url
stringlengths 33
62
| title
stringlengths 10
115
| author
stringclasses 2
values | content
stringlengths 61
41k
| comments_count
int64 0
1.68k
| scraped_at
float64 1.76B
1.76B
|
|---|---|---|---|---|---|
https://www.phoronix.com/news/Njk3Mw
|
VIA Pushes Out A Few More Driver Patches
|
Michael Larabel
|
Late last month the open-source community was presented with Chrome 9 series DRM support as the first step in providing 3D acceleration for these VIA IGPs atop a free software stack. Today we have been greeted with more patches from VIA's Bruce Chang.
The patches submitted to the DRI-devel mailing list fix a system hang with multi X support, fix a system hang issue caused by 3D scaling+ACPI, and address a segmentation fault when playing video with AGP after resuming from the system suspend mode.
These VIA driver fixes will likely find their way into the Linux 2.6.29 kernel.
| 1
| 1,760,738,472.927761
|
https://www.phoronix.com/news/Njk1OQ
|
LLVM Back-End For Gallium3D Almost There
|
Michael Larabel
|
While Tungsten's Gallium3D architecture is still a ways out from being used by most open-source graphics drivers and then being picked up by end-users, it continues to pickup new technical features. Corbin Simpson and Stephane Marchesin that work on the Radeon and Nouveau projects, respectively, have been working to building LLVM back-ends for Gallium3D. Corbin is a step closer to getting his LLVM compiler working: it now builds, but it ends with a segmentation fault.
LLVM, or the Low-Level Virtual Machine, is a compiler infrastructure written in C++ but can handle building programs in other languages too. LLVM is being used by Gallium3D for optimization purposes and using a real compiler to compile the shaders for the GPU. Corbin's LLVM back-end has been for an R300 vertex shader. We first talked about LLVM and Gallium3D back in February with the possibility of a GPGPU API in Gallium3D.
There's still plenty of work ahead, but this is a step in the right direction. Some other recent activity with the Gallium3D project includes the Nouveau driver being merged in Gallium3D v0.2, Generic Gallium3D Video Decoding, and new APIs coming to Gallium3D.
Confirmation of Corbin's initial R300 LLVM back-end building can be found on dri-devel. More on Gallium3D's LLVM can be found in this header file. For more on the Low-Level Virtual Machine Compiler Infrastructure, check out its web-site.
| 11
| 1,760,738,473.464061
|
https://www.phoronix.com/news/Njk2MQ
|
Silicon Motion Driver Gets RandR 1.2, EXA
|
Michael Larabel
|
If you happen to be using a Silicon Motion GPUs, there's good news with its latest open-source driver update. A Mandriva engineer has released xf86-video-siliconmotion 1.7.0, which adds in some significant features. The capabilities of RandR 1.2 (including the rotation mode) are finally supported by this graphics driver along with dual-head support.
If display configuration doesn't interest you though, the Silicon Motion X.Org driver now has support for EXA acceleration. Should you still be interested in XAA, there are a few fixes in this new driver for the XFree86 Acceleration Architecture. The 1.7.0 driver release also introduces support for the SMI 50x ASICs.
Besides these major features there are a number of other small additions and many bug-fixes found in xf86-video-siliconmotion 1.7.0. In total there are over 120 changes in this driver update alone.
To look at the change-log and release announcement, head on over to the X.Org mailing list archives.
| 7
| 1,760,738,473.472419
|
https://www.phoronix.com/news/Njk1OA
|
No X Server 1.6 Release Candidate For Christmas
|
Michael Larabel
|
The X Server 1.6 release schedule called for the first release candidate to come about two weeks after the second beta. We ended up getting a third beta but more than two weeks have now passed and there is no X Server 1.6 release candidate available. It should have arrived earlier this week, but there hasn't even been any Git activity to the server-1.6-branch in nine days.
Since the X Server 1.6 Beta 3 release there has been a handful of commits mostly pertaining to XQuartz, X Input, and RandR. Hopefully the X Server 1.6 release will still arrive in early January, but it doesn't look like it will meet the planned January 5th release date.
X Server 1.6 is bringing RandR 1.3, Direct Rendering Infrastructure 2, X Input 1.5 Device Properties, and Predictable Pointer Acceleration.
To those of you celebrating Christmas or other holidays, Merry Christmas.
| 3
| 1,760,738,474.000333
|
https://www.phoronix.com/news/Njk1NQ
|
VIA Releases Chrome 9 Series DRM
|
Michael Larabel
|
A month ago VIA had published 2D and 3D documentation (along with video register guides) for some of their newer IGPs and they had also announced a partnership with the OpenChrome project. As we have come to find out, some OpenChrome developers are under NDA with VIA Technologies already and they'll be looking at improving their ASIC support, delivering RandR 1.2 support, and making other fundamental improvements to this open-source VIA X.Org driver. Today though VIA has stepped forward once more and they have now released the DRM code for their Chrome 9 series.
Through a set of three patches published on the DRI mailing list, VIA's Bruce Chang published the Direct Rendering Manager code that will be required for bringing up 3D support on the Chrome 9 IGPs. This Chrome 9 DRM amounts to about 5,000 lines of code and should build against the Linux 2.6.28 kernel. This code will likely enter the Linux 2.6.29 kernel permitting that the initial feedback from the kernel / X.Org developers is incorporated quickly.
Among the Chrome 9 chipsets from VIA are the CN896, K8M890CE/K8N890CE, P4M900, and VN896. This DRM is the first step of the puzzle in enabling open-source 3D support for VIA hardware, with the next step being the Mesa code.
| 3
| 1,760,738,474.008633
|
https://www.phoronix.com/news/Njk1Mw
|
S3's New OpenGL 3.0 Linux Driver Still M.I.A.
|
Michael Larabel
|
A month ago S3 announced the Chrome 500 GPU along with what we called a magical Linux driver. In the press release for this budget graphics card, S3 Graphics mentioned this product can handle Blu-Ray playback, DirectX 10.1, and OpenGL 3.0 applications on both Windows and Linux. Their previous binary Linux drivers have been less than pleasing and there isn't even an official Blu-Ray player on Linux. NVIDIA has been the only manufacturer to deliver OpenGL 3.0 support on Linux thus far.
Following that we had heard back from a press representative at S3 Graphics and in fact they are working on a new Linux driver and that by mid-December there would be a beta OpenGL 3.0 Linux driver on their web-site.
Well, it's nearing the end of December and this morning a new press release hit our inbox entitled "S3 Graphics Unleashes Full OpenGL 3.0 API Support with Beta Driver for Chrome 500 Series GPUs." Sure enough, they have a beta OpenGL 3.0 Linux driver as well as a version of CyberLink PowerDVD software that works with the Chrome 400/500 series, but it's Windows-only.
Checking out the S3 Graphics web-site there still is no driver at all for the Chrome 500 series. The Linux driver they offer for the S3 Chrome 400 series is crippled to say the least. The Chrome 400 Linux driver was updated a month ago (v14.02.01), but its OpenGL support only extends to version 1.5 (Editor's Note: on the Chrome 20 series), there is no support for CPU architectures besides Linux x86 (not even x86_64), Compiz isn't supported, and its only compatible with older Linux distributions. That's just a few of the problems.
We're now waiting back to see what S3 Graphics has to say about their Linux support.
| 3
| 1,760,738,474.542533
|
https://www.phoronix.com/news/Njk0OQ
|
A Horde Of X.Org Drivers Get Updated
|
Michael Larabel
|
In time for the holidays and the forthcoming release of X Server 1.6 next month, a horde of X.Org driver updates were released. David Airlie has announced new versions of xf86-video-trident, xf86-video-apm, xf86-video-ast, xf86-video-chips, xf86-video-glint, xf86-video-neomagic, xf86-video-sis, xf86-video-tseng, xf86-video-tdfx, xf86-video-s3, xf86-video-ark, and xf86-video-s3virge. Peter Hutterer announced a new version of xf86-input-vmmouse.
There aren't many changes to these less popular drivers besides removing dead code, supporting new ASICs, making ISA optional, and various bug-fixes. The release announcement and source download links for these new drivers can be found on the X.Org mailing list.
| 4
| 1,760,738,474.550269
|
https://www.phoronix.com/news/Njk0Mg
|
RandR 1.3 Officially Gets Properties Support
|
Michael Larabel
|
Late last month RandR 1.3 with panning support arrived and was committed to the forthcoming X Server 1.6 release. Many drivers have picked up the RandR 1.3 support presented thus far, but now the properties support has finally landed for version 1.3 of the Resize and Rotate extension.
Properties support for RandR 1.3 has been discussed for quite some time and was discussed again in October with an initial proposal. This RandR 1.3 properties support has now been committed to master in the respective Git trees and the RadeonHD driver has picked up support for the required properties.
The known properties for RandR 1.3 include EDID data, signal format, signal properties, connector type, connector number, compatibility list, and clone list, but only the signal format and connector type are mandatory to be a RandR 1.3 compliant driver.
The latest RandR protocol specification can be read about here. Additional information about this new functionality that's exposed to RandR clients can be read about on Matthias Hopf's blog (the one responsible for much of the RandR 1.3 work).
| 0
| 1,760,738,475.059628
|
https://www.phoronix.com/news/NjkyNw
|
VMware Acquires Tungsten Graphics
|
Michael Larabel
|
There seemed to have been little buzz generated by this announcement when it first came about, but Tungsten Graphics has been acquired by VMware. They were acquired in late November for undisclosed terms and their only news mention of this acquisition is below (from their website).
"November 26, 2008: Tungsten Graphics has been acquired by VMware. Open source graphics technology development will continue as part of VMware's engineering team."
There isn't even any mention of the Tungsten Graphics on the VMware news page, which is odd considering this virtualization company is publicly traded on the NYSE.
Tungsten Graphics, of course, is the company behind Mesa, the TTM memory manager, Gallium3D and other Linux graphics innovations.
This move will hopefully be beneficial to the Linux graphics scene and allow Tungsten Graphics to continue in a steadfast manner while at the same time improving the 3D support within virtualized environments.
| 0
| 1,760,738,475.081495
|
https://www.phoronix.com/news/NjkzOQ
|
Kernel Mode-Setting Gets Ready, libdrm Updated
|
Michael Larabel
|
With the grunt of the development efforts on the Linux 2.6.28 kernel winding down, soon the merge window for Linux 2.6.29 will open and that's where Intel is prepared to (finally) commit the kernel mode-setting framework and code that supports their hardware to the mainline Linux kernel. With this milestone finally coming, Intel's Jesse Barnes has updated libdrm with the needed files.
Found in the Git master code for libdrm are several new files and modifications to existing files that provide some of the support needed to provide video mode-setting within the Direct Rendering Manager. Additionally, a few basic mode-setting tests were added to the libdrm code-base. This libdrm update is just one piece of the puzzle in obtaining kernel mode-setting support.
Kernel mode-setting support for Intel hardware is now primed and will be committed with the Linux 2.6.29 kernel while the KMS support for ATI hardware will likely appear in a later release. Likewise, the kernel mode-setting support for NVIDIA hardware through the Nouveau driver won't enter the mainline kernel anytime soon.
| 9
| 1,760,738,475.560186
|
https://www.phoronix.com/news/NjkzMA
|
XFree86 4.8.0 Released
|
Michael Larabel
|
For those still living in XFree86 land, version 4.8.0 of this X Window System implementation is now available. This is the first major update to XFree86 in about a year and a half, but with changes that aren't too exciting and have already been available through X.Org.
The X Server found in XFree86 4.8.0 has PCI-X and PCI Express fixes and a few other minor changes. The various ATI drivers in XFree86 have picked up support for new Radeon GPUs, fixes a memory leak, and contains a few other changes. The Intel i830 driver has one bug fix.
The release notes for XFree86 4.8.0 can be read here. Sources and binaries for XFree86 4.8.0 can be found in this directory.
| 9
| 1,760,738,476.267298
|
https://www.phoronix.com/news/NjkwMA
|
X Server 1.6 Beta 2 Now Available
|
Michael Larabel
|
Following a slight delay, Keith Packard has now tagged the second beta for X Server 1.6 in its git repository branch. Recent the 1.6 server has received updated documentation (just man pages), an upgraded dependence on dri2proto for the continued DRI2 work, and various X Input and XQuartz improvements. The commits to the server 1.6 branch are available through cgit.
Later this month we should start seeing the release candidates for X Server 1.6 followed by the final release in January -- if all goes according to plan. X Server 1.6 brings DRI2 support, RandR 1.3, and X Input 1.5 with device properties.
To obtain X Server 1.6 Beta 2, clone the xorg/xserver git repository from FreeDesktop.org and switch to server-1.6-branch.
| 10
| 1,760,738,476.755852
|
https://www.phoronix.com/news/NjkxNw
|
X Server 1.6 Beta 3, New Drivers Get RandR 1.3 Panning
|
Michael Larabel
|
Keith Packard pushed in 22 new commits to the server-1.6-branch of the X Server Git repository last night and then tagged X Server 1.6 Beta 3. This update is mostly made up of RandR 1.3 work along with a few commits related to X Input and DRI2.
The xf86-video-radeonhd driver was first to pickup RandR 1.3 support with panning, but committed to their respective Git repositories last night was RandR 1.3 panning support for both the xf86-video-ati and xf86-video-intel drivers.
The next development release for X Server 1.6 should be a release candidate later this month. The final release of X Server 1.6 is expected in early January.
| 0
| 1,760,738,476.770263
|
https://www.phoronix.com/news/Njg5NQ
|
X.Org Talks @ FOSDEM 2009
|
Michael Larabel
|
Luc Verhaegen has announced today that the X.Org Foundation will once again have a development room at next year's FOSDEM. The Free Open-Source Developers' European Meeting is taking place on the 7th and 8th of February in Brussels, Belgium.
There is room for 13 X.Org talks, and hopefully they will all be filled. Sadly though, they will not be having any X.Org event prior to the FOSDEM event like what was originally anticipated. With there no longer being an XDC (X Developers' Conference) and XDS (X Developers' Summit) each year like in the past, there will just be this two-day FOSDEM affair plus a planned X.Org summit next September in Portland, Oregon.
Information on the X.Org room at FOSDEM 2009 can be read about on the mailing list. Additional information can be found on the X.Org Wiki.
Like last year, Phoronix Media will be providing recordings of all X.Org talks during FOSDEM. If you want to watch the videos from last year, they can be found on our RadeonHD.org site. For FOSDEM 2009, we may provide HD recordings along with better audio channels. We'll also be publishing articles and live news items from the event.
If you're attending FOSDEM 2009, be sure to stop by the X.Org room if you want to hear more about kernel mode-setting, Gallium3D, the Graphics Execution Manager, Direct Rendering Infrastructure 2, and other great topics.
| 0
| 1,760,738,477.361523
|
https://www.phoronix.com/news/Njg4OA
|
X Generic Event (XGE) Protocol Specification
|
Michael Larabel
|
As part of his work on Multi-Pointer X, Peter Hutterer had developed the Generic Event Extension for X.Org, or commonly referred to as XGE. The X Generic Event Extension makes it possible for clients to reuse a single event opcode, which is needed with MPX since the X Server is currently limited to supporting only 64 opcodes between all X extensions.
