would have thought of it already, but you know... it'd be realy kewl. :)

[blindfrog]

Oh dear bridgman are you still forced to use 56kbps modem? That's not right at all! You should call to amnesty :D and through them demand at least some form of 3g network there or something? How about satellite connection anyway?

Mankind still forces someone to use 56kbps modems. It's just plain wrong that someone has to read these forums and answer to our (sometimes appropriate ;)) whining with a modem. I think part of me just died...

[bridgman]

I could live with 56K, but being >20Km from the central office I get 26.4 KB/s on a good day and 13-21Kb/s on a not-so good day. On a bad day I don't get a dial tone :D

Satellite connections are easy to get around here, but not if you live in a pine forest.

[grantek]

Satellite connections are easy to get around here, but not if you live in a pine forest.
"Pine forest", or "forest of unused satellite dish towers"? ;)

[bridgman]

I checked. Red Pine (aka Norway Pine). 70' tall and about 9" wide, on a 6' x 8' grid. Lots and lots of trees. Roughly 170 of them between the satellite and the highest mounting point I can find (other than a 70' tower). Maybe 40-50 if I don't mind the treetops blowing in front of the dish when it's really windy.

[mtippett]

...

Whats wrong with xrandr? In my limited experience with fglrx (Yes I tried it a while ago on Ubuntu 8.10 but it didn't last long on my system) multiview didn't work very well at all, especially with compiz which is what makes dual screens on Linux these days much more usable (windows actually maximize to the right size properly when using two screens of different sizes across a large virtual desktop etc, which I often do at work). xrandr under xorg 1.6.2, intel 2.7 and compiz 8.2 and the latest 2.6.30 kernel patch is very solid for me, Also according to most benchmarks I have seen crossfire performance is not great compared to Windows. You don't plan to implement crossfire in the open source driver? Sorry but I have a hard time believing fglrx is so perfect that people actually want to use the most powerful ATi cards on Linux over other options for enterprise rendering workstations, even though you surely know better than me.

...



Hi,

You seem to have a few things mixed up in the above.

Multiview was in very early development around 8.10, the complete multiview (Xinerama) work was completed this year. Feel free to revisit it this year. Note that Multiview is only really useful for multi-GPU configurations. If you have a single GPU, RANDR is the better solution.

The drivers support RANDR, but realistically that isn't the feature you are looking for. RANDR is primarily related to the dynamic enablement of monitors and the arbitrary placement over a virtual desktop.

The maximization behaviour you are looking at is actually the XINERAMA extension, not RANDR directly. XINERAMA provides the output location over the X server. The Window manager uses the XINERAMA to ensure that maximization is located correctly relative to the monitors.

RANDR is supported with the proprietary driver on all OSes with X Server 1.3 and later.

Crossfire with the proprietary driver is automatically enabled only for a subset of games. Crossfire *only* gives good scaling where the GPU is the limiting factor.

Regards,

Matthew

[L33F3R]

oh? phenoms suck with linux?
No, phenoms suck as a whole. yes they do have good price/performance but at the end of the day it helps to not run a 2nd rate CPU from a company who pumps jobs to india just to pay the bills.
I posted on lkml because of a usb bug in sb700 - and amd devs were very quick in responding.
That doesnt have anything to do with making decent hardware, irrelevant.
Amd's amd64 architecture was even developed with linux devs together:
http://marc.info/?l=linux-kernel&m=107763851825114&w=2
yea, back in the day. I applaud that amd for that and i myself owned 2 athlon 64 systems. Im saying AMD has taken a turn for the worst in recent time.

and what have the cpus to do with the graphics? Nothing that is.

Generally speaking one would want both a CPU and a GPU in a home setup so that is also irrelevant.

But hey, buy intel and tell us how much 3d sucks - or how poulsbo graphics are working out for you.
AMD releases documentation, intel pumps alot of money into X, take your pick. And you wont see me buy intel because I am one who buys a system I can actually use. For the time being that is nvidia. My route is not for everyone but unless your goal is open source then its illogical to go another route. My argument was how people dont donate to what runs their system but will pump money into a video card simply out of faith and judging by your lack of response in that specific regard I will assume I am correct.

[AdrenalineJunky]

No, phenoms suck as a whole. yes they do have good price/performance but at the end of the day it helps to not run a 2nd rate CPU from a company who pumps jobs to india just to pay the bills.



i disagree - for the strong majority of people and purposes anything beyond a phenom 955 is largely overkill. and as for the india crack - intel has tons of overseas jobs as well.

[RealNC]

Don't Intel chips have an "assembled in Malaysia" or something printed on them?

