Coming just a day after AMD had opened up their production microcode from their proprietary drivers for the R100 to R600 GPUs, a significant milestone has been reached in the road to open-source 3D graphics capabilities for the Radeon X1000 (R500) series. We now have hardware-accelerated glxgears!
In the next step towards open-source 3D support for the R500 and R600 GPUs (Radeon X1000 and Radeon HD 2000/3000), AMD has just pushed its production microcode into the Mesa/DRM git tree. This is the microcode found in the fglrx driver and it covers the Radeon R100 to R600 product families.
For Linux distribution vendors, right now is proving to be an awkward time for them as they decide which ATI driver will ship as the default choice in their spring distribution refresh. The problem used to be whether to ship a binary-only driver in the distribution in order to provide "out of the box" support for all available graphics hardware, but on the ATI/AMD side the software distributors are now facing the challenge of which open-source driver they should call the de facto standard. In this article we are briefly looking at the matter of the xf86-video-ati vs. xf86-video-radeonhd drivers, the highly political issue of AtomBIOS, and what some of the popular Linux distributions are deciding to use this spring.
Last month right before FOSDEM 2008, the 3D programming documentation for the R500 GPUs (Radeon X1000) series was released. This documentation consisted of a register reference guide for the R500 GPUs as well as a programming guide covering such areas as the command processor, vertex shaders, and fragment shaders. While the register reference guide for the R600 series is still being worked on, for those with older ATI graphics processors, AMD has went back and created a register reference guide for the R300 series.
Following a period of relative inactivity in the xf86-video-radeonhd git tree over the past few weeks, this afternoon Novell's Egbert Eich had pushed forward 55 changes to this open-source R500/600 driver. The AMD Radeon HD 3400 and 3600 series are now supported and there are a number of other significant changes. The RV620/635 mode-setting support has required the most significant work since the R500 series, since all output blocks have been altered due to the introduced DisplayPort capabilities. We have already tested out this latest code and were left with a positive impression.
On the same day as the limited open-source support arriving in the xf86-video-nv driver, NVIDIA's binary display driver for Linux has been updated to officially add support for the GeForce 9600GT graphics card. This new driver update is version 171.06 (Beta) and its only change is the added 9600GT support, but that's compared to the 171.05 driver that was targeted specifically for the Tesla S870.
Last month with the Catalyst 8.02 Linux driver we mentioned just how low on the end-user changes the fglrx driver has been in recent months, and today the Catalyst 8.3 Linux driver has been released with the release train slowly chugging along. The only new feature in this month's proprietary Linux driver is X-Video support for the Xpress 1200 hardware, but there are a few bug fixes that may satisfy some Linux users.
With FOSDEM 2008 taking place about two weeks ago, the 36th edition of the Nouveau Companion contains a special feature with FOSDEM-Nouveau coverage as well as in-depth thoughts from the developers that had attended this Linux conference in Brussels. Also covered in this development newsletter is new Gallium3D progress in their mainline Nouveau branch, MMioTrace is still planned for the Linux 2.6.26 kernel, and more RandR 1.2 improvements within this free software NVIDIA driver.
It was a year ago that AMD had replaced its aging FireGL Control Panel inside its Linux driver with the AMDCCCLE, or AMD Catalyst Control Center Linux Edition. Since that time, this Catalyst Control Center for Linux continues to mature with a few new features being added here and there, and version 2.0 could in fact be introduced in an upcoming release. At the same time, the control panel utility that ships with NVIDIA's binary driver, nvidia-settings, has stayed more or less the same for the past few years with only a few minor revisions. How do these two Linux control panels compare though when it comes to the features? For this article we have put AMDCCCLE and nvidia-settings side-by-side to compare and contrast both utilities.
We know NVIDIA has been working on a new Linux display driver release, with the GeForce 9 series having been introduced just this past week, and today a new driver release has arrived for Linux, Solaris, and FreeBSD. However, this new 169.12 display driver doesn't introduce support for the GeForce 9600GT and its official changes are limited to maintenance items.
This afternoon at FOSDEM 2008 in the X,Org development room, Keith Whitwell of Tungsten Graphics had provided a status update on the Gallium3D graphics driver architecture followed by Jerome Glisse talking about his work in porting the Radeon driver from DRM to Gallium3D. Jerome had also commented on other Radeon/X.Org topics such as kernel mode-setting. In this article are Gallium3D notes from these two talks.
