Slashdot Mirror

You might want to rethink your opinion on AMD, they are getting there: http://www.x.org/wiki/RadeonFeature

Since when does http://www.x.org/ require Flash?

The ATI drivers for Linux were never perfect, but they worked decently. But ATI/AMD would drop support for older chips that were still in use. The open source community never provided a shim to let these older drivers work with newer builds of X.

Does open sourcing the drivers really fix the compatibility problem? To me, not building a shim suggests a general lack of caring about ATI drivers. Do we really need the source to give a future to aging ATI/AMD chips?

As of January 19 phoronix, puts the average speed of the latest available open-source driver at roughly 70% the speed of the Catalyst driver before the pre-R600 support was discontinued in early 2009. This is using composite results from the ATI Radeon X1800XL, Radeon X1800XT, and X1950PRO graphics cards being benchmarked on Nexuiz, Warsow, OpenArena, World of Padman, and Urban Terror. These cards use the R300g driver. Newer cards using the R600g driver (cards with HD in the name) are not currently anywhere near these results.

A bit of history:

A long time ago, documents describing the specifications of graphics cards were generally available under NDA to XFree86 developers. Then Nvidia started releasing binary only drivers. ATI eventually followed suite. The last series with docs available from this era is ATI's R200. The R300 released in 2002 did not have docs released for it. Following the lack of docs, driver development stagnates.

April 6, 2004 XFree86/X.org fork.
After Keith Packard was kicked out of the XFree86 core group and XFree86 switched to non-gpl compatible license a fork ensues. Project Leadership of XFree86 had been basically hostile to developers and had retarded increases in the developer base and improvements in the graphics stack for literally years. Following the fork a renaissance in X Server development begins.

July 24, 2006 AMD acquires ATI.
Speculation about open drivers begin.

May 10, 2007 Red Hat Summit
AMD's Henri Richard says something about improving the open source drivers. Speculation becomes flood of rumors.

September 06, 2007 ATI/AMD's New Open-Source Strategy Explained
AMD announces plans to contribute specification documents and code to the open source drivers. By this time successive X.org releases have seen:

• Removal of XIE, PEX and libxml
• Window translucency, XDamage, Distributed Multihead X, XFixes, Composite
• EXA, major source code refactoring. Switch to autotools build system instead of Imake
• EXA enhancements, KDrive integrated, AIGLX
• Removal of LBX and the built-in keyboard driver, X-ACE, XCB, autoconfig improvements
• Input hotplug, output hotplug (RandR 1.2), DTrace probes, PCI domain support.

September 11, 2007 XDS2007 Program
The "softpipe" talk by Keith Whitwell of Tungsten Graphics is the earliest reference I can find to Gallium3D. References to Gallium3D show up on Tungsten Graphics website at approximately the same time according to internet archive. Apparently Tungsten Graphics released a softpipe driver (gallium driver for cpu) at this time, along with a "proof of concept" i915 driver.

September 12, 2007 AMD Releases 900+ Pages Of GPU Specs
RV630 Register Reference Guide and M56 Register Reference Guide.

January 04, 2008 AMD Releases Additional R600 GPU Programming Documentation
M76 and RS690 register guide weighing in at 458 and 422 pages respectavly. Contains LVTMA and i2c information not found in previous docs. LVTMA is the second digital output block on the ATI R500/

No, OpenGL 2.1 is the highest supported, but this is because Mesa - the open source implementation of OpenGL - doesn't support anything higher. Somebody needs to implement OpenGL 3 and 4 before there can be drivers for it.

The GPU in rs690 is actually a 4xx variant, not 2xx.

Not according to the X driver.

Are you using the power saving features described at the bottom of http://www.x.org/wiki/RadeonFeature ?

It's the CPU power-saving that AMD did not contribute to Linux, not the GPU. I can't USE the GPU long enough under linux to get to the point where I need power saving.

The Linux graphics drivers are in a state of extremely heavy development at the moment, you can follow the fun here, here and here. The main Phoronix website also has a lot of articles.

The Linux graphics drivers are in a state of extremely heavy development at the moment, you can follow the fun here, here and here. The main Phoronix website also has a lot of articles.

The X.Org team has been planning X12 for years. http://www.x.org/wiki/Development/X12 is the roadmap. We will probably start X12 development as soon as all X11 bugs are fixed. (Haha, only serious.)

Also, what do you mean by "resources or balls?" Plenty of challengers have shown up over the years. DirectFB, Fresco, Berlin, Y Windows, etc. None of them displaced X because *X is a hell of a lot better than you give it credit for*. Wayland's developer realizes this; he's not trying to replace X, but to work alongside it.