[L33F3R]

i disagree - for the strong majority of people and purposes anything beyond a phenom 955 is largely overkill. and as for the india crack - intel has tons of overseas jobs as well.

for sure. but intel has also pumped a fair chunk of change to bring them back as a recession goal. I cant tell you how many companies regret moving R&D overseas and it makes alot of sense.

And no, in todays environment it is not overkill. Right now I am running XP and OSX visualized ontop of mint which is an already bloated linux distro. I could only wish i had more power. Not everyone is like me tho. Also you must think in terms of how quick your hardware will go obsolete. Sometimes it doesnt hurt to pay a little more if your system will take you that little bit longer. The strong majority of people dont have AAA games to play on linux either, doesnt help either argument but its worthy to note. Overbuying is an investment, maybe its overkill for a while but it doesnt take long for the software to catch up.

[AdrenalineJunky]

Don't Intel chips have an "assembled in Malaysia" or something printed on them?

some of them might, i dunno, all of the assembly and testing plants are outside the us, as well as many of thier wafer creation plants, and 5 of thier 9 research plants.

[L33F3R]

old news. last i heard they were being shifted around.

[AdrenalineJunky]

for sure. but intel has also pumped a fair chunk of change to bring them back as a recession goal. I cant tell you how many companies regret moving R&D overseas and it makes alot of sense.

And no, in todays environment it is not overkill. Right now I am running XP and OSX visualized ontop of mint which is an already bloated linux distro. I could only wish i had more power. Not everyone is like me tho. Also you must think in terms of how quick your hardware will go obsolete. Sometimes it doesnt hurt to pay a little more if your system will take you that little bit longer. The strong majority of people dont have AAA games to play on linux either, doesnt help either argument but its worthy to note. Overbuying is an investment, maybe its overkill for a while but it doesnt take long for the software to catch up.

i don't really consider future proofing a good reason to overbuy - let me put it this way, if you spend 2000 on a computer you will probably have more hardware then you need for at least the first year, then say 2 at more or less a comfortable level, and a year behind the curve.

if you buy a 1000 dollar computer every two years, you'll have a comfortable hardware level till you replace it, and never be behind the curve.personally that seems like a better solution to me.

and actually as far as R&D centers go the majority of AMD's are in the US, 8 out of 14.

[AdrenalineJunky]

old news. last i heard they were being shifted around.

took that information from intels website as of a couple days ago. i looked it up for a discusion i was having at another forum.

[L33F3R]

not as of what intel is currently doing. You cant make a,production plant in 2 days and call it square. Maybe L33F3R can :D.

http://www.dbtechno.com/computers/2009/02/11/intel-invests-billions-to-create-new-chip-plants-in-usa/

Similar to what i seen on CNN. Not sure if its the same time however.

[AdrenalineJunky]

not as of what intel is currently doing. You cant make a,production plant in 2 days and call it square. Maybe L33F3R can :D.

http://www.dbtechno.com/computers/2009/02/11/intel-invests-billions-to-create-new-chip-plants-in-usa/

Similar to what i seen on CNN. Not sure if its the same time however.

hmm, hadn't heard that - chip plants are different then R&D centers.

read a couple more stories and they are shutting down 3 asian plants and 2 american ones, citing that they were using outdated technology.

is cool that they are opening them all up in the us though.

[L33F3R]

personal preference tells me that it doesnt matter where something is made so much as it is good quality, my point was that amd sold 20% of their stock to pay the bills. This was some time ago and doesnt seem to be working for them.

[energyman]

for sure. but intel has also pumped a fair chunk of change to bring them back as a recession goal. I cant tell you how many companies regret moving R&D overseas and it makes alot of sense.

And no, in todays environment it is not overkill. Right now I am running XP and OSX visualized ontop of mint which is an already bloated linux distro. I could only wish i had more power. Not everyone is like me tho. Also you must think in terms of how quick your hardware will go obsolete. Sometimes it doesnt hurt to pay a little more if your system will take you that little bit longer. The strong majority of people dont have AAA games to play on linux either, doesnt help either argument but its worthy to note. Overbuying is an investment, maybe its overkill for a while but it doesnt take long for the software to catch up.

so you should buy amd - because all amd cpus have hardware virt and nested pages, speeding up virtualization. unlike intel were only selcted cpus have hardware virt - and you have to choose between raw speed, sse versions and virtualization...

[curaga]

so you should buy amd - because all amd cpus have hardware virt and nested pages, speeding up virtualization. unlike intel were only selcted cpus have hardware virt - and you have to choose between raw speed, sse versions and virtualization...
Semprons don't have AMD-V, at least the currently available don't.

