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/Nzg1Mg
|
Pull Request Goes In For X's udev Input Handling
|
Michael Larabel
|
A month ago we reported on news regarding the X.Org plans to move away from HAL considering the FreeDesktop.org Hardware Abstraction Layer project is no longer being developed. Since then patches have emerged to support a xorg.conf.d directory for storing some device-specific options and some new xorg.conf configuration options have emerged for filling in some of the gaps previously covered by HAL.
The most recent development in this process to get the X Server no longer using HAL for input hot-plugging and device handling is the pull request submitted this week for the udev input-hotplug branch. This work now uses libudev on Linux for handling input device hot-plugging. Julien Cristau, the developer behind this udev input-hotplug branch, has said "I think this is now ready for prime time." However, when building the X.Org Server with this newest code it still is defaulting to off.
Here is the pull request for this work that should land in X Server 1.8, which is scheduled to be released in March.
| 10
| 1,760,738,438.096924
|
https://www.phoronix.com/news/NzgyNw
|
X Server 1.8 Snapshot 2 Released
|
Michael Larabel
|
With there being just more than three months left until the planned release of X Server 1.8, Keith Packard has just issued the second snapshot for those interested in trying out this developmental X.Org server. The first X Server 1.8 snapshot came two months ago, but this second snapshot is arriving later than expected after having to deal with some bugs.
What is important to note with this second snapshot is that it breaks the X Server ABI for input, video, and extension interfaces. This means that users of X Server 1.8 that depend upon the binary NVIDIA and ATI/AMD graphics drivers will need to wait for an appropriate driver update that brings ABI compatibility with this new X Server. NVIDIA should be supporting the interface changes soon, while if based upon AMD's past actions, it will be some months before the Catalyst Linux driver delivers on X Server 1.8 support.
There are just under 150 changes in X Server 1.8 Snapshot 2 that touch the DIX, EXA, XQuartz, and other areas. The release announcement for X Server 1.8 snapshot 2 can be found on xorg-announce. X Server 1.8 under their new release process is planned for release at the end of March and is expected to have xorg.conf.d support, removal of HAL support, and other changes.
| 5
| 1,760,738,438.648209
|
https://www.phoronix.com/news/NzgyNg
|
A Video To Show Off X.Org's Multi-Touch Support
|
Michael Larabel
|
Back in June we shared a nifty video that was produced by the French-based ENAC Interactive Computing Lab that showed off some multi-touch capabilities on Linux. At that time and in the video they were using the Linux kernel and then recognizing gestures from input events and then sending specialized commands over D-Bus straight into Compiz. Since that time though, X Server 1.7 / X.Org 7.5 has been released, which fully delivers on Multi-Pointer X support.
As a result of the MPX support in the mainline X.Org Server, the Interactive Computing Lab has produced a new video to show off this support as they had informed us this morning. In the video (below) they are using Fedora 12 with its X Server 1.7 and Linux 2.6.31 kernel.
A variety of touch-based devices are shown in this video, among which are the Broadcom 5974, Stantum, NTrig, 3M, and DiamondTouch surface. While their system configuration was largely stock from Fedora, they had used a modified xf86-input-evdev driver and a multi-touch management client (multitouchd), which is available from the computing lab's web-site.
| 47
| 1,760,738,438.656625
|
https://www.phoronix.com/news/NzgxNw
|
The xorg.conf.d Patches Emerge
|
Michael Larabel
|
One of the features being worked on for X Server 1.8 is the removal of HAL support. The FreeDesktop.org Hardware Abstraction Layer project is nice in that is multi-platform, but the HAL project has largely been abandoned and is being replaced by UDisks and similar projects.
HAL is currently being used by the X Server for input device detection with hot-plugging support, mapping, and device option handling, but all of that is in the process of being gutted. Instead the X Server will revert to using platform-specific libraries and code for handling these responsibilities. To replace the device option handling that was previously done through HAL FDI files, the X Server is picking up support for reading configuration files from a directory rather than just the conventional /etc/X11/xorg.conf.
Being worked on right now (with the patches already having gone through multiple iterations) is support for storing .conf files within /etc/X11/xorg.conf.d. The X Server will continue to support reading options from the traditional xorg.conf and these options will take precedent over any values set within xorg.conf.d, where the other options can be stored. Within the xorg.conf.d directory could be a file for your mouse, a file for your input tablet, joystick, and any other devices. If there is no xorg.conf or any configuration files within this new directory, the X Server will fall-back to its usual auto-detection routines.
The latest version of the xorg.conf.d patch can be found on xorg-devel, but it should end up being merged into the X Server shortly.
HAL FDIs support matching a configuration file to a specific device that's installed or to not load the configuration at all if the device isn't found. This works by matching the FDI file to either the manufacturer/product ID, device path, or a string. Through other patches, the xorg.conf.d files will gain similar support. Dan Nicholson (who also wrote the xorg.conf.d support) has submitted patches that introduce InputAttributes and InputClass configuration options.
InputAttributes stores the product/vendor name, device path, and also some device attributes -- such as whether the device has a pointer and/or input keys. The InputClass support then exposes this to the X configuration options so that configuration options can be set for specific vendors/products, a path-name pattern, or specific types of devices (keyboard, pointer, joystick, tablet, touch-pad, or touch-screen). This then allows X.Org input drivers to install their own configuration file within /etc/X11/xorg.conf.d, but these options will only be applied by the X Server if the actual device is present.
The InputAttributes and InputClass patches can be found here and here, respectively.
| 6
| 1,760,738,439.281574
|
https://www.phoronix.com/news/Nzc5Nw
|
X Server 1.7.4 RC1 Brings Few Fixes
|
Michael Larabel
|
X Server 1.7.2 was released at the end of November and then X Server 1.7.3 was released at the start of this month to fix two show-stopping issues in the earlier point release. However, the first release candidate for X Server 1.7.4 has now been made available by Peter Hutterer.
As all new development work is going into X Server 1.8, which will be released in March, these 1.7.x releases are just carrying various bug-fixes. With X Server 1.7.4 RC1 there is only a handful of bug-fixes that touch EXA, X-Video, and DIX, but nothing particularly exciting.
The X Server 1.7.4 RC1 release announcement can be read on xorg-announce. The official release of X Server 1.7.4 is expected in late December or early January, but at least another release candidate should be out before that time.
| 0
| 1,760,738,439.289686
|
https://www.phoronix.com/news/Nzc1OQ
|
X.Org Development Room @ FOSDEM 2010
|
Michael Larabel
|
Luc Verhaegen, the longtime X.Org developer who was previously employed by Novell to work on their X11 stack and the RadeonHD driver and continues to hack on his VIA driver, has announced that at FOSDEM 2010 there will be an X.Org development room. Luc has been organizing X@FOSDEM annually for the past five years.
X@FOSDEM 2010 though will not be the usual two-day event filled with X talks, but sadly, it has turned into a one-day affair. Over concerns of not having enough talks to fill the two days this year and Luc becoming interested in the CoreBoot project since being laid off at Novell, he has decided to host the X.Org room as the CoreBoot development room on Saturday.
X@FOSDEM 2010 will just be on Sunday, the 7th of February, in Brussels, Belgium. A one day X conference will allow for seven talks that span the course of the day, but on the previous night there may be the usual X.Org dinner at a local restaurant. Available now is the X.Org FOSDEM 2010 Wiki along with Luc's announcement.
Phoronix will be providing live coverage from the event both written and video recordings. We have recorded X@FOSDEM for the past two years and then have hosted them at RadeonHD.org and Phoronix.com. We will be doing it again for 2010, but this time with HD video recordings and hopefully no audio problems like last year. That Saturday we might end up providing CoreBoot recordings and notes too, if there are no other talks going on of greater interest.
It's a bit sad though that X@FOSDEM is now a one-day event, especially considering there is only one other X conference per year (XDS) compared to in year's past when there was the two-day X@FOSDEM, the X Developers' Conference, and then the X Developers' Summit.
| 5
| 1,760,738,439.915992
|
https://www.phoronix.com/news/Nzc3Nw
|
What Kind Of People Use X.Org's VESA Driver?
|
Michael Larabel
|
Yesterday we shared the first numbers from our 2009 Linux Graphics Survey that showed the open-source ATI driver is now more popular than ATI's official Catalyst driver. The full results from this survey that concluded last month will be published tomorrow, but in yesterday's graph that we showed there was also something else interesting: the X.Org VESA driver usage. Approximately 1.5% of the survey respondents were using the xf86-video-vesa driver.
This seems like a rather high percentage considering the xf86-video-vesa driver just implements the basic standard VESA core to provide very basic user mode-setting on most graphics cards and monitors, but half the time it won't even mode-set to the monitor's native resolution. The VESA driver does not implement any hardware acceleration besides support for a shadow frame-buffer (ShadowFB) on the CPU, does not support X-Video, there is no power management, and is basically only good for starting the X Server in effectively a fail-safe mode if you happened to break your proper X.Org driver or xorg.conf configuration and are need to get back into an X Server as you restore your system's graphics stack.
With interest as to what the 315 people are thinking that participated in our survey and just outright use this driver, we looked at just their numbers and are sharing a few of them today. The first item we looked at was for those using the VESA driver, what sort of hardware they stated they were using. This is where it gets even odder.
Were these just 315 people using old legacy hardware that has no maintained X.Org driver any longer? Actually, for a majority, this was not the case. In fact, 83% of the people that were said to be using the VESA driver were using either Intel, NVIDIA, or ATI/AMD hardware! The other 17% were using SiS / XGI, VIA, Matrox, or something else. When it came to the percentages of the big three vendors, Intel was at 26%, NVIDIA was at 29%, and ATI was at 28%.
There should be no reason to pick the xf86-video-vesa driver over xf86-video-intel, which is Intel's main Linux graphics driver that continues to be certainly maintained and picks up new features almost quarterly. The xf86-video-intel driver supports all of Intel's current IGPs that are on the market, though some of the older Intel vintage hardware may no longer be supported or is broken, but still these numbers should be smaller. Granted, there is also Poulsbo, but at least a new driver coming.
For the 29% of the respondents using the VESA driver with NVIDIA hardware, again, it's rather odd. NVIDIA has their well-maintained binary-only driver that supports so many features and NVIDIA also has their three legacy drivers for supporting their older graphics hardware that no longer works with their latest mainline driver. For those not wishing to use NVIDIA's binary driver, there is NVIDIA's xf86-video-nv open-source driver that at least provides mode-setting and 2D acceleration. The xf86-video-nv driver may be obfuscated and crippled in terms of features, but it supports most NVIDIA hardware out there and most of the time at least mode-sets to a better resolution than xf86-video-vesa does with most monitors. Of course, the other option for NVIDIA users is using Nouveau for kernel mode-setting, 2D, and 3D acceleration (through Gallium3D). The Nouveau project also aims to support every NVIDIA graphics card available.
Again, for the ATI hardware its even more of a mystery why anyone would be running the VESA driver. Unless you just purchased a Radeon HD 5750, Radeon HD 5770, or another Evergreen graphics card and are waiting on open-source support as you don't wish to use the Catalyst driver, there are both open and closed drivers available that are better than the basic xf86-video-vesa driver.
The ATI Catalyst driver supports the R600, R700, and R800 (Evergreen) GPUs while their older Catalyst 9.3 driver supports the R300/400/500 series with distributions running older kernels and X Servers. This driver provides a near feature parity to the ATI Windows driver and allows the customer to use their hardware to pretty much the fullest extent. On the open-source side there is then the xf86-video-ati and xf86-video-radeonhd DDX drivers along with the ATI kernel mode-setting support found in recent kernels. For the open-source ATI drivers there is 2D, X-Video, and 3D support (currently through Mesa but Gallium3D is on the way). The open-source ATI stack is quite nice and only continues to get better.
For the other 17% of the VESA-using survey participants, well, VIA is sort of toying with their Linux drivers at the moment and for those with any Volari hardware the XGI Linux driver is dead. Assuming those VESA users are using their Linux systems for a desktop, they should just read our graphics card reviews and buy a new graphics card so you can fully utilize your Linux desktop rather than being impaired by a graphics driver that is really only good for system recovery. Or at least bug some X.Org driver developers to try to get a hardware driver to at least function.
Other interesting notes from the VESA-using population is that most of them were on X.Org 7.4/7.5, so it's not like they are running some old Linux distributions where the new drivers are not available or will not build. To no surprise, most of those using the X.Org VESA driver were running a resolution of 1280 x 1024 or less.
If you actively use the VESA driver on your system, please stop by the Phoronix Forums and let us know why.
| 70
| 1,760,738,439.92563
|
https://www.phoronix.com/news/Nzc2Mw
|
The X.Org Plans For Moving Away From HAL
|
Michael Larabel
|
To address the questions that have been coming up frequently regarding the X.Org Server and the plans to stop using HAL, Sun's Alan Coopersmith has created a new Wiki page called XorgHAL.
On this Wiki page, Alan explains how this hardware abstraction layer for Linux and other operating systems is currently in use for finding input devices, being notified of input device hot-plugging, mapping system input devices, and setting input device options. X.Org has been using HAL since X Server 1.4 and it continues to be used in the most recent X Server 1.7 series, but the migration away from HAL will likely be completed in time for X Server 1.8. This next version of the X.Org Server is planned for release in March.
In replace of HAL in the X Server will be a lot more of OS-specific code to the operating system's respective libraries for device enumeration and device notification. On Linux this means connecting a lot of libudev directly into the X Server. Instead of the input device options being stored in HAL FDI files, the X Server will begin supporting a xorg.conf.d directory for handling input driver options. These new files will utilize a new syntax for matching the options with the respective devices. While the X Server will be picking up this directory support, editing the xorg.conf will remain supported and a valid option.
This code is not yet in place, but it's the current plans for dropping HAL from the X Server for input devices, since the upstream project itself is stopping development in favor of supporting DeviceKit and other projects.
| 29
| 1,760,738,440.578739
|
https://www.phoronix.com/news/Nzc2Ng
|
X Server 1.7.3 Released
|
Michael Larabel
|
It was just one week ago that X Server 1.7.2 was released, but X Server 1.7.3 is now available today, which is far earlier than expected as it was not anticipated until the end of December. However, this X Server 1.7.3 release is out early because of two show-stopping issues that were found in the week-old 1.7.2 release.
X Server 1.7.3 corrects a break in the build system and the second bug is an ABI compatibility problem. X Server 1.7.4 will then end up being what 1.7.3 was supposed to be, about five weeks from now.
The X Server 1.7.3 release announcement can be read on the xorg mailing list.
| 0
| 1,760,738,440.587277
|
https://www.phoronix.com/news/Nzc1MA
|
Our 2009 Linux Graphics Survey Ends Tonight!
|
Michael Larabel
|
Our 2009 Linux Graphics Survey, which annually occurs to help and better inform both developers and end-users of the Linux desktop, ends tonight! Right now we are just shy of 14,000 submissions for this year's survey that has been running for the month of November. Take the survey now if you have not already done so. Results will be published in approximately one week.
| 2
| 1,760,738,441.08589
|
https://www.phoronix.com/news/Nzc0Nw
|
VIA's DRM Patches Finally Get Some Feedback
|
Michael Larabel
|
A month ago we shared that VIA tried again to push some new DRM code into the mainline Linux kernel. This was months after VIA Technologies had already tried multiple times pushing new Direct Rendering Manager code for its hardware into the kernel, but failed for various reasons. With this latest attempt, the patches received no comments nor were they accepted into the mainline tree. However, this morning they have finally received some comments.