Starting off the X.Org talks at FOSDEM 2008 was AMD's John Bridgman, who has been leading the AMD GPG open-source efforts. John had talked briefly about the history of their open-source efforts for the R500+ series and the evolution of AMD GPU hardware. Among the interesting comments made were that TexturedVideo/X-Video support for the R500/600 may be coming quite soon, DRM and Mesa work forthcoming, open-source multi-GPU CrossFire support is on the road-map, and they will be re-releasing R200 programming documents. The RV620 and RV635 documentation is expected in the near future.
For the past several weeks we have been referencing AMD's "tcore" in several of our articles, which is a user-space software suite that has been developed and used internally at ATI by engineers to work on various aspects of their binary drivers. Tcore is primarily used for testing prior to the availability of the actual silicon for their forthcoming graphics processors. John Bridgman and Alex Deucher have been working tediously to get this tcore source-code sanitized and cleared for public release, and finally they have reached this milestone. AMD has just published the first bits of open-source 3D programming documentation for ATI GPUs. This 3D programming documentation covers the R500 series and even goes back with information on the R300/400 series as well. The R600 3D programming guide will also be out soon. This information available today is what will foster the growth of open-source R500/600 3D support for the Radeon and RadeonHD drivers as well as R600 2D acceleration.
In the 35th edition of the Nouveau Companion, these open-source 3D NVIDIA driver developers talk about being a better bug reporter, X-Video improvements, PowerPC fixes, and MMioTrace being delayed until the Linux 2.6.26 kernel. At least three Nouveau developers will be at FOSDEM 2008 next week in Brussels, Belgium.
Whether it is a big or small update, every month AMD releases a new Catalyst package for both Linux and Windows for their supported ATI Radeon products. Last month in the Catalyst 8.01 Linux driver the changes had just consisted of a few bug fixes and nothing more. Today the Catalyst 8.02 Linux driver has been released, and like last month, it's short on the change-log. The fglrx changes are very brief and the driver version is still in the 8.45 release stream and wasn't even bumped to 8.46.x.
Since last September when AMD's new open-source strategy was unveiled, their xf86-video-radeonhd driver and NDA-free documentation efforts have largely been to improve its Linux customer base. However, with this open community enablement, it has also led to an improved ATI hardware status in Sun's Solaris Operating System. Introduced in Build 80 of OpenSolaris was the RadeonHD display driver and this R500/600 driver has now appeared in Solaris Express Developer Edition 1/08 and Project Indiana Preview 2.
Last night NVIDIA quietly uploaded a new Linux display driver to their FTP server. This new driver is tagged 171.05, while the latest public driver has been 169.09. Having already three releases in the 169.xx series, this is a moderate update to 171.xx, but according to NVIDIA it's not for everyone. There is no official change-log that NVIDIA has published for the 171.05 driver, and the change-log that ships with the driver hasn't been updated (whether it be intentional or not). The only word that has come out of the NVIDIA camp on this new driver is from Christian Zander and he has said that this driver is only intended for use with the Tesla S870 GPU Computing Systems. The legacy NVIDIA Linux drivers have also been updated this week.
At the Linux.Conf.Au conference today, Intel has announced NDA-free programming documentation covering the 965 Express and G35 Express integrated graphics processors (IGPs). Intel's display driver has long been open-source, but up until now, they have not been releasing the programming documentation for these products to the public. This move comes months after AMD announced their new open-source strategy and began releasing register documentation on their R500 and R600 GPUs. These newly released documents by Intel even cover 3D and video programming for their IGPs.
The public release of AMD's "tcore" sample code is imminent, thereby steering the open-source development efforts toward R600 2D acceleration and the basis of the 3D support for the RadeonHD driver. However, in somewhat of a surprise, this afternoon XAA and EXA support was added to the RadeonHD driver for the R500 series. For those that aren't X enthusiasts or Linux veterans, XAA and EXA are architectures for providing 2D graphics acceleration. With this accelerated 2D support, we have benchmarked both XAA and EXA on the xf86-video-radeonhd driver as well as with the xf86-video-ati driver and the binary fglrx driver.
The 34th edition of the Nouveau Companion is now available for your reading pleasure. The Nouveau crew this time around debates the issue of whether they should push a 2D-only release of the Nouveau driver out the door while continuing the 3D work, integrating MMioTrace into the mainstream kernel, the status of RandR 1.2, TV-Out support getting underway, and last but not least is the status of the Nouveau Gallium3D driver.