[aidanjt]

First of all, I'd like to applaud AMD for the fantastic work they've done with the radeon drivers, and you personally, bridgman, for the work and community relations you've maintained. It's quite frankly, brilliant. That said:

Our largest Linux market is still the professional workstation market, which is almost entirely based on enterprise distros and relatively stable hardware platforms. That's where fglrx comes from, and where it continues to be essential. As long as we are making a proprietary driver, we are also trying to use it as a vehicle to deliver neat new features to consumer users, such as MultiView and Crossfire.

I simply cannot agree with this assertion, you have absolutely no reliable metric for measuring the volume of desktop radeon usage. Furthermore, even if this were true, it is in AMD's best interests to have an fglrx which supports the latest X.org and Linux kernel releases, so when a workstation customer buys a FireGL card, the drivers will already be stable and well tested on whatever 'Enterprise' distro they go with. Not that all workstation customers *will* use an 'Enterprise' distro either. And I'm sure they wont be amused when their $2,000 paperweight wont work on their machine. I'm sure you can agree that kind of situation is not in AMD's best interests since it gives away custom to nVidia.

[mtippett]

I simply cannot agree with this assertion, you have absolutely no reliable metric for measuring the volume of desktop radeon usage. Furthermore, even if this were true, it is in AMD's best interests to have an fglrx which supports the latest X.org and Linux kernel releases, so when a workstation customer buys a FireGL card, the drivers will already be stable and well tested on whatever 'Enterprise' distro they go with. Not that all workstation customers *will* use an 'Enterprise' distro either. And I'm sure they wont be amused when their $2,000 paperweight wont work on their machine. I'm sure you can agree that kind of situation is not in AMD's best interests since it gives away custom to nVidia.

We do have a reliable metric for enterprise sales. We sell to OEMs who sell to Workstation customers. There is also marketing research in the Workstation market that indicates which market segments have what OS mix.

Unfortunately, for the consumer side, we have haphazard OEM engagement and no quanitifiable data for how may home of office PCs are running Linux.

Our distro mix is based on direct market information from OEMs and direct WS customers. Yes, there are a percentage of users who don't fit that market profile.

As has been said a few times before, the 2.6.29, 2.6.30 and 2.6.31 kernel bumps have been painful due to some of the upstream changes. Likewise some of the XOrg changes previously were painful too. We ultimately had to make the choice to bloat the installer with a new X driver module, rather than investing further time to combine some of the modules.

Regards,

Matthew

[aidanjt]

As has been said a few times before, the 2.6.29, 2.6.30 and 2.6.31 kernel bumps have been painful due to some of the upstream changes. Likewise some of the XOrg changes previously were painful too. We ultimately had to make the choice to bloat the installer with a new X driver module, rather than investing further time to combine some of the modules.
Agreed, the last year has been pretty rough with the major re-factoring of the F/OSS graphics stack. But I brought this up just because I wanted certain people to stop and think before making claims like "WS users are our primary customers". Aside from being a strawman argument, it's non-verifiable in factual accuracy.

[mtippett]

Agreed, the last year has been pretty rough with the major re-factoring of the F/OSS graphics stack. But I brought this up just because I wanted certain people to stop and think before making claims like "WS users are our primary customers". Aside from being a strawman argument, it's non-verifiable in factual accuracy.

Okay.

Workstation is one of the few quantifiable revenue streams. We *can* say this money coming into the company is tainted with Linux. We can't say that about consumer.　　Other vendors have other reasons for being in Linux. Of course over time the reasons may dilute and triggers may change, but John Bridgman is correct in that we historically have been involved in Linux primarily driven by the WS requirements.

[aidanjt]

...but John Bridgman is correct in that we historically have been involved in Linux primarily driven by the WS requirements.
When worded like that, I can both understand and agree with it. :)

Anyway, thanks again, to you guys for your efforts. Even though I can't make full use of my 4870 I'm grinning and bearing with it out of appreciation and understanding of the technical problems involved with rewriting a full video driver series and the recent graphics stack changes.

[mtippett]

When worded like that, I can both understand and agree with it. :)

Anyway, thanks again, to you guys for your efforts. Even though I can't make full use of my 4870 I'm grinning and bearing with it out of appreciation and understanding of the technical problems involved with rewriting a full video driver series and the recent graphics stack changes.

In particular, where aer your issues with the 4870, not that I can prioritize your issues over other work, but I am interested none the less.

Regards,

Matthew

[aidanjt]

In particular, where aer your issues with the 4870, not that I can prioritize your issues over other work, but I am interested none the less.

Mostly older bugs which are now fixed according to your changelogs, and kernel/xorg-server incompatibilities, then there was the gentoo fglrx maintainer, who lost interest (and obviously that isn't AMD's fault) but that's changed now. Although I haven't tried fglrx 9.7 yet, but I plan to do so whenever I have the time and am in the mood to experiment.