VMware's Thomas Hellström who previously wrote a new VIA 3D Linux stack was the one that commented on this newest set of patches from VIA. Thomas ended up acknowledging one of the patches that fixes a video hang issue (1), expressed four key problems with the VIA DRM patch that adds suspend/resume support through ACPI (2), and then the third patch also had technical issues with it (3).
With the Linux 2.6.33 kernel merge window soon opening, it's unlikely we will see any revised patches from VIA up for review again prior to Linux 2.6.34.
| 16
| 1,760,738,441.757989
|
https://www.phoronix.com/news/Nzc0NQ
|
X Server 1.7.2 Released
|
Michael Larabel
|
Just as planned, X Server 1.7.2 was released this morning to encompass all of the bug-fixes that have went into the X Server 1.7 branch over the past five weeks. The X Server 1.7.2 release message can be read on xorg-announce. Peter Hutterer is planning for an X Server 1.7.3 release in five weeks.
The next major X Server release will be version 1.8 that is currently scheduled for release in late March.
| 5
| 1,760,738,442.334839
|
https://www.phoronix.com/news/Nzc0Ng
|
Plymouth Gets Tighter Integration With GDM, X
|
Michael Larabel
|
For about two years now Red Hat has been working on the Plymouth project to replace RHGB with this graphical boot program that leverages kernel mode-setting and other newer Linux innovations to provide a clean, flicker-free boot experience. Over the course of the past few Fedora releases, Plymouth has continued to pickup new features and is also now being used by Mandriva. While Plymouth already does a great job at mode-setting to the display's native resolution and then showing the selected Plymouth plug-in and then to switch over to GNOME's GDM quite smoothly as the X.Org Server starts up, this process is getting even smoother now.
Committed to the Plymouth master code-base and a branch of GDM is support for very nicely handing off control of the display(s) between Plymouth and GDM. With the Plymouth code found in Fedora 12 and earlier, it sort of assumes and hopes that the X Server starts up nicely from where Plymouth left off, but if that doesn't go smoothly, the user is left in a troubled state. Now though this has changed where Plymouth receives a deactivation signal, which rather than just killing Plymouth, it leaves the plug-in idling on the screen until the X.Org Server takes over and is properly initialized. The GDM then signals to Plymouth when to properly shut itself off and is switched over to showing the X Server.
The new work that hit the Plymouth tree can be found in this Git commit, which was previously part of Plymouth's no-fbcon branch. The Plymouth integration work for the GDM can be found in the plymouth-integration branch. Lastly, Red Hat's Ray Strode who has been the principal engineer of Plymouth has written a blog post covering this nice Plymouth-to-X hand-off process.
| 31
| 1,760,738,442.343451
|
https://www.phoronix.com/news/NzcyMw
|
X Server 1.7.2 RC2 Released With 15 Fixes
|
Michael Larabel
|
X Server 1.7.2 is scheduled to be released a week from Friday, so in preparations for that, Peter Hutterer has just pushed out the second release candidate for this minor point release. X Server 1.7.2 RC2 is made up of 15 fixes (mostly for XQuartz) since the RC1 release that arrived earlier this month.
There's nothing real exciting in this release beyond addressing bugs as most developers are now focused on X Server 1.8, which will be released in March. The X Server 1.7.2 RC2 release announcement with a link for the source download can be read om xorg-announce.
| 1
| 1,760,738,442.838131
|
https://www.phoronix.com/news/Nzc0Mg
|
The X.Org Survey Is Ending Soon
|
Michael Larabel
|
Our 2009 Linux Graphics Survey is ending at the end of the month. If you haven't already taken the survey, we would encourage you to do so now. The results from this survey will help both developers and end-users of the Linux desktop. It should only take a few minutes to complete, so why not take it now?
| 0
| 1,760,738,442.846804
|
https://www.phoronix.com/news/NzY3NQ
|
XvMC Comes To xf86-video-unichrome Driver
|
Michael Larabel
|
Earlier this year Luc Verhaegen, one of the key contributors to the RadeonHD graphics driver, was laid off from Novell after a round of cutbacks at their German facility. While remaining unemployed, Luc has contributed to the CoreBoot project with ATI graphics card flashing support and native VGA text mode support, among other work. Additionally, he continues to dabble with his own open-source VIA driver, xf86-video-unichrome.
This driver, which competes alongside xf86-video-via and xf86-video-openchrome, now offers XvMC VLD support. However, it's not a traditional implementation of X-Video Motion Compensation, but rather Luc ended up designing his own X extension in the X protocol for sending MPEG slices to the hardware. This was done due to not liking the design of traditional XvMC implementations, which he describes in this blog post.
In the xf86-video-unichrome repository is this XvMC addition for the VT3122 and VT3108 ASICs with MPEG2 slice decoding support. XvMC is also supported by the OpenChrome driver.
| 21
| 1,760,738,443.456457
|
https://www.phoronix.com/news/NzY3OQ
|
X Server 1.7.2 Is A Step Closer To Release
|
Michael Larabel
|
At the end of October X Server 1.7.1 was released and at that time Peter Hutterer had said that the 1.7.2 build should arrive in about five weeks. It hasn't been five weeks yet, but it looks like he's still on track to delivering this bug-fix release on time or earlier. Peter has just announced the first pre-1.7.2 development release.
The xorg-server 1.7.1.901 release carries 18 changes atop the 1.7.1 release that can be found in X.Org 7.5. These changes include various fixes for EDID, X Input, EXA, X Quartz, and other areas. Expect X Server 1.7.2 to be properly released with these changes (plus likely a few more) later on in November or early December.
Meanwhile, the bulk of the development work is being focused on X Server 1.8, which should be released in March of 2010.
| 0
| 1,760,738,443.46444
|
https://www.phoronix.com/news/NzY2Ng
|
VIA Keeps Trying For Kernel Inclusion Of Its DRM
|
Michael Larabel
|
Last December the Linux folks at VIA Technologies had released their Chrome 9 series DRM code, which is needed for Linux 3D support with these newer-generation VIA IGPs, but this initial version ended up getting rejected from inclusion into the mainline kernel on the basis of the rest of VIA's 3D stack for the Chrome 9 being closed-source and some problems with the code itself. The situation was similar to that of Intel's Poulsbo DRM being rejected from reaching the mainline Linux kernel earlier this year.
This July the Chrome 9 DRM was re-released with aspirations of getting it in the mainline Linux kernel, but it was virtually the same as December's version and it too got knocked down for inclusion on the basis of no open-source "clients" using this Direct Rendering Manager driver and security issues with the code itself.
In August there was then another new VIA 2D driver released, but unlike the various other VIA Linux drivers out there, this one actually uses the new DRM code. This makes the DRM code at least useful by an open-source 2D driver, but within this Chrome 9 DRM remains the 3D bits too, which is not used by an open-source code at this time nor has VIA even released any code samples to test the 3D portion of this DRM for sanity and security.
VIA though is now pushing once again for getting its Chrome 9 DRM in the mainline Linux kernel. Last week VIA's Bruce Chang sent in a request to the dri-devel mailing list and to the official kernel DRM maintainer, David Airlie, but it was met by no responses at all -- positive or negative. This VIA Unichrome DRM adds support for an ACPI suspend/resume issue, an interfaces for communication with ddmpeg and V4L, and a fix for a video hang issue caused by a verification function. With no responses, Bruce this morning pinged everyone again. No developers have yet to respond to this latest request, and it's quite up in the air whether this DRM will actually get pulled in this time, but we will continue to monitor the situation.
| 20
| 1,760,738,444.076235
|
https://www.phoronix.com/news/NzY0OA
|
X.Org Development Discussion Continues
|
Michael Larabel
|
In late September there was a call by Peter Hutterer for a new X.Org release process that consisted of a six-month release cycle for the X Server, all development work to be done in feature branches and not Git master, and a three-stage development cycle. The agreed upon version was pretty much the same as Peter's version, but it also called for the X.Org drivers to be pulled back into the X Server (around version 1.10).This process was immediately adopted with X Server 1.8 getting a release date just six months out, and the first X Server 1.8 snapshot already being available. So everything is all set, right? Well, a discussion regarding the X Server release and development process has continued for the past several days among key X.Org developers.
Many messages have been accumulating about this matter in this mailing list thread, but here's the summary. While a six month release cycle right now is good for the X Server, and its current schedule of new versions that would be out in March / September just in time for the seasons of Linux distribution refreshes, this may not work in the future. Once the input/video drivers are merged back into the X Server, some developers are calling for a three month release cycle of the X Server.
Moving the X.Org drivers back into the X Server -- even though they were merged out of the X Server a few years back -- is wanted so that invasive API/ABI changes can be carried out without causing hell on the user. Lots of dead, legacy code could be deleted from both the drivers and the X Server, without making it messy on distribution maintainers and end-users with having certain version requirements for the different components with each X Server release. Instead, users would just need to grab the latest X Server, which would include the compatible drivers and need not worry about any API/ABI breakage. Having the drivers in the X Server would also cause more users to test out the latest X Server development code, which isn't happening too frequently right now.
However, with an X Server + drivers combo, some -- including Keith Packard -- are saying that a six month release cycle is not frequent enough. Intel ships their X.Org driver updates quarterly and many of the other open-source X.Org drivers ship more releases than two a year too. With what the current model looks like now, a user could be waiting up to six months for a new driver that supports the latest graphics hardware, new features, bug-fixes, etc.
Depending upon the policies of the Linux distribution you are using, you may need to wait for an entirely new update to your distribution before the new X Server / drivers are pulled. Though as it stands right now, some distributions (such as Ubuntu) end up having older drivers and will not push down new major xf86-video-* driver releases until their next bi-annual release, which forces the users already to build from source or use a bleeding-edge package repository in order to gain new hardware support or important features. These users end up just seeing two driver releases per year already, though distributions like Fedora update their graphics drivers much more frequently.
When it comes to having an even shorter release cycle, there becomes the problem of getting the actual releases out the door and in a working, quality state. The X.Org project is strapped for developers as it is, while putting out for X Server releases per year would place an even larger burden on their shoulders. Beyond just getting the releases out there, with a short release cycle it also becomes a matter of how legacy and stable branches would be maintained and for how long, which again is more work on the X.Org developers, and decisions to be made about back-porting work in the X Server, etc.
Other ideas have also come up on the mailing list with only allowing API/ABI breakage on even versions of the X Server and even splitting out graphics drivers on a per-chipset/series basis. Instead of just having say the "radeon" driver module, there would be a separate driver for the R300, R500, R600, R700 series, etc. Albeit, much of the code between the different families is already shared.
Of course, the drivers could just not be merged back into the X Server, but then cleaning up the X Server and drivers becomes more difficult with API/ABI breaks. There's a few decisions to ultimately be made, and we'll keep tabs on this thread to see what comes about.
| 2
| 1,760,738,444.091499
|
https://www.phoronix.com/news/NzY1OA
|
Our Linux Graphics Survey Is Off With A Bang
|
Michael Larabel
|
Merely eight hours ago we launched our 2009 Linux Graphics Survey to collect some data about the popular graphics drivers and hardware being used by our Linux readers along with other metrics such as the common ways one goes about installing their driver, what X server is being used, etc. The results of this survey are interesting in their own right, but they also help developers better understand what their users are most interested in with regard to the Linux graphics stack / X.Org and provide other statistics. This is our third time running this survey after successful runs in 2007 and 2008, but for 2009 it's looking like we will have a record number of survey responses.
As of right now we just crossed 2,100 responses in the eight hours that this year's survey has been active. If this pace kept up, we could see well over 100,000 responses this year -- in comparison to over just 20,000 last year. Being conservative though, I think this year we will reach at least 50,000 results before the graphics survey ends at the end of November. This generated data should be especially interesting, which will be published on Phoronix in early December.
If you haven't already, take our 2009 Linux Graphics Survey.
| 27
| 1,760,738,444.600122
|
https://www.phoronix.com/news/NzY0Mg
|
A Hackfest To Improve Linux Video Playback
|
Michael Larabel
|
When it comes to video playback on Linux, the premiere choice for video acceleration is currently using VDPAU with its CPU-efficient, GPU-accelerated capabilities that even has no problems playing 1080p video files with extremely low-end hardware. However, VDPAU is not yet widespread in all Linux video drivers, and other free software developers have been working on improving other areas of the Linux video stack too. One of these developers is GNOME's Benjamin Otte who has been working on using Cairo/Pixman for raw video in GStreamer. Additionally, he has organized a Linux video "hackfest" that will take place next month in Barcelona, Spain to further this Linux video playback work.
Faced by high CPU usage with video decoding and just a poor experience overall when playing back video files with GStreamer, Benjamin has been hacking away at using Cairo/Pixman with GStreamer and providing better hardware acceleration for this open-source multimedia framework. His mailing list messages regarding the details of what exactly he is doing can be found here and here. Benjamin's work has come to fruition and has now vastly improved the desktop experience when watching videos in WebKit, after making changes to GStreamer, Cairo, Pixman, and the X Server. This work though hasn't yet hit the mainline trees nor is it complete, but this is the focus of the Barcelona video hackfest next month.
Developers from Cairo, GStreamer, Pixman, and Intel will be taking part in the video hackfest at Collabora's Barcelona offices. Also what will be discussed is using OpenGL or X Render for video handling, adding support for more image formats (YUV and un-multiplied RGBA) to Pixman/Cairo/X Render, improving SIMD acceleration in Pixman, integrating GStreamer threading better in OpenGL/X/Cairo, and discussing other hardware-accelerated video decoding such as what is offered by XvMC, VA-API, VDPAU, and XvBA.
More information on this work to improve Linux video playback can be found in this mailing list message, a blog post, and a Wiki page.
| 22
| 1,760,738,445.347868
|
https://www.phoronix.com/news/NzY0NQ
|
X.Org 7.5 Now Released
|
Michael Larabel
|
X Server 1.7 was released earlier this month thereby paving the way for the release of X.Org 7.5, which is finally out there this afternoon. X.Org 7.5 consists of X Server 1.7.1 and all of the latest drivers, libraries, and other modules.
The key features introduced since X.Org 7.4 last year are Multi-Pointer X, X Input 2.0, input device properties, RandR 1.3, and DRI2. If you're not already familiar with these new features, well, you are not reading Phoronix enough. The X.Org 7.5 release announcement can be read on the mailing list with additional information being available from the X.Org web-site.
The next release, X.Org 7.6, will likely come next October. The next X Server release, X Server 1.8, is planned to come out in March.
| 12
| 1,760,738,445.856789
|
https://www.phoronix.com/news/NzY1Ng
|
DRI2 Sync + Swap Extensions Near Reality
|
Michael Larabel
|
When running a modern Linux graphics driver stack in a composited environment there is a lot less tearing -- particularly with regard to video playback, but OpenGL applications too -- now than there was in the past, but there is still room for improvement. One of the ways to improve this is by properly controlling the display of buffers with how often the swaps occur and to sync them with the monitor's refresh rate or the rate at which the compositor is running. Unfortunately, the X Server and open-source graphics stack have lacked such functionality even though some OpenGL/GLX extensions for this support have been around for years. Intel's Jesse Barnes though has been working on proper DRI2 support for synchronization and swapping.
Jesse has added support to Mesa and the X Server that will allow for the appropriate OpenGL/GLX extensions that expose syncing and frame counting capabilities (SGI_swap_control, SGI_video_sync, SGIX_swap_barrier, etc). This code though is currently housed in his personal Git repositories as the patches await review and feedback (mailing list message). This DRI2 synchronization and buffer swaps code is also dependent upon revisions to dri2proto and libdrm along with the latest DRM code that will eventually be pushed into the Linux 2.6.33 kernel due to its vblank event support.