Linux is Unix, and X is a part of any proper Unix.

INVALID

I use Linux exclusively. At work, where I'm required to use Windows, so I run several Linux virtual machines. My kids run Linux. My servers run Linux. 90% of my Linux machines do not have X or any other windowing system. In my previous job, 100% of all Linux systems (several thousand) had no X server, hell, 99.99% of them were headless.

Linux is not Unix, it's LIKE Unix. X is not Linux - X runs in many environments - including OSX and Windows

I don't know what the performance is with the proprietary fglrx driver (ATI/AMD stopped supporting my X1600 Pro a while ago), but I think the current status of the open-source driver is "works, but don't expect anything amazing".
Last I knew about the proprietary driver was that it performed better than the OSS ones, but still not as good as Windows.

For reference: http://wiki.x.org/wiki/RadeonFeature and the pages linked to from there.

Like x.org?

Other way around; AMD has always released specs and started releasing ATI specs after ATI was acquired. You may notice that http://www.x.org/docs/AMD/ is lacking docs for the r200 and earlier; that's because AMD made the acquisition during the r400 era, and the docs for older chipsets were more or less lost forever at that point.

Right now, the open-source drivers are called radeon, r300, r600, etc.; one developer committed his code as "amd" instead at one point. (It got changed to avoid end-user confusion.)

According to the Radeon feature page, not yet. The R600 and R700 cards has it, but Evergreen doesn't.

spoke too soon:
http://lists.x.org/archives/xorg-devel/2010-August/012037.html

And also http://lists.x.org/archives/xorg-devel/2010-August/012048.html

Acceleration of H.264 is different than OpenGL acceleration. You can have a card with full GL acceleration that doesn't accelerate H.264 decoding. Indeed many older cards were like this. The original GeForce 8800s didn't have full H.264 acceleration, despite their massive amount of 3D hardware.

However, you can use GPGPU techniques to run almost any software you want on the 3D engine, including video decoding. I understand this is a planned feature of the opensource driver for older Radeon cards: "Video Decode (XvMC/VDPAU/VA-API) using the 3D engine"

AMD and Intel need to have a contest on the shittiest driver category. I have one of each. Each revision of xserver-xorg-video-intel bricks my laptop in a new and exciting way. And AMD's fglrx is a steaming pile of rendering errors, inconsistent performance, and crashes.

On the other hand, both Intel and AMD have released specs and participate in open source development. So in the long run, either one is a better choice than NVidia. So I'll continue to complain about them and submit bug reports. It's the open source way.

ATI@OSS drivers: Using it right now, running good.
more info about supported features

I have profiled lots of things with Valgrind, and what you say is true. However, to the best of my knowledge, the xorg-server is still single-threaded. Mind you, I haven't been in the source in a long time, so maybe the docs are out of date?

http://www.x.org/wiki/Development/Documentation/Performance#InteractivePerformance

But I thought the open source way of doing things was better, isn't that what we always hear on Slashdot? What is it that the FLOSS advocates are always saying over and OVER and over? "Release the specs, we'll take care of the rest"?

Well they have released the specs, so the ball is in your court now. But the GP was spreading FUD by saying even the Windows drivers don't work with ATI cards, which I called him out on because since being bought by AMD their drivers have been nothing but stable and as a PC builder and repairman I move a lot of ATI product. In fact ever since Bumpgate I have sold nothing but ATI cards and since the "bang for the buck" has gotten so high I have moved to selling AMD builds exclusively and my customers couldn't be happier with stability and performance.

But from what I understand the open source driver is coming along quite quickly and you can go up to the 4xxx series now and considering the fact that you can get 4xxx hardware for under $50 it won't cost you hardly anything to try it out.

• xserver-xorg-video-ati 1:6.12.6-1
• libdrm2 2.4.18-2
• libgl1-mesa-dri 7.7-4 (and the other mesa stuff)
• linux-image-2.6.32-trunk-amd64 2.6.32-5

I can play tremulous, urbanterror, and openarena normally, but nexuiz crashes the X server and the commercial ETQW and quake4 crashes missing some higher OpenGL functionality, so YMMV.
It is my opinion that this risky "develop everything anew" Gallium3D strategy will pay off, because the AMD/ATI, Intel, Nouveau and VMware teams can then bundle their efforts on the exciting higher-level "state tracker" layers (such as more recent OpenGL with GLSL for games, and OpenCL!, and maybe some kind of video acceleration or at least DCT also if they agree on which one) and only need to write modesetting and Gallium driver compiler stuff themselves.
But nobody can say for sure if all the temporary instabilities and incompatibilities will all be behind us at the end of 2010. It's good enough for me :-)

http://www.x.org/docs/AMD/

r700 specs came out some months after the chips did, the same is expected for r800 (should be soon now).