[L33F3R]

so you should buy amd - because all amd cpus have hardware virt and nested pages, speeding up virtualization. unlike intel were only selcted cpus have hardware virt - and you have to choose between raw speed, sse versions and virtualization...

http://en.wikipedia.org/wiki/X86_virtualization

just for reference

[energyman]

from that side:

>Not all recent Intel processors support VT-x as Intel uses the feature to segment their market.[6] The following Intel processors include support for VT-x:[7]

[L33F3R]

yes. the need for visualization goes down as you get into the celeron/sempron markets.

[Ant P.]

Unfortunately, for the consumer side, we have haphazard OEM engagement and no quantifiable data for how many home or office PCs are running Linux.

Have you asked any distros about this? Ubuntu already has an optional survey, and the results are made public (http://popcon.ubuntu.com/). Their restricted driver stats (http://popcon.ubuntu.com/restricted/misc/by_inst) suggest that there are about as many desktop fglrx users as there are nVidia users.

I wouldn't know how that compares to actual FireGL card users though. I'm stuck with a R100 at work :/

[cutterjohn]

In particular, where aer your issues with the 4870, not that I can prioritize your issues over other work, but I am interested none the less.

Regards,

MatthewSince you're "interested", let's fix these freeze on video playback after awake from sleep bug AND freeze on awake from sleep with desktop effects(compiz/beryl) enabled too.. They've been around since 9.2 (or 3) AFAIK and likely longer.

[Panix]

Hey, sorry to interrupt. I'm confused about the issues of ATI cards and drivers in Linux still. I suggest someone presenting a list of what 'works' and doesn't work regarding use of fglrx/binary drivers and open source drivers.

If one wants to play games, the popular ones like Call of Duty 4, WOW and Crysis or whatever, you have to use the fglrx drivers, right? Would a Radeon 4870 card be usable or do you have major issues so you cannot play games? I consider you cannot use the card if you run into issues and have to spend time compiling drivers or otherwise configuring stuff (after the drivers have been installed).

Secondly, if one wants to watch a movie or video that requires the necessary codecs but the codecs are installed and the movie is in divx mode (.avi) or DVD H.264 (for e.g.), is the movie playable without issue when using fglrx drivers? Is there anything you have to do? I read you need to disable composite/compiz (desktop effects). Is that right? Why?

Perhaps, AMD/ATI should pick either the binary or open source driver and focus entirely on it until one or the other is workable and not plagued with issues? In other words, 'fix' one so the problems are minimized or at least no worse than the situation when using a Nvidia card and Nvidia drivers. Then concentrate on the other driver option (be it, open source or binary). Trying to optimize both or dividing the workload on both seems to be very difficult since many people have been disappointed in the current status for ATI hardware so far (even if it has come a long way in Linux). It's one thing to say that development have been made at a faster pace and improvement is at a greater significance but if buyers of cards that cost around $200 and more still have extreme issues, that is a major problem. Buyers of these cards don't care if ATI drivers in Linux are improving. They want to use their card without major obstacles.

If they have to wait months before playing a recent game or deal with tearing with videos/movies they have, they won't tolerate it. If the alternatives (Nvidia, Intel etc.) allow for avoiding the isues, they will switch or demand sufficient improvement.

Anyway, I would like more specifics or a list of what works and doesn't and which driver path (open source or binary/fglrx) you should use and what to expect when using it. I'd like to get an ATI card still because they are a good price, especially ATI Radeon HD4850 or 4870 cards. I would probbably have a Windows partition so I could use it in Windows until significant improvements happen since I'm planning on having a second machine/build. But, I would like to know whether there would be any major progress in Linux so that I could boot up Linux at some point and use ATI hardware to watch movies/play games or what not.

Anyway, sorry for rambling on. I hope I didn't disturb the flow of the discussion too much.

[RealNC]

Hey, sorry to interrupt. I'm confused about the issues of ATI cards and drivers in Linux still. I suggest someone presenting a list of what 'works' and doesn't work regarding use of fglrx/binary drivers and open source drivers.

It's mostly a matter of how it works. That's why it's confusing I guess. With fglrx, everything works. But if you try, you see that even while it does work, it does so in a very, very crappy way. So many questions can't be answered simply with "works" and "doesn't work". That's for fglrx though. I don't know about the open drivers, they don't support my card yet.

[BlackStar]

Since you're "interested", let's fix these freeze on video playback after awake from sleep bug AND freeze on awake from sleep with desktop effects(compiz/beryl) enabled too.. They've been around since 9.2 (or 3) AFAIK and likely longer.