Besides cutting down on tearing and other visual defects, proper synchronization of swaps can provide a smoother experience with a consistent frame-rate and can enhance the performance of some applications. More information on the technical end of this support can be found on the composite swaps Wiki page. This support isn't yet living in Git master, but it's surely on its way, which will benefit Linux desktop users.
| 17
| 1,760,738,445.866129
|
https://www.phoronix.com/news/NzYzMw
|
X Server 1.7.1 Released, 1.7.2 In 5 Weeks
|
Michael Larabel
|
Just as planned, X Server 1.7.1 was released this morning by Peter Hutterer. This bug-fix update to X Server 1.7 contains a fair amount of fixes and other minor improvements since the 1.7.0 release that came less than a month ago. Compared to X Server 1.7.1 RC2 from earlier this week, there are some XACE updates.
Peter is planning on having X Server 1.7.2 be released in about five weeks with the first of those release candidates coming in two weeks. The X Server 1.7.1 release announcement can be found on the X.Org mailing list. X.Org 7.5 is planned for release later today.
| 0
| 1,760,738,446.475364
|
https://www.phoronix.com/news/NzYyNA
|
X.Org 7.6 May Be Released In October 2010
|
Michael Larabel
|
On the same day of the first X Server 1.8 snapshot being released, Sun's Alan Coopersmith who has been leading the X.Org 7.5 katamari release efforts has shared his thoughts going forward for X.Org 7.6. X Server 1.7.1 should be released by week's end and X.Org 7.5 final should be out shortly thereafter. However, it's not likely that X Server 1.8 will be greeted by X.Org 7.6.
Alan is suggesting that X.Org 7.6 targets X Server 1.9. If the new X Server release process is followed, X Server 1.9 should be released in September or October of 2010. This would be just under a year since the release of X.Org 7.5 / X Server 1.7. X.Org 7.6 may also be the first release to carry a hard requirement against XCB.
The details regarding these very early X.Org 7.6 plans can be found on the xorg-devel mailing list.
| 1
| 1,760,738,446.483404
|
https://www.phoronix.com/news/NzYyMQ
|
The First X Server 1.8 Snapshot Released
|
Michael Larabel
|
A month ago there were proposed development changes to the X Server to utilize more feature branches in Git, a three-stage development process, and to deliver predictable time-based releases. A version of this proposal has been agreed upon by key X.Org developers and it's already being applied towards the X Server 1.8 work. X Server 1.7 was released less than a month ago and X.Org 7.5 isn't even out yet, but the first development snapshot for X Server 1.8 is now available.
Keith Packard, who is the release manager for the X Server 1.8 branch and in control of the X Server Git master branch for the next six months per this new release process, declared xorg-server 1.7.99.1 this morning. This first development snapshot has 137 changes from 28 different developers. While no major features have yet to be merged for X Server 1.8, some of the key changes in this first snapshot include fixed D-Trace support on Mac OS X, XSELinux and XACE updates, GLX 1.4 enabled on DRI2, many XQuartz changes, EXA fixes, and even a fix for the Hurd operating system.
The release announcement with source download links for this X Server 1.8 snapshot and the change-log can be found on the xorg mailing list.
The X Server 1.8 release schedule puts the official 1.8.0 version out in March of next year. Among the expected features are XKB2, improved EXA performance, X Input 2 updates, and other new features.
| 7
| 1,760,738,446.980514
|
https://www.phoronix.com/news/NzYwNA
|
X Server 1.6.5 & 1.7.1 RC1 Released
|
Michael Larabel
|
Less than two weeks after X Server 1.7.0 was released, Peter Hutterer has announced the first release candidate for X Server 1.7.1. This release carries a handful of fixes, including addressing some important issues like potential memory corruption and infinite loops, along with build improvements. Peter is asking anyone who still wants to submit patches for X Server 1.7.1 that they be submitted by this Friday. The release announcement for X Server 1.7.1 RC1 can be read here.
Minutes after the X Server 1.7 RC1 announcement, Keith Packard released X Server 1.6.5. This stable release makes it so that the X Server no longer crashes when a driver tries to initialize (the broken) DGA and there is also one Xephyr improvement. The X Server 1.6.5 announcement is also on the mailing list. With X Server 1.7 being out there, Keith has no plans at this time to work on a X Server 1.6.6 release.
| 1
| 1,760,738,446.988559
|
https://www.phoronix.com/news/NzYxOQ
|
X Server 1.7.1 Release Candidate 2
|
Michael Larabel
|
Shipping with X.Org 7.5 when its to be released shortly will be X Server 1.7.1, which is the first point release in the X Server 1.7 series that introduced X Input 2.0, Multi-Pointer X, and other new capabilities. For some last minute testing before going gold, X Server 1.7.1 Release Candidate 2 was made available over the weekend.
This release candidate carries some fixes that will make the X Server work again on the Alpha architecture, patches for Xming and Xwin, and fixes for other bugs. The final release of X Server 1.7.1 is expected this Friday.
Check out the release announcement for more details.
| 1
| 1,760,738,447.328958
|
https://www.phoronix.com/news/NzYxOA
|
X.Org 7.5 Release Candidate 1 Is Now Available
|
Michael Larabel
|
Back on the 2nd of October, X Server 1.7.0 was released after several delays with getting this major, MPX-bearing update released. However, the X.Org 7.5 release of all of the updated X packages was not released at the same time as X Server 1.7, in fact, it's still not released. Early this morning though the first release candidate of the X.Org 7.5 packages were made available.
Since the release of X Server 1.7.0 there has been a flow of X package updates that did not have stable releases available that incorporated the bug-fixes and new features that accumulated in many of the packages since the X.Org 7.4 release more than a year ago. Most of the X.Org packages are now ready for the 7.5 release, but there are still a few more expected before X.Org 7.5 is officially released. Additionally, X Server 1.7.1 will be released in time for X.Org 7.5.
The X.Org 7.5 Release Candidate 1 announcement was made by Sun's Alan Coopersmith on the X.Org mailing list. X.Org 7.5 should be out within a few weeks.
| 5
| 1,760,738,447.524318
|
https://www.phoronix.com/news/NzU5Nw
|
X Server 1.8.0 Gets A Release Date
|
Michael Larabel
|
X Server 1.7 was released one week ago, but the release schedule for X Server 1.8 has now been decided by its release manager, Keith Packard. According to what was just committed to the X.Org Wiki, X Server 1.8 is planned for release on the 31st of March, 2010. In order to meet this release date, the end of the merge window is 31st of December and the end of the bug-fix window is on the 28th of February.
X Server 1.8 is going to be the first release that follows the new X.Org development process. Last month prior to the X Developers' Conference, Peter Hutterer called for a new development process in which each developer must now develop their code within their own branch and not the X.Org Server master branch, where as -- similar to Linux kernel development -- the release manager must pull in your code. Further following the Linux kernel development process, the X.Org process now has a feature merge, bug-fix, and release freeze stages. Lastly, Peter called for predictable time-based releases on a six month cycle. The agreed upon version of this release process also calls for X.Org drivers to be merged back into the X Server core by the X Server 1.10 release (to come in 2011). This new release process is coming after numerous failed attempts of delivering an X.Org / X Server release on time, but ideally this new process will put them in line for delivering a quality release on schedule.
If this release is made on time, X Server 1.8.0 will also be released on the same day as the planned GNOME 3.0 release; what an exciting day! The Ubuntu 10.04 LTS release schedule has the feature freeze for the Lucid Lynx closing a month before the X Server 1.8 release, but hopefully the development will be on track that a development release gets pulled into Ubuntu 10.04 prior to the end of February. This X Server 1.8 release plan should allow it to land with sufficient time for Fedora 13 and others landing in mid-to-late Q2'2010.
X Server 1.8 should incorporate XKB2, offer improvements to X Input 2, improved EXA performance, and more.
| 1
| 1,760,738,447.942002
|
https://www.phoronix.com/news/NzU5Mg
|
New OpenChrome Driver Release (v0.2.904)
|
Michael Larabel
|
It's been a while since the last OpenChrome driver release for VIA hardware (14 months), but in order to offer compatibility with X.Org 7.5 / X Server 1.7, the team quickly rushed out a new release.
Some of the improvements made to this VIA DDX driver that competes with xf86-video-via and xf86-video-unichrome include VX800 support, basic VX855 support, CX700/VX800 integrated TMDS support, ARGB cursor support, and oanel mode-setting code rework. There's also been assorted bug-fixes and other work piling up in this driver over the past year too.
The release announcement for the xf86-video-openchrome 0.2.904 driver can be found on the X.Org mailing list.
| 2
| 1,760,738,448.740823
|
https://www.phoronix.com/news/NzU4MA
|
Getting X.Org 7.5 Out The Door
|
Michael Larabel
|
X Server 1.7 was released yesterday, but the X.Org 7.5 release containing this server package plus other updated X bits was not done at the same time. While it's less important having a new official X.Org release immediately since all of the critical packages can already be downloaded independently, the X.Org 7.5 release is expected soon and Sun's Alan Coopersmith began a thread entitled finishing the X11R7.5 katamari.
There are a number of packages since the X.Org 7.4 release last year that have since been abandoned and/or are no longer maintained and then there are some other packages that haven't had new releases, which is needed for X Server 1.7 compatibility. There's also a nice chart comparing the package versions between X.Org 7.4 and 7.5. For those interested, check out his mailing list message for all of the details. Expect X.Org 7.5 to make it out the door in the coming days.
| 0
| 1,760,738,449.230224
|
https://www.phoronix.com/news/NzU3OQ
|
X Server 1.7 Released With XI2/MPX Glory
|
Michael Larabel
|
It's late, but it's finally here. X Server 1.7 has been released. This major X Server update finally delivers on X Input 2.0 and Multi-Pointer X support along with other input improvements, cleaned up XKB code, VGA Arbitration support, EXA improvements, XDMX support, working XQuartz for Mac OS X, and many other changes.
The lengthy mailing list release announcement detailing this work can be found on X.Org. X Server 1.8 should come in about six months.
| 9
| 1,760,738,449.249092
|
https://www.phoronix.com/news/NzU3Ng
|
New X.Org Release Process Has Been Reached
|
Michael Larabel
|
Last week we talked about a new X.Org release process proposal for improving the consistency and quality of X Server releases through taking a number of relatively simple steps. Well, this week from XDS2009, a revised proposal has been agreed upon now making it policy for X Server 1.8 / X.Org 7.6 and later.
With this new process, there will be consistent six-month releases that should be very predictable. This is going to begin next Monday once X Server 1.7 is released, so hopefully X Server 1.8 will make it out in March, which would allow it to be pulled into the Q2'2010 round of distribution updates.
As part of this process, the master X Server Git repository is no longer open to all, but is only to the release manager. All X.Org developers should have their own X Server branch(es) and then make pull requests to the release manager if they want code to touch the mainline code-base.
The developers also want to pull the X.Org drivers back into the X Server core, which was done previously before their modularization, but they want to put the drivers back into core in order to come back to a more coherent API. This though will not be happening until early 2011 or so (around X Server 1.10).
Daniel Stone concludes his mailing list message detailing this new policy with, "Phoronix: Yes, we're really trying this time." Bravo! Let's hope this new policy really becomes a reality as these changes will surely benefit quite a number of users, distribution vendors, and developers.
| 21
| 1,760,738,449.761878
|
https://www.phoronix.com/news/NzU3MQ
|
X Developers' Conference 2009 Notes
|
Michael Larabel
|
Last year the X.Org developers raided the Edinburgh Zoo for their annual summit and earlier this year there was FOSDEM for the usual X.Org event in Brussels, while this week is the 2009 X Developers' Conference taking place in Portland, Oregon. While we are not providing live coverage (or audio/video recordings) from this event like those in the past (European events are the best and there happens to be a collision with Munich's Oktoberfest), the X.Org Wiki is being updated with (short) notes from the event.
The notes from XDC2009 are available on this Wiki page. As of right now, the topics to be covered so far today include reviewing the Google Summer of Code work, an X.Org Foundation board talk, and rootless X with Composite.
We will be monitoring this page and reporting any interesting items that may come out of the conference this week, which runs through Wednesday. The talks coming up later today include debugging Gallium3D drivers, X.Org 7.6 planning, and power management in X. Coming up tomorrow are talks on Mesa and OpenGL 3.x, GL extension for memory managers / kernel mode-setting / FBOs, OpenGL shaders for Shader Model 3.0/4.0, synchronization and presentation, future of X.Org DDX drivers, and ending off the day is a talk on the Wayland Display Server.
Ending out XDC2009 on Wednesday includes talks on the PCI rework (VGA Arbitration), an XACE update, why X.Org sucks and what can be done, X on embedded platforms, "why can't I bisect", and X on alternative tool-kits.
| 5
| 1,760,738,449.770844
|
https://www.phoronix.com/news/NzU2OA
|
X.Org 7.5 Gets Closer While X Server 1.6.4 Is Out
|
Michael Larabel
|
Just in time for the start of the X Developers' Conference this week in Portland, there are two new X Server releases to report. X Server 1.6.4, which is a bug-fix release in the X Server 1.6 series for X.Org 7.4, has been released. Additionally, X Server 1.7 RC3 has been released, which is the final release before X Server 1.7 and X.Org 7.5 are to be released.
X Server 1.6.4 brings seven fixes, which is slightly less than some of the earlier bug-fix releases, but it's an improvement nevertheless and can be used by the distribution vendors that will be shipping with X Server 1.6 in their next round of updates due to the late X Server 1.7. None of the bug-fixes stand out in particular, but the 1.6.4 release announcement can be found on the mailing list along with the source download links.
X Server 1.7 RC3 is the third and final test release for this new X Server that brings Multi-Pointer X, X Input 2.0, VGA Arbitration, and other new features. There are quite a number of bug-fixes in this release candidate compared to RC2. As a small note, X Server 1.7 will actually be the first release in quite a while where there is a usable XQuartz server from an official X.Org release. This release announcement can be read here.
Peter Hutterer expects to release X Server 1.7 final with X.Org 7.5 within the next week pending any changes to the blocker bugs. Cheers (or rather, Prost!) to this hopefully being the last much-delayed X Server release, thanks to the proposed process changes currently being discussed for X.Org Server development.
| 1
| 1,760,738,450.262856
|
https://www.phoronix.com/news/NzU2Mw
|
Proposed Process Changes For X Server 1.8
|
Michael Larabel
|
While the X.Org developers are responsible for a lot of critical code and much of it is quite old and massive, they are often challenged by hitting a release on time and often face multiple release schedules before coming close to delivering a new X Server / X.Org release. Just take a look at X.Org 7.4 Finally Released, X.Org 7.5 Released. Wait, Nope!, or X Server 1.4.1 Is Released, No Joke as recent examples. With the X Server 1.7 / X.Org 7.5 release cycle that we are finally getting close to ending, it took multiple tries and is coming out over six months later than expected. Granted, there is Multi-Pointer X support and other new improvements, but concerns about the degrading quality of releases have been voiced for years. Fortunately, this situation may finally turn around.
Peter Hutterer, who has been working on the X.Org input code for some time and is the developer behind MPX, late into the X.Org 7.5 release cycle decided to step up to the plate and personally get this important X.Org update out the door. Peter is not stopping after X Server 1.7, but has already made a proposal regarding the release and development changes going forward for X Server 1.8 and into the future.