As for the Intel GMA500, its driver is shit and always will be, because is it actually a PowerVR mobile chip.

The other modern Intel chips are underpowered, but not really that bad (the old ones had bugs and serious misfeatures).

?
(lameness filter: activate!)

And here is the page that describes current 3d support in Gallium3D: http://www.x.org/wiki/GalliumStatus

Notice the sea of grey under the nVidia chipsets. So if you want games, keep using the binary for now.

I will agree that open-source drivers for ATI cards are fantastic (and binary drivers are truly terrible). I'm using the new (using release candidates of kernel 2.6.32) r600 hardware acceleration support, and it's already working very well for me (mostly for Google Earth and Kwin desktop effects, both of which work flawlessly and very smoothly).

However, I would caution that support for the chip mentioned in this article (Radeon Evergreen) is marked as "TODO". Presumably, it should progress relatively fast, because AMD is basically being helpful.

Nvidia deserves some credit for updating their binary driver regularly, and making helpful changes very fast when alphas of KDE 4 started showing up performance issues in some previously rarely-used features, but AMD has done rather better by actually providing documentation to freedesktop people (even if ATI never maintained their own binary driver very well at all).

Introducing the RadeonHD driver, a community-developed driver supporting the R500-R700s.

It is certainly good news that AMD is opening up the specs for ATI graphics but ultimately i'm going to base my buying descisions on what works best now not what may work best in a few years time.

http://wiki.x.org/wiki/radeonhd has the following claims

The following subsystems have not been implemented yet or show some limitations:

        * 3D acceleration is active by default only on R5xx and RS6xx right now. Experimental support for R6xx and R7xx is available, but not for the faint of heart. Also, there is an experimental 3D bringup tool for testing on 6xx/7xx.
        * 2D acceleration is active by default now, except on RV740.
        * No TV and Component connector support so far.
        * Suspend & Resume isn't completely tested, but works on a variety of hardware. Your mileage may vary. Note that typically you need some BIOS workarounds on the kernel command line, ask your distribution for that.
        * Powermanagement has to be enabled explicitely. Depending on your hardware, the fan might run at full speed. This turned out to be really tricky.

See also RadeonFeature and RadeonProgram for a features and supported 3D program lists.

The following known bugs have not been resolved yet (ordered by severity):

        * Digital output on PCIEPHY (RS780) doesn't light up unless connected at boot time. Affects mostly displays connected to laptops thru DVI/HDMI. It is a problem with the AtomBIOS byte code parser which is used at the moment. The only work around is to boot with this output connected at boot time.
        * Bug 14500: External monitor does not display native resolution
        * Some cards seem to provide broken connector tables. We're constantly fixing those. Please report if you have one.

Seems like it's still a work in progress to me.

Hey, whatever gets us more page views.

(If you haven't been to http://x.org/ , you might not get the joke.)

Since ever? X.org for example has been around quite awhile.

Varying color depth, widescreen display ratios, scaling elements to varying resolutions, transparency and composite effects, GEM, KMS, DRI, hardware and software acceleration, modern font rendering, sub-pixel hinting, LCD displays, anti-aliasing, etc. On the input front you have mice that with 8 buttons, multiple mice, touchpads, joysticks, touchscreens, mice that poll at varying rates with varying "resolution", etc.

That's quite the list; many of the non-input related items aren't the job of the X server though. Multiple-cursor support went in this release, I believe.

So sure, the type of work the X server is being asked to do has changed, but development has been keeping up. Ever since the switch to Xorg things have been ticking along. The new modular build system has made working on the project easier, hardware autoconfiguration works very well, and the deprecation of old cruft is continuing quite aggressively - see http://www.x.org/releases/X11R7.5/doc/RELNOTES.html#AEN620 .

We've come a long way in the last 5 years.

Bug fixes accepted
Writing 3d video drivers isn't exactly like banging out "hello world"

http://www.x.org/wiki/radeonhd

[quote]Explore the fearsome depths of the labyrinthine cursed dungeon /etc in an attempt to find the ancient lost artifact, A Fucking Working Configuration![/quote]

If you want something that genuinely deserves having the Indiana Jones music on in the background, check out the website for Latex. ;)

If you want something for making analogies with the Necronomicon, we also must mention X. Perverted? Check. Blasphemous? Check. Likely to send you howling, barking, raving insane if you delve too deeply into its' unholy innards? Check.