While MPX went mainline earlier this year, it's being disabled in X Server 1.6 since X Input 2 isn't ready. As a result, there won't actually be anything using the X Generic Event Extension in X Server 1.6. The XGE protocol specification though will ship with the server 1.6 documentation.
If you're interested in learning more about XGE, the XGE protocol specification (which is quite brief) can be read on the X.Org mailing list and there is more information about XGE on the X.Org Wiki.
| 0
| 1,760,738,477.370497
|
https://www.phoronix.com/news/Njg4Mw
|
RandR 1.3 Arrives With Panning Support
|
Michael Larabel
|
RandR 1.3, the first major update to the X.Org Resize and Rotate extension since it picked up support for display hot-plugging and other goodies, has been finalized. RandR 1.3 has been in planning since last year, but over the past few weeks has really come together and is now ready for introduction in X Server 1.6.
Back in January there was talk of GPU object support for RandR 1.3 and then in late October the RandR properties support started to come together. Today though, Novell's Matthias Hopf has sent thirteen patches to the X.Org mailing list that introduce RandR 1.3 for xrandr, randrproto, libXrandr, and the X Server.
With this RandR 1.3 support comes panning support. To use the panning option through the command-line xrandr utility is the --panning argument. Patches have already been supplied that make the RadeonHD driver compatible with RandR 1.3 and its panning capabilities.
While RandR 1.3.0 is available today through these patches, it won't become mainstream for a few months until the various distribution vendors have adopted the respective packages following the release of X Server 1.6 in January.
| 14
| 1,760,738,477.99231
|
https://www.phoronix.com/news/Njg3Ng
|
X Server 1.6 Beta 1 Has Been Delayed
|
Michael Larabel
|
X Server 1.6, the update to the X Server for X.Org 7.4 that Intel had called for by year's end, has just experienced a minor setback. The release schedule that came about two weeks ago had put the release branching and initial test release to occur on the 24th of November. However, the beta 1 release has been pushed back by a few days.
Keith Packard shared on the development mailing list this week that the render matrix operations into pixman, which is great, except that X Server 1.6 now relies upon a new unreleased version of pixman. Until a new version of pixman is out, X Server Beta 1 can't be pushed out. This is similar to the situation we had with X Server 1.5 where the developers were dependent upon the release of Mesa 7.1.
Keith Packard will be pushing out a new pixman release within a few days to clear the way for X Server 1.6. In addition though, X Server 1.6 is dependent upon a newer version of inputproto that also has yet to be released, but Peter Hutterer hopes to have that ready by the week's end.
Permitting there are no other roadblocks, X Server 1.6 Beta 1 will likely be released next week. However, the second beta release was planned for the 8th of December, which could be in jeopardy depending upon when Beta 1 actually makes it out, thereby pushing back the entire release schedule. With there being critical development releases around the holidays, it won't be surprising if this release cycle gets dragged out longer -- just hopefully not as long as the wait for X Server 1.4.1.
Among bug fixes and other features, some of the noteworthy additions in X Server 1.6 are Predictable Pointer Acceleration, 2D performance improvements, a revised version of Direct Rendering Infrastructure 2, RandR 1.3, and X Input 1.5 with device properties. Multi-Pointer X and X Input 2 were pushed back from this release into X Server 1.7.
The final release of X Server 1.6 is expected in early January.
| 2
| 1,760,738,478.002235
|
https://www.phoronix.com/news/Njg3NQ
|
Linux Graphics Survey Reminder
|
Michael Larabel
|
Ten days ago we launched our 2008 Graphics Survey to poll Linux users on the graphics hardware they use, which of the drivers they depend upon, and X.Org-related features they are most interested in. This survey data is used not only for us, but for the respective developers as well to get a better understanding of how the Linux community at large is shaping. This is in continuation of our 2007 Linux Graphics Survey with those results being available here.
To take this short survey, visit this page. This year's survey is going on until the 15th of December.
| 0
| 1,760,738,478.505055
|
https://www.phoronix.com/news/Njg3Mg
|
S3 Graphics Is Working On New Linux Driver
|
Michael Larabel
|
Yesterday we shared that S3 Graphics launched their Chrome 500 Series graphics cards, with the initial model being a sub-$50 USD OpenGL 3.0 capable solution. It's no power workhorse, but in the press release they shared: "Today's users will now be able to enjoy the latest Blu-ray playback, streaming HD video, DirectX 10.1, and OpenGL 3.0 applications on Microsoft Windows and Linux platforms."
As we shared in the article yesterday, the last Linux driver from S3 Graphics was released last year and it's riddled with limitations. This Linux driver doesn't even support the earlier Chrome 400 series. Now though they mention Linux support in the same sentence as OpenGL 3.0 support, streaming HD video, Blu-ray playback, and DirectX 10.1 support. Because of this, we set out to get the situation clarified by S3 Graphics. Just minutes ago the US PR contact for S3 Graphics, Benson Tao, had responded:
Hi Michael,
Thanks for the questions!
We will put our Chrome Linux drivers on the website by the middle of December. Currently, we already have some customers designing us in on Linux platforms using our current drivers.
We will also have the OpenGL 3.0 beta drivers available at that time also.
Thanks,
Benson It seems though that the S3 Chrome 500 series will indeed be supported on Linux and there will even be OpenGL 3.0 support. What wasn't answered though was about any video acceleration support. When we obtain additional information or when the driver is actually released, we'll be sure to take a closer look.
| 22
| 1,760,738,478.513614
|
https://www.phoronix.com/news/Njg2NA
|
X.Org EvDev 2.1 Driver Released, New Features
|
Michael Larabel
|
In addition to being responsible for Multi-Pointer X, Peter Hutterer has also been working a fair bit on the evdev driver as of late. Evdev is the generic X.Org input driver and in the most recent release that occurred yesterday it has picked up a few new key features.
The xf86-input-evdev 2.1 driver has added support for axis inversion, touchscreen support, runtime calibration for absolute devices, axis swapping, mouse wheel emulation, and drag lock. In addition, this driver has runtime configuration property support, but this capability won't yet be in a released version of the X Server until version 1.6.
The release announcement (and source download links) for the evdev 2.1 driver can be read on the X.Org mailing list. In addition, more information about this release can be found on Peter's blog. To build the evdev 2.1 driver X Server 1.5.3 or manually patch this driver due to the EVIOCGRAB change.
| 7
| 1,760,738,479.034662
|
https://www.phoronix.com/news/Njg2OA
|
OpenChrome Gets A TODO List, Clarifies Work
|
Michael Larabel
|
Following VIA's press release this morning that they released 2D/3D/video documentation and have joined forces with OpenChrome, Xavier Bachelot has notified the OpenChrome users of what's taking place and he has provided the developers with a TODO list of items they wish to accomplish.
In a message to OpenChrome users, Xavier shares that some of the OpenChrome developers are now under NDA with VIA Technologies (read: not all documentation will be in the public domain), they are working on 3D support for the Chrome 9 series IGP, and they are working on bringing some of the features found within VIA's open-sourced driver (xf86-video-via) into the OpenChrome code-base.
In a message to OpenChrome developers, Xavier detailed what he believes should be the TODO list going forward for the OpenChrome driver. The items of high priority include support for the VIA VX800 ASIC, DVI support (integrated TMDS), improved integrated and hardwired LVDS support, and dual-head support. Of medium priority are TV output, better VT1625 TV encoder support, support for multiple X-Video ports, better hardware cursor support, RandR 1.2, and XvMC VLD for Unichrome Pro II.
Not as important to OpenChrome is iDCT / MoComp acceleration, MPEG-4 acceleration, more TMDS encoder support, more LVDS encoders support, and additional TV encoders support. In a follow up message it was also learned that VIA is already resisting to provide support that would assist in XvMC (X-Video Motion Compensation) due to MPEG licensing concerns over the publishing of any documentation or code.
| 1
| 1,760,738,479.746193
|
https://www.phoronix.com/news/Njg1Nw
|
MPX, X Input 2 Not Ready For X Server 1.6
|
Michael Larabel
|
Two days ago we shared the X Server 1.6 release schedule that would place the final release of this quickly-developed X Server release on the 5th of January. Keith Packard originally hoped to ship X Server 1.6 by the end of this year, but that won't happen due to the forthcoming holidays.
One of the major features that was going to make up X Server 1.6 was Multi-Pointer X, which had entered the mainline X Server back in May and allows for multiple input devices. This afternoon though it's been stated by Peter Hutterer that MPX won't appear in this next release.
In addition, X Input 2 also won't be ready for X Server 1.6. In place of X Input 2 in the 1.6 server release will be X Input 1.5 with device properties, which landed into git master after the X Server 1.5 / X.Org 7.4 release.
Peter's confirmation of no X Input 2 / MPX can be found on the X.Org mailing list. What will be ready for X Server 1.6 is Predictable Pointer Acceleration, Direct Rendering Infrastructure 2, and RandR 1.3.
If the X.Org development community can stick to a release schedule, X.Org 7.5 and X Server 1.7 should be out in April. If the X Input 2 and Multi-Pointer X code has matured enough, we'll finally see that input hotness in Q2'09.
| 3
| 1,760,738,480.268349
|
https://www.phoronix.com/news/Njg1Mw
|
X Server 1.6 Gets A Release Schedule
|
Michael Larabel
|
Back in September when the X developers raided the Edinburgh Zoo for the 2008 X Developers' Summit, Intel's Keith Packard made the rather dramatic announcement that he intended to ship X Server 1.6 and he would step up as the release manager.
This announcement was significant in that it had taken a year to go from X Server 1.4 / X.Org 7.3 to X Server 1.5 / X.Org 7.4 (and there was six months just for a point release), but Keith would throw X Server 1.6 together in just about three months.
Keith's intentions for X Server 1.6 is strictly a time-based release cycle as there are some features not yet in a released version of the X Server that he would like to have available to Intel's customers. X Server 1.6 is scheduled to introduce Predictable Pointer Acceleration, a revised version of DRI2 that hadn't made it in time for X Server 1.5, RandR 1.3 (the Resize and Rotate extension), and X Input 2 or X Input 1.5 with device properties. If X Input 2 is primed for X Server 1.6, this server release will also arrive with Multi-Pointer X support.
Since XDS we haven't heard any more on the schedule of X Server 1.6, but this afternoon Keith Packard had provided an update on the X.Org mailing list. Keith is hoping to branch X Server 1.6 from git master on November 24 and then create an X Server 1.6 Release Candidate 1. Following that, X Server 1.6 Release Candidate 2 is scheduled to ship on December 8 and a third release candidate on the 23rd of December. Around the time period of X Server 1.6 RC3, efforts will turn from new code development to bug fixing. With the current schedule, X Server 1.6.0 would then be released on the 5th of January.
This is just days after the end of 2008, but shouldn't be bad as long as the release schedule is met and the bug blocker list at least semi-cleared. If the original X.Org 7.5 release schedule remains in tact, the release of that with X Server 1.6 or X Server 1.7 will be occur on the first of April.
When it comes to RandR 1.3 support, just minutes after announcing this release schedule, Keith Packard proposed a series of patches (32 in total) that add projective transform support. Discussed recently have also been new properties for RandR 1.3.
| 2
| 1,760,738,480.277871
|
https://www.phoronix.com/news/NjgzNA
|
X Server 1.5.3 Released, Brings 21 Fixes
|
Michael Larabel
|
X Server 1.5 / X.Org 7.4 was released at the end of September. X Server 1.5.1 was quickly released around the same time and X Server 1.5.2 was released in early October. A month later we now have another bug-fix release for X Server 1.5. Adam Jackson has announced the release of X Server 1.5.3 with about 21 fixes.
The fixes in X Server 1.5.3 range from EXA to EDID to RANDR areas. The release announcement and change-log can be read on the X.Org mailing list.
| 0
| 1,760,738,480.889036
|
https://www.phoronix.com/news/Njg0MA
|
Intel Pushes Out New GEM, KMS Patches
|
Michael Larabel
|
Developers at Intel's Open-Source Technology Center have been busy with a number of projects, and Jesse Barnes in particular has been active with a few kernel mode-setting and Graphics Execution Manager tasks.
Among other recent work, Jesse has put out new patches for KMS and GEM on the DRI development mailing list. One patch adds support for mapping of GEM objects if the underlying graphics driver supports it. The second patch is revised kernel mode-setting for the DRM (Direct Rendering Manager) and Intel driver. Compared to earlier patches, the size has been reduced and there's a few fixes, but more work is still ahead.
While not directly related to Intel or KMS/GEM, there are a few improvements in Mesa going on as well. Tungsten's Brian Paul has added support for the centroid qualifier and invariant keyword in the GLSL 1.20 implementation of this open-source OpenGL stack. Some of the other commits also add in AND/OR/NOT/XOR instructions and GL Shading Language support for DP2/DP2A/NRM3/NRM4 instructions. This work can be found in the Mesa git repository for what will go on to form Mesa 7.3 (and Mesa 7.4 once stabilized).
| 3
| 1,760,738,480.897772
|
https://www.phoronix.com/news/NjgyNA
|
Getting KMS Ready For Linux 2.6.29 Kernel
|
Michael Larabel
|
It will be a while before the Linux 2.6.29 kernel merge window opens, considering we are just at the second release candidate for Linux 2.6.28, but Intel's Jesse Barnes is beginning to prepare the patches for kernel-based mode-setting support.
There's still quite a bit of work left before kernel mode-setting will be stable, but the patches can be found on the DRI development mailing list. These patches need to sit atop the recent GTT (Graphics Translation Table) mapping and EXA pixmap management patches.
Jesse has also sent along the Radeon driver that adds DRM/KMS mode-setting support (mailing list message), but it's currently out of date.
For more information on kernel mode-setting support for Linux, checkout A Preview of Kernel-based Mode-Setting and The State of Kernel Mode-Setting. Additionally, you may be interested in Red Hat's Plymouth and Fedora 10, which are KMS friendly.
Introduced in the Linux 2.6.28 kernel is GEM support for providing a kernel memory manager to the X.Org graphics drivers, which is a pre-requisite for kernel mode-setting. Intel had developed the Graphics Execution Manager after having a falling out over Tungsten's TTM.
Also available is the patches for core KMS support and the Intel i915 KMS driver.
| 1
| 1,760,738,481.528307
|
https://www.phoronix.com/news/NjgyNg
|
Compiz Fusion, Multi-Pointer X, & Input Redirection
|
Michael Larabel
|
Back in July we talked about Compiz support for MPX. MPX, or Multi-Pointer X, is the technology now in the mainline X Server for the X Server 1.6 release that allows multiple pointers to function independently on a single system. Today a set of patches were pushed into a Compiz Fusion git repository that allow all Compiz plug-ins to work nicely with all input devices, fixes for Input Redirection, and Input Redirection support in shift, scale, freewins, shelf, and ring. More on this work can be read in this blog post.
| 0
| 1,760,738,481.536508
|
https://www.phoronix.com/news/NjgyMw
|
New DMX Server Provides More Features, Less Complex
|
Michael Larabel
|
David Reveman, a key developer of XGL and Compiz, has announced new work he has done on the DMX (Distributed Multi-head X) Server. His work is based upon the original DMX Server but he characterizes the changes as being closer to a rewrite than a simple update. With his development branch, dmx-2, the server is reported to be less complex and more maintainable, but for end-users it adds in a number of new features.