Peter acknowledges the current key problems with the X.Org release schedule being unpredictable, Git master is unusable at times, and there is a late testing cycle. Mr. Hutterer is proposing three fundamental changes and they include the use of feature branches, three stages to the development cycle, and predictable time-based releases. With regard to feature branches, Peter proposes that all work that is large and potentially disruptive to the X Server stability should first be done in a branch of the X Server rather than in Git master. Once the branch is developed, tested, and fairly sane it could then be pushed along into the master branch. Peter cites his Multi-Pointer X development work last year as a good example of a feature branch working out well.
The three stages of the proposed development cycle for X.Org would be feature merge, bug-fix, and release freeze. The merge window for the X Server would be open for about three months at a time followed by two months of bug-fixing and then a one month period for the X Server to be frozen and well-tested prior to a release. This would allow consistent six-month releases of the X Server / X.Org. Albeit with different time lengths, this would put the X.Org development process much closer to that of the Linux kernel. Feature branches ready for the next release would be merged into Git master during that time, followed by bug-fixing afterwards, and then at the end it's time to cut the release and then branch the X Server. The X Server would not be branched in advance of a release, but rather afterwards and prior to the next release's merge window opening.
Lastly, Peter calls for predictable time-based releases and not to delay any releases based upon half-complete features. As part of getting X.Org released on time, he also calls for monthly snapshots of Git master for easier testing and then towards the end of the cycle to provide weekly or bi-weekly snapshots until the release candidates are pumped out. Having predictable time-based releases would please quite a lot of people, especially the distribution vendors that have been challenged in the past by deciding whether they can switch to the next X.Org / X Server in their development cycle or whether delays will jeopardize the move and postpone it to their next release.
Peter's mailing list message that outlines this proposal for X Server 1.8+ can be read on xorg-devel. As of right now there are no replies to his thread, but hopefully that will change and a version of this proposal will become policy for X.Org going into the future.
| 21
| 1,760,738,450.271855
|
https://www.phoronix.com/news/NzU0OA
|
X Server 1.7 Release Candidate 2
|
Michael Larabel
|
Just about a week after X Server 1.7 RC1 made it out, a second release candidate for this next X.Org Server is available. This second release candidate brings bug fixes and other improvements, but the code is currently within a feature freeze. Expect a third release candidate next week.
The change-log along with source download links for this X.Org 7.5 server can be found on the X.Org mailing list.
| 4
| 1,760,738,450.816812
|
https://www.phoronix.com/news/NzU1MQ
|
VIA's Improved Framebuffer Driver In Linux 2.6.32
|
Michael Larabel
|
While VIA's Chrome 9 DRM has yet to be accepted into the mainline Linux kernel since its mostly used by VIA's binary-only driver and then recently an updated 2D driver, with the viafb driver outside of X.Org, this frame-buffer driver has picked up many improvements with the Linux 2.6.32 kernel.
The viafb driver in the Linux 2.6.32 kernel has rewritten 2D engine support to offer improved support along with added support for the VIA VX800 Chipset too. In another commit is support for the VX855 hardware in this kernel frame-buffer driver. There's also a number of other VIA-related commits, but these are the main changes that will be found in the Linux 2.6.32 kernel.
Of course, this graphics related change isn't as prominent as the VGA Arbitration code, KMS page-flipping ioctl, Intel improvements, or R600/700 KMS and 3D support that all can be found in this next kernel release.
| 1
| 1,760,738,450.825376
|
https://www.phoronix.com/news/NzUzNA
|
X Server 1.7 Finally Reaches Release Candidate
|
Michael Larabel
|
It's a few months late, but the first release candidate for X Server 1.7 is now available. Peter Hutterer announced the release this morning along with freezing the X Server ABI for the 1.7 series. A X Server 1.7 snapshot was made available earlier this month, but a few more bug-fixes can be found in today's release candidate.
Peter is planning at least a second release candidate before X Server 1.7.0 is released and then the release of X.Org 7.5. Notably, X Server 1.7 brings support for Multi-Pointer X.
The X Server 1.7 RC1 release announcement with download links can be found on xorg-announce.
| 6
| 1,760,738,451.310463
|
https://www.phoronix.com/news/NzU0Nw
|
DRM Changes For Linux 2.6.32 Kernel
|
Michael Larabel
|
David Airlie has just called for the first DRM pull request for the Linux 2.6.32 kernel. To no surprise, R600/700 3D support can be found within the Direct Rendering Manager code as with kernel mode-setting support for these newer ATI Radeon graphics cards.
Also found within the Linux 2.6.32 DRM is the much-worked VGA Arbitration code, the KMS page-flipping ioctl, Radeon TV-Out, and the Intel DRM code now supports dynamic clocking for improved power management.
The list of highlights for the Linux 2.6.32 DRM so far can be found on the dri-devel mailing list. Sadly, it looks like the Nouveau kernel mode-setting DRM will not be ready for this forthcoming kernel.
| 27
| 1,760,738,452.008957
|
https://www.phoronix.com/news/NzUyMg
|
Kudos To Peter Hutterer With X.Org 7.5
|
Michael Larabel
|
X.Org 7.5 with X Server 1.7 will be arriving months late once it's released after failing to meet the original April release schedule and then failed twice with two more proposed releases during the summer. However, the latest release schedule, which puts the final release in late September or so, might actually work out this time -- in good part thanks to Peter Hutterer.
Peter Hutterer is the university student from Australia that developed Multi-Pointer X (to be found in X Server 1.7) and is responsible for a great deal of other work with X.Org when it comes to input and who now works for Red Hat. As of late, Peter has stepped up to handle many of the responsibilities with the X.Org 7.5 release and finally getting it out the door.
Peter had took care of issuing the first X Server 1.7 snapshot (and is working on getting X.Org 7.5 RC1 out shortly), pushed out new driver releases that he is responsible for, checking on the status of all X.Org drivers (mailing list message) to verify their X.Org 7.5 compatibility that in turn so far caused new snapshots of the Intel and ATI drivers, providing a large patch set that fixes DMX support after being broken for months (mailing list message), and has done a lot of other X.Org work lately too.
Peter's working on code and at the same time stepping up to see that X.Org 7.5 / X Server 1.7 makes it out into the world. Kudos to his work and the work of all other X.Org contributors.
| 10
| 1,760,738,452.514019
|
https://www.phoronix.com/news/NzUyNw
|
The X.Org 7.5 Super Module
|
Michael Larabel
|
X.Org 7.5 with X Server 1.7 will hopefully make it out into the world later this month and to provide for easy testing of these new X packages, Peter Hutterer continues his work by providing an X.Org 7.5 Super Module.
The X.Org 7.5 Super Module is simply a Git repository that contains all of the latest X packages going into X11R7.5. Among the many packages are X Server 1.7, xf86-video-intel 2.8.99.901, xf86-video-nv 2.1.15, xf86-video-ati 6.12.4, and libdrm 2.4.13. Many old and obsoleted packages have also been removed from X.Org 7.5.
Look for X.Org 7.5 with its Multi-Pointer X support and other new features and fixes to arrive later this month. However, outside of Fedora 12, most desktop Linux distributions will not be pulling these new packages into their stable releases until their 2010 updates.
| 3
| 1,760,738,452.522608
|
https://www.phoronix.com/news/NzQ5OA
|
X.Org 7.5 Is Frozen, XKB2 Gets Postponed Again
|
Michael Larabel
|
X.Org 7.5 with X Server 1.7 was supposed to be released earlier this month, but that didn't happen on schedule nor their earlier release plans. However, this evening, Daniel Stone has laid out the new plans for X.Org 7.5 / X Server 1.7.
The key points from Daniel's email are:
- XKB2 has been postponed, again. Peter Hutterer and Daniel are now looking at having version 2.0 of the X Keyboard Extension ready for X Server 1.8, which we anticipate will come in the first half of 2010. Fortunately, X Server 1.7 will still be home to X Input 2.0 with Multi-Pointer X support.
- The code for X Server 1.7 in the master branch is now frozen. Branching of X Server 1.7 will likely come in a few days.
- The first development snapshot of X Server 1.7 (v1.6.99.900) will be rolled over the weekend. A second release (v1.6.99.901) should be out shortly thereafter.
- Daniel is hoping for a final X.Org 7.5 / X Server 1.7 release around the time of the 2009 X Developers' Conference, which is taking place from the 28th to the 30th of September. Basically, we are now looking at a final release in what will hopefully be one month.
| 0
| 1,760,738,453.042357
|
https://www.phoronix.com/news/NzUwOA
|
X Server 1.7 Has Its First Snapshot
|
Michael Larabel
|
It's coming late, but it looks like we may finally see X.Org 7.5 in the next month. The biggest component making up X.Org 7.5 is X Server 1.7, which is finally getting underway with test releases after its code was frozen earlier this week.
Peter Hutterer this time around had announced the release of X Server 1.6.99.900, which is the first testing snapshot of X Server 1.7 and will be followed by a few release candidates. He has also codenamed X Server 1.7 the "Depressed Dodo".
Among the features that X Server 1.7 delivers is revised input handling, X Input 2 with support for Multi-Pointer X, VGA Arbitration support, EXA mixed pixmaps support, and then many bug fixes and other work.
The xorg-server 1.6.99.900 release announcement with the short change-log can be read on the mailing list.
| 6
| 1,760,738,453.051377
|
https://www.phoronix.com/news/NzQ4MQ
|
X Server 1.6 Lives On With Another Release
|
Michael Larabel
|
X.Org 7.5 that will bear X Server 1.7 has already been delayed multiple times. The latest release schedule promised it a week ago, but there hasn't even been a single pre-release or development snapshot to date. Originally X.Org 7.5 was supposed to be out in April, but it looks like we are still at least a few months from seeing this important update that introduces Multi-Pointer X and other new features and changes.
With these delays and lack of communication regarding a definitive timetable for the next X.Org / X Server release, distributions like Ubuntu and OpenSuSE are having to revert their plans to release their next round of distribution updates with these new X.Org components. With the X.Org 7.4 with X Server 1.6 user-base remaining large -- and will likely remain that way until H1'10 when X.Org 7.5 may finally start appearing in new distribution releases -- Keith Packard has gone ahead and is now preparing a new X Server 1.6 point release.
There have already been three point releases in the X Server 1.6 series, which is more than any other server series in recent times, but X Server 1.6.4 will now soon see the light of day. The first release candidate for X Server 1.6.4 is now available and it contains 11 fixes atop X Server 1.6.3.
The changes and source download links for X Server 1.6.4 RC1 can be found on the X.Org mailing list.
| 4
| 1,760,738,453.556027
|
https://www.phoronix.com/news/NzQ4Nw
|
VGA Arbitration Code Hits The X Server
|
Michael Larabel
|
The Linux VGA Arbiter, which has long been talked about and gone through several revisions, is finally approaching the point of entering the mainline kernel and as of last night was committed to the X Server and PCI access library. The VGA Arbiter fixes the problem where multiple graphics cards with legacy VGA interfaces could get sent the wrong data if there are multiple X Servers in use. The arbiter on the other hand is able to control which one is accessed and when.
The VGA Arbiter is ready to enter the Linux 2.6.32 kernel in early September once the Linux 2.6.31 kernel is out the door and the merge window for 2.6.32 opens. As a result, David Airlie pushed the VGA arbitration code that resides in the X Server into its Git master branch (commit), which will be part of X Server 1.7 once released. The libpciaccess Git master code has also received its needed changes.
This is particularly good news for those interested in a proper multi-seat system but have been affected by the lack of VGA arbitration.
| 4
| 1,760,738,453.570199
|
https://www.phoronix.com/news/NzQ3Mw
|
VIA Releases A New 2D Linux Driver
|
Michael Larabel
|
Earlier this month we shared that VIA would be releasing a new 2D graphics driver for Linux and this morning they have done just that. While previously VIA Technologies had thrown their weight behind the OpenChrome driver, and there are other VIA drivers out there like the UniChrome driver, they have been working on their own xf86-video-via driver. Back in August they had released a xf86-video-via driver, but today's release of this open-source driver now uses their new kernel DRM for providing 2D (EXA) acceleration.
This new VIA 2D driver can be downloaded here, but there really is no information about it on their Linux web-site. The release notes in the driver mention support for X.Org 7.4, ACPI suspend-to-ram and suspend-to-disk support, CRT/LCD/DVI support, dual display support, X-Video extension support, EXA 2D acceleration, and RandR 1.2 support.
The VIA chipsets supported by this driver include the Chrome 9 series, but specifically the CX700M, VX700, CN700, CN896, VN896, VX800, and VX855. This support from VIA is limited to 2D/video support and they have publicly said already that they will not release an open-source 3D driver for the Chrome 9 hardware even though they want a Gallium3D driver. VIA wants the open-source community to make a driver for them, but to date they still haven't released quite enough documentation to make that a reality and there would be a lack of development manpower.
The new xf86-video-via driver includes the kernel DRM source-code with the package too. VIA is actually shipping with six separate copies of the DRM source-code. There is separate Direct Rendering Manager code to support the CN700/CX700M/VX700 and CN896/VX800/VX820 ASICs. Additionally, there is separate DRM folders for the Linux 2.6.24, 2.6.27, and 2.6.28 kernels.
VIA will likely push for its Chrome 9 DRM to enter the mainline Linux kernel, but even with this 2D driver being out there, it's uncertain whether it will pass free software developer acceptance. Previously their DRM was rejected on the basis of partial open-source GPU drivers and at the time the only client of their Chrome 9 DRM was VIA's binary-only driver that they will not open-source. Now there is the xf86-video-via driver that uses this DRM, but still with it just using this kernel interface for the 2D acceleration, its only touching a subset of the code. Well, let's see how it goes this time.
| 9
| 1,760,738,454.157668
|
https://www.phoronix.com/news/NzQ3Nw
|
Xi2, Multi-Pointer X Support For FreeGLUT
|
Michael Larabel
|
Support for X Input 2.0 with Multi-Pointer X capabilities can currently be found in the latest Git master code for the X Server and related libraries, while it will appearing in desktop Linux distributions once there is the X.Org 7.5 release. To become more useful, Xi2/MPX support needs to be picked up by the different tool-kits and other libraries so that the applications can become more multi-pointer aware. There has been work underway on supporting Xi2/MPX in GTK+, but that work hasn't yet landed in a mainline tree. For those that use FreeGLUT for their OpenGL Utility Tool-kit Library, it's looking like it soon will receive proper Multi-Pointer X support.
Florian has shared on the X.Org mailing list his latest FreeGLUT patch that adds MPX/Xi2 functionality to this GLUT project. This looks like it will land in time for the release of FreeGLUT 2.6.0 in the near future.
| 3
| 1,760,738,454.166926
|
https://www.phoronix.com/news/NzQyNA
|
X Server 1.6.3 Released With More Fixes
|
Michael Larabel
|
With Keith Packard going on a week-long holiday, he decided to christen the X Server 1.6.3 release just before leaving. X Server 1.6.3 is now available and is a collection of minor fixes that have built up since X Server 1.6.2. The next feature release, which will bring items like Multi-Pointer X, is X Server 1.7 and that will be released in conjunction with X.Org 7.5, which is likely to come in September or October.