First of all, a "GUI" refers to what the user sees, not how that's implemented. There is no piece of software anywhere that provides a "GUI". What you're talking about is a toolkit. Toolkits have converged already in terms of functionality and appearance.

If you want to port to "Linux" (or really, Unix-like systems in general), you use GTK or QT these days. Period. You can argue about which toolkit is better, but the difference is arbitrary, and both are available under the permissive LGPL license.

Also, the X server is not licensed under the GPL, but under the X11 license, which permits practically anything, much like the BSD license. There are no "politics" involved with the X11 license. If your ignorance is so profound that you've never heard of the X11 license, one of the major Free Software licenses, you have no basis for commenting on the windowing system.

Apple, had it decided to extend X, would not have started using the Linux kernel, or any of the existing toolkits, but instead created an Aqua that simply talked to the X server the same way GTK and QT do today.

Emperor Joseph II:My dear young man, don't take it too hard. Your work is ingenious. It's quality work. And there are simply too many notes, that's all. Just cut a few and it will be perfect.
Mozart: Which few did you have in mind, Majesty?

If only Video Card makers would open up their standards so open source drivers can be used for them.

*cough*. Oh wait, you expect them to write the drivers for you as well, using their own specs? Actually they are doing that too, but they need lots of help because most of their resources are tied up in the closed source FPS battle with nVidia. But your excuse is certainly outdated....

Last thing I want to do is come home and monkey around with making X work with Y.

That seems rather pointless. Why would you want to run X and Y at the same time?

The free open source drivers for ATI card have been supporting tear free video for a long time now.

Thank you for pointing that out. I must say, I was impressed that 0100010001010011 (652467) had all of his personal files on X.org before I realized that he meant example.com.

You know, X.Org would probably appreciate it if you didn't appropriate their website name. Use "example.org". That's what it's there for.

Disk imaging probably is an option for Linux, however I don't think the tools are as developed as they are for Windows, so I imagine you would not be able to do many hardware configurations with a single image.

Uhh, you realize Linux distros just package their kernels with every single open source driver there is and that every single piece of hardware is detected fresh on each and every boot, right? Linux doesn't configure itself only for the particular hardware it sees at install time like Windows does.

Xorg was the last remaining component to do that, and that was fixed with Xorg 7.2 and released in Ubuntu 7.10 a year and a half and three releases ago.

How do I find out which version of Xorg I have?

A piece of software I wanted to install recently had two procedures - one for Xorg 7.0 and above, and one for earlier versions.

Xorg -version reports a lot of crap, but never a useful/comparable version number.

X.Org X Server 1.5.2
Release Date: 10 October 2008
X Protocol Version 11, Revision 0
Build Operating System: Linux 2.6.24-19-server i686 Ubuntu
Current Operating System: Linux UbuntuViaBox 2.6.27-11-generic #1 SMP Wed Apr 1 20:57:48 UTC 2009 i686
Build Date: 09 March 2009 10:48:54AM
xorg-server 2:1.5.2-2ubuntu3.1 (buildd@rothera.buildd)
        Before reporting problems, check http://wiki.x.org/
        to make sure that you have the latest version.
Module Loader present

I assumed I had at least 7.0...

Have you tried the open source ati drivers? I understand the opengl support is not as complete, but perhaps 2d stuff will work better?

No, I haven't since according to http://wiki.x.org/wiki/RadeonFeature most of the basic features are not implemented yet for R700 HD.

Like everything else in the open-source world, we move in increments. Don't expect breakthroughs, expect progress. :3

http://nouveau.freedesktop.org/wiki/FeatureMatrix
http://wiki.x.org/wiki/RadeonFeature

~ C.

the kernel application binary interfaces are a moving target.

That's why we have glibc, which abstracts that ABI from applications.

Kernel driver interface - the horse was already beaten to death many times ( see here ).

a consistent configuration system, to enable distribution;

Windows tried that with Registry - and it didn't worked. And it will never work since "one size never fits all" requirements of all applications.

native file versioning;

Was tried many times before and failed miserably. As long as majority of files are blobs, versioning on level of file system makes no sense. Versioning on level of applications is implemented already more or less everywhere it was needed and SVN/git is there for the rest of applications.

audio APIs;

See ALSA and its user-space libraries.