David's revised DMX Server uses D-Bus for configuration, added Composite support, provides X-Video support, adds in RandR 1.2 support, and a few other notable changes.
The Distributed Multi-head X server was originally designed to be a proxy server that in turn interfaces with one or more X Servers that aren't necessarily running on the same physical computer. With David Reveman's latest work is to use DMX (also known as Xdmx) as a proxy server on virtual machines and remote desktops. A VNC viewer is then able to connect directly to the DMX Server.
Additional details on this work can be found on the X.Org mailing list.
| 1
| 1,760,738,482.050548
|
https://www.phoronix.com/news/NjgxOQ
|
Properties For RandR 1.3 Discussed Again
|
Michael Larabel
|
RandR 1.3 has been in planning for some time as the first update to the Resize and Rotate extension since the prominent 1.2 release that added output hot-plugging support and other features. RandR 1.3 wasn't finished in time for X Server 1.5 / X.Org 7.4, but Keith Packard has called for it in the next X Server 1.6 release.
When the X.Org developers raided the Edinburgh Zoo for their X Developers' Summit in September, Keith had intended to close RandR 1.3 in roughly one month. It's late, but we're now approaching November and discussion surrounding the much talked about properties support for the RandR 1.3 protocol has been rejuvenated by Novell's Matthias Hopf.
Matthias is trying to set the standard properties for RandR 1.3. The known properties in his second draft include standard properties for EDID data, signal format, connector types, connector numbers, maximum bandwidth for the selected signal format, a list of all output and signal format pairs, clone list, and panning area.
The RandR 1.3 properties addition is described in the protocol specification as:
"Properties are used for output specific parameters, and for announcing static or rarely changing data. Announced data is typically immutable. Properties are also used for evaluating new parameters before adding them to the RandR protocol." The discussion surrounding RandR 1.3 properties can be read on the X.Org mailing list. For those looking to learn how to use the command-line xrandr utility, check out A Newbie's Guide To RandR 1.2.
| 2
| 1,760,738,482.059979
|
https://www.phoronix.com/news/Njc4NQ
|
DRI2 Protocol Gets Updated, Again
|
Michael Larabel
|
Plans for version 2 of the Direct Rendering Infrastructure, or better known as DRI2, had come about last year at the 2007 X Developers' Summit. DRI2 allows for a number of technical improvements within X.Org graphics drivers, but for end-users it provides accelerated direct rendering to redirected windows. Today though much of DRI2 is now stabilized, but it didn't come without a lot of hard work and a few key revisions.
Kristian Høgsberg began merging his DRI2 code in February and then he merged the remaining bits in late March that provided the direct rendering support. However, then came Intel's Graphics Execution Manager.
The Graphics Execution Manager (or commonly known as GEM) is a kernel memory manager developed by Intel for their graphics driver and a complete replacement for Tungsten's TTM. Up to that point, TTM was set to become the de facto memory manager for use by open-source X.Org drivers. With that said, the initial design of DRI2 had required TTM's memory manager to be used. When Intel jumped ship on TTM it presented the need for DRI2 to be reworked. As a result of this, DRI2 support was dropped from the X Server 1.5 / X.Org 7.4 release.
Since then DRI2 has been reworked to accommodate GEM. Those bits had entered the different development trees in late August. At this year's X Developers' Summit (X Developers Raid The Zoo), a few more improvements were agreed to between Kristian and the other developers. Today this latest batch of improvements have been merged.
Prior to leaving on a holiday, Kristian committed the new DRI2 protocol and documentation along with updates to the X Server, Mesa, and Intel driver.
What's been added in this latest Direct Rendering Infrastructure 2 push is a DRI2CopyRegion request and the framework that allows DRI2 to be extended in future versions. With the CopyRegion request it's mostly a rename from SwapBuffers. No vsync handling or page-flipping support has yet been added to DRI2.
After today's commits, Kristian is now comfortable letting DRI2 be merged to master in Intel's X.Org driver. No other drivers, however, have yet to fully adopt this latest version. Intel users though will need to enable the DRI2 support through their xorg.conf. Kristian hopes to have most of the DRI2 work settled by next month. With X Server 1.6 being planned for a December release, the timing should be ideal. The announcement regarding today's DRI2 update can be read on the X.Org mailing list.
| 8
| 1,760,738,482.556141
|
https://www.phoronix.com/news/NjgwMw
|
More DRM Patches For Linux 2.6.28 Kernel
|
Michael Larabel
|
Last week Intel's Graphics Execution Manager (GEM) had entered the Linux 2.6.28 kernel. As most of you should be aware, GEM is the new kernel memory manager for graphics processors that was developed by Intel as a replacement for TTM. Not only was there this batch of new code, but new graphics-related work continues to be pushed into the 2.6.28 kernel.
David Airlie has called out to Linus to pull in another series of changes. In this most recent patch set there are a number of GEM stabilization fixes and other kernel DRM (Direct Rendering Manager) fixes. Included in this code is the vblank improvements, which Eric Anholt talked about this week. The kernel mailing list message can be read here.
While the 2.6.28 kernel will go without kernel mode-setting (KMS) support, there's a fair amount to be excited about already in this next Linux kernel -- particularly if you're using the open-source Intel driver.
In related Intel graphics news, Jesse Barnes has pushed forward patches that add GTT mapping support for GEM and fence register management on Intel 965 hardware and later (message) and GTT mapping support for libdrm (message).
| 0
| 1,760,738,483.262185
|
https://www.phoronix.com/news/Njc3Mw
|
CyberLink Announces Linux HD Video Player
|
Michael Larabel
|
CyberLink's proprietary PowerDVD player has been available on Linux for sometime -- and can even be purchased through the Ubuntu store -- but today they have kicked their Linux support up by a notch or two. They have announced this morning that PowerDVD Linux and PowerCinema are now available for Linux-powered netbooks (such as the ASUS Eee PC 901) and nettops. CyberLink is now supporting the playback of high-definition digital media content on these mobile Linux devices.
CyberLink PowerDVD Linux is their HD video player and CyberLink PowerCinema Linux is their media player that supports DVDs, video files, music, and more. PowerCinema supports playback on Linux of ASF, WMV, MPEG-1, MPEG-2, DAT, and AVI formats.
These two Linux solutions are available to OEM customers. Hopefully we'll see them appearing on netbooks soon so we can begin watching HD content. This announcement was made today in a CyberLink press release.
In related news, it's expected that this month AMD will bring their UVD2 support to Linux for accelerating more multimedia content on the GPU.
| 16
| 1,760,738,483.765425
|
https://www.phoronix.com/news/Njc3OA
|
X Server 1.5.2 Released, Brings Bug Fixes
|
Michael Larabel
|
X Server 1.5.1 was only released two and a half weeks ago as a quick bug-fix release to X Server 1.5.0, which is the server component for X.Org 7.4, but today we have a third release. Adam Jackson has announced the release of X Server 1.5.2 with 14 changes.
X Server 1.5.2. brings a few int10 fixes (including a memory leak fix), an EDID fix, only building DRI2 when DRI2 is enabled, an array-index based devPrivates implementation, and a few other fixes. The change-log and source download links can be found off the X.Org mailing list.
| 4
| 1,760,738,483.774712
|
https://www.phoronix.com/news/Njc2Nw
|
Mobile GPU Switching With X.Org & Linux?
|
Michael Larabel
|
One of the most recent innovations on the mobile front has been integrating two graphics processors into a notebook but not for binding them together via SLI or CrossFire but for real-time GPU switching. This technology though isn't supported on Linux, at least not yet.
Intel's Centrino 2 and AMD's Puma platforms support having an integrated graphics processor and discrete graphics processor. The idea behind this is that when the notebook is running on battery power it can switch to using the IGP and turn the other GPU completely off in order to save power. While the performance of the Intel GMA X4500 HD isn't that bad, when these newest notebooks are connected to a power adapter the main graphics processor can switch to the discrete performance-oriented GPU. You get the best of both worlds by having a long battery life but still having the graphics capabilities to perform well with the Phoronix Test Suite or whatever 3D goodness catches your fancy.
This dynamic switching and powering down of unused GPUs isn't supported though by X.Org at this time. There were some comments made during XDS 2008 that Intel's X.Org team soon could get underway in supporting this new technology on Intel notebooks, but not much was said. Today though Red Hat's Adam Jackson has commented on the matter. Adam Jackson was the X.Org 7.4 / X Server 1.5 release manager and shares responsibility with David Airlie for the state of X in Fedora and Red Hat Enterprise Linux.
In this blog post, Adam talks about GPUs and the different ways through with these dual GPU notebooks are configured. Some of these notebooks have BIOS options for controlling the GPU to use while others expose both GPUs on the PCI bus the entire time. So far they haven't experienced any luck in benefiting from ACPI in this work. Aside from just recognizing the GPU that is currently communicating with the display, extensive work is also required within X.Org to make this real-time switching even feasible. Adam concludes with, "Getting this to work well should actually be a lot of fun, and there's lots of opportunity to sweep away old bad design and come up with something good."
| 5
| 1,760,738,484.397151
|
https://www.phoronix.com/news/Njc0Mg
|
X Server 1.5.1 Has Been Released
|
Michael Larabel
|
The X Server 1.4.1 bug-fix release came out 212 days later than anticipated and X Server 1.5.0 was released months later than scheduled. X.Org 7.4 in fact hasn't even been released yet and that was supposed to come back in May. However, just twenty days after X Server 1.5.0 was released, we now have X Server 1.5.1.
Red Hat's Adam Jackson has just announced the X Server 1.5.1 release on the X.Org mailing list. This point release just includes conditionalizing Composite-based backing store on pScreen->backingStoreSupport (compliments of NVIDIA), moving RELEASE_DATE below AC_INIT, disabling shared pixmaps for EXA, and fixing panoramiX request and reply swapping.
The source-code can be downloaded immediately and expect this to be quickly adopted by Linux distributions that are already tracking X Server 1.5 / X.Org 7.4. With this fixer-up release now out there, hopefully X.Org 7.4 will make it out before the end of the month.
| 1
| 1,760,738,484.405979
|
https://www.phoronix.com/news/Njc0MA
|
X Input 1.5 Gets Device Properties
|
Michael Larabel
|
X Server 1.5 was originally slated to receive XKB 2 and Xi 2, but these major input improvements ended up getting postponed until at least X Server 1.6. While X Server 1.5 was only released earlier this month (and X.Org 7.4 still has yet to be found), Intel is already calling for X Server 1.6 by year's end.
With X Server 1.6 being a strictly time-based release cycle, it's unknown whether X Input 2 will be completed in time or if Multi-Pointer X will be enabled by default. We may not end up seeing X Input 2 until X.Org 7.5 / X Server 1.7, which would be well into next year.
While X Input 2 isn't ready yet, Peter Hutterer has back-ported his device properties protocol into the X Input 1.5 extension. Last night on the X.Org mailing list he proposed his patches for the X Server that will remove the Configure/Query property functionality and replace it with his device properties protocol. This work comes in the form of three patches for the X Server plus patches that will be required for the evdev and inputproto support.
This work hasn't yet been merged to git master, but can be found in Peter's mailing list messages. No input API breakage has occurred from this change. For those interested in the specification to the device properties protocol, the first draft of it can be read here (the final version of the documentation will be pushed with the patches).
The good news here is that until X Input 2 is completed, at least we'll have X Input 1.5 with device properties.
| 0
| 1,760,738,485.041911
|
https://www.phoronix.com/news/NjczMg
|
The Multi-Seat Display Manager
|
Michael Larabel
|
On behalf of the C3SL Multi-Seat Team, Paulo R. Zanoni has announced today on the X.Org mailing list the MDM, or Multi-Seat Display Manager. The Multi-Seat Display Manager isn't a new X Display Manager at all, but in fact it's a wrapper for the other managers out there such as GDM, KDM, and XDMCP. The MDM then configures the real display manager to use a multi-seat configuration, which dramatically simplifies the process for this once daunting process. The only burden placed upon the graphics card and X.Org graphics driver is support for RandR 1.2.
For those interested in configuring a multi-seat system with multiple monitors, keyboards, and mice, they can build the Multi-Seat Display Manager from git, use the provided Debian packages, or try out their Ubuntu 8.04 LiveCD with MDM pre-installed. Their X.Org mailing list message can be read here. They have also established a MDM Wiki. Lastly, the MDM LiveCD can be found on this Wiki page.
| 6
| 1,760,738,485.050622
|
https://www.phoronix.com/news/NjcxNQ
|
X.Org 7.5 May Be An April Fool's Joke
|
Michael Larabel
|
X.Org 7.4 was supposed to be released on September 10, but that didn't happen. Though our notes on the X.Org 7.4 release can be read here (the article was automatically published as I was traveling all day without Internet access, so I was wagering on not another delay). Today though Daniel Stone has laid out the plans for X.Org 7.5.
Beyond what is planned for X Server 1.6, X Server 1.7 will be the release due as part of X.Org 7.5. The feature freeze for X Server 1.7 is planned for the 21st of January 2009 and the 31st of February will mark the code freeze for this cycle with an X Server 1.6.99.901 / X.Org 7.5 Beta 1 release. A second beta and release candidate are both planned for a March release. Then if all goes according to plan, X.Org 7.5 / X Server 1.7 will be released on April 1, 2008. However, Daniel Stone notes in his mailing list message:
1st April 2009: Release 7.5, makes good April Fool's story. If not, Phoronix gets to run '7.5 released ... nah, X.Org didn't release on time again! April Fool's! Ha!'. Either way, everyone's a winner.
| 4
| 1,760,738,485.549459
|
https://www.phoronix.com/news/NjcwOA
|
X.Org 7.4 Release Planned For Sep 10
|
Michael Larabel
|
Mesa 7.1 was released near the end of August and X Server 1.5.0 was released just earlier this week, but the release of X.Org 7.4 consisting of all the latest X packages has yet to make it out the door.
While the developers attending the X Developers' Summit had taken the afternoon off to entertain themselves within the zoo's gift shop, Daniel Stone had announced X.Org 7.4 will be released on September 10, 2008. He announced this date on the X.Org Wiki. This means X.Org 7.4 is arriving one year and four days after the release of X.Org 7.3... So much for this six-month release cycle. But if everything goes according to Keith Packard's plans, a major update (X Server 1.6) will be out this year already as an update to the 1.5 release.
| 1
| 1,760,738,485.568987
|
https://www.phoronix.com/news/NjcwNw
|
XDS 2008 Day 3 Audio Recordings
|
Michael Larabel
|
Below are the Ogg recordings from the third and final day of the 2008 X Developers' Summit. Most notably on this day, Keith Packard talked about X Server 1.6 features and his plans to release it this year as well as clarifying his GEM/UXA work and making other comments. Also discussed were DRI2 and Red Hat's Plymouth.
Red Hat's Plymouth Direct Rendering Infrastructure 2 Intel / X Server Planning
| 0
| 1,760,738,486.065413
|
https://www.phoronix.com/news/NjcwNQ
|
Closing X Developer Summit 2008 Notes
|
Michael Larabel
|
Aside from the short X Server 1.6 release plans and clarifying UXA+EXA, there are a few other notes to share from this afternoon's X.Org/Intel talk.