The link to download the source to xorg-server-1.6.3 along with the release announcement can be found on xorg-announce.
| 2
| 1,760,738,454.666737
|
https://www.phoronix.com/news/NzQ2NQ
|
KMS, GEM Comes To Linux Mobile Phone
|
Michael Larabel
|
The Neo FreeRunner that was developed by the OpenMoko project and manufactured by FIC, now has kernel mode-setting support and GPU memory management via the Graphics Execution Manager. The Neo FreeRunner smart-phone has a S-Media Glamo 3362 graphics accelerator and an independent developer decided to write the necessary kernel DRM, libdrm, and xf86-video-glamo DDX driver to introduce this support.
Up to this point kernel mode-setting has largely been talked about for Intel and ATI hardware (along with support continuing to emerge for NVIDIA hardware through the Nouveau project), but now KMS is running on this open-source smart-phone, which is complete with GEM memory management. If this isn't enough, a Mesa driver for the S-Media Glamo 3362 to provide OpenGL acceleration is also being worked on, but that driver is still very young.
The story of a GEM+KMS driver coming to the OpenMoko Neo FreeRunner can be found in this blog post.
| 11
| 1,760,738,455.401541
|
https://www.phoronix.com/news/NzQ0OA
|
VIA Soon To Release Its DRM-Using 2D Driver
|
Michael Larabel
|
Back in June, VIA Technologies rolled out its Chrome 9 DRM (for a second time) in hopes of pushing it into the mainline Linux kernel. At that time, VIA's DRM was again rejected and it led to a discussion over partial open-source drivers since the only user of this interface was their binary-only driver.
In that discussion back then, VIA's Bruce Chang shared that they had a new 2D driver in the works and that it would utilize this DRM. The driver was supposedly coming soon, even though they had been working with and using the OpenChrome driver before and even announced their love affair.
This morning Bruce has written on the dri-devel list that they should soon be releasing this new 2D driver. Bruce is hoping that this 2D driver's source-code will be released in about two weeks. Right now the driver is in the hands of a few community testers for feedback, but they are hoping that the driver will be released very soon unless any new issues are found.
Having a new open-source 2D driver for VIA graphics hardware will be good, but still doesn't mean we will see their DRM going into the mainline kernel. There is still the 3D portion of their (undocumented) DRM interface that isn't used by any open-source drivers nor even any simple tests. VIA has said they will not be providing any open-source 3D driver for the Chrome 9 series hardware. They would like the community to come up with a 3D driver, but they haven't released enough documentation yet, and no developers really seem interested at this time.
| 2
| 1,760,738,455.915531
|
https://www.phoronix.com/news/NzQ2Mg
|
The X.Org 7.5 Release Day, But With No Release
|
Michael Larabel
|
The latest X.Org 7.5 release plans placed the 17th of August as the release date for this next major X.Org update along with the release of X Server 1.7. However, to no surprise, this is not going to happen today. X.Org 7.5 / X Server 1.7 continues to be delayed without any revised release schedule.
X Server 1.7 was supposed to be branched more than a month ago, but that never occurred, nor was there any beta release. In effect, X.Org 7.5 / X Server 1.7 is currently running eight weeks behind schedule right now, but still, it's unlikely to see this release come together in eight weeks from now. This is in fact just the latest delay, with X.Org 7.5 originally being scheduled for release back in April and then July.
With X Server 1.7 running behind schedule it's already been rejected from Ubuntu 9.10 and from OpenSuSE 11.2. It will likely be dropped from other distributions refreshing this fall too. Once release and picked up by the major distributions, X.Org 7.5 / X Server 1.7 will bring Multi-Pointer X and many other changes to the Linux desktop.
X.Org 7.4 was released in September of 2008 (after facing delays too), while this February there was the X Server 1.6 release available as an update. With the way things are currently going for X.Org 7.5, we will likely not see its release before late October or November, which will put about 13 or 14 months between these major releases.
| 14
| 1,760,738,455.930738
|
https://www.phoronix.com/news/NzQxMw
|
X Server 1.6.3 RC1 Makes It Out
|
Michael Larabel
|
With X.Org 7.5 running behind schedule, X Server 1.6 may end up seeing a bit more light before it is replaced by X Server 1.7. If the X.Org 7.5 release schedule drags on much long, it could be in jeopardy for making the rounds of distribution updates this fall, which would leave X Server 1.6 being used for many until early next year.
X Server 1.7 introduces Xi2, Multi-Pointer X, and other features, but still X Server 1.6 has been pretty good release. Keith Packard, the release manager for this series, is in fact working on putting out another point release. On Sunday afternoon Keith had put out the first release candidate for X Server 1.6.3.
X Server 1.6.3 so far includes 12 fixes for different areas from RandR to EXA. The change-log with source download links for X Server 1.6.3 RC1 can be found on the X mailing list. X Server 1.6.3 will be released once none of the other developers have new patches they wish to see integrated in the server 1.6 branch.
| 4
| 1,760,738,456.5609
|
https://www.phoronix.com/news/NzQyMg
|
Nearly Two Dozen X.Org Drivers Get Updated
|
Michael Larabel
|
In time for the X.Org 7.5 release (whenever that may come), David Airlie has put out new driver releases for nineteen of the X.Org video drivers. These aren't updates to the mainline ATI/AMD, Intel, or even NVIDIA drivers, but some of the drivers for less common graphics hardware.
The X.Org drivers that were updated include those for APM, AST, Cirrus, Fbdev, Glint, i128, i740, Mach64, MGA, NeoMagic, r128, s3virge, S3, Savage, SiS USB, SiS, TdFx, VESA, and Voodoo hardware. The updates to these drivers range from simple bug fixes and support for xextproto 7.1 to other compatibility fixes.
Yesterday there was also an update released to the evdev driver as well as the VMware video driver for virtualized guest Linux operating systems.
The release announcements for each of these driver updates can be found from the xorg-announce July archives.
| 12
| 1,760,738,456.569481
|
https://www.phoronix.com/news/NzM5Ng
|
X.Org 7.5 / X Server 1.7: No Branching, No Beta
|
Michael Larabel
|
Last week we reported it looked like X.Org 7.5 would be released late, and sadly, we are now behind on two important milestones in the road to releasing X.Org 7.5 / X Server 1.7. X.Org 7.5 has already been challenged by a few delays and was supposed to be released in April, but it took a month until the release manager even released the release didn't happen. There were then two more release schedules, with the latest one calling for an August release.
Sadly, it doesn't look like X.Org 7.5 will be able to make it out in August. In order to hit an August 17th release date, X Server 1.7 was supposed to be branched from master back on the 6th of July. Over two weeks later, that has yet to even occur. Two days ago was the second X.Org 7.5 milestone, but that fell through too. Throughout this time there still is no revised release schedule available or any communication from Daniel Stone (manager of the X Server 1.7 schedule) regarding these delays.
On the 20th of July there was supposed to be the "Code freeze 1.7, release 1.6.99.901, 7.5 beta 1; no automatic updates beyond this point." Well, there is no X.Org 7.5 Beta 1 let alone any code freezes or stopping the automatic updates, considering that X Server 1.7 is still living in master.
X.Org 7.5 is just the latest release that has been challenged by multiple delays that end up spanning months. This problem has been going on for quite a while now due to having not enough developers for this critical piece of free software, and at the same time, X.Org has been undergoing radical changes.
Among the features that can be found in X.Org 7.5 once released are X Input 2.0, Multi-Pointer X, E-EDID support, and it will also be the first official X.Org release with RandR 1.3 support. Input master Peter Hutterer has been getting ready for this next release with several new input-related packages having been announced in recent days, including libXi, xextproto, libXtst, inputproto, and libXext.
Maybe we will see X.Org 7.5 in September?
| 14
| 1,760,738,457.195985
|
https://www.phoronix.com/news/NzQxMg
|
Proper Multi-Seat X Support Is On The Way
|
Michael Larabel
|
While multi-seat computing has been available on Linux for years, it's often been a chore to setup and required some time. Beyond just being time consuming and an unnecessary hassle, the way of setting up a multi-seat computer through an X Server with multiple nested Xephyr servers is not pleasant. There have been several attempts at improving the multi-seat Linux experience by creating a multi-seat display manager and taking various other steps, but to date this is still a challenge to setup. The good news though is that this may soon change.
Red Hat was wanting to deliver a good multi-seat implementation with Fedora 11, but that feature ended up being recalled and postponed to the next Fedora release, Fedora 12. Rather than mixing X.Org and Xephyr, a cleaner way of handling this would be to just start up multiple X Servers so that each seat would have its own X Server. If you do that right now though, you could run into some issues when you have multiple X Servers and multiple graphics cards, etc.
To work around this issue of multiple X Servers colliding with one another when talking to the graphics cards, there was a revival of the VGA Arbiter back in May. The VGA Arbiter was first designed nearly five years ago and its goal was to allow multiple X Servers to run with multiple graphics cards having the legacy VGA interface without causing any problems. Tiago Vignatti has now revised this VGA Arbiter work from earlier this year (along with the help of David Airlie) and he has now submitted a new VGA Arbiter implementation to the Linux kernel mailing list.
The VGA Arbiter hasn't made its way into the mainline tree yet and there are also some other patches that need to work their way upstream when it comes to the X Server and libpciaccess. However, once done, this should be a large step forward in improving the Linux multi-seat experience. Tiago has briefly talked about this work on his blog.
| 9
| 1,760,738,457.204492
|
https://www.phoronix.com/news/NzM4Ng
|
VIA Rolls Out Chrome 9 DRM, Pushes For Kernel
|
Michael Larabel
|
It has been a while since we last had any major to report on VIA with their open-source efforts, but this morning they have finally published DRM code that supports their Chrome 9 IGP hardware. The announcement regarding this new Chrome 9 DRM was made on the dri-devel list and was made up of three patches.
This new Direct Rendering Manager code from VIA amounts to about 5,000 lines of C code and about 1,500 lines of code for the header files. Of course, this DRM alone isn't too useful without a Mesa driver to take advantage of it. VIA previously expressed interest in a Gallium3D driver for the Chrome 9, but as of yet no code has been released for a Gallium3D driver or a traditional Mesa driver. It was back in January that Tungsten Graphics published a new DRM and Mesa driver for older VIA hardware.
VIA is looking to get this Chrome 9 DRM upstream within the Linux kernel, but some parts of it have already come under scrutiny by the DRM maintainer, David Airlie. However, we will likely see this new DRM kernel in the Linux 2.6.32 kernel at the latest. There are several different VIA chipsets that use the Chrome 9 integrated graphics processor and among them are the CN896, K8M890CE/K8N890CE, P4M900, and VN896.
| 2
| 1,760,738,457.716456
|
https://www.phoronix.com/news/NzM4NA
|
It Looks Like X.Org 7.5 Will Be Released Late
|
Michael Larabel
|
Last week we asked the question of whether the X.Org 7.5 release schedule will be met. This X.Org / X Server update has already been delayed multiple times and it is currently running behind with their current release plans. X Server 1.7 was supposed to be branched from their master code-base ten days ago, but as you can see, at the time of publishing this still hasn't happened. In just four days there is supposed to be the next X.Org 7.5 milestone.
This next milestone is a code freeze for X Server 1.7 and will mark the first beta release of X.Org 7.5. However, with this code not even been branched yet from master, do you think this next milestone will be on time with the 20th of July? A month from tomorrow is when the official release of X.Org 7.5 is scheduled to be. Had the earlier release schedule been met, X.Org 7.5 with X Server 1.7 would have been released yesterday.
As of yet, there is no official communication whether X.Org 7.5 is being delayed or not, which leaves some distribution vendors and others in the dark. The X.Org developers (many of whom are unpaid) are doing a great job evolving this critical part of the Linux desktop, but the ever-changing release schedules and lack of communication at times is certainly an area that can be improved. Hopefully though X.Org 7.5 will still make it out well in time for the Q3'09 round of distribution updates.
| 46
| 1,760,738,457.725664
|
https://www.phoronix.com/news/NzM3Mg
|
X Server 1.6.2 Gets Released
|
Michael Larabel
|
The first pre-release for X Server 1.6.2 came a month ago followed by a release candidate about one week ago, but this afternoon Keith Packard has announced the final release of X Server 1.6.2.
In X Server 1.6.2 are a handful of bug fixes and improvements that should be stable for the X Server 1.6 series while the grunt of the development work has been focused on X Server 1.7 / X.Org 7.5, which will hopefully be released in late summer.
The X Server 1.6.2 release announcement can be read on the xorg-announce list.
| 1
| 1,760,738,458.3368
|
https://www.phoronix.com/news/NzM3Mw
|
Will The X.Org 7.5 Release Schedule Be Met?
|
Michael Larabel
|
While the X Server is a very critical component to the Linux desktop and free software ecosystem, the X.Org project has had problems with delivering releases on time. Almost any X Server / X.Org release from recent times can be looked at and generally there were multiple delays in getting that release out that often spanned multiple months. In fact, for a simple point release last year it was over 200 days late. The most recent example is X Server 1.6, which was meant to be strictly a time-based release, but that arrived almost three months late. Beyond just being late, the quality of these X.Org releases have been degrading. Currently under development is X Server 1.7 and that too is running behind schedule.
X Server 1.7 / X.Org 7.5 was originally scheduled for release in April, but that wound up being an April Fool's Joke as nothing was actually accomplished in getting this update ready for release in time. In late April, the release schedule was then revised for a July release, but that was quickly backed to an August release schedule where it stands today. The most recent delay was to allow X Input 2.0 to be merged (so that we can finally have Multi-Pointer X) prior to the feature freeze.
This most recent release schedule put the X Server 1.7 code branching on the 6th of July, but if you head on over to the xorg/xserver branches page, you can see that as of right now there is no X Server 1.7 branch. Daniel Stone who has been heading up the X.Org 7.5 management is currently at the Gran Canaria Desktop Summit, which may explain this latest delay, but if that is the reason it will be at least a week that this release schedule will be behind.
Granted the simple branching of the X Server code from Git master is not a huge milestone worth celebrating, but it simply didn't happen yet and hopefully this will not impact the rest of the release schedule once again. The first beta release of X.Org 7.5 is planned for the 20th of July, which is less than two weeks away.
If this latest release schedule is to be followed, X Server 1.7 / X.Org 7.5 is to be completed and out in the wild on the 17th of August, which is just a bit more than one month away. It would be wonderful if no further delays are pushed onto X Server 1.7, but at this point the bet would probably be on September.
| 27
| 1,760,738,459.038374
|
https://www.phoronix.com/news/NzM2MA
|
A Root-less X Server Nears Reality
|
Michael Larabel
|
One of the benefits of moving the different graphics hardware drivers over to using kernel mode-setting, an in-kernel GPU memory manager (whether it be GEM or TTM), and other newer X innovations is the possibility of now running the X Server without root privileges. By doing so, this of course improves the security since this very large chunk of code is no longer running with all of these high-privileged rights.
Due to now living in a KMS-enabled world, at least on the Intel and ATI side (the NVIDIA side is still slowly but surely coming via Nouveau), it's rather easy to get the X Server running without any special rights. Intel's Jesse Barnes explains on the X.Org mailing list that only a small patch is needed for the X Server and then a trivial one to the Direct Rendering Manager in the kernel. Right now, however, the X Server patch is a bit "hackish", but he and other developers are currently collaborating on the best approach to implement this new capability.