I can confirm both issues (4850, amd64).

[m4rgin4l]

Hey, sorry to interrupt. I'm confused about the issues of ATI cards and drivers in Linux still. I suggest someone presenting a list of what 'works' and doesn't work regarding use of fglrx/binary drivers and open source drivers.

If one wants to play games, the popular ones like Call of Duty 4, WOW and Crysis or whatever, you have to use the fglrx drivers, right? Would a Radeon 4870 card be usable or do you have major issues so you cannot play games? I consider you cannot use the card if you run into issues and have to spend time compiling drivers or otherwise configuring stuff (after the drivers have been installed).

Secondly, if one wants to watch a movie or video that requires the necessary codecs but the codecs are installed and the movie is in divx mode (.avi) or DVD H.264 (for e.g.), is the movie playable without issue when using fglrx drivers? Is there anything you have to do? I read you need to disable composite/compiz (desktop effects). Is that right? Why?

Perhaps, AMD/ATI should pick either the binary or open source driver and focus entirely on it until one or the other is workable and not plagued with issues? In other words, 'fix' one so the problems are minimized or at least no worse than the situation when using a Nvidia card and Nvidia drivers. Then concentrate on the other driver option (be it, open source or binary). Trying to optimize both or dividing the workload on both seems to be very difficult since many people have been disappointed in the current status for ATI hardware so far (even if it has come a long way in Linux). It's one thing to say that development have been made at a faster pace and improvement is at a greater significance but if buyers of cards that cost around $200 and more still have extreme issues, that is a major problem. Buyers of these cards don't care if ATI drivers in Linux are improving. They want to use their card without major obstacles.

If they have to wait months before playing a recent game or deal with tearing with videos/movies they have, they won't tolerate it. If the alternatives (Nvidia, Intel etc.) allow for avoiding the isues, they will switch or demand sufficient improvement.

Anyway, I would like more specifics or a list of what works and doesn't and which driver path (open source or binary/fglrx) you should use and what to expect when using it. I'd like to get an ATI card still because they are a good price, especially ATI Radeon HD4850 or 4870 cards. I would probbably have a Windows partition so I could use it in Windows until significant improvements happen since I'm planning on having a second machine/build. But, I would like to know whether there would be any major progress in Linux so that I could boot up Linux at some point and use ATI hardware to watch movies/play games or what not.

Anyway, sorry for rambling on. I hope I didn't disturb the flow of the discussion too much.

You'll hear a lot of people swearing that fglrx works and it is the best, but then you'll also hear people saying that it's crap. I'm in a third group, the ones that happen to use an unsupported distro.

So that's the thing. If you're hoping to get an acceptable fglrx experience on Linux, you'll have to stick to whatever AMD chooses to support. It won't be perfect, though.

In regard of the oss drivers, yes, they are progressing really fast, but they're still pretty far away from an acceptable set of features. It'll be at least a year (and a half?) until the useable drivers reach the most popular distros.

My advise is this: If you're a serious gamer and a Linux user, avoid AMD altogether (as I'm going to do until they get their shit together). If you don't mind using Windows to play, then an AMD card is not such a bad idea, your Linux experience will suck, but at least you'll be able to play the games in all their intended glory.

One more thing, forget about using Intel to play games. Until Larrabee hits the market, Intel cards are only good for running WordStar and Lotus 1-2-3 on a text console.

[Panix]

You'll hear a lot of people swearing that fglrx works and it is the best, but then you'll also hear people saying that it's crap. I'm in a third group, the ones that happen to use an unsupported distro.

So that's the thing. If you're hoping to get an acceptable fglrx experience on Linux, you'll have to stick to whatever AMD chooses to support. It won't be perfect, though.

In regard of the oss drivers, yes, they are progressing really fast, but they're still pretty far away from an acceptable set of features. It'll be at least a year (and a half?) until the useable drivers reach the most popular distros.

My advise is this: If you're a serious gamer and a Linux user, avoid AMD altogether (as I'm going to do until they get their shit together). If you don't mind using Windows to play, then an AMD card is not such a bad idea, your Linux experience will suck, but at least you'll be able to play the games in all their intended glory.

One more thing, forget about using Intel to play games. Until Larrabee hits the market, Intel cards are only good for running WordStar and Lotus 1-2-3 on a text console.
Well, I haven't heard a lot of people swear up and down that fglrx works! At least, not yet. :)

What are the supported distros? Fedora and Ubuntu are included, right? I use those two most often although I use Kubuntu instead of Ubuntu. Not sure if the graphics/video experience is different with the KDE desktop (I have read people here ask for reviews using KDE distro or Kubuntu and I agree, btw).