See SDL.

and the integration of X11 with apps.

As was shown by FreeDesktop initiative not really needed nor X folks want to be bothered by all the end user bells and whistles.

Finally, he argues that Linux needs a committee to insure that all GUIs work consistently and integrate better on the back-end with the kernel.

Committee?? Buahahhahahaha!!!1!!cos(0)!!!!!!!

All what he says was tried before (see (11)) and generally can be described as "failed".

I wasn't aware that ATI had released any information more helpful than that which NVIDIA had. ATI's drivers are still closed source.

ATI now has three drivers for Radeons

1. The official proprietary driver (fglrx)
2. The Free driver co-created from the specs released several years ago, which were then dropped (radeon)
3. The new Free driver created from the new specs released last year (radeonhd)

I wasn't aware that ATI had released any information more helpful than that which NVIDIA had. ATI's drivers are still closed source.

ATI now has three drivers for Radeons

1. The official proprietary driver (fglrx)
2. The Free driver co-created from the specs released several years ago, which were then dropped (radeon)
3. The new Free driver created from the new specs released last year (radeonhd)

It's Unix-ish. Try compiling X11 (or any of hundreds of other POSIX compliant software packages) from source on a Mac. I'll wait.

X11 compiles just fine.

http://www.xfree86.org/current/Darwin.html
http://developer.apple.com/opensource/tools/X11.html
http://ftp.x.org/pub/X11R6.9.0/doc/html/Darwin.html

My primary complaint is that most OSS developers expect all Unix systems to be Linux systems. Which means that I have to let Linux software get its hooks into my OS X system in order to get anything compiled. Since OS X is NOT Linux, this is quite an unpleasant process.

It's capable of running its own proprietary OS that is specifically designed to not run on any otherwise capable hardware

OS X runs Unix software. Period. I usually get a host of tools installed first thing on my Mac. Thankfully, this has become less and less necessary over time as Apple has started including many of the most useful utilities up front.

I don't see why hardware mfgrs don't make drivers for Linux if they're already doing Unix drivers.

Because the only thing "Unix" about Mac OS X is Darwin. If something is not in Darwin, porting isn't as easy as the recompile and debug cycle that you're probably thinking of. For example, Mac OS X doesn't use X.Org's X11 server at the base of the graphics stack like *BSD and *Linux do. Instead, it runs X11 on top of Quartz, and the driver model differs between X.Org X11 and Quartz.

I got her onto thunderbird to access the email and it just kept timing out.

It might keep timing out if the e-mail provider doesn't offer IMAP or POP3. The more proprietary e-mail providers tend to do that.

Without going into NDA territory, there are acceleration docs for the r3xx-r5xx at http://www.x.org/docs/AMD/ , and r6xx documentation is currently pending legal review. Xvideo and EXA are done in Radeons in the 3D engine.

nVidia and ATI are currently being investigated by SEC and DoJ for price fixing and antitrust violation, yes.

If you really want to know why is it bad to have closed source drivers, check Nvidia's and ATI's linux driver forums. Example1, example2, example3.
Closed source development, compared to the open source one, sucks the monkey's ass.
ATI released the specs, at least partially, and this is the result. That's why I didn't buy Nvidia.
I'm currencly using the binary driver from ATI, while waiting for the open source radeonhd to be completed.

By current units sold market share VIA is small. My understanding is that it's roughly 40% Intel, 30% NVIDIA, 20% AMD and everyone else is crammed into the remainder 10% (that's total shipments of both desktops and platforms, discrete and integrated cards). (Rummages around web) Here's a link to GPU units sold in the second and third quarters of 2007. It looks like VIA sold almost 3 times less than ATI (but they seem to be on an upward progression).

It's the timing that makes this more interesting because Intel have pushed so much work on their open source drivers they are now the easiest "current" GPUs to get going on systems like Linux. Intel have done this by hiring engineers to work on Linux AND releasing specs which is kinda a double whammy. If this turns out to be the only way companies can achieve a similar smooth out of the box operation on FOSS OSes it's not going to cheap for others. It's also interesting that AMD had also taken big steps in specs and drivers direction before this move by VIA. Some would argue VIA's hand has been forced into this if they wish to remain relevant in the FOSS playground. Others point out that this is a process that can only be started by a willing company.

This is a brave play by VIA but there are more challenges to come. The next question is what they do with regard to the OpenChrome and Unichrome drivers and how to integrate the work they've done into the xorg development process. Judging from their Linux kernel integration it looks doable so long as VIA have some help.