It looks like we may be seeing root-less X Servers in the near future.
| 11
| 1,760,738,459.568714
|
https://www.phoronix.com/news/NzMyMA
|
DisplayLink Linux Driver Continues To Mature
|
Michael Larabel
|
It was just one week ago that there was an update released to the DisplayLink X.Org driver and its kernel frame-buffer module, but there is now yet another update available. This time around the xf86-video-displaylink driver is at version 0.3 and it adds preliminary support for RandR and eliminates its ShadowFB support. Supporting the Resize and Rotate extension will make it easier to configure multiple display heads that are using a DisplayLink USB adapter, but so far this support is still very early. The shadow frame-buffer support has been eliminated in favor of using the frame-buffer kernel module's back-buffer, which means less system RAM is now used.
The main developer behind this DisplayLink Linux work, Roberto De Ioris, who has been working on this code since DisplayLink became Linux friendly, plans to continue enriching this driver. Next up he will be adding DisplayLink support within the xf86-video-intel driver so that a virtual CRTC can be created for each DisplayLink adapter on the given system. In turn the Intel X.Org driver can then use the Intel IGP to draw the screen contents in an accelerated fashion before handing it over to the monitor attached via the DisplayLink USB device.
The latest DisplayLink Linux driver release announcement can be read here.
| 4
| 1,760,738,459.576933
|
https://www.phoronix.com/news/NzMyMg
|
RandR 1.4 & Per-Output DPMS Controls
|
Michael Larabel
|
Adam Jackson, a well known X.Org developer at Red Hat, has proposed adding per-output DPMS controls to the Resize and Rotate extension, which would ultimately go on to be part of RandR 1.4. Adam has proposed a patch on the X.Org development list to add per-output DPMS support to the RandR protocol. In other words, DPMS modes of on, stand-by, suspend, or off could be forced to individual monitors (rather than all displays) with drivers that implement this new RandR support.
However, as not much else has been discussed yet in regards to RandR 1.4, it would be unlikely for any of this to appear with the next release (X.Org 7.5 / X Server 1.7). Today's patch just defines the per-output DPMS portion of the RandR protocol but does nothing in regards to implementing the support within any driver. We are also just under two weeks away from the X Server 1.7 feature freeze. Also still missing from the Resize and Rotate extension is GPU object support.
The current version is RandR 1.3 and it provides panning support and other minor improvements. This version was introduced with X Server 1.6 and most of the open-source drivers now implement the RandR 1.3 panning support. For more information on the X.Org Resize and Rotate extension, read The Newbie's Guide To RandR 1.2.
| 7
| 1,760,738,460.082569
|
https://www.phoronix.com/news/NzM1Mg
|
X Server 1.6.2 RC2 Brings More Bug Fixes
|
Michael Larabel
|
Keith Packard has been the release manager for the X Server 1.6 series and back in early May he issued an X Server 1.6.2 release cadndidate to close a few more X.Org bugs in this critical piece of software. After pulling in another heaping of bug fixes, Keith has now announced X Server 1.6.2 Release Candidate 2, which will go on to become the official 1.6.2 release assuming no bugs or regressions appear.
In X Server 1.6.2 are over 36 bug fixes that address a variety of different problems from DRI2 to EDID to fixing up other problems. X Server 1.6.2 final will hopefully be out soon. The 1.6.2 RC2 announcement can be read on the X.Org mailing list.
Coming out soon after this will be the start of X Server 1.7 development releases, with the latest schedule having it planned for an August release.
| 0
| 1,760,738,460.090828
|
https://www.phoronix.com/news/NzMxMQ
|
Writing At Length About X Input 2.0
|
Michael Larabel
|
X Input 2 was finally merged into the X Server this past week that brought a much needed update to this input device extension for X.Org. Most prominently, X Input 2 can handle Multi-Pointer X much better.
For those interested in more details surrounding the X Input 2.0 extension, Peter Hutterer has written at length about this input extension update. Peter, of course, is the mastermind of MPX and is the one largely responsible for X Input 2.0. In XI2 Recipes, Part 1, Peter discusses the changes in X Input 2, MPX in XI2, X Input 2 initialization, and other attributes. This is a good read for anyone interested in the input side of X.Org.
Both MPX and X Input 2.0 will officially be introduced with X.Org 7.5 / X Server 1.7. The latest release plans place X.Org 7.5 on the table for a final release in August.
| 2
| 1,760,738,460.592659
|
https://www.phoronix.com/news/NzMxOA
|
TTM Memory Manager Gets Ready For Release
|
Michael Larabel
|
With the release of the Linux 2.6.30 kernel, the merge window for new features to enter the next Linux 2.6.31 development cycle is about to open. There's been much speculation whether TTM and Radeon kernel mode-setting would enter this next mainline kernel release or if it would be dragged on for another three months, but it looks like TTM at least is getting very close to entering the mainline tree.
Thomas Hellstrom of VMware/Tungsten Graphics has signaled the state of the TTM memory manager by issuing an RFC on the dri-devel list. Thomas is requesting developers comment on this Tungsten-originated patch that would add the Translation Table Maps support to the DRM within the kernel. Assuming no critical issues get brought up, we could then see TTM enter the Linux 2.6.31 kernel.
The generic TTM code adds about 5,800 lines of code to the Linux kernel. TTM is the GPU memory management subsystem that was previously gaining much steam among Linux driver developers until Intel had developed the Graphics Execution Manager and switched its driver over to using this code for memory management. GEM had entered the Linux 2.6.28 kernel and one of its main advantages is being simpler than TTM at the time. However, GEM was designed all around the needs of Intel and their hardware, so a GEM-only memory solution hasn't proved to be ideal for those developers working on NVIDIA and ATI/AMD hardware. The Nouveau and Radeon developers have been working on a GEM-ified TTM manager that uses TTM internally but exposes the GEM interface. TTM will also be used for part of VIA's new DRM driver. In-kernel memory management is a prerequisite for kernel mode-setting.
TTM is a GPU memory manager subsystem designed for use with GPU devices with various memory types (On-card VRAM, AGP, PCI apertures etc.). It's essentially a helper library that assists the DRM driver in creating and managing persistent buffer objects.
TTM manages placement of data and CPU map setup and teardown on data movement. It can also optionally manage synchronization of data on a per-buffer-object level.
TTM takes care to provide an always valid virtual user-space address to a buffer object which makes user-space sub-allocation of big buffer objects feasible.
TTM uses a fine-grained per buffer-object locking scheme, taking care to release all relevant locks when waiting for the GPU. Although this implies some locking overhead, it's probably a big win for devices with multiple command submission mechanisms, since the lock contention will be minimal.
TTM can be used with whatever user-space interface the driver chooses, including GEM. It's used by the upcoming Radeon KMS DRM driver and is also the GPU memory management core of various new experimental DRM drivers. TTM has been a long time coming, but it looks like it may finally enter the limelight! Now we just need to hope the Radeon kernel mode-setting driver will also be ready to enter the Linux 2.6.31 kernel.
| 14
| 1,760,738,460.629384
|
https://www.phoronix.com/news/NzMwOQ
|
DisplayLink's Frame-buffer and X.Org Drivers
|
Michael Larabel
|
Last month we shared the news about DisplayLink providing USB GPU support under Linux that was open-source and backed by the company while the Linux drivers were being developed in conjunction with the Linux Driver Project. Today, on our birthday, they have released udlfb 0.2.2 and xf86-driver-displaylink 0.2. The udlfb is the DisplayLink Linux Frame-buffer driver while xf86-driver-displaylink is the accompanying X.Org driver for use with DisplayLink video devices.
New in the DisplayLink Linux driver world is preliminary support for double buffering to provide smooth video playback, RLE compression, real screen blanking, code refactoring to support 24-bit outputs (coming in the next release), and the X.Org driver should have reduced flickering on multiple damaged regions. Just three days ago was actually the first public release of this DisplayLink driver that is written by Italian programmer Roberto De Ioris. This driver currently lacks RandR support and various other features, but right now in fact the X.Org driver is just about 640 lines of code while the Linux kernel frame-buffer for it comes in at over 1,000 lines.
The release announcement along with a source download link for both Linux drivers can be found on the X.Org mailing list.
| 1
| 1,760,738,461.107476
|
https://www.phoronix.com/news/NzMwNg
|
X Input 2.0 Merged Into The X Server
|
Michael Larabel
|
Just as planned, the X Input 2 protocol for X.Org has entered the master X Server code-base. X Input 2 (or Xi2 for short) is a significant update over the original X Input extension and allows for Multi-Pointer X support and other enhanced input features. With this code finally entering the mainline server it means that MPX / X Input 2.0 will indeed be features of X Server 1.7 / X.Org 7.5, which are scheduled to be released around August.
The message announcing this code merger was made on the xorg-devel mailing list.
| 11
| 1,760,738,461.122131
|
https://www.phoronix.com/news/NzI5MQ
|
X Input 2.0 Hitting Master In Seven Days
|
Michael Larabel
|
X Input 2.0 is officially set to arrive in the X Server in just seven days. This input extension update for X.Org has been a longtime in the making and is needed for Multi-Pointer X and also improves other areas to enhance the input subsystem. X Input 2 was supposed to hit X Server 1.6 but then that got pushed back and now the X.Org 7.5 / X Server 1.7 schedule has been pushed back yet again to allow Xi2 to finally merge.
Peter Hutterer announced on the X.Org development list early this morning that X Input 2.0 will be pulled into the X Server master branch a week from today, on the 4th of June. While this input code is finally hitting the master code-base, the API or ABI is not finalized yet and is subject to change in the coming weeks. Peter also shared that X Input 2.0 is quite simple while the next version of Xi, X Input 2.1, will be more feature-complete.
Some of the features of X Input 2.0 include sub-pixel precision for coordinates, 32-bit keycodes, maniulation of the device hierarchy, support for multiple pointers (MPX), and simplified event selection. X Input 2.0 will hopefully officially make it out with X Server 1.7 in X.Org 7.5 this August.
| 2
| 1,760,738,461.628945
|
https://www.phoronix.com/news/NzI4OQ
|
X.Org 7.5 Release Schedule Slips Again
|
Michael Larabel
|
X.Org 7.5 with the X Server 1.7 update has been delayed, again. X.Org 7.5 was originally scheduled for release in early April but that ended up being an April Fool's Day joke. In late April the X.Org 7.5 / X Server 1.7 release schedule was then revised for a July release. The feature freeze for X Server 1.7 was supposed to occur last week, but Daniel Stone has announced this morning that X.Org 7.5 will be set back by another month.
Rather than a 22nd of May feature freeze, that freeze is now not coming until the 22nd of June. All of the other milestones for this next X Server update have also been pushed back by approximately one month. Rather than delivering this important software update on the 15th of July, it is now planned for release on the 17th of August. The first beta release will not even come until the 20th of July, five days after the final release was to be expected.
Besides the feature freeze on the 22nd of June, the X Server 1.7 branching is set for the 6th of July, the code freeze and first beta on the 20th of July, a second beta on the 3rd of August, and the release candidate on the 10th of August. X Server 1.7 and X.Org 7.5 will then hopefully make it out the door a week later on the 17th of August.
This latest delay is due to X Input 2 not yet being ready for general usage. However, input expert Peter Hutterer expects to merge X Input 2 and XKB2 into the master X Server code-base by the 14th of June. Therefore, X.Org 7.5 is being delayed so that it can incorporate these major input improvements that have a long time in the making. Multi-Pointer X will also now be possible for the masses of Linux desktop users.
It is unfortunate that there is already another delay in the X.Org 7.5 / X Server 1.7 release schedule, but at least there is better communication with this latest delay. With the previous delay the revised release schedule was not even determined until weeks after the update was supposed to be released. This time around when there is a delay, there is now communication just days later informing developers, end-users, and other X.Org stakeholders about the change. At least this is a step in the right direction.
With that said, the feature freeze for Ubuntu 9.10 is now just a week after this X.Org update is supposed to be released. With another delay, we may not see X.Org 7.5 in this next Ubuntu release -- though X Input 2 is already available via a PPA.The announcement regarding the updated X.Org 7.5 release plans can be found on the X.Org mailing list.(PS: Hi Daniel!)
| 15
| 1,760,738,462.347215
|
https://www.phoronix.com/news/NzI4OA
|
Ubuntu Picks Up X Input 2 Support In PPA
|
Michael Larabel
|
With the release of X Server 1.7 as part of X.Org 7.5 hopefully coming in July, X Input 2 will finally be entering the limelight after being long in development. This updated X Input Extension is critical in allowing Multi-Pointer X support in this next X Server release.
As X Input 2 could potentially pose a few problems for applications that depend upon this server extension, Thomas Jaeger has assembled a Launchpad PPA (Personal Package Archive) for Ubuntu that provides the latest development code that supports Xi2. In the xorg-xi2 PPA is an early snapshot of X Server 1.7, a pre-release of the xf86-input-evdev 2.3 driver, and newer versions of X Input, x11proto-input, wacom-tools, Mesa, libXi, and asciidoc.
Jaeger's announcement of these X Input 2 testing packages for Ubuntu can be read on the Ubuntu X.Org mailing list. Assuming the release of X.Org 7.5 does not get pushed back any further, the newest packages should land in time for the release of Ubuntu 9.10 thereby providing plenty of goodness for allowing multiple pointers on the Linux desktop.
| 5
| 1,760,738,462.981619
|
https://www.phoronix.com/news/NzI3NA
|
DisplayLink Provides USB GPU Support On Linux
|
Michael Larabel
|
Besides Intel, VIA, and ATI/AMD cooperating with X.Org and Linux developers by providing source code and documentation to help with the enablement of their hardware under Linux, another major company has come to the open-source table. No, sadly it is not NVIDIA. DisplayLink is the company and it has now provided an open-source library so that products using their technology will eventually work with Linux.
DisplayLink developed the technology that allows graphics processors to operate over USB to power high resolution displays. Companies like LG Eletronics, Hewlett-Packard, ASUS, Samsung, and Acer use the technology from DisplayLink. This Linux library for DisplayLink that has been released under the GNU LGPLv2 license is not immediately usable as first programs/drivers need to be written to take advantage of the library.
As is announced in DisplayLink's Linux press release, they have partnered with Novell's Greg Kroah-Hartman and his Linux Driver Project to begin working on the support. We were also told by Bernie Thompson, DisplayLink's VP of Software Platforms, that Keith Packard is also involved in bringing DisplayLink support to Linux.
Nice job DisplayLink for releasing this open-source library that will ultimately mean USB graphics support on Linux.
| 11
| 1,760,738,462.989519
|
https://www.phoronix.com/news/NzI3MQ
|
The Linux VGA Arbiter Has Been Revived
|
Michael Larabel
|
Tiago Vignatti has announced he has revived work on the VGA Arbiter for Linux and will be attempting to push this code upstream in the Linux kernel, just four years after this arbitration code was first hashed out. The VGA Arbiter seeks to address an old problem where having multiple graphics cards that use the the legacy VGA interface with multiple X Servers could cause havoc with the same command being sent to both graphics cards instead of just to the intended adapter. This though is not a problem when the driver and/or hardware disable the legacy interface. However, for those affected, the VGA Arbiter fixes this problem for those graphics cards using the legacy VGA interface by controlling which one is accessed.
The patches that provide the Linux VGA arbitration code have been sent to the X.Org development list with an explanation of this latest work by Tiago. He shares that running two X Servers in parallel now work (in a multi-seat setup) and secondary graphics cards work. The VGA Arbiter requires changes to the Linux kernel, X Server, and libpciaccess.