I could live with firing up Windows (for games or what else) in the meantime although not proud to say it. However, I'm saying that because I'd have more than one machine. Furthermore, I watch movies on my computer so I was wondering how the experience is if you have an ATI video card. If I have to use Windows for that, too, then that is pushing me towards Nvidia again.

I liked ATI cards, the 4870 and 4890 since they are shorter cards than Nvidia comparable cards (in terms of performance) so that I can use any case with them easily (maybe even my Antec Solo?!?) and supposedly, the temps (power consumption, too?) are a bit better?

However, like you, I am not sure I want to play the waiting game and be forced to use Windows for very long when I could just go to Nvidia and use whatever OS I want and do what I want without worrying too much about graphics/video issues. Although, I must say, I have some strange video behavior with LiveCDs with my EVGA 7950 GT card. Even the latest K/Ubuntu 9.04 LiveCD presented the issue and I still have no explanation for it. :) I assume once the OS is installed and the drivers are installed, things improve, however.

Thanks for the replies, guys! :D

[cutterjohn]

I can confirm both issues (4850, amd64).Ok, in my case MSI GT725-074US, Ubuntu 9.04 x86-64. I didn't bother posting my specs as in EVERY single cat release thread I've mentioned that they're still there, and usually leave my specs there.

I had a shocker yesterday though, this bug is apparently, intermittent, as I tried to play a video with mplayer after sleep and it actually played. Although now that I'm thinking about it this behavior was exhibited in the past as well. I still think like I mentioned elsewhere that the state is not being corrected stored on sleep or possibly restored when the machine is awakened.

In any event my CTRL-PRTSCR + RSEINUB keys are getting worn out.


Windows, heh, the dreaded atikmdag has stopped working has started showing up when playing 3 of my games now, which prompted me to use mobility modder & installed cat 9.7s, and install SP2 for Vista. They seem to have helped slightly, but it's very odd that this problem only seems to show up in 3 of my games, Drakensang, UFO: Aftershock(buggy anyways), and rarely in Mount & Blade(vanilla). I'm guessing a combination of weird DirectX programming(all of those games are by small studios) and possibly one of the recent M$ updates(last month). It's hard to tell though as it first reared it's ugly head with Aftershock which is horribly buggy anyways. Happen at any temps according to ATI Tray Tools, plus furmark didn't trigger it after running fullscreen @ various resolution for 20m or so, passes multiple memtest+ passes, so I'm inclined to believe some strange rendering error or something fishy in Vista itself.

Running Oblivion, STALKER SoC, and some other games didn't seem to trigger this, so I'm also inclined to doubt hardware problems in addition to the above furmark & memtest+ results.

SP2 also ate my wifi drivers a while later, but the drivers needed to be update anyways(195MB?! WTF?!), and I see that the problem is common after a quick Google, but it's a good thing that I had linux installed otherwise I'd've been SOL as I'm not sure where my driver + restore disks(useless anyways as they're the kind that want to make the notebook look like it did when shipped apparently) are ATM.

[Pulfer]

However, like you, I am not sure I want to play the waiting game and be forced to use Windows for very long when I could just go to Nvidia and use whatever OS I want and do what I want without worrying too much about graphics/video issues.

Using nVidia is good idea. I recently replaced ATI 4850 card with nVidia 250 card and I was impressed with nVidia's proprietary driver quality. ATI's hopeless Cataclysm sucks in everything, ATI's promising open source ati/radeon driver is very good in 2D and video playback, but with nVidia you get all features (2D, 3D etc) at the same high level of quality.

Most likely I will buy ATI card in future, when open source driver gets full-feature support for modern cards. But for now buying expensive modern ATI card (like 4850) is a waste of money if you use Linux...

[RealNC]

I agree with everything except the part about the 4850 being expensive.

[Pulfer]

I agree with everything except the part about the 4850 being expensive.

I mean it's relatively expensive (compared to 4650 etc or some nVidia cards).

It makes sense to buy ~4850 if you
1. use Windows to play modern games
2. and Linux only to watch movies, browse Internet and do other things that don't require 3D support.

But if you don't play in Windows you don't need powerful card for (2) then.

And if you need 3D in Linux, you better avoid ATI for now.

[NeoBrain]

I can confirm both issues (4850, amd64).

I can confirm that S2RAM never worked for me with Compiz enabled, never since 8.42 (oh well, anything earlier couldn't even run Compiz, 8.41 doesn't count)

[cutterjohn]

Using nVidia is good idea. I recently replaced ATI 4850 card with nVidia 250 card and I was impressed with nVidia's proprietary driver quality. ATI's hopeless Cataclysm sucks in everything, ATI's promising open source ati/radeon driver is very good in 2D and video playback, but with nVidia you get all features (2D, 3D etc) at the same high level of quality.