X-Plane is marketed as the "most thorough, flexible, and realistic flight simulator available for personal computers" and ships not only for Windows and Macintosh platforms but also for Linux. Laminar Research produces X-Plane and while it's currently not part of our testing suite, we recently took X-Plane v9 Beta 18 for a test flight. Previously their community leader had classified using ATI Linux drivers with X-Plane as an "unusable disaster" with "insurmountable problems", but is that really the case? We explored the situation in this article.
Back in June of 2005 with the ATI Linux 8.14.13 driver release was a new installer to more easily facilitate the installation of this binary graphics driver using a graphical interface for a generic setup or generating distribution-specific packages (at that time Red Hat was the only officially supported distribution). With time, this installer has evolved by gaining new features and more distributions are being supported through their --buildpkg command for generating custom driver packages. These packaging scripts are now even hosted in the open for more community interaction. With two new driver options that will be formally introduced next month in Ubuntu's packaging scripts for the Catalyst 8.02 Linux driver, the installation process of the ATI fglrx driver on Ubuntu will become several steps easier.
One month and one day after the NVIDIA 169.07 display driver was released, the Santa Clara folks have released a minor driver update for Linux and Solaris. No new features have been introduced, but just a couple of bug fixes.
AMD has today released the 8.01 Linux Catalyst package, which contains the fglrx 8.45 driver. New in this release are updated packaging scripts and a few fixes, which many have been hopeful for in this release. Compared to our usual Linux driver articles, this one is a bit shorter due to the KDE 4.0 release event going on at the Googleplex, which we are in the process of covering.
The Nouveau crew is out with their first Nouveau Companion of the year, which details the recent progress made with this reverse-engineered open-source NVIDIA display driver. Among the items talked about in this issue include PowerPC fixes, RandR 1.2 for older NVIDIA graphics cards, MMioTrace being broken with the Linux 2.6.24 kernel, and a variety of other topics.
In the second NDA-free documentation dump, AMD has just released programming data on the M76 and RS690 graphics processors. While the RadeonHD developers have already had these documents, this information will help the free software community in understanding the internal workings of AMD's graphics processors. In this article, we have information on this just-released data as well as what else the community can expect in the way of documentation in the near future.
AMD is on the heels of releasing the next set of GPU programming documentation to aide in the development of the open-source R500/600 drivers (xf86-video-ati and xf86-video-radeonhd). It's already been discussed what this NDA-free documentation release will have, but one of the questions that have repeatedly come up is if/when AMD will release information on accelerated video playback. AMD's John Bridgman has now stated what they plan to release in the video realm as well as a new requirement for their future graphics processors: being open-source friendly while avoiding DRM.
Last week the RadeonHD v1.1 driver was released, which (among other changes) had introduced extended monitor detection, RS600 support, and preliminary support for the RV670-based ATI Radeon HD 3850 and Radeon HD 3870 graphics cards. This v1.1 driver release combined with a recent git commit for adding additional TMDSA/B electrical values has led to quite a pleasant experience already with these new midrange graphics cards that have been on the market for less than two months.
At the end of our ATI Year in Review for 2006, we had stated, "next year will be a very interesting time for ATI/AMD Linux users." Looking back upon that statement, it has certainly turned out to be true, but perhaps an understatement for all of the AMD Linux work that has actually went on this year. The Catalyst Control Center was finally ported to Linux; there is now AIGLX support for use with Compiz, and the most substantial improvement being a brand new code-base for their proprietary Linux driver. Aside from their binary driver, they have developed a strong interest in better enabling the open-source community through releasing GPU documentation to the public (without NDAs!) and collaborating with Novell on the development of the RadeonHD driver. It has been one hell of a year for AMD, and in this article, we are going to look back at their twelve major driver releases from the year as well as re-benchmark all of these different versions.
In 2005 we started our annual AYiR (A Year in Review) articles for looking at the progress of the proprietary ATI and NVIDIA Linux display drivers over time. Now in our third year of doing this, it's time to see how NVIDIA's binary driver has panned out over the past year. NVIDIA had introduced GeForce 8 support this year as well as a few other features, but nothing major like in past years, and how has the OpenGL performance changed? With benchmarks in hand, we have the scoop for you today.
770 display drivers articles published on Phoronix.