More information on the Linux VGA Arbiter can be found on the X.Org Wiki and is described in detail on this page.
| 6
| 1,760,738,463.596107
|
https://www.phoronix.com/news/NzIzMg
|
X.Org 7.5 Release Schedule Revised For July
|
Michael Larabel
|
X.Org 7.5 with the X Server 1.7 release was planned for release at the start of this month, but sadly they far from made it -- no test releases of the new X Server are even available yet. Though after reading Phoronix, Daniel Stone remembered this release schedule and has decided to give another try at X.Org 7.5. Daniel now hopes the X.Org developers will be ready to deliver X.Org 7.5 with X Server 1.7 on the 15th of July.
This revised schedule puts the feature freeze towards the end of May, the X.Org 7.5 Beta 1 release on the 19th of June, a second beta and release candidate in early July, and then the final release of X.Org 7.5 / X Server 1.7 to come in the middle of July.
The big improvement with X Server 1.7 will be the integration of X Input 2, XKB2, and finally the enabling of Multi-Pointer X. Besides that there is plenty of other input and graphics work going on too, as you know from reading Phoronix.
There is a bad track record within the X.Org community of delivering new releases on time, but let's hope they are able to commit to this schedule and that we see a release in July and not September. Daniel's revised X.Org 7.5 release schedule can be found on xorg-devel.
| 20
| 1,760,738,463.611354
|
https://www.phoronix.com/news/NzIzOA
|
Radeon KMS, New TTM Code Works But Needs Testing
|
Michael Larabel
|
It has been a while since Jerome Glisse last had any major public announcements, but this morning he is calling for those using the open-source ATI Linux graphics stack to checkout the latest code. Kernel mode-setting for ATI Radeon hardware should now be working atop the new TTM-based memory management code.
Jerome along with Alex Deucher, David Airlie, and other X.Org developers have been busy getting ATI kernel mode-setting ready for mainline inclusion hopefully at some point this year (perhaps 2.6.31 or 2.6.32). For their in-kernel memory management they are not using solely the Graphics Execution Manager that was developed by Intel, but instead they are using a GEM-ified TTM manager. Internally they are using a form of Tungsten's TTM but to the user-space they expose the GEM API. The TTM code that the open-source ATI developers are using is based upon the newttm work which hopefully will be ready to be pushed into the mainline Linux kernel soon. The Nouveau developers with memory management are working in a similar way.
Jerome reports that this latest DRM, Mesa, DDX driver, and Mesa Radeon Rewrite driver are working quite well. With kernel mode-setting he has even experienced proper suspend-and-resume when suspending while playing Quake 3 in a composited environment.
Similar to Intel's performance problems after all of the invasive changes, this new code is not currently running as fast as their traditional stack. There are already a few items, however, on their agenda to hopefully boost the performance beyond their original performance levels by improving buffer tiling and swapping along with other memory optimizations.
For information on how to test out this new Radeon kernel mode-setting code with new TTM memory management support can be found on Jerome's blog.
| 55
| 1,760,738,464.246724
|
https://www.phoronix.com/news/NzI1NQ
|
An X Server 1.6.2 Pre-Release Made It Out
|
Michael Larabel
|
With the release of X.Org 7.5 with X Server 1.7 not coming until at least July, Keith Packard is readying a second update in the X Server 1.6 series. Early this morning Keith announced the release of an X Server 1.6.2 pre-release.
This X Server 1.6.2 pre-release known as v1.6.1.901 contains just over a dozen fixes since the release of X Server 1.6.1 back in April. Generally we have not encountered X Server pre-releases for new point releases, but this time around there is one and hopefully it will lead to a few more fixes coming about prior to the official release of X Server 1.6.2.
If you are interested in grabbing the source code to the latest X Server 1.6.2 pre-release or want to view the 15 or so changes that make up xorg-server 1.6.1.901, you can find it on the X.Org mailing list.
| 11
| 1,760,738,464.255409
|
https://www.phoronix.com/news/NzE1Nw
|
A Working X Input 2 Implementation
|
Michael Larabel
|
X Input 2, a major rework to the X.Org input extension, is a step closer to reality. Originally it was hoped that X Input 2 would be ready for X Server 1.6, but that didn't end up happening and also resulted in Multi-Pointer X being disabled in this latest X Server release. This morning, however, Peter Hutterer (of MPX fame) has his first working X Input 2 implementation.
This code has not yet landed in the mainline code-bases of the X Server, libXi, inputproto, or xinput, so for now he has this working implementation running in his own Git repositories. Peter though has re-based this work off the latest X.Org Git code and "it works without issues" while implementing the X Input 2.0 draft specification. While X Input 2 is working, there are currently a number of limitations and there are several features still missing.
In Peter's mailing list are all of the details along with how interested parties can help out and where X Input 2.0 is going to go. What is not clear at this point, however, is whether this revised input extension will be ready for X Server 1.7 / X.Org 7.5.
| 3
| 1,760,738,464.875617
|
https://www.phoronix.com/news/NzIwMw
|
X Server 1.6.1 Released With 15 Fixes
|
Michael Larabel
|
While there is currently little work going on in the way of concerted development efforts towards X.Org 7.5 / X Server 1.7, a point update to X Server 1.6 has been released.
While Intel had wanted to release X Server 1.6.0 by the end of 2008, its release did not end up happening until late February when it brought DRI2, RandR 1.3, and numerous other features. This afternoon though, Red Hat's Adam Jackson has released X Server 1.6.1, which packs a couple of bug-fixes atop the X Server 1.6.0 code-base.
X Server 1.6.1 has a few fixes for SELinux usage, RandR, GLX, X Input, XKB, and Xext. In total there are 15 changes for X Server 1.6.1.
The release announcement concerning X Server 1.6.1 can be read on the X.Org mailing list.
| 10
| 1,760,738,464.883739
|
https://www.phoronix.com/news/NzE0Mw
|
xf86-video-ati 6.12.0 Driver With R600/700 Support
|
Michael Larabel
|
The xf86-video-ati 6.11 driver was released less than a month ago, but coming out today is now the xf86-video-ati 6.12.0 release. What this release brings is most notably EXA and X-Video support for the ATI R600 and R700 generations of graphics processor. The R600/700 GPUs include the Radeon HD 2000, HD 3000, and HD 4000 series. In order to benefit from this R600/700 acceleration, however, an updated kernel DRM is required.
Beyond the enhanced R600/700 support are various other fixes. The complete list of changes with source download links for xf86-video-ati 6.12 can be found on the X.Org mailing list.
| 78
| 1,760,738,465.379221
|
https://www.phoronix.com/news/NzE0Mg
|
VDPAU + OpenGL 3.0 On Gallium3D This Summer?
|
Michael Larabel
|
This is the fourth year that the X.Org project will be participating in Google's Summer of Code project. In the past we have seen work like Generic Gallium3D Video Decoding using a graphics card's shaders to assist in video decoding no matter the hardware vendor. This year there are a few interesting possibilities.
While no student developers have yet stepped up, among the possible projects include bringing VDPAU (NVIDIA's excellent video API) to Gallium3D, a pure OpenGL 3.0 state tracker for Gallium3D, an R600 Gallium3D driver, true multi-touch support, finishing X Input 2, and merging xkbcomp into the X Server.
The current list of Summer of Code ideas for X.Org can be found on this Wiki page. Additionally, there is also a active mailing list thread.
| 9
| 1,760,738,466.051134
|
https://www.phoronix.com/news/NzEyMA
|
S3 Graphics Linux Driver Faster Than Windows?
|
Michael Larabel
|
Back in November we saw the launch of the S3 Graphics Chrome 530 GT and with that they talked up a new magical Linux driver that would provide HD video acceleration support along with OpenGL 3.0 capabilities. But no driver was released, however, a day later it was confirmed by S3 Graphics that they were working on a new Linux driver. Their PR representative said the driver was to be released in December, but that didn't happen. In February they continued to talk up their Linux support but months later there still was no driver. However, that changed in late February when S3 Graphics did in fact roll out a new Linux display driver.
The S3 Graphics driver for Linux is still proprietary (though its kernel module claims to be GPL) but it did introduce OpenGL 3.0 support along with VA-API acceleration for GPU-assisted video playback. We have not been able to test this new Chrome 500 series Linux driver, but a site called "Chrome Center" has tested out the Linux and Windows drivers.
Chrome Center's Stefan had ran Nexuiz on both Linux and Windows to compare the performance of a S3 Graphics Chrome 430 GT. His processor was an AMD Phenom II. His results show that the S3 Graphics Linux driver is actually faster than their Windows driver.
Stefan concludes, "I don't actually know if the Linux driver is that good or if the Windows driver is that bad... I will do a few more benchmarks to investigate this further." In an email to us, Stefan has mentioned that he is looking to use the Phoronix Test Suite for additional S3 Graphics tests.
| 6
| 1,760,738,466.537141
|
https://www.phoronix.com/news/NzEyMw
|
RandR 1.3 Has Officially Been Released
|
Michael Larabel
|
Present in X Server 1.6 is the server component to RandR 1.3, but today version 1.3 of the Resize and Rotate extension has been officially released. RandR 1.3 introduces panning support and other changes that we have talked about previously. The components released today were randrproto 1.3.0 and libXrandr 1.3.0. There are also some new options within the xrandr utility.
RandR 1.3 support can already be found within the Intel and ATI drivers.
Cheers to Matthias Hopf and all developers involved.
| 0
| 1,760,738,466.546017
|
https://www.phoronix.com/news/NzA3Mw
|
X Server 1.6 Gets A Bit Closer To Release
|
Michael Larabel
|
X Server 1.6 was supposed to be released by the end of 2008 as Intel had called for it to be a strictly time-based release. However, the release schedule ended up putting it as a early January release (the 5th of January to be exact). The release schedule was being closely followed up until late December and then come January there was no X Server 1.6 release in sight. To this date, X Server 1.6 still hasn't been released, but it's slowly getting there.
Last night Keith Packard had tagged X Server 1.5.99.903, which is the third release candidate for X Server 1.6. This version is coming after 24 new commits were made to the X Server 1.6 branch. While the release is getting closer, there are still five open bugs that block the release.
Once released, X Server 1.6 brings RandR 1.3, Predictable Pointer Acceleration, Direct Rendering Infrastructure 2, and other features to the desktops of Linux users.
| 14
| 1,760,738,467.114683
|
https://www.phoronix.com/news/NzA4Nw
|
The Voodoo Linux Graphics Driver Gets Updated
|
Michael Larabel
|
In the event you are still using any 3Dfx Voodoo 1 or Voodoo 2 graphics cards, the X.Org Linux driver for them has been updated. No new features have been implemented for these outdated graphics cards, but the xf86-video-voodoo 1.2.1 driver just brings a couple of fixes to ensure the driver still works with the most recent X Server bits. The release announcement for the Voodoo Linux driver update can be found here.
The xf86-video-voodoo driver wasn't the only antiquated X.Org driver to get updated today, but so did the xf86-video-rendition driver for the old Rendition graphics cards. The Rendition driver just has two fixes, as mentioned in the release announcement.
Lastly, the unaccelerated xf86-video-sisusb driver for SiS chips connected via a USB adapter has too been updated. The xf86-video-sisusb 0.9.1 driver just has some build fixes and other minor fixes like the other two X.Org driver updates this morning.
| 3
| 1,760,738,467.12377
|
https://www.phoronix.com/news/NzA3MQ
|
S3 Graphics Responds About Linux Support
|
Michael Larabel
|
Last week S3 Graphics had released the Chrome 540 GTX, which is their newest and fastest PCI Express graphics card. Similar to when announcing the S3 Chrome 540 GT, in the Chrome 540 GTX press release they once again mention Linux support along with OpenGL 3.0 capabilities. However, they talk up Linux support, but fail to provide the support. We have just heard back though from S3 Graphics' Benson Tao, which is the one that previously told us there would be Chrome 500 Linux support in December along with a beta OpenGL 3.0 driver. What though did he have to say this time? His email is below.
Hi Michael,
Thanks for getting back to me in the other email.
We are doing an internal build of the Chrome 500 Linux driver to incorporate some of the additional hardware features and upgrades (over the Chrome 400 Series GPUs).
If you want to test the Linux now, the Chrome 400 Series drivers also support the Chrome 500 Series since it is a unified driver architecture.
Thanks,
Benson
So now there is no ETA for when an S3 OpenGL 3.0 Linux driver will arrive. This time though he claims the Chrome 400 Linux driver (partially) supports the Chrome 500 series. On the Chrome 400 driver page, however, it lists the Chrome 430 GS, Chrome 430 GT, Chrome 440 GTX, Chrome 430 ULP, and Chrome 435 ULP as being the supported ASICs. The Chrome 400 Linux release notes also don't mention any support for the Chrome 500 series.
Benson also hadn't mentioned any Chrome 500 support in the Chrome 400 Linux driver when he originally talked up their new Linux driver. Their Chrome 400 Linux driver was last updated in November when they originally pushed the driver out, so it would be a bit hard to believe they had in all of the Chrome 500 PCI IDs at that time even for the just-released Chrome 540 GTX.
To make things more interesting, the kernel module for the Chrome 400 Linux driver claims its GPL licensed.
$ modinfo s3g.ko
filename: s3g.ko
license: GPL
srcversion: D1EEA2B5E900753E01FD4D1
depends:
vermagic: 2.6.24-19-generic SMP mod_unload 586
Included with the Linux driver package though is no real source-code (just a few source files for building the DRM), no GNU GPL license attached on any of it, and the LICENSE file shipping with the Chrome 400 Linux driver still very much reflects this driver being a binary blob. A mistake? Is S3 Graphics utilizing GPL-only symbols? Do they really mean for their kernel driver to be GPL?
Well, the S3 Graphics Linux situation still isn't clear. When we have any other information to report, we will be sure to do so. Share your thoughts in the forums.
| 14
| 1,760,738,467.632436
|
https://www.phoronix.com/news/NzA2Ng
|
X@FOSDEM 2009: RandR 1.3, GEM, Gallium3D, Etc
|
Michael Larabel
|
All of the recordings from the X.Org meetings that took place during FOSDEM 2009 are now available on Phoronix. There were nine topics in total from RandR 1.3 to shader compiler optimization strategies.
Novell's Matthias Hopf had explained RandR 1.3 with all of the work involved in this update to the Resize and Rotate extension for the X Server that now has panning support and other new-found capabilities. An RandR 1.3 demonstration was also done by Keith Packard. Later on, Matthias had provided a 3D understanding of the ATI R600/700 series.
Beyond demoing RandR 1.3, Keith Packard of Intel had talked about rebuilding the Linux desktop with all of the work that has been done recently from the GEM memory manager to kernel mode-setting and more. Intel's Eric Anholt had talked about the work that's still left to be accomplished: largely DRI2 v-blank support and writing a new GLSL compiler. During this talk we also learned that UXA will not be merged back into EXA and Intel is considering VDPAU support for their driver.
In traditional FOSDEM fashion, Stephane Marchesin had provided a Nouveau status update on where this community-spawned open-source NVIDIA driver is at. The Nouveau team still hasn't issued a stable 2D or 3D driver release, but they are had at work on kernel mode-setting support and once that has stabilized we may see a stable 2D driver release. On the 3D front they are continuing to make good progress with Gallium3D.
As a non-graphics talk but still related to X.Org, Helge Bahmann talked about bringing multimedia and audio extensions into the X Server. Helge has written patches already that would add compressed image and audio support to the X protocol along with playback synchronization capabilities.
Back on to the topic of GPUs, Red Hat's Matthew Garrett talked about GPU power management. He is hoping to decrease the power bill for Linux users by conserving power through different strategies for the GPU, memory, outputs, and displays.