- This shouldn't come as a surprise, but the release of xf86-video-intel 2.5 driver is planned for later this month. The first release candidate of this new open-source Intel driver is already available, but more test releases are expected before it's officially released. This driver will be GEM-capable, support KMS / kernel mode-setting, DRI2, and more Render acceleration improvements.
- Following the release of xf86-video-intel 2.5.0, work will be underway in preparing for a 2.6 release in December.
- Intel's Linux and Windows driver development teams are working together in sharing their video code between the two platforms. Intel video playback on Linux should improve as a result, but first they're waiting on permission to release some of the Intel 965 video code that's more structured on the Windows side than their current Assembly-based implementation.
- Keith Packard has expressed interest in cleaning up the X/Mesa build process by avoiding duplicated files (primarily C code and header files) between multiple components. In addition, it's his goal to have no auto-generated files -- that have been made by any scripts -- to be housed within git.
| 5
| 1,760,738,486.775198
|
https://www.phoronix.com/news/NjcwMw
|
Intel Calls For X Server 1.6 This Year
|
Michael Larabel
|
While X Server 1.5.0 was finally released this week without X.Org 7.4, Keith Packard is calling for the release of X Server 1.6 this year. Due to Intel's customers needing some of the newest X features, Keith Packard has stepped up to be the release manager for X Server 1.6 and he will be running this release cycle on a strictly time-based schedule.
X Server 1.6 will ship with the revised DRI2 infrastructure, RandR 1.3, and X Input 2 (if it's ready in time) or X Input 1.5 with device properties. X Server 1.6 will also ship with Predictable Pointer Acceleration and other features.
Multi-Pointer X was already merged to master, but Keith Packard will be disabling MPX in X Server 1.6 if Xi2 isn't completed in time.
The new input thread work has also been requested as one of the X Server 1.6 features.
The X Server 1.6 release will come ahead of the X.Org 7.5 release, which he hopes will be ready by March of 2009. With that said, we're now looking at X Server 1.7 for X.Org 7.5. No release manager has volunteered at XDS 2008, but a release team may be formed to take the burden off a single person.
Keith Packard also mentioned during his XDS talk that he intends to close RandR 1.3 in roughly a month. He is hoping to have standard properties ready for this Resize and Rotate extension update, but that is dependent upon Matthias Hopf completing the work.
A tracker bug for X Server 1.6 and additional information should be available shortly.
| 11
| 1,760,738,487.307769
|
https://www.phoronix.com/news/NjcwMg
|
XDS 2008 Day 2 Audio Recordings
|
Michael Larabel
|
Now available are the audio recordings from the second day of XDS 2008 where a Gallium3D status update was provided along with Intel's Graphics Execution Manager and a variety of other topics such as open-source Radeon graphics, Intel community testing, and GLSL.
These files are available for download in Ogg format and should also be of higher quality than yesterday's recordings. If you run into any problems, load it into Audacity and use the amplify option to improve the audio level. If you run into any other problems, let us know in the forums.
Gallium3D Status Update
Gallium3D For Intel i915 Hardware
Intel Community Graphics Testing
Open-Source ATI R100-700 Radeon
GL Shading Language
Intel Memory Management / GEM
| 0
| 1,760,738,487.316356
|
https://www.phoronix.com/news/NjcwMQ
|
X.Org Developers Raid The Zoo
|
Michael Larabel
|
Activities for the second day of the 2008 X Developers' Summit are now over. The day ended with the X developers in attendance receiving a private tour of the Edinburgh Zoo, the venue for this event.
The schedule for today's talks that mostly consisted of open-source graphics were also thrown off when the developers decided to visit the Gentoo Penguins. They were let loose for a rather odd parade.
| 4
| 1,760,738,487.934473
|
https://www.phoronix.com/news/NjY5OQ
|
Intel Provides Update, Plans For GEM
|
Michael Larabel
|
Intel's Eric Anholt just finished speaking at the 2008 X Developers' Summit about video memory management. Specifically, Eric was talking about GEM, or the Graphics Execution Manager, that came about as a result of concerns that arose about Tungsten's TTM. GEM is now the kernel memory manager they are focusing their open-source development work on for the xf86-video-intel driver and is what they hope will become the de facto standard for memory management.
In this Graphics Execution Manager talk, Eric started talking about their past memory management methods (i830_memory.c, exa_offscreen.c, and texmem.c) that imposed many limitations along with the shortcomings they experienced with TTM. A new memory manager was needed for composited OpenGL, EXT_framebuffer_object, EXT_texture_from_pixmap, reduced memory consumption, and private back-buffers.
In this talk the next plans they have for GEM are fence register management, hardware contexts, user-land cache management, rectangular pwrite, GIT mmap, and fault-based flushing.
During this talk, Intel's Keith Packard called for a common API for GEM that can be used across multiple drivers. "I would love to see the GEM API extended to support discrete graphics cards," noted Packard. Keith though is quite confused over David Airlie's decision to use a GEM-ified TTM memory manager for the open-source ATI driver rather than fully using their Graphics Execution Manager.
| 1
| 1,760,738,487.943148
|
https://www.phoronix.com/news/NjY5OA
|
XDS 2008: GLSL, Radeon, Graphics Testing
|
Michael Larabel
|
This morning Tungsten Graphics was speaking at XDS 2008 about the status of Gallium3D. However, the rest of the day is filled with a variety of other OpenGL and graphics related talks. Intel's Gordon Jin had talked about Intel's community testing process, Jerome Glisse had talked about a few aspects of the open-source Radeon drivers, and Ian Romanick just gave a talk on GLSL (GL Shading Language).
Intel Graphics Testing: - AutoBuildAutoTest - Intel's suite for building from source the driver components and running its tests.
- Call for increased quality of bug reports.
- Intel wants more community testers. Intel community team created to have test coverage for their old hardware.
Radeon: - R100-200: No Gallium support planned and limited development restricted to kernel mode-setting and memory management.
- R300-400: Gallium, Vertex shader, pixel shader, memory management, and kernel mode-setting planned.
- R500: Vertex shader, pixel shader, support loop with limited depth, kernel mode-setting through AtomBIOS, memory management, and power management.
- R600/700: Unified pixel/vertex shader, complex, drawing triangle, power management is critical.
- R500 3D support should be the same as the R300/400 support.
- Command Submission is easier with modern ATI hardware where it takes a hardware formatted stream that can be sent directly to the hardware.
- Producing hardware shader code from a high-level shading language like GLSL is one of their biggest tasks ahead.
- Kernel mode-setting for R100 to R400 was ported from their DDX code and for the R500 series and lter it's using the new AtomBIOS parser.
- The ATI support will use the GEM API but underlying code will be powered by TTM.
GLSL: - Language extensions allow many language intrinsics to be written in GLSL: __constructor, _operator, __asm.
- GLSL 1.30 support not even started.
- Really Bad State: one-off parser generator, linker can't support multiple compilation units on the same shader target, ARB_fragment_program is the IR.
- Desired infrastructure: sensible parser infrastructure, back-end independent IR, parser re-written using flex and bison.
- Language features they want: A compliant linker, real integer support, and full GLSL 1.20 and 1.30 support.
- Other features needed: geometry shader support, code-generator generator.
- Assembly shaders to MIR (modeled after GCC).
- Current GLSL infrastructure will continue on for at least a year.
There wasn't anything pertinent shared during the Intel graphics talk and much of the Radeon information isn't new if you're a faithful Phoronix reader staying up to date with the latest news, but the GLSL talk had shed some new light on Ian's (and Intel's) plans for GLSL within Mesa. Quite a few improvements for GLSL are planned including support for GLSL 1.30, which came as part of OpenGL 3.0.
| 7
| 1,760,738,488.56212
|
https://www.phoronix.com/news/NjY5NQ
|
X Server 1.5.0 Is Now Released!
|
Michael Larabel
|
Adam Jackson has just announced the release of the much anticipated X Server 1.5.0. This is the key component that will make up X.Org 7.4, which we expect to see released (hopefully) this week during the X Developer Summit. The X Server 1.5.0 release announcement can be read on the mailing list, while we'll have a full run-down on X.Org 7.4 as soon as it's released.
| 14
| 1,760,738,488.570906
|
https://www.phoronix.com/news/NjY5MQ
|
X Developer Summit 2008 Starts
|
Michael Larabel
|
The 2008 X Developer Summit has started this morning. However, as there is no Internet connection here other than a single 3G cell connection for 30 people, the audio recordings and Phoronix content will not be uploaded until later today.
In the mean time though, here are a few notes from the talks thus far:
X.Org Foundation:
- $110,000 in the bank
- Starting next year just one X conference plus weekend FOSDEM. Next year's X Developer Summit may be in Portland, Oregon at the Portland State University and will be a two or three day event. Will likely take place in September 2009.
- Still working on the 501C3 for the X.Org Foundation. Using the Software Freedom Law Center to do the legal-end of reorganization
- Board elections. Who knows what's going on?
- Bi-weekly board meetings on IRC
- No idea what's going on with X membership
X.Org Infrastructure Report:
- X.Org, FreeDesktop.org, and other pages upgraded to MoinMoin 1.6
- 22,000 emails moderated. A lot of spam on Wiki and e-mail lists.
- BugZilla upgraded to latest version
- Building redundancy into the server infrastructure
- Replace gitweb with cgit for FreeDesktop git
| 3
| 1,760,738,489.214292
|
https://www.phoronix.com/news/NjY4OQ
|
XDS 2008, With Input & GL Happiness Days
|
Michael Larabel
|
There's just about thirteen hours until the 2008 X Developers' Summit gets underway at the Edinburgh Zoo in Scotland. While attendance is low and the program was just getting thrown together at the last minute, it looks like everything will work out.
This three day X.Org summit will kick off with the "input" day tomorrow that covers various aspects of X.Org input mechanisms. There will be an overview talk by the Multi-Point X mastermind, Peter Hutterer, and a pointer acceleration talk by Simon Thum (X.Org very recently received Predictable Pointer Acceleration). Covering tomorrow afternoon are talks on improving input latency, run-time GPU power management (not exactly to do with input), and live X Input 2 specification drafting.
The second day of XDS 2008 is what Daniel Stone describes as the "GL-related happiness" day. Tungsten Graphics will be providing a new status update on their Gallium 3D architecture (their Q1'08 status update), a talk about Gallium 3D with Intel 915 hardware, and DRI2 version 2 (the revised DRI2 design following the fallout from the TTM dependences with the original design) to take up the morning. In the afternoon there will be a talk on kernel mode-setting for ATI Radeon hardware, an OpenGL 3.0 talk, and another form of an OpenGL talk is also planned.
The third and final day of XDS 2008 is planned to be just a mix of other uncovered topics. During this time will be Red Hat's Plymouth, Intel graphics testing, an Intel Keith Packard talk, and ending out the day we may just see the release of X.Org 7.4 / X Server 1.5, which is long overdue. Well, the Wiki-based schedule says "Lock doors, no-one leaves until 7.5 released", but that's really for X.Org 7.4 unless they intend everyone to live at the Edinburgh Zoo until the middle of 2009 when X.Org 7.5 will perhaps be ready.
| 3
| 1,760,738,489.223702
|
https://www.phoronix.com/news/NjY4Ng
|
Mandriva Improves Silicon Motion Driver
|
Michael Larabel
|
Along with VIA releasing a new open-source X.Org video driver, there is work underway on improving the status of another open-source graphics driver. Silicon Motion is perhaps more obscure than VIA Technologies when it comes to integrated graphics, but they primarily specialize in low-power graphics chips for tablet PCs.
The xf86-video-siliconmotion driver usually only receives a handful of commits a year and from a few different developers such as Adam Jackson, Egbert Eich, and Alex Deucher. However, this month there has been some significant improvements to this tablet PC driver.
The xf86-video-siliconmotion driver most notably is picking up EXA acceleration, X-Video, and RandR 1.2 support. The RandR support does include Rotation support. Code for the SMI501 Chipset has also been added.
This new Silicon Motion driver work is being done by Paulo Cesar Pereira de Andrade, Arnaud Patard, and Francisco Jerez. These developers work for Mandriva and have managed to get their hands on Silicon Motion documentation and register guides, which explains this revitalized work. Aside from the commits hitting the xf86-video-siliconmotion tree last night, there were only six commits made to the driver this year. Following these most recent commits, the Silicon Motion X driver has been bumped to version 1.6.1.
| 5
| 1,760,738,489.709646
|
https://www.phoronix.com/news/NjY4NQ
|
XDS 2008 Quickly Approaching
|
Michael Larabel
|
XDS 2008 is happening next week in Edinburgh, Scotland as the X.Org Foundation's second conference for the year. Since last week when last mentioning XDS 2008, there have been a few more attendee confirmations but it still looks like this event will be slightly underpopulated. The good news though is that the program is still getting more talks towards the last minute.
This three day affair will have at least three talks on X.Org input (improving input latency, introduction to pointer acceleration, and a summary), an X.Org Foundation update, a Radeon talk by Jerome Glisse, DRI2 version 2, Red Hat's Plymouth boot experience, the state of Gallium 3D, Gallium 3D on the Intel 915 series, run-time power management for graphics hardware, and graphics testing.
Jerome Glisse, the mastermind of the original open-source R500 (xf86-video-avivo) driver, will be talking about kernel-based mode-setting (KMS), memory management, cache coherency, command submission, and other topics as it pertains to open-source and ATI.
The DRI2 (Direct Rendering Infrastructure 2) talk is on the improvements made to DRI2 to make it not dependent on TTM and the removal of SAREA and and locks. This DRI2 work was committed to Mesa and the X Server yesterday.
For those out of the loop on Red Hat's Plymouth, it's a replacement for RHGB (Red Hat Graphical Boot) that will premiere with Fedora 10. Plymouth is much-improved over RHGB and it also takes advantage of kernel mode-setting to provide a pleasant experience for the end-user.
When it comes to the graphics testing talk, Intel's Gordon Jin will be talking about end-user testing with what Intel needs. I may also decide to give a talk on the Phoronix Test Suite 1.2.0 and performance profiling / testing.
The XDS 2008 program in its continually-revised form can be found on the X.Org Wiki. While the program is coming along, there is still under 30 people scheduled to be at this development summit. AMD's Alex Deucher will not be attending this event now that they're getting close to open-source R600/770 3D and John Bridgman will also not be attending. Additionally, Daniel Stone, who organized most of this event, won't be attending either.
During the X Developers' Summit next week expect coverage, photographs, and audio/video recordings on Phoronix.
| 3
| 1,760,738,490.412378
|
https://www.phoronix.com/news/NjY3Ng
|
A GEM-ified TTM Manager For Radeon
|
Michael Larabel
|
Back in May when X.Org developers were voicing concerns about Tungsten's TTM as being the kernel memory manager used for graphics drivers, Keith Packard had unveiled the work Intel had been doing for an alternate kernel memory manager. This memory manager they call GEM, or the Graphics Execution Manager, is a competing solution but it has some advantages such as being simpler to develop drivers around (A Technical Explanation of Intel's GEM). Intel then continued in throwing out their TTM code and merging GEM to master.