Most likely I will buy ATI card in future, when open source driver gets full-feature support for modern cards. But for now buying expensive modern ATI card (like 4850) is a waste of money if you use Linux...Should've gotten the GTX260-216 as the 250 is a re-badged 9800GTX+. The 260 eats 4850s for breakfast, and holds its own against 4870s. (Even the GTX260M is pretty good... but so is the lessermobile model GTS160M IIRC almost @ 260M performance. Well AMD had their window of opportunity, but it looks like it's passed them by again(K8, granted they had less time with their mobile GPUs but given that they had years of dominance with the K8 and couldn't manage to keep it up... but that's what happens when you fail at R&D, letting suits run an engineering firm...).)

[Pulfer]

Should've gotten the GTX260-216 as the 250 is a re-badged 9800GTX+.

I've been thinking about that but... Actually, GTX260-216 is too expensive for my needs. It costs ~230$ here while ASUS 1024Mb "ENGTS250 Dark Knight/HTDI" is ~170$. So I decided to buy 250 which is equal to 4850 in performance, I guess.

[naraki]

Do the HD2000+ only Catalyst drivers still suck?

[cutterjohn]

I've been thinking about that but... Actually, GTX260-216 is too expensive for my needs. It costs ~230$ here while ASUS 1024Mb "ENGTS250 Dark Knight/HTDI" is ~170$. So I decided to buy 250 which is equal to 4850 in performance, I guess.Huh? A few weeks ago they were about $170 or so at newegg...

I can already see several right now, $150 after MIR... $180 before...

Must be MIR season on nVidia cards as it looks like ALL of them have them... even though the GT250 is slightly cheaper, ~$20 for decent ones, I'd personally still go for the 260-216...

[Pulfer]

Huh? A few weeks ago they were about $170 or so at newegg...

People who live in countries with such low hardware prices are lucky. :-)

[Qaridarium]

Do the HD2000+ only Catalyst drivers still suck?

for me.. i do not have any problem...

i use a catalyst9.9 beta

kernel 2.6.30 works, wine works in oblivion an other direktX9 games.

but yes i do not use Composit manager.. .. (full composit support witout patch your xserver will come on ubuntu 9.10 becourse intel put a bug in the xserver only affect amd-carts)

the wine spezific part of openGL3.2 is also in the beta driver..

only the geometric shader part dosn't exist right now.

[NeoBrain]

And if you need 3D in Linux, you better avoid ATI for now.

I can only speak for myself here ofc, but ATI's 3D support with fglrx has been working perfectly for me for some releases new (actually, I can't remember any issue with it apart from the famous memleak in 8.42).
The only problem was Wine, but even this work reliable for me now.

My other issues come down to the damn blur plugin in Compiz/KWin not working, starting a second X server hardlocking my system (xinit xterm -- :2), Wine taking 10 seconds to initialize since 9.7 and S2RAM/S2Disk not working reliably.

[Pulfer]

I can only speak for myself here ofc, but ATI's 3D support with fglrx has been working perfectly for me for some releases new (actually, I can't remember any issue with it apart from the famous memleak in 8.42).
The only problem was Wine, but even this work reliable for me now.

I wish I could say the same... But Catalyst never worked well for me. It was freezing my system here and there. The last time I installed 9-6, opened KDE 4.2.4 settings (wanted to check out if I could turn the compositing on with XRender) and got system freeze. After that I reverted back to open source ati/radeon. In general, I don't remember any Catalyst release that didn't give me hard locks. In one release you get freeze here, in another you get freeze there...

[cutterjohn]

Pulfer, I really don't thinkit hard locks the entire machine, but only X. Well, that's my suspicion at least since I still, ATM, don't have a second machine handy to attempt to ssh in and see whats going on(if I could).

Also, can't you find somewhere in Europe(outside of Russia) to maybe order your hardware from? Somewhere in Europe they must also have stores with decent prices even if they usually already tack on the 20% or VAT, which you should be able to get back anyways...

[Panix]

for me.. i do not have any problem...

i use a catalyst9.9 beta

kernel 2.6.30 works, wine works in oblivion an other direktX9 games.

but yes i do not use Composit manager.. .. (full composit support witout patch your xserver will come on ubuntu 9.10 becourse intel put a bug in the xserver only affect amd-carts)

the wine spezific part of openGL3.2 is also in the beta driver..

only the geometric shader part dosn't exist right now.
So, if you use a 'Catalyst' beta (aka experimental ATI driver?), 2D and 3D work flawlessly in Linux using a newer ATI card? Oh yeah, you need a recent Linux distro with at least 2.6.30 kernel?

