As illustrated today by the release of Intel's "Sandy Bridge" CPUs there is a new desire by Linux users: open-source drivers "out of the box" at launch. Over the years the expectations of Linux users have gone from simply wanting Linux drivers for their hardware to wanting open-source Linux drivers (read: no binary blobs) to now wanting open-source drivers in the distribution of their choice at the time the hardware first ships. This is a great problem to now be experiencing, as since starting Phoronix seven years ago, the Linux hardware experience has improved a great deal where it's no longer a question if there will be Linux support but when. Some hardware vendors, such as Intel, are now working towards this goal of same-day open-source Linux support -- and in some cases achieving it -- but for open-source Linux drivers for graphics it's a particularly tall hurdle to jump.
This week at the Consumer Electronics Show in Las Vegas (I'll be there again looking out for Linux), Intel will officially launch their next-generation Sandy Bridge micro-architecture and CPUs. The NDA though expired at midnight on these first CPUs so there is now a stream of reviews coming out. Is there any Linux graphics test results for the Core i3 2100, Core i5 2400, Core i5 2500K, and Core i7 2600K? Unfortunately, there is not.
Intel began working towards Sandy Bridge support (the Intel HD graphics found on their next-generation CPUs to be launched next month) since this past February and in the months since and it's now their open-source Linux drivers are nearly ready for the first early-adopters of these soon-to-be-released Intel Core i5/i7 processors.
Intel's next-generation MID (Mobile Internet Device) platform to succeed Moorestown is codenamed Medfield and is slated to be released next year. However, in usual Intel fashion, open-source patches for supporting this next-generation platform under Linux are beginning to make their way out there months in advance of the hardware's public availability.
Intel is gearing up to release their xf86-video-intel 2.14 DDX driver in the coming weeks, which will be their quarterly open-source X.Org driver update for their Intel IGPs. In preparations for this release and the forthcoming release candidates, Intel's Carl Worth has tagged the xf86-video-intel 2.13.901 driver in Git, which is an intermediate development snapshot.
The developers behind the MeeGo operating system, the Linux-based distribution that married Moblin and Maemo, have just released version 1.1. MeeGo 1.1 is available for Intel Atom and ARMv7 architectures and comes in the netbook, IVI (in-vehicle information/entertainment systems), and handset editions. Besides the core MeeGo OS hitting 1.1, the MeeGo SDK 1.1 has reached beta and will be released before the MeeGo Summit next month in Dublin.
While Intel has not even rolled out their Sandy Bridge processors yet, their OSTC developers have been working on support for this next-generation micro-architecture with integrated graphics core under Linux for many months. It was back in February when we originally reported on Sandy Bridge GPU support coming to Linux.
Intel's Poulsbo Linux support is a bloody mess. It has been for nearly two years now and the situation has really not improved at all. While Intel IGPs are generally well supported under Linux with an open-source driver stack (besides being very slow), the Poulsbo hardware on Linux is notorious and does not have a fully open-source driver because the GMA 500 chipset is designed around the PowerVR SGX 535 graphics core from Imagination Technologies rather than being brewed in-house. The situation is really bad.
Intel's Ian Romanick has just written an e-mail message entitled What I'm working on to the Mesa development list. With Intel's new GLSL compiler being used by Mesa and can be found within the Mesa 7.9 release, Intel's open-source graphics developers have worked onto working on some other areas of their 3D driver stack.
With Mesa 7.9 on the way, Intel has just released the xf86-video-intel 2.13 release candidate as part of their quarterly updates to their open-source Linux driver stack. There aren't any major features as part of the Intel 2.13 DDX driver update, but there are a good number of bug-fixes.
While the open-source Radeon DRM/KMS (along with the closed-source Catalyst) drivers have had support for audio over HDMI / DisplayPort, patches are finally moving along by Intel's Zhenyu Wang for bringing up audio on Intel chipsets over these newest display interfaces.
While Ubuntu 10.10 will have no i8xx driver fix for those with this vintage Intel hardware that's been plagued with stability problems and other issues since Intel introduced their Linux kernel mode-setting and GEM driver, there is now a workaround upstream for this issue. Originally the plan was to add back user-space mode-setting support to the Intel X.Org driver that would not use the Graphics Execution Manager (GEM) and this code-path could be enabled by i8xx customers to workaround the cache coherency issues while losing KMS support, but a new workaround was devised.
Back in February we reported on the first signs of open-source support for Intel's Sandybridge, a.k.a. their sixth-generation Intel graphics processor integrated on their upcoming CPUs that succeed the Clarkdale/Arrandale CPUs. The Sandybridge hardware still has not launched nor will it until late this year or early next year, but the open-source support has been underway for months and from time to time we see new Linux code patches related to Sandybridge.
The Intel Linux driver has been challenged by stability problems and other issues for owners of i8xx hardware since they rolled out kernel mode-setting and Graphics Execution Manager (GEM) support