This unexpected introduction of GEM has caused the other driver developers to rethink their memory management situation and it has also generated some additional headaches within the X.Org community. For instance, a new X acceleration architecture has been introduced (albeit based off the EXA API) but with hooks for GEM. In addition, the forthcoming release of X Server 1.5 / X.Org 7.4 has lost DRI2 support since it depended upon the TTM interface.
Red Hat's David Airlie, who mostly works on the xf86-video-ati driver, has a brief update on some of the work he has been doing with GEM. In the DRI development mailing list, Airlie shares that he has removed all the TTM interfaces and ioctls from the kernel API, a DRM hardware lock file, object zeroing for the video memory, and has built a Radeon GEM interface that resides on top of the TTM internals. Some developers have voiced concerns about GEM being designed with just Intel in mind, but now David has an interface that's compliant with the GEM, but at it's heart is still TTM. This work can be found in his modesetting-gem git branch.
| 19
| 1,760,738,490.910895
|
https://www.phoronix.com/news/NjY4NA
|
VIA Releases New X.Org Driver
|
Michael Larabel
|
Back in April at the Linux Foundation Collaboration Summit, VIA had announced an open-source initiative. However, not all open-source developers have been pleased with these actions by VIA seeing as they have let down the community in the past. It's even admitted VIA has been trying to copy some aspects of AMD's open-source strategy and have stopped open-source drivers in the past. To date though VIA has released a kernel frame-buffer driver and three programming guides (but one of them wasn't exactly new). Back in July though, VIA Technologies had appointed Harald Welte as its open-source liaison after they were evaluating their open-source role. Now VIA has released its first open-source X.Org driver.
This driver, which has been named xf86-video-via, isn't complete though. This open-source driver supports the VIA CLE266, PM800, CN400, CN300, P4M800CE, P4M800Pro, CN700, VN800, CX700, VX700, K8M890, P4M890, CN896, P4M900, and VX800 IGPs. It has full mode-setting support for CRT, LCD, and DVI devices along with 2D acceleration, cursor acceleration, and even X-Video but it goes without 3D acceleration. When it comes to the 2D acceleration, it also lacks support for EXA acceleration. The driver also goes without TV support.
The xf86-video-via driver package amounts to about 113,800 lines, but it isn't all new code. The code is copyrighted by VIA and S3 Graphics and dates back to 1998. Some portions of the code looks like it may be derived from its closed-source X.Org driver for VIA hardware.
What isn't clear though at this time is whether they intend to develop this driver openly or what features they intend to provide with this driver. Will they be adding EXA and 3D support? Right now this driver can only be downloaded from VIA's Linux website (direct download). This xf86-video-via driver is not being housed currently in the X.Org git repository where all of the other xf86-video-* drivers are nor has this driver even been announced on the X.Org mailing list.
While at first the xf86-video-radeonhd that was created by AMD's open-source initiative in conjunction with Novell had only mode-setting support with no acceleration or other features, at least they had shared their plans. With VIA right now we have no idea where they intend to stop. The RadeonHD developers are also active in IRC and on mailing lists.
What is also a bit disheartening with this code drop is that they apparently haven't worked with the community-spawned Unichrome or OpenChrome projects. These projects came about as efforts by the community to have open-source drivers for VIA hardware when VIA Technologies didn't step up to the plate. Now that VIA is trying to step up to the plate, it looks like they are just ignoring the work done by these development groups for the past few years. The OpenChrome driver is open-source and even has 3D support.
| 9
| 1,760,738,490.931125
|
https://www.phoronix.com/news/NjY1Ng
|
X Gets Predictable Pointer Acceleration
|
Michael Larabel
|
X.Org 7.4 / X Server 1.5 has experienced an incredibly long delay in getting out the door. It was originally supposed to ship in February, then May, and now its stagnate until Mesa 7.1 ships. It looks like it will be a late August or early September release, which is almost a year after X.Org 7.3 had shipped.
If the delays alone weren't bad enough, this X update will be shipping with a slimmer set of features than what was originally expected. Most recently DRI2 was dropped from X.Org 7.4 due to the TTM vs. GEM kernel memory manager situation, but it also lost out on Multi-Pointer X (MPX) and other features.
Permitting these features don't get further delayed, X.Org 7.5 / X Server 1.6 will hopefully end up being a nice release. We don't expect this next update to ship until the middle of next year or so, but it already contains the MPX integration, input improvements, and should contain a fixed up version of the Direct Rendering Infrastructure 2. In addition, a new acceleration method has been announced for the X Server.
Last week the UXA acceleration architecture was announced by Keith Packard as a GEM-ified architecture based upon EXA acceleration, but this isn't to be confused with this new announcement. In fact, this announcement by Simon Thum is for Predictable Pointer Acceleration.
The predictable part of this mouse pointer acceleration is to tell whether the pointer will move. Its features include user-selectable profiles control pointer acceleration, adaptive and constant deceleration, acceleration becomes predictable, and no overshoot when X blocks for a short time. Planned for the future is sub-pixel position and velocity and velocity estimation. However, most users likely won't notice this change when it goes into effect. This has been proposed for X Server 1.6.
The announcement for this new pointer acceleration can be read on the X.Org mailing list. Detailed documentation surrounding X's Predictable Pointer Acceleration can be found on the X.Org Wiki.
| 0
| 1,760,738,491.462198
|
https://www.phoronix.com/news/NjY3Mg
|
XDS 2008 In Less Than Two Weeks
|
Michael Larabel
|
As a reminder, there's about a week and a half now until XDS 2008. The 2008 X Developers' Summit is taking place in Edinburgh, Scotland from the 3rd to 5th of September. This is the first X.Org conference since XDC 2008 back in April. Unfortunately though, as of today the attendance count is still a bit low and there are few planned talks.
Registered thus far (Wiki-based registration) are just 28 attendees and six of those aren't yet confirmed. The XDS 2008 program right now just lists talks on the X.Org Foundation status, a future releases session, and an input summary on Multi-Pointer X and other improvements. Over the course of the three days though is room for 21 talks.
In comparison, last year at XDS 2007 were 50 attendees (their maximum capacity allowed) and even a waiting list had formed. The conference last year was three days long as well and was booked with talks, it's also where AMD had launched their open-source strategy. Hopefully we will see a last minute surge for XDS 2008...
The usual X.Org leaders will be there such as Daniel Stone and Keith Packard along with other Intel X developers (Eric Anholt, Ian Romanick, Gordon Jin, Jesse Barnes, etc). Planned for representing the ATI/AMD side is just Alex Deucher and hopefully John Bridgman. There from Novell's xf86-video-radeonhd crew is Luc Verhaegen and Egbert Eich. No NVIDIA engineers are scheduled to be there whether it be from NVIDIA Corporation or the open-source Nouveau group.
Keith Whitwell with Tungsten Graphics will be there as well, who lately has been working on Gallium 3D. Sun's Alan Coopersmith will also be there as well as Peter Hutterer, the mastermind of Multi-Pointer X.
We will be covering the three-day affair, which will hopefully end with several interesting talks. We also have permission to record audio/video of these talks, so expect those recordings as well.
For more information on XDS 2008 or for the free registration, visit the X.Org Wiki.
| 7
| 1,760,738,491.471308
|
https://www.phoronix.com/news/NjY1NA
|
SIGGRAPH 2008 Graphics Coverage
|
Michael Larabel
|
Taking place this week in Los Angeles, California is SIGGRAPH 2008, which is one of the best and most well known graphics conferences. We aren't attending this conference, but the biggest news to have come out of it so far this week has been the OpenGL 3.0 (and GLSL 1.30) release. There is quite a bit of negative feedback surrounding OpenGL 3.0 as it's failed to deliver on what was previously promised by the Khronos Group and those involved with the OpenGL design process. However, plenty of other events have taken place at SIGGRAPH too.
At SIGGRAPH, NVIDIA has unveiled a new Quadro Plex system, the Quadro Plex D Series VCS, which is a $10,000 super computer when it comes to visual computing for styling and design, geo-sciences, and scientific visualization. The NVIDIA Quadro Plex 2200 D2 VCS is made up of dual Quadro FX 5800 GPUs. For more information, check out the NVIDIA press release.
If you're interested in coverage of the technical talks from SIGGRAPH, we would recommend you check out Ian Romanick's talk. Ian was previously on the OpenGL ARB, worked on the open-source XGI Linux driver while at IBM, and most recently has joined Intel to work on Linux graphics as part of Keith Packard's Portland team and to work on optimizing Intel's Mesa 3D stack.
Ian's coverage from the first day covers Intel's Larrabee GPU architecture, interactive photo tourism, and much more.
On the second day of SIGGRAPH he mentions NVIDIA's talk on realistic hair in real-time, "Logarithmic Perspective Shadow Maps", "Multiresolution Texture Synthesis", and many other graphics-related papers were also delivered. Last but not least, the day ended with the much-anticipated OpenGL 3.0 talk. According to Ian, not many there were that angered about Longs Peak, but free beer was involved (though probably not opened the proper way).
| 2
| 1,760,738,491.966176
|
https://www.phoronix.com/news/NjYzNw
|
X.Org 7.4 To Lose DRI2 Support
|
Michael Larabel
|
It's six months late and X.Org 7.4 still hasn't shipped as its being held up on the release of Mesa 7.1. Hopefully though we'll see the release of Mesa 7.1, X.Org 7.4, and the X Server 1.5 in the very near future. However, there has been some last minute bloodshed before this first major X Server release in nearly a year. It appears that DRI2, which was first proposed back at 2007 X Developers' Summit, will be dropped from the X Server 1.5 series.
The design of DRI2, or the Direct Rendering Infrastructure 2, was a modest improvement over DRI (DRI2 design page) and noticeably to end-users it allows (accelerated) direct rendering to redirected windows. This becomes noticeable (or there the lack of) when using a compositing manager (Compiz) and watching a video or running an OpenGL application. DRI2 though has been dropped from the X Server 1.5 series due to its dependence on the TTM memory manager.
With Intel having jumped ship on TTM earlier this year and X developers complaining about TTM, the days of this hyped kernel memory manager from Tungsten Graphics look limited. Intel threw all of their weight into developing the Graphics Execution Manager as a new kernel memory manager. Some have been concerned about the GEM API being designed solely around Intel's needs, but it looks like it will be merged into the Linux kernel and will become the memory manager for new graphics driver work.
This abandoning of the TTM memory manager in favor of GEM has caused some problems for the Direct Rendering Infrastructure 2. DRI2 was designed around the TTM API and supporting this memory management API is a prerequisite for drivers supporting this infrastructure. As a result, DRI2 -- to some extent -- is forced back to the design table.
The DRI2 mastermind, Kristian Høgsberg, has announced on the X.Org mailing list that some of the DRI2 design decisions now need to be re-evaluated when switching from TTM to GEM. As a result, the current DRI2 code will be stripped away from the X Server 1.5 branch. We've been in a feature freeze for X Server 1.5 and this redesign work will take some time, so there's no hope in seeing a patched up DRI2 before X.Org 7.4 ships.
Input Hotness was already pushed back from X Server 1.5 as well as Multi-Pointer X (MPX). DRI2 is now the latest feature to enter the X Server 1.6 party. Also on the table for 1.6 is Distributed Multi-head X (DMX) and RandR 1.3. X Server 1.6 will ship with X.Org 7.5 sometime in 2009.
| 38
| 1,760,738,491.981527
|
https://www.phoronix.com/news/NjYyMg
|
VIA Publishes Three Programming Guides
|
Michael Larabel
|
Earlier this week we shared that VIA had appointed an open-source liaison to work on providing the community with documentation and source-code for their products and work to improve VIA's image within the Linux and open-source communities. Since VIA announced their open-source strategy earlier this year, all they had provided was a simple kernel frame-buffer driver. However, VIA has now made available three programming guides that cover their PadLock, CX700, and VX800/820 products. In total these three documents amount to about 800 (PDF) pages.
PadLock is VIA's hardware-based security engine embedded into their recent x86 processors (among them are the VIA Eden, C3, and C7 processors). The VIA PadLock provides a random number generator, advanced cryptography engine, and RSA algorithm computation. The VIA PadLock programming guide appears to cover all areas of this security engine.
The VX800 chipset was VIA's first IGP (Integrated Graphics Processor) with Microsoft DirectX 9.0 support and contains a variety of other features and enhancements. The VIA VX800/VX820 documentation is 494 pages and it provides a register overview for both Northbridge and Southbridge functionality from shadow RAM controls and SATA links to power management. This documentation was last updated last month and still contains VIA's confidential watermarks, even though this documentation is distributed freely with no Non-Disclosure Agreement.
The CX700 documentation is similar to that of the VX800/820, but instead it covers VIA's single-chip system media processor designed for the mobile market. The VIA CX700 uses the UniChrome Pro graphics processor and supports other features such as Serial ATA 2.0, DDR2, etc. This register reference guide is just over 200 pages.
These three documents along with other source-code and binary drivers can be downloaded from the VIA Linux page.
| 13
| 1,760,738,492.579768
|
https://www.phoronix.com/news/NjYzMw
|
X Developers' Summit 2008 Details
|
Michael Larabel
|
Last week from OSCON 2008 we reported that the 2008 X Developers' Summit (XDS) would be taking place from the 3rd to 5th of September, not the 10th to 12th as was shared at XDC. Now this week Daniel Stone has shared more details on XDS2008, which now is only about a month away.
There is now a XDS2008 Wiki page, registration page, and program. So far on the program is a talk on the X.Org Foundation and a session on future releases of X.Org.
The first week of September we will be in Edinburgh for this X conference, and of course we'll be reporting on the interesting happenings from this event. Permitting there is interest and we're allowed to do so, we'll also willing to provide audio/video recordings of the different talks.
| 1
| 1,760,738,492.588339
|
https://www.phoronix.com/news/NjYxNA
|
VIA Appoints An Open-Source Liaison
|
Michael Larabel
|
VIA's commitment to the open-source community has been everything but stellar. VIA Technologies has taken advantage of the open-source community before, and many are saying VIA is doing another open-source bluff. Back at the Linux Foundation Summit earlier this year, VIA had announced they were going open-source with documentation, open-source drivers, and the works. However, to this date most of what we have seen opened is just a simple fbdev driver.
Today though from the Ottawa Linux Symposium, VIA has announced they've appointed an open-source liaison. Who's taking up this job? Harald Welte. Welte has been involved with OpenMoko, netfilter/iptables, and projects to enforce the GPL license. Harald will be responsible for refining VIA's open-source strategy and improving their Linux drivers.
The press release for VIA appointing Harald Welte as its open-source liaison can be read here.
| 24
| 1,760,738,493.106581
|
https://www.phoronix.com/news/NjYxMQ
|
X.Org Server 1.4.99.906 Released
|
Michael Larabel
|
X.Org 7.4 / X Server 1.5 is running months behind schedule, but now holding up its release is Mesa 7.1 or there the lack of. However, Adam Jackson has pushed out a new X Server 1.5 development release with a few more changes while waiting for this updated Mesa release.