Stephane Marchesin and Jerome Glisse had ended out X@FOSDEM 2009 by talking about GPU shaders. Stephane Marchesin talked about mixing the LLVM compiler with Gallium3D to optimize GPU shaders before they are sent to the driver/hardware. Jerome talked about different shader optimization strategies.
Some of these recordings have some audio feedback or other distortions due to a faulty microphone adapter. The good news, however, is that the audio can be easily cleaned up through Audacity. For those interested in MPEG recordings for this reason or those not interested in the Flash videos, they can be found in the Phoronix Blip.TV channel.
| 15
| 1,760,738,467.64134
|
https://www.phoronix.com/news/NzA2MQ
|
S3 Graphics Still Talks Up Linux Support, But Fails
|
Michael Larabel
|
S3 Graphics has announced this afternoon the release of the Chrome 540 GTX, which they advertise as "The World's Most Connected Hi-Def Card" with its HDMI, DisplayPort, and DVI connections. The Chrome 540 GTX runs at 850MHz, uses GDDR3 memory, and shares other features to the Chrome 530 GT that was introduced in the fourth quarter of 2008. In the press release announcing the S3 Graphics Chrome 540 GTX they once again mention Linux support... But is there any Linux support?
When it comes to Linux, mentioned in the Chrome 540 GTX press release is:
"DirectX 10.1 and OpenGL 3.0 applications on Microsoft Windows and Linux platforms can be effectively run using the Chrome 540 GTX programmable shader cores to speed up 3D games, non-gaming rendering applications, and other visual processing functions using the built-in GPGPU engine." They continue to talk up OpenGL 3.0 on Linux, and with the Chrome 540 GTX press release they even mention their GPGPU engine. This statement is similar to the announcement they made in the press release for the Chrome 530 GT. Back in November they mentioned Chrome 530 GT Linux support complete with OpenGL 3.0 and Blu-Ray / HD video playback. They even mentioned DirectX 10.1 and Linux platform support in the same sentence. However, they hadn't provided any Linux driver on their web-site.
A day later we heard back from S3 Graphics and was told they are working on a Linux driver. Benson Tao, the US PR representative for S3 Graphics, had shared they would publish a Chrome 530 GT Linux driver on their web-site by the middle of December and it would be accompanied by a beta OpenGL 3.0 driver. Well, in late December their Linux driver was still missing (not even an OpenGL 1.x or 2.x limited driver) and no sign of it coming.
After today's press release that once again mentions Linux support for the Chrome 500 series, is there finally a driver? Nope. There is a Linux driver for the Chrome 400 and 20 series, but nothing for either of their Chrome 500 series GPUs. Their Linux drivers for the older S3 Graphics ASICs are limited and for the Chrome 400 series their latest release is in fact their initial driver release ever made.
It's nice to see Linux mentioned in corporate press releases, but when in fact it's backed up by actual support. We have just sent off a note to S3 Graphics asking for another update on their Chrome 500 series Linux driver -- and whether it will still include OpenGL 3.0 and HD video playback support.
The Chrome 540 GTX press release can be found in the S3 Graphics press room.
| 7
| 1,760,738,468.135165
|
https://www.phoronix.com/news/NzA0OQ
|
Bringing Multimedia, Audio Into The X Server
|
Michael Larabel
|
Helge Bahmann talked about bringing multimedia and audio extensions into the X Server this morning at FOSDEM. Helge talked about this topic last year at FOSDEM, but in the past year he has made more progress in this area. No multimedia extensions are yet present in the mainline X Server, but his code is working to varying degrees based upon the situation. This code is also utilizing some existing X functionality already.
Helge's work would add support for compressed images and audio to be passed through the X protocol. By doing so, playback synchronization would be handled by the X Server and would allow for some network transparent media players.
If this code ends up entering the mainline X Server, it doesn't look like it will be committed anytime soon since there is still quite a bit of work left. The current work is very low-level with a higher-level interface still being needed, there is currently no hardware-accelerated decoding, there are severe latency problems with the X Server until that becomes multi-threaded, and an audio/image de-compressor would need to enter the X Server.
Below is Helge's X multimedia talk.
| 9
| 1,760,738,468.143567
|
https://www.phoronix.com/news/NzA0NA
|
X.Org Development Talks @ FOSDEM 2009
|
Michael Larabel
|
The Free Open-Source Developers' European Meeting is taking place this weekend in Brussels, Belgium. There is a series of dedicated development talks taking place on both days for X.Org on topics ranging from kernel mode-setting to LLVM and Gallium3D to memory management. At Phoronix we will be providing recordings of these talks that will be accessable online. We hoped to have these recordings up in a near real-time manner, but due to the Internet connection at FOSDEM not being ideal this year, that isn't taking place. We hope to have up the video recordings and notes from our talks within a couple days (but hopefully not that long).
| 1
| 1,760,738,468.716866
|
https://www.phoronix.com/news/NzAyOA
|
X Server 1.6.0 RC2 Finally Makes It Out The Door
|
Michael Larabel
|
X Server 1.6 was supposed to be out nearly a month ago, but today we are finally getting closer to nearing this much anticipated release. Intel's Keith Packard has announced the second release candidate for X Server 1.6.0. The RC2 release is arriving almost three weeks after the first release candidate was made available. Fixed in this release is a few RandR and input changes, but a majority of the work deals with Apple's XQuartz in the server.
It looks like we will now see the final release of X Server 1.6.0 in early February. X Server 1.6.0 introduces RandR 1.3, Direct Rendering Infrastructure 2, Predictable Pointer Acceleration, and other improvements.
| 4
| 1,760,738,469.395475
|
https://www.phoronix.com/news/NzAzNQ
|
Synaptics X.Org Driver Reaches Version 1.0.0
|
Michael Larabel
|
The xf86-input-synaptics driver, which is used as the touch-input driver for many notebooks, has reached version 1.0.0. Worth noting in this release is improved device auto-scaling and support for input device properties. The improved device auto-scaling should give supported touch-pads a more unified feel without needing to do any manual configurations. Should you need to configure your Synaptics touch-pad, the input device properties support makes it possible without having to enable SHM config.
The release announcement for xf86-input-synaptics 1.0.0 can be found on the X.Org mailing list. A new release of the xf86-input-evdev driver has also been announced with a few fixes, which can be read about here.
| 0
| 1,760,738,470.001799
|
https://www.phoronix.com/news/NzAzMQ
|
XRandR Gets Support For RandR 1.3 Features
|
Michael Larabel
|
RandR 1.3 has come together for the X Server 1.6 release with namely new properties and panning support. Besides this update to the Resize and Rotate extension for the X.Org Server, xrandr, the command-line utility for controlling RandR-capable drivers/hardware, has been receiving improvements as well.
A new development release of xrandr has been announced this morning with bug-fixes an updated manual, and support for new RandR 1.3 properties. New xrandr options have been added for --current, --nograb, --primary, and --noprimary. The --current option will return the current state of screen configurations, --primary will set an RandR output to be the primary output, and --noprimary will not define a primary output. With the existing RandR 1.2 drivers, there is now also support for adjusting the gamma through the --gamma argument and then specifying the RGB values. Lastly, the --transform and --scale options are now documented within the man page.
The release announcement for this new test release of xrandr can be found on the X.Org mailing list. Around the time of releasing X Server 1.6, we should see the final release of xrandr 1.3.0. For those interested in controlling their RandR outputs through an RandR GUI may need to wait longer for the support of version 1.3 features.
| 4
| 1,760,738,470.009983
|
https://www.phoronix.com/news/NzAyNQ
|
What Do You Want In Linux Drivers This Year?
|
Michael Larabel
|
NVIDIA has already released quite a few Linux drivers this year already that improve their VDPAU support and stabilize their OpenGL 3.0 implementation. Yesterday AMD had then released its first proprietary Linux driver of 2009 that brought OpenGL 3.0 support. While both sides are off to a good start, what else do you want to see from them and their drivers in 2009?
On the open-source side there is a lot to get excited about. Gallium3D will soon be landing in Mesa, more hardware will receive kernel mode-setting support, more of the X.Org drivers will turn to using a kernel memory manager like GEM or TTM, improved video decoding support, and there's many other features that we may see this year in the open-source drivers. But what will we see this year within the high-performance proprietary drivers from ATI/AMD and NVIDIA?
AMD will certainly be introducing X-Video Bitstream Acceleration (XvBA) this year, but over the course of last year they are reaching a feature parity with their Windows Catalyst driver now that the fglrx driver supports OverDrive, CrossFire, and other features. On the NVIDIA side, what else is there to come? More CUDA enhancements? OpenCL? Finally, a graphical installer that can run within X? More video playback improvements?
Tell us in the Phoronix Forums what you hope to see from the proprietary and open-source X.Org display drivers in 2009.
| 75
| 1,760,738,470.497473
|
https://www.phoronix.com/news/Njk5Nw
|
X.Org Talks Planned For FOSDEM 2009
|
Michael Larabel
|
It's less than a month away until FOSDEM, the annual Linux European meeting for developers, is held in Brussels, Belgium, but the list of X.Org-related talks is beginning to be finalized.
Among the talks planned so far are about RandR 1.3, the open-source Nouveau driver, multimedia (audio) extensions for the X Server, GPU power management, r600_demo, LLVM + Gallium3D, and shader compiler optimization strategies.
There still is, however, room for a few more talks. Hopefully an Intel employee (say Keith Packard) will talk about the Graphics Execution Manager, kernel mode-setting, or another exciting X topic. We would also hope for another update about Gallium3D by Tungsten Graphics.
The current schedule of X.Org talks during FOSDEM 2009 can be found on the X.Org Wiki. Stay tuned to Phoronix for coverage of FOSDEM.
| 0
| 1,760,738,470.509164
|
https://www.phoronix.com/news/NzAxOA
|
The State of X Server 1.6: Still Missing
|
Michael Larabel
|
For those that haven't noticed or have forgot, X Server 1.6, which was the update Intel wanted out by the end of 2008, still hasn't been released. The latest schedule for X Server 1.6.0 called for a January 5th release, but that had simply passed by and there wasn't even a release candidate.
As it stands right now, no code has been committed to X Server 1.6 since the first release candidate was tagged over two weeks ago. On the X Server 1.6 tracker bug there are still nine open bugs, which means just one bug has been closed since the 5th of January. Hopefully though we will end up seeing a X Server 1.6 release in February.
Among the features that this new X Server update will carry once released are DRI2, RandR 1.3, Predictable Pointer Acceleration, and X Input 1.5 with Device Properties.
X.Org 7.5 / X Server 1.7 was talked about for release in April, but that now looks even more unlikely. Chances are the next X.Org update will not be out until the second half of 2009, unless it receives some significant development help.
| 8
| 1,760,738,471.010457
|
https://www.phoronix.com/news/NzAxMQ
|
TTM-based OpenChrome In A Working State
|
Michael Larabel
|
With VIA Technologies delivering on their promises by finally releasing 2D/3D documentation and driver code, and Tungsten Graphics creating a new VIA 3D stack for a client, there has been a lot to report on in the VIA Linux scene. Tungsten Graphics and VIA are both interested in creating a Gallium3D driver for the Chrome 9 series, Tungsten already created a feature-rich DRM and Mesa driver, and there is a lot of other work going on too. What's new this week is a build-able TTM-based OpenChrome driver.
Thomas Hellström, the same TG employee working on the VIA 3D code, has announced that his branch of the OpenChrome driver now implements the TTM (Translation Table Maps) memory manager. TTM was set to become the de facto standard for in-kernel GPU memory management until Intel had conceived the Graphics Execution Manager (a.k.a. GEM). This open-source VIA driver using the TTM memory manager will run, but XvMC and EXA (when using the latest X Server 1.6 bits) acceleration are both lacking.
Using this driver requires checking out the branch of DRM and libdrm that provides the new TTM management support. For more information on what's needed if you want to check out this TTM-based driver, follow this thread on the OpenChrome mailing list. Unlike the open-source ATI and NVIDIA driver implementations that are using a GEM-ified TTM manager whereby TTM is used internally to manage the memory but externally it implements the GEM API, this isn't the case with VIA. However, Tungsten Graphics is the company that developed TTM, which is why they may be a bit hesitant to play with GEM.
| 7
| 1,760,738,471.048498
|
https://www.phoronix.com/news/Njk4OQ
|
X Input 2.0 Protocol Draft Specification
|
Michael Larabel
|
Peter Hutterer, the mastermind behind Multi-Pointer X, has released the draft specification for the X Input 2.0 protocol.
Over X Input 1.5, X Input 2 brings support for multiple input master devices and the master/slave device hierarchy being visible to the client. All events in X Input 2 will also be backed by GenericEvents. For those interested in reading the specification, it can be read in this email.
Peter also mentioned that X Input 2 should be ready for X Server 1.7, which will possibly be out later this year and coincide with X.Org 7.5. Originally X Input 2 was set to be included with X Server 1.6, but that didn't happen and before that it was even on the table to be included with X Server 1.5. Now though it looks like X Server 1.7 will be the magic release and will also allow Multi-Pointer X to be enabled by default.
| 3
| 1,760,738,471.592982
|
https://www.phoronix.com/news/Njk4MQ
|
X Server 1.6 Sees 34 New Commits Today
|
Michael Larabel
|
The outlook turned grim for meeting the X Server 1.6 release schedule when no release candidate was out by Christmas (and we still have yet to see any tagged RC) and then when the final release date passed with no X Server 1.6 in sight. Nothing had been committed to the X Server 1.6 repository since the middle of December, but we are finally seeing some activity this afternoon.
X Server 1.6 received 34 new Git commits in the past two hours. These commits largely relate to XQuartz, but there are a few other noteworthy patches too. For instance, with XAA acceleration off-screen pixmaps are now disabled by default.
All of the latest work on X Server 1.6 can be found by checking out the X Server Git tree and switching to the server-1.6-branch branch.
| 0
| 1,760,738,471.602123
|
https://www.phoronix.com/news/Njk3MA
|
No X Server 1.6 Release In Sight
|
Michael Larabel
|
During the last X Developers' Summit that took place at the Edinburgh Zoo, Keith Packard called for X Server 1.6 to be released by the end of 2008. Once the release schedule was set though, the final ship date for X Server 1.6.0 was the 5th of January. Well, today's the day and there is no release.
The X.Org development community was relatively on time with their Beta 1, Beta 2, and Beta 3 releases, but all work has virtually seized since that point. The first release candidate was supposed to be out before Christmas but that never happened and we have yet to see any release candidate even weeks later.
Nothing has been committed to the server-1.6-branch of the X Server Git repository since the 16th of December. The X Server 1.6 tracking bug has ten open bugs (of 17 total).
The sad reality though is that this setback shouldn't be surprising considering the numerous delays that have plagued the X.Org camp in recent times. X Server 1.4.1, which was a bug-fix update for X.Org 7.3, was released 212 days late. Most recently, X.Org 7.4 / X Server 1.5 was released months late. The list of other delays and problems continues on.
The X.Org developers are stretched very thin, but recently they have had the added burden of handling some invasive changes such as the Graphics Execution Manager for providing GPU memory management within the kernel, several revisions of DRI2, drivers turning to kernel mode-setting, and other radical work.
X.Org 7.5 was talked about for an April 1st release but nothing has yet amounted when it comes to even a release schedule and planning.
When the release of X Server 1.6 does occur, which we hope is just weeks away, it will introduce RandR 1.3, X Input 1.5 with device properties, Predictable Pointer Acceleration, and Direct Rendering Infrastructure 2.
| 18
| 1,760,738,472.236172
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.