Wouldn't people be jumping on this, then? An official announcement must be forthcoming?

I think I would really consider an ATI card then. At least, I could get one to replace my Geforce 7950 GT which works fine (I think) in any distro I've tried. There are issues when running a LiveCD or DVD though and it could be partly Nvidia-card related or just not being able to handle any card with whatever drivers are pre-loaded on the CD (this is with KDE4 desktop distros, btw).

I find it preplexing why ATI would still have so much trouble when some older Nvidia hardware works fine with any distro regardless of the kernel version.

[snogglethorpe]

the hilarious thing is:
AMD releases docs - and people still complain. Funny. Or sad. Depends on POV.

One of the most basic attributes of humans is that they love to whine. Also, they hate to think too hard. And they want a pony.

This thread is a perfect example.

[energyman]

So, if you use a 'Catalyst' beta (aka experimental ATI driver?), 2D and 3D work flawlessly in Linux using a newer ATI card? Oh yeah, you need a recent Linux distro with at least 2.6.30 kernel?

Wouldn't people be jumping on this, then? An official announcement must be forthcoming?

I think I would really consider an ATI card then. At least, I could get one to replace my Geforce 7950 GT which works fine (I think) in any distro I've tried. There are issues when running a LiveCD or DVD though and it could be partly Nvidia-card related or just not being able to handle any card with whatever drivers are pre-loaded on the CD (this is with KDE4 desktop distros, btw).

I find it preplexing why ATI would still have so much trouble when some older Nvidia hardware works fine with any distro regardless of the kernel version.

take it with a bit of salt. 2d is fine with current drivers too - as long as you don't use composite. And he claims to not use it... so it is not clear what happens if you use composite with the upcoming drivers.

They will probably out in the next 10 days, so we will all knew soon.

[BlackStar]

take it with a bit of salt. 2d is fine with current drivers too - as long as you don't use composite. And he claims to not use it... so it is not clear what happens if you use composite with the upcoming drivers.

They will probably out in the next 10 days, so we will all knew soon.

2d is fine even with composite with 9.7, slightly slower than without but quite smooth nonetheless. 4850, x86_64 mode.

[energyman]

2d is fine even with composite with 9.7, slightly slower than without but quite smooth nonetheless. 4850, x86_64 mode.

fine is a wide field. It is a bit slow, but bearable in my opinion, but too slow for a lot of people

[RealNC]

The problem isn't the speed. At least not the biggest one. It's mainly the glitches and bugs when composite is enabled.

[BlackStar]

The problem isn't the speed. At least not the biggest one. It's mainly the glitches and bugs when composite is enabled.

Other than the intermittent black rectangles in Firefox that go away as soon as you scroll, what other issues are you seeing with composite enabled? Seems to be working pretty nicely here (and I have a feeling that most composite glitches can be traced back to Compiz rather than fglrx...)

[AdrenalineJunky]

The problem isn't the speed. At least not the biggest one. It's mainly the glitches and bugs when composite is enabled.

as of installing 9.7 in 64 bit arch with kwin compositing the only buggs i have are related to playing videos.

with the opengl-ati redering in smplayer the a video will pause with transparency on move and wobbly windows while i'm moving it, with xv it does pause, but xv still has some pretty bad tearing issues.

in windowed every once in a while, with video playing, it will sort of loose track of where the borders of the video should be - so the video is playing over part of the area that should be reserved for smplayers controls, this has only happened a couple times.

in full screen mode if i move the mouse to where the control bar at the bottom of the screen pops up, the desktop image will flash on the screen for a fraction of a second, same thing happens if i hit the volume +/- quick-keys.

high def videos still have significant tearing issues with compositing on, Standard definition seems to work fine.

[Qaridarium]

So, if you use a 'Catalyst' beta (aka experimental ATI driver?), 2D and 3D work flawlessly in Linux using a newer ATI card? Oh yeah, you need a recent Linux distro with at least 2.6.30 kernel?

Wouldn't people be jumping on this, then? An official announcement must be forthcoming?

I think I would really consider an ATI card then. At least, I could get one to replace my Geforce 7950 GT which works fine (I think) in any distro I've tried. There are issues when running a LiveCD or DVD though and it could be partly Nvidia-card related or just not being able to handle any card with whatever drivers are pre-loaded on the CD (this is with KDE4 desktop distros, btw).

I find it preplexing why ATI would still have so much trouble when some older Nvidia hardware works fine with any distro regardless of the kernel version.

on my system 9.6 and 9.7 works fine to!

but the 9.8 works much better becourse i can use the newest stable kernel.

and there are improvments in wine and others.

why sould i use stable driver if the experimental/unstabke works perfekt for me witout problems?