This new release, xorg-server 1.4.99.906, has about two dozen bug fixes but nothing too extraordinary. The X Server 1.4.99.906 release announcement can be read on the X.Org mailing list.
| 0
| 1,760,738,493.817575
|
https://www.phoronix.com/news/NjU3Nw
|
An Update On Generic GPU Video Decoding
|
Michael Larabel
|
One of Google's Summer of Code projects this year is to bring hardware-based video acceleration to Linux with Gallium3D. The advantage of this design is that the implementation is designed to be universal to any driver using Gallium3D, which for now is largely just the Nouveau driver and an experimental Intel version.
With many of the open-source drivers currently lacking any form of GPU-based video decoding acceleration (such as XvMC or the forthcoming VA-API), this will be a terrific feature as it will provide this functionality once the drivers make the switch to Tungsten's Gallium3D as this method doesn't require any hardware/driver-specific work. This Summer of Code work is focusing upon an XvMC (X-Video Motion Compensation) front-end, which right now is limited to MPEG-2 acceleration, but more video standards may be added later.
Anyhow, the latest news (albeit a week late) on this development work can be found on the developers website. He has posted a few screenshots from test clips now that he has made some additional progress with field-based prediction working. The way this young developer is achieving this feat is through writing shaders with Gallium3D. Right now he's relying upon the Gallium3D soft-pipe driver, but according to his road-map he hopes to have it running with the open-source NVIDIA "Nouveau" driver by the end of the month. If all goes according to plan, by the end of August there should be performance tuning, bug fixes, documentation, and the rest of what's needed to make it a viable piece of software.
| 10
| 1,760,738,494.348955
|
https://www.phoronix.com/news/NjYwNg
|
The Status Of DRI2, GEM, KMS, Etc...
|
Michael Larabel
|
Later this week at OSCON 2008, Intel's Keith Packard will be talking about the Linux desktop when it comes to X.Org, Mesa, and related areas. Unfortunately, it doesn't look like the talk will be that technical (unless of course you haven't been staying up to date on our graphics articles), but today he has provided an in-depth blog posting. In this post, Keith provided his take on the current status of X when it comes to output technologies.
In Keith's blog post he covers output hot-plugging, initial mode selection, kernel mode-setting, the Graphics Execution Manager, Composite Acceleration, and DRI2. From the post, Keith seems confident that GEM (Intel's answer to the TTM for GPU memory management in the kernel) will be merged into the kernel in the coming weeks, there is now improved Composite acceleration, and DRI2 with GEM may work in the near future.
For all of the details, visit Keith's blog. Should Keith say anything especially interesting during his OSCON talk on Wednesday (or responds to heated questions), we'll be sharing the information.
| 6
| 1,760,738,494.357201
|
https://www.phoronix.com/news/NjYwMw
|
X Devs Drop NVIDIA Auto-Config Support
|
Michael Larabel
|
Sparking a heated Sunday afternoon debate, NVIDIA's Aaron Plattner had commited a trivial change to the X Server that resulted in several key open-source X developers becoming disgruntled. Ultimately, this NVIDIA-spawned patch ended up being recalled just hours later.
This aim of this patch was quite simple. Graphics processors with a vendor ID of 0x10de (which means one from NVIDIA Corporation) it would attempt to auto-configure the server to use NVIDIA's binary driver (if installed) before defaulting to the open-source xf86-video-nv driver. This patch had nothing to do with making the X Server require this binary blob or anything along those lines, but if the proprietary NVIDIA driver is present, the server would use it. This would allow the NVIDIA driver to be used without having a xorg.conf file present or manually specifying the driver to load. The xorg.conf file will eventually be removed in favor of greater auto-configuration and other persistent driver options. Some Linux vendors have already begun shipping their distributions without this configuration file by default.
Aaron Plattner's reasoning for this patch was that if the user has chosen to install the binary driver from NVIDIA for enabling 3D acceleration, it should be automatically used. It'd work out nice for a new user who wishes to use the binary NVIDIA driver to play games or use Compiz, but does it teach them anything about free software? David Airlie was the one that proposed this patch be reverted on the basis that there should be no auto-configure support for drivers that don't ship with X.Org.
Daniel Stone had also added in that this patch is "partially pointless" and adds a greater burden onto the X.Org developers. Daniel's standing is that this patch shouldn't belong in the master development branch, considering the common ABI breaks with new X Server versions and the lag time it takes NVIDIA to support these updates. However, in the event the NVIDIA driver does break, it should proceed to load the (open-source) "nv" driver.
To the desktop Linux user, the problem with this official open-source NVIDIA driver is that it contains only very basic card functionality and is limited to 2D acceleration. The code to this driver has also been intentionally obfuscated by NVIDIA engineers. Unfortunately, the community reverse-engineered Nouveau driver isn't yet in a state to become the open-source default for X.Org.
Following additional bickering, this NVIDIA patch ended up being eliminated. The thread regarding this auto-configure patch can be read on the X.Org mailing list.
| 48
| 1,760,738,494.971106
|
https://www.phoronix.com/news/NjU4MA
|
Compiz Support For Multi-Pointer X
|
Michael Larabel
|
In May we shared that Multi-Pointer X (or MPX for short) was entering the mainline X server. While it was merged to master that month, X Server 1.5 was already branched out and therefore it won't appear in X.Org 7.4, but it will appear in X Server 1.6 (X.Org 7.5) until next year.
While it's now in the mainline branch, Peter Hutterer, the chief developer of MPX, hasn't stopped there. One of his most recent accomplishments was modifying Compiz to support Multi-Point X. In his personal git repository, he now has a custom version that does support multiple inputs and multiple events occurring simultaneously while still enjoying the wobbly windows and other desktop effects presented by this window manager. To do this, Peter had to rewrite portions of the Compiz Core to support XI2 (X Input 2) by replacing some functions and listening to X Input events instead of internal Compiz core events. Additional work was then also required to support the Compiz plug-ins on MPX.
In a blog post by Peter explaining this Compiz+MPX work, he mentions that many details are left in an unfinished state. Unfortunately, due to time commitments, he won't be continuing this work with an MPX-aware XI2-supportive version of Compiz. Perhaps someone will step up to the plate and finish it off and down the road be able to merge it into the mainline Compiz.
| 8
| 1,760,738,494.979954
|
https://www.phoronix.com/news/NjU2MA
|
libdrm Updated For X.Org 7.4, Mesa 7.1
|
Michael Larabel
|
As we mentioned yesterday when X Server 1.5 RC4 was released (only to be outdone by 1.5 RC 5 that same day), this new version of the X Server (and therefore X.Org 7.4) requires building against the latest Mesa code that ultimately will become known as Mesa 7.1. In addition to needing the latest-and-greatest from Mesa, a new version requirement for libdrm is also in place. The Direct Rendering Manager library is now at version 2.3.1 and that or a newer version (to come) will be needed for X Server 1.5 / X.Org 7.4. David Airlie's change-log for libdrm 2.3.1 just says "Stuff changed - you need this for Mesa 7.1, and Xorg 1.5 Deal with it." No GEM or TTM code is present in this DRM point-release. However, Intel will be rolling the Graphics Execution Manager into libdrm 2.4 very shortly.
The (very brief) libdrm 2.3.1 release announcement can be read on dri-devel.
| 0
| 1,760,738,495.501077
|
https://www.phoronix.com/news/NjU1OQ
|
X Server Oops, Another Release Today
|
Michael Larabel
|
The much-delayed X Server 1.5 RC4 (v1.4.99.904) release made it out today, but it's already being replaced by a newer version. No, there wasn't a big development marathon today or anything spectacular, but X Server 1.4.99.904 got pushed out with a show-stopping bug. The XInput ABI (for the non-programmers, the Application Binary Interface) number went with version 3.1, when it was supposed to be 2.1. Keith Packard had also squeezed in a change for wrapping AddTraps in EXA and Damage.
X Server 1.5 RC5 (v1.4.99.905) was therefore pushed out just a few hours after 1.5 RC4 with this ABI numbering fix and Keith's change. The announcement can be found on the X.Org mailing list.
| 1
| 1,760,738,495.509162
|
https://www.phoronix.com/news/NjU1Ng
|
X Server 1.4.99.904 Released
|
Michael Larabel
|
The development cycle for X.Org 7.4 has stretched on for months longer than anticipated, and earlier this month it sounded like X.Org 7.4 would end with a quick release, but still it's being dragged out longer. This morning the release manager for X.Org 7.4, Adam Jackson, had announced the release of X Server 1.4.99.904. This is another development release -- equivalent to a Release Candidate 4 -- before X Server 1.5.0 is ready for X11R74.
In total, X Server 1.4.99.904 has 81 changes, which consist of CVE security fixes, fixes so the X Server can build properly on the Alpha platform, dropping the GLCore dependency (and move it to dynamic loading), memory leak fixes, and more. With that said, the X.Org 7.4 bug tracker is now cleared out! Unfortunately though, not all of the original bugs were addressed as some were moved to being cleared at a later time (circa X.Org 7.5 in 2009). Building X Server 1.4.99.904+ now requires building against Mesa 7.1 RC1 or later.
It looks like X.Org 7.4 and X Server 1.5.0 may finally make it out the door in July, not in May as once planned. For the features coming with X.Org 7.4, check out The Progress Of X.Org 7.4. You may also be interested in reading X Server 1.4.1 Is Released, No Joke where we look at the chilling issue of the degrading quality of X.Org releases.
The X Server 1.4.99.904 release announcement (with source download links) can be found on the X.Org mailing list.
| 1
| 1,760,738,496.07608
|
https://www.phoronix.com/news/NjUyMg
|
VIA Evaluating Its Open-Source Role
|
Michael Larabel
|
Two months ago VIA Technologies had announced their intentions on joining the open-source bandwagon, but since that time we've seen little more than talk. The VIA Linux Portal was announced as a place for hosting "drivers, technical documentation, source code, and information regarding the VIA CN700, CX700/M, CN896 and the new VIA VX800 chipsets." However, this "Linux Portal" isn't anything more than a binary dungeon. There are a few binary-only VIA Linux drivers on that website and nothing more. There are bug tracking and forum sections on this website too, but they aren't yet established.
Last month we shared the views of several open-source developers that VIA's open-source efforts may be a bluff. Though, however, last month VIA proceeded to provide a 16,434 line kernel frame buffer driver.
So one might assume VIA has just been busy working on sanitizing documents and code, right? Well, their silence hasn't been because they are hard at work preparing information, but they are still back at the planning stage.
Since yesterday's announcement of X.Org 7.4 coming soon, there has been a discussion within the mailing list to determine whether the UniChrome or OpenChrome driver should be the default for X.Org. UniChrome and OpenChrome are both developed by third-parties unaffiliated with VIA as their official xf86-video-via driver has its share of issues. Between UniChrome and OpenChrome, each driver has its advantages when it comes to which one has already adopted libpciaccess, supports the most VIA ASICs, etc. But that's not the point of this article.
In this mailing list thread, Xavier Bachelot -- who is one of the OpenChrome developers -- had provided a brief update on VIA's open-source strategy. The OpenChrome developers had a two-hour talk with VIA concerning this strategy, which happened a month ago. Since then, all that VIA has told these open-source developers is "a one liner saying they are still 'collecting the idea inside the VIA Linux group' and they still don't know what they 'can do for the first step'."
So if you had hoped to see a plethora of code and documentation from VIA early this summer, guess again. As we shared in an earlier article, VIA has been planning a strategy since at least last October when they had contacted Luc Verhaegen (the UniChrome developer, and one of the Novell developers working on RadeonHD) looking for ways that VIA could improve its open-source image.
We have no idea how far VIA Technologies now is in their "idea collection" process internally or what options they are looking at for their "first step", but it will likely be a ways out before anything comes to fruition. AMD's open-source strategy was being worked on internally for three or four months before it was publicly announced last summer. AMD has provided a number of documentation dumps covering the mode-setting to 2D to 3D for their GPU product families since last September. AMD's most recent documentation dump was yesterday when they released the R600 ISA document, but the community is still waiting on all of the 3D R600 information and sample code (to come in the form of TCore and another package they'll soon be announcing). Soon as their next-generation GPUs are released, the community will even have more on their plate that they're waiting for. AMD has two dedicated engineers working on this open-source sanitizing process and even still it takes quite a while. They have to pour through thousands of lines of source-code, type up documentation, and ask their design architects to fill in voids within their internal documentation. This is then followed by having all of the information approved by a review board before it's ready for release.
With VIA still thinking of what to do for the open-source community, we're looking at some time before the community sees any action -- even if VIA just goes through half the protocols that AMD is going through with their open-source process. Long story short, the silence isn't because VIA is busy working on their open-source contributions, but they don't even know yet what they want to do. By that time, NVIDIA could be opened up and AMD could be working on open-source R800 specifications.
| 4
| 1,760,738,496.084876
|
https://www.phoronix.com/news/NjUxNw
|
X.Org Releases X Server 1.4.2
|
Michael Larabel
|
It took more than 200 days for X Server 1.4.1 to be released, but X Server 1.4.2 is coming just about 24 hours after yesterday's release 1.4.1. This release is coming along quickly as Matthieu Herrb had shared multiple vulnerabilities in the X Server extensions. Three of these CVEs have to do with the RENDER extension while the other two fixed in X Server 1.4.2 are for memory corruption with the RECORD and Security extensions and an MIT-SHM arbitrary memory read. The X Server 1.4.2 release announcement can be read on the xorg mailing list.
| 0
| 1,760,738,496.721487
|
https://www.phoronix.com/news/NjUxOA
|
Coming To A Desktop Near You: X.Org 7.4
|
Michael Larabel
|
The last day and a half has been filled with X.Org articles at Phoronix... Yesterday was X Server 1.4.1, today was X Server 1.4.2, and in just a few days we will now see X Server 1.5.0. X Server 1.5.0 and X.Org 7.4 were expected to be released in tandem earlier this year, but its February then May release date had slipped with no official word of when a release would be out. There were a total of four development releases planned prior to X Server 1.5.0, and we've only had two of them so far. However, Adam Jackson has made it known on the X.Org mailing list that X.Org 7.4 will soon be released.
In this message Adam mentioned he wanted to have a last look at the bug list and address anything that looks terrifying or trivial and then make a few last technical changes. Permitting there isn't a big push by many developers to clear out the blocker bugs, we're looking at just short of three dozen blocker bugs being left open before X.Org 7.4. Due to glcore changes, Mesa 7.1 will be a requirement for X.Org 7.4 / X Server 1.5.0. Before X.Org 7.4 is declared, Adam (and any volunteers) will be running through the drivers one last time and the evdev-input 2.0 driver will also be released.
Once X.Org 7.4 is out the door, planning begins for X.Org 7.5 / X Server 1.6.0 (more Phoronix coverage at that time).
| 4
| 1,760,738,497.424395
|
https://www.phoronix.com/news/NjQ4Ng
|
Multi-Pointer X Merged To X.Org Master
|
Michael Larabel
|
Earlier this month we reported at Phoronix that Multi-Pointer X was going mainline with the planned merge to master date being the last week of May. Peter Hutterer has remained on track with this goal and as of last night, Multi-Pointer X (or MPX for short) can now be found in the X Server master branch. However, as Daniel Stone was busy with the Debian-OpenSSL fiasco (and maybe working on X Server 1.4.1), the new version of XKB has yet to reach master.
It was broken for a while, but input device coordinate scaling has been fixed with Multi-Pointer X, which allows tablet PCs to use this MPX code found in the mainline git tree. In the X.Org mailing list, Hutterer goes over the details on the API/ABI bumps and what is needed if you immediately want to get your hands on this X technology that allows for multiple keyboards and mice to be used simultaneously.
With MPX being in master, this will be a feature for X.Org 7.5, as while X.Org 7.4 is running behind schedule it already has been branched off.
| 1
| 1,760,738,497.971116
|
https://www.phoronix.com/news/NjQ3Ng
|
Intel's Graphics Execution Manager
|
Michael Larabel
|
Amid the on going discussion right now surrounding the TTM memory manager, Intel's Keith Packard has announced another new project that picks up another acronym. The Graphics Execution Manager, or gem for short, that as Keith explains, "It takes the lessons we've learned from TTM and constructs just the API we need to implement the dri_bufmgr interface." That may not sound interesting to an end-user, but through the use of this Intel Graphics Execution Manager, the Intel i915 Linux performance has been improved by 50% in OpenArena and glxgears is running 60% faster. However, this Intel gem doesn't take advantage of features found in newer Intel IGPs (GMA 3000 / i965 series) quite yet. In Keith's mailing list announcement he goes on to explain the technical workings of the Graphics Execution Manager as well as its API.
| 6
| 1,760,738,497.979343
|
https://www.phoronix.com/news/NjQ3NQ
|
Developers Voice Concerns About TTM
|
Michael Larabel
|
If you're interested in the internal workings of X.Org and Linux graphics drivers, you may want to read the latest discussion going on the DRI mailing list that concerns the TTM memory manager. Thomas Hellstrom of Tungsten Graphics had asked what is stopping TTM from going in the mainline kernel, which led David Airlie to chime in on its current lack of open-source drivers utilizing this memory management system (really just the Intel driver at this point) and thoughts from other developers. Some feel TTM is too oriented towards its usage on Windows and they have other technical hard-feelings towards this Tungsten Graphics creation. Read the thread on dri-devel.
| 0
| 1,760,738,498.488187
|
https://www.phoronix.com/news/NjQ2Nw
|
VIA Gives 16,434 Lines Of OSS Code
|
Michael Larabel
|
Back at the Linux Foundation Austin Summit, VIA had announced plans to develop a new open-source initiative in a similar fashion what AMD has been doing. However, in the weeks following that they haven't done much for the open-source community. As was highlighted in VIA's Open-Source Efforts A Bluff?, their Linux website just contains two binary drivers right now and not much of anything else -- not even bug tracking software or a mailing list. This has upset some, but fortunately VIA has stepped up to the plate and shown they are actually doing more than a media blitz.
VIA has released over 16,000 lines of code that provides a frame-buffer driver in the Linux kernel. This code is licensed under the GNU GPLv2 and appears to be crafted by VIA's Joseph Chan. Supported by this driver is VIA's Unichrome CLE266, K400, K800, PM800, CN700, CX700, K8M890, P4M890, P4M900, and VX800 IGPs. We're still pouring over the code, but it seems to be in pretty good shape and does support digital connections (and does seem to support HDMI already) -- in other words it appears to be further along than when the RadeonHD driver started out.
This current work comes as nine patches presented on the linux-fbdev-devel mailing list (a few VIA patches).
Kudos go out to VIA Technologies this morning for this code dump, but the work isn't over. They still have a lot of work left to do to mend relations with the Unichrome and OpenChrome projects and focusing upon 3D and video playback work, etc. However, this is a step forward in showing that VIA may actually come around this time and play ball with the open-source community.
| 0
| 1,760,738,498.504658
|
https://www.phoronix.com/news/NjQzOA
|
X Developers To Meet At The Zoo
|
Michael Larabel
|
The 2008 X Developers' Conference wrapped up today at the Googleplex in California. With it being the last day of the conference, there isn't much information to share judging by the XDC Notes, compared to the first two days. There is, however, tentative information regarding this year's X Developers' Summit. XDS 2007 was in Cambridge (UK), but XDS 2008 will be taking place at the conference facilities found inside the Edinburgh Zoo. This X.Org conference is tentatively scheduled from the 10th to 12th of September. Fitting well with Linux developers, this Scottish zoo does have both King and Gentoo Penguins.
Additional details surrounding XDS 2008 are expected in a few weeks.
| 0
| 1,760,738,499.062852
|
https://www.phoronix.com/news/NjQzNQ
|
XDC 2008: Radeon, RandR 1.3, & More
|
Michael Larabel
|
The second day of XDC 2008 was filled with quite a bit of technical information (as you can see from the Wiki notes page) ranging from the DRI2 infrastructure and talking about its event ring buffer and simpler DDX driver requirements to discussing video playback APIs. However, there are a few bits of information that are relevant to the community at large.
- Radeon kernel mode-setting is in progress. The new Radeon driver that does kernel mode-setting will be using the AtomBIOS. Jerome Glisse realizes now though that rewriting the driver was probably a mistake and should have started out with the existing DDX (xf86-video-ati) driver code. AtomBIOS is being used for leveraging the engineering work of AMD, mode-setting is "uninteresting", and reducing the time to support the new hardware so that these developers can focus a majority of their time on "fun" features. Right now there is an ongoing debate of AtomBIOS usage versus communicating directly with the hardware registers, which is the large difference in the Radeon and RadeonHD drivers. Future work includes moving to DRI2 and writing a Gallium3D driver.
- David Schleef had talked about video playback on X and his work on Dirac, which is a codec designed to be competitive with MPEG-4. He had went over the pros and cons of X-Video and OpenGL playback and says the future as having a pluggable pipeline, extending X-Video, and offering GPGPU (General-Purpose computing on Graphics Processing Units) codecs.
- Alex Deucher had talked extensively about the R600 GPU architecture, but he had also commented that the TCORE programming SDK will be out soon as well as another programming guide. This TCORE code should be out very soon -- there was a 50%+ probability that it would have been released yesterday. We have talked about what TCORE means in an earlier article.
- Last but not least, Keith Packard had talked about RandR 1.3. The proposed features for this update to RandR 1.2 includes DPMS events, per-output DPMS, panning rectangle, projective transforms, GPU objects, CRTC properties, and standard output properties.
More details can be read on the XDC Wiki.
| 0
| 1,760,738,499.071786
|
https://www.phoronix.com/news/NjQzMQ
|
XDC 2008 Kicks Off This Morning
|
Michael Larabel
|
The 2008 X Developers' Conference (XDC) is kicking off this morning at the Googleplex in Mountain View, California. The XDC (not to be confused with XDS, the X Developers' Summit) is made up of a variety of presentations and lightning talks surrounding X.Org and upcoming work done by this important free software project.
Initially the program ideas for this year's conference were rather limited, but they have started to fill up the schedule for this three-day affair. Among the topics that will be discussed at this Google-sponsored XDC include the status of X.Org 7.4, mobility support, suspend and resume support (or there the lack of), Mesa / Gallium3D technical overview, DRI2, TTM memory manager driver optimization, ATI R600 architectural overview, state-of-the-art Radeon driver overview, and WINE and X.
About 50~60 attendees are expected for the 2008 X Developers Conference. Among these attendees are all of the notable X.Org developers and other stakeholders, such as Daniel Stone, Alex Deucher, John Bridgman, Matthew Tippett, Adam Jackson, Keith Packard, Jerome Glisse, Zack Rusin, and others. ATI/AMD has three people at this conference (Deucher, Bridgman, and Tippett) from Canada while Andy Ritger seems to be the only NVIDIA representative in attendance. This is odd seeing as NVIDIA's Santa Clara offices are just a stone's throw away from Mountain View.
Unfortunately, we weren't able to make it out to this event, but when anything interesting occurs through monitoring the Wiki pages, IRC channels, or we hear it from one of the attendees, we'll be sure to pass it along. There's a possibility that AMD could be releasing additional code/documentation at this event.
More information on XDC 2008 can be found on its Wiki page.
| 2
| 1,760,738,499.611647
|
https://www.phoronix.com/news/NjQzMw
|
XDC 2008 Day 1 Notes: X.Org 7.4 & More
|
Michael Larabel
|
As was mentioned yesterday, the X Developers' Conference (XDC) is taking place this week at the Googleplex. There's only around 50 developers in attendance, but for everyone else there still is a flow of information to gather. A few Google employees had recorded each of the talks from the first day, however, they aren't yet available on You Tube or Google Video. Unfortunately, it doesn't look like Google will be recording the X videos for the second and third days of this conference. Red Hat's Adam Jackson, however, has been summarizing the highlights of each talk in near real-time and has been uploading it to the X.Org Wiki.
On the first day of XDC 2008, Adam Jackson (X11R74 release manager) had talked about the X.Org 7.4 release status, Kevin Martin on various aspects of X.Org, a Mesa / Gallium3D overview by Tungsten's Zack Rusin, Matthew Garrett on suspend/resume support, and Ben Byer with Xquartz.
From the XDC 2008 Notes, it was discussed whether XAA (XFree86 Acceleration Architecture) support could be dropped altogether from X.Org, but it was decided against for this release as EXA for 2D acceleration still isn't ready for prime-time. Adam's talk then turned into a discussion over the quality of X.Org releases and need we say the degrading quality. The talk then turned to X.Org testing and the need for a new testing infrastructure, which the Phoronix Test Suite could very well handle.
Turning to Kevin Martin's talk, on the X.Org goals list are embracing the dynamic hot-plugging world, presenting a smooth and flicker-free experience (kernel mode-setting), seamless integration of Composite, redirected OpenGL and X-Video, and a secure X environment. Some of the major X accomplishments of recent times include RandR 1.2, DRI2, DRM and kernel mode-setting, Render acceleration, Cairo 1.6, and driver work.
Zack's Gallium3D talk appears to be mostly a repeat of what was said back at FOSDEM 2008 during the Gallium3D status update.
All of the highlights from the XDC 2008 conference can be read on the Notes Wiki page.
| 0
| 1,760,738,499.619895
|
https://www.phoronix.com/news/NjQxMQ
|
DRI2 Direct Rendering Now Available
|
Michael Larabel
|
Back on the 4th of February, Kristian Høgsberg began merging his new DRI2 components. This initial DRI2 (Direct Rendering Infrastructure 2) work was made up of the DRI2 module, glxdri2 to GLX, and DRM/Mesa patches. Today Kristian has merged his last major part of DRI2 and that is the direct rendering support. With this new code, it's now possible to do directed rendering to redirected windows! This directed rendering to redirected windows support even works with Compiz and other OpenGL window managers using the GLX_EXT_texture_from_pixmap extension. However, to reach this state the DRI interface was broken. Kristian's announcement can be read on the DRI devel mailing list. The only X.Org video driver using DRI2 right now is the Intel batchbuffer branch for xf86-video-intel.
| 22
| 1,760,738,500.141342
|
https://www.phoronix.com/news/NjM1Nw
|
Features Of A Happy Linux Desktop?
|
Michael Larabel
|
During Keith Packard's talk at FOSDEM 2008, he had went in-depth on talking about what he wants to see for a "happy desktop" for Linux users. Most of these issues are commonsense or have been described in earlier talks (such as from Linux.Conf.Au), but below is his list. Of course, all of his issues are X related.
- Fully Composited Desktop
- No Tearing
- Integrated Video, 3D, 2D, APIs
- Flicker-free Boot
- Fast User Switching
- Hot-plug Everywhere
- Lower Power
- Faster. Everywhere.
- Reducing Root Code
Keith had described each of the above topics and then ways that these issues will be addressed. He hopes most of these issues will get worked out this year. For kernel mode-setting, he feels the Linux 2.6.26 kernel is too early for its merge but will come later. You can view the X.Org @ FOSDEM 2008 talks on the RadeonHD.org server.
What would you like to see for a "happy" X.Org or Linux desktop? Tell us in the forums.
| 9
| 1,760,738,500.882139
|
https://www.phoronix.com/news/NjM2MA
|
X.Org 7.4 Release Planning
|
Michael Larabel
|
With Red Hat's Adam Jackson being interested in shipping X.Org 7.4 for Fedora 9, he has stepped up as becoming the release manager of this next X.Org release and has put together a schedule that will allow this to become possible.
The X.Org 7.4 release schedule places the X Server 1.4.99.901 release this coming Friday, X Server 1.4.99.902 on March 14 or 21, X Server 1.4.99.903 on April 4 and would also mark a freeze where only fixes would be allowed into X Server 1.5, X Server 1.4.99.991 with an X hard-freeze, and on April 25 is planned the final release of X Server 1.5 / X.Org 7.4. Permitting everything goes as planned with this release schedule, it will be released ahead of May.
While there have already been features planned for X.Org 7.4, some of the items in particular that Adam Jackson wants to see addressed is pciaccess, input improvements, RandR 1.2 initial configuration changes, and many X.Org bugs just need to be eliminated. There still is nearly two dozen X drivers not utilizing libpciaccess, but Adam would like to see this change in time.
An X.Org 7.4 tracker/blocker bug has been created and is Bug 10101. Comments and discussion surrounding this X.Org 7.4 news and Adam's announcement can be read on the X.Org mailing list.
| 0
| 1,760,738,501.372029
|
https://www.phoronix.com/news/NjM3Nw
|
X Server 1.4.99.901 Released
|
Michael Larabel
|
Thanks in large part to the release leadership of Adam Jackson, a new development X server is now available that ultimately will become known as X Server 1.5.0. This new xorg-server release, v1.4.99.901, has a plethora of changes ranging from fixing memory leaks to EXA improvements. In total there are hundreds of changes to be found within xorg-server 1.4.99.901, with the complete list being available in the X.Org mailing list. More information on the plans for X.Org 7.4 can be found in this article.
X.Org 7.4 / X Server 1.5 is planned for release in May, meanwhile X Server 1.4.1 still has yet to be released.
| 0
| 1,760,738,501.387029
|
https://www.phoronix.com/news/NjM1NA
|
Metisse X Window System Roadmap
|
Michael Larabel
|
Remi Cardona is currently talking at FOSDEM 2008 in the X.Org development room about Metisse. For those that don't know, Metisse is an open-source X Window System that was conceived as a French research project (and is incorrectly considered competition to Compiz). Remi's talk is on the features of Metisse, various desktop demonstrations, and bringing Metisse and X.Org together. Some of the items he shared for the Metisse road-map include adding a configuration interface, Expose-like features, FreeDesktop.org integration, and then X.Org integration. This research project is also looking at collaborative features such as window sharing between computers or other mobile devices, etc.
The video of the FOSDEM 2008 Metisse talk will be up soon on our other server. All other talks from yesterday and this morning are already available for download as Ogg Theora.
| 4
| 1,760,738,501.881846
|
https://www.phoronix.com/news/NjM1Ng
|
XvMC To Support More Video Standards?
|
Michael Larabel
|
While the XvMC (X-Video Motion Compensation) extension is reliable for offloading MPEG video decoding to the GPU, its limitation is that it only supports MPEG video formats and nothing more. We had expected XvMC not to be around much longer, since Intel has been devoting resources in creating a new video extension for X.Org. This new video work of Intel's is known as VA-API, or Video Acceleration API, and is still quite early in development. VA-API, however, will be able to handle offloading more tasks along with support for all of the latest video standards (MPEG-4, H.264, VC-1, etc). VA-API is not based upon XvMC but is written from scratch.
While VA-API will eventually prevail, it appears that Intel isn't yet ready to bring VA-API on full-force and that there is still life left to XvMC with another revision. During Keith Packard's X.Org talk at FOSDEM 2008, while talking about 2D, 3D, and video APIs, he had on his agenda bringing more video standards support to XvMC then merely MPEG. VA-API wasn't mentioned at all during this talk. It turns out that Intel at least still sees life left in XvMC as VA-API still has plenty of work to be accomplished and this new extension will take some time to be adopted. In the meantime, it looks like we'll see XvMC advancements for video playback improvements under Linux. Eventually, however, VA-API will prevail and become the new standard.
With it looking like there is still some new blood left in the X-Video Motion Compensation extension, perhaps this will force NVIDIA to re-evaluate their decision of dropping XvMC support for the GeForce 8+ series. Likewise, perhaps this will encourage AMD to finally deliver XvMC support into their fglrx Linux driver. XvMC support for recent Intel IGPs is available in an xf86-video-intel XvMC branch.
| 4
| 1,760,738,501.904413
|
https://www.phoronix.com/news/NjMzNg
|
X.Org 7.4 To Be Released In May?
|
Michael Larabel
|
Not that you really would have expected X.Org to ship next month as planned, but X.Org 7.4 looks like it may now be released in May. The original plan for X.Org 7.4 was to release it in February, but that was when X.Org 7.3 was planned for August though it didn't make it out the door until September. With a six month release cycle, March was the new strike point for this X.Org update. We're now running into late February and X Server 1.4.1 hasn't even been released yet, which is a maintenance update that was supposed to arrive last November. There are still five bugs blocking X Server 1.4.1. Red Hat's Adam Jackson has updated the 7.4 Wiki page to reflect that X.Org 7.4 is now planned for a May release. Hopefully that time-frame is met, but that could turn into a June release. If meeting their ideal six month release cycles, X.Org 7.5 would then be released in November or December, but because of the holidays that is likely a Q1'09 release.
While their release schedules aren't often met, the X.Org developers do have great ideas and planned work for these coming releases and none of their hard work should be discounted. Some of the planned features for X.Org 7.4 and X.Org 7.5 are mentioned in this news posting.
The 2008 X Developers' Conference (XDC2008) has been announced and will be taking place from April 16th to the 18th at the Googleplex. The program/talks for it have yet to be determined and the attendance is capped at 70. Phoronix will likely be covering this event.
| 0
| 1,760,738,502.497481
|
https://www.phoronix.com/news/NjMxNw
|
DRI2 Comes Closer To Its Debut
|
Michael Larabel
|
Kristian Høgsberg, the creator of AIGLX, has announced on the X.Org mailing list that he is beginning to merge his new DRI2 components. This work includes the new DRI2 module, glxdri2 to GLX, and DRM/Mesa patches. The DRI2 work is ready for early adopters, but is able to safely reside next to XF86DRI. With an Intel DDX patch in the intel-batchbuffer git branch, the xf86-video-intel driver will have a DRI2 option to manually enable the DRI2 module.
DRI2 is the new DRI (Direct Rendering Infrastructure) design that was originally drafted at the X Developer Summit (XDS) last year. One of the requirements for DRI2 is the use of the TTM memory manager. More technical information is available on the X.Org Wiki. More information on the DRI2 merger plans can be read in Kristian's message.
| 9
| 1,760,738,502.517046
|
https://www.phoronix.com/news/NjI5OA
|
The Latest On RandR 1.3
|
Michael Larabel
|
For those not satisfied by RandR 1.2 or just wish to live on the cutting-edge of X.Org developments (like us), this week on the X.Org mailing list has been a discussion among driver developers surrounding RandR 1.3. RandR 1.3 is the next update to the Resize and Rotate (RandR) extension that allows for resizing, rotating, and reflecting of the X screen. With the RandR 1.2 update it had introduced display hot-plugging support. When it comes to features, RandR 1.3 is most notably expected to introduce GPU object support. This GPU object support is another layer between the X screen and the CRTCs. Ultimately, this should allow multiple GPUs to be merged into a single X screen.
Jesse Barnes had initiated the RandR 1.3 discussion this week by proposing two additions that would reduce mode-setting flickers and an output property method that would allow other properties to stay up-to-date with events that bypass the driver (i.e. back-light changes from hot-key events). Alex Deucher had then proposed an additional level of abstraction for encoders and connectors, rather than just outputs. Other items brought up by developers include exposing the i2c bus in RandR, support for monitor calibration controls, and multi-card support. Read more on the X.Org mailing list.
RandR 1.3 is targeted for X Server 1.5 / X.Org 7.4, but it could be delayed (v1.5 has already lost the "input hotness"). The X Server 1.5 release is planned for March of this year, but a delay is expected. For more on the current capabilities of RandR, check out our newbie's guide to RandR 1.2.
| 2
| 1,760,738,503.015127
|
https://www.phoronix.com/news/NjI0Ng
|
X Server 1.4.1 Pre-Released, Finally!
|
Michael Larabel
|
It's coming a month late, but today X Server 1.4.1 has been pre-released. The xorg-server is now at version 1.4.0.90 and contains over two dozen fixes since X Server 1.4.0. The X Server 1.4.1 pre-release notes and download link is available from the mailing list announcement. On the xorg-server 1.4.1 blocker bug are still five outstanding bugs.
| 0
| 1,760,738,503.02334
|
https://www.phoronix.com/news/NjIzOQ
|
Linux Joystick Driver Gets EvDev Support
|
Michael Larabel
|
Sascha Hlusiak has announced the release of xf86-input-joystick v1.3.1, which is for providing joystick input support on X.Org 7.3. What makes this joystick driver release newsworthy is that it now provides evdev integration. If you're not familiar with evdev, it's the generic input driver for X.Org, and commonly is used for accessing USB input devices. This new joystick driver supports evdev but has fall-back support for the Linux joystick interface. Thanks to the evdev support, the xf86-input-joystick driver can now be auto-loaded with hot-plug support through HAL. A few other changes also make up the xf86-input-joystick v1.3.1 release, which is detailed in the release announcement.
| 0
| 1,760,738,503.597548
|
https://www.phoronix.com/news/NjIzOA
|
No "Input Hotness" For X Server 1.5
|
Michael Larabel
|
There's still one bug left before the release of X Server 1.4.1, but Daniel Stone has announced on the X.Org mailing list that X Server 1.5 will no longer contain any "input hotness". Specifically, XKB 2 and Xi 2 were planned but that has been postponed until at least X Server 1.6. This delay has occurred because these changes will require quite a bit of work with how input events are processed and related MPX changes. Expect new X.Org input hotness then towards the end of 2008.
| 3
| 1,760,738,504.286263
|
https://www.phoronix.com/news/NjIyOQ
|
Two Bugs Left Before X Server 1.4.1
|
Michael Larabel
|
X Server 1.4.1 was originally slated for release on the first of November, then Daniel Stone had pushed the release back to November 11, but that didn't happen either. The X Server 1.4 Wiki page still reflects a release date of November 11, but that's passed by nearly a month.
For the server 1.4.1 blocker bug, back on November 11 it had five dependent bugs with three still open. As it stands today, there are 10 bugs on the list with all of them being closed except for two. These two bugs holding up the release of xorg-server-1.4.1 is for SetKeySymsMap's memmove writing too many bytes (Bug 9180) and a client shutdown caused by EINTR in select (Bug 9240). Both of these bugs have been open since the end of 2006.
Perhaps X Server 1.4.1 will arrive as a Christmas present this holiday season?
| 1
| 1,760,738,504.888429
|
https://www.phoronix.com/news/NjE4NA
|
X Server 1.4.1 Delayed, Again?
|
Michael Larabel
|
X.Org 7.3 was released two months ago, and scheduled to be released today was X Server 1.4.1. This release is supposed to address some of the issues that had come up with X Server 1.4.0; however, it looks like this release will be delayed once again. The X Server 1.4 Wiki page still shows the release date as November 11, but there are a number of open bugs remaining. The xorg-server-1.4.1 blocker bug has five dependent bugs open with only two being resolved. These five remaining bugs are for memmove in SetKeySymsMap, EXA corruption, EXA negative tile coordinates, a KDE Konqueror X11 issue, and input events being duplicated across different windows. Nothing has been brought up on the X.Org mailing list yet, but it doesn't look like X Server 1.4.1 will be released today. In the meantime, you may want to read The Degrading Quality Of X.Org Releases?
| 0
| 1,760,738,504.898185
|
https://www.phoronix.com/news/NjA4OA
|
Planned Features For X.Org 7.4, 7.5
|
Michael Larabel
|
Last month at the X Developer Summit in Cambridge, Eric Anholt, Adam Jackson, and Daniel Stone had talked about the future of X.Org releases for the next year. Over the weekend, Daniel Stone had updated the XDS 2007 Notes at X.org with the latest plans for X.Org 7.5. The current schedule for X.Org include the X server 1.4.1 release coming out on the first of November followed by X server 1.5.0 in March of 2008. Planned for X.Org 7.4 and X server 1.5 is XGE, XACE, RandR 1.3, PCI rework, XKB 2, _X_EXPORT, DRI memory manager, GLX 1.4, and Glucose.
XGE is the X Generic Event extension, which is required for MPX support, but these events can also be used by other extensions. The XACE framework is for handling security policy extensions. Parts of XACE have been in the server since X.Org 7.2 but it's not fully implemented and lacks the loadable security modules. For X.Org 7.4 there are XACE modules planned for at least SELinux and Solaris Trusted Extensions.
RandR 1.2 is currently the latest and greatest Reside and Rotate extension, but RandR 1.3 is planned for the future with GPU object support. RandR's GPU object is designed to be another layer between the X screen and the CRTCs. The PCI rework involves using the libpciaccess library, and some of the open-source X.Org drivers (such as the Avivo driver) already utilize this new library.
Next up is Glucose, which is an OpenGL based acceleration architecture that will be hooked into the X server and its acceleration is done using the same paths as XGL. Glucose is designed to work with AIGLX as it provides a convenient way of accelerating common rendering primitives.
Previously we reported that Multi-Pointer X Support would arrive for X.Org 7.4, however, according to the Wiki page that is now on the table for X.Org 7.5. Also planned for X.Org 7.5 is pushing Distributed Multihead X (DMX) up to DIX. Multi-Pointer X Support provides multiple cursor and keyboard support in X. MPX does allow for multi-pointer interaction with legacy applications while also allows for new and more innovative applications.
Like any software project, these current features planned for X.Org are tentative and subject to change and is just some of the work going into X.Org. At the same time, there has been concerns over the degrading quality of X.Org releases so hopefully X.Org 7.4/7.5 will mark the beginning of greater quality assurance work prior to the release. Additional details can be found from the XDS 2007 Notes.
| 17
| 1,760,738,505.415658
|
https://www.phoronix.com/news/NjExMA
|
X Server 1.4.1 Delayed To November 11
|
Michael Larabel
|
As opposed to being pushed out on the first of November, Daniel Stone (the release manager), has delayed the release of X server 1.4.1 by ten days. X server 1.4.1 will now be released on November 11 (permitting no more delays) and will contain a variety of bug-fixes and input fixes for extended events currently found in X server 1.4.0. The list of waiting and merged patches as well as other X server 1.4 details can be found on the X.Org Wiki. With X.Org currently being criticized for its degrading release quality, this X server 1.4.1 delay is hopefully good news and will result in a more polished release.
Daniel had also mentioned on the X.Org mailing list that the X server 1.5.0 release schedule has yet to be finalized and will likely not be known until the grunt of the libpciaccess work is completed.
| 0
| 1,760,738,505.429719
|
https://www.phoronix.com/news/NjA2NA
|
X.Org RadeonHD Gets AtomBIOS Parser
|
Michael Larabel
|
The open-source RadeonHD driver that is even supports the Radeon HD 2900XT has continued over the past few days with a number of changes. Among these changes has been adding identification support for new GPUs (including the Apple MacBook Pro 2.2), removing some experimental checks, adding the conntest utility, and this morning AMD's AtomBIOS parser was merged into the xf86-video-radeonhd git tree. This AtomBIOS parser was written internally at AMD but is now available as open-source code. Check out the RadeonHD git code if you want to try out an open-source driver for ATI's R500 and R600 graphics processors.
| 1
| 1,760,738,505.995145
|
https://www.phoronix.com/news/NjA3Mg
|
The Degrading Quality Of X.Org Releases?
|
Michael Larabel
|
On the X.Org mailing list, Alan Coopersmith had raised concerns over the release criteria for X11 and how with recent releases (namely X.Org 7.3), the de facto standard for making a release was far from being met. Alan, Sun's X engineer, had listed the release criteria as the blocker bug list being cleared, the complete tree/release modules being build-able on at least one platform, XTS successfully running on at least one platform, and the documentation being updated and then released. X.Org 7.3 was released eight days late and failed to fully meet any of the requirements -- some X.Org 7.3 driver packages didn't even build against the respective X server.
Alan Coopersmith's points are certainly valid, and in a rush to meet deadlines and push out software, the quality has lessened. But don't get the wrong impression, X.Org developers do work hard, are doing a great job, and that isn't to be discounted! This problem doesn't exclusively apply to the X.Org development community, but largely across the free software spectrum. To succeed, however, something must be done and we must push for quality-driven releases as we prepare for the next wave of Linux desktop users. We must have timed releases but at the same time we need to expect at least the same level of quality.
What are your thoughts on X.Org 7.3? How can open-source software releases be rejuvenated to prevent degradation? Share your thoughts in the Phoronix Forums.
| 8
| 1,760,738,506.003348
|
https://www.phoronix.com/news/NjAzOA
|
X.Org 7.3 Gets Released
|
Michael Larabel
|
Well, the delay we told you yesterday ended up being a one-day delay. As of this morning, X.Org 7.3 and X server 1.4 have been officially released. You can find the download details and other information on the X.Org Wiki.
| 4
| 1,760,738,506.50381
|
https://www.phoronix.com/news/NjAzNw
|
X.Org 7.3 Release Gets Delayed, Again
|
Michael Larabel
|
X.Org 7.3 and X server 1.4 were originally scheduled for release towards the end of August, but bugs had pushed back its release to September 5. Well, the day is now over and X.Org 7.3 is not to be found. The 7.3 release tracker bug still has eleven bugs currently outstanding. The X.Org mailing list hasn't announced a new release date yet. Additional information on X.Org 7.3 is available from its Wiki page and our X.Org 7.3 Preview.
| 1
| 1,760,738,506.546091
|
https://www.phoronix.com/news/NTk4OA
|
X.Org 7.4 To Get Multi-Pointer X Support
|
Michael Larabel
|
X.Org 7.3 will be released later this month, and now Daniel Stone has updated the X.Org Wiki with a page for the X.Org 7.4 release. With the six month release cycle, X.Org 7.4 is planned for release in February of 2008. Some of the features so far on their planning page includes Multi-Pointer X which provides multiple cursor support, support for UnixWare, SELinux security module, and a Solaris Trusted Extensions security module. The two security modules will use XACE (X Access Control Extension). Find out more on the Wiki.
| 0
| 1,760,738,507.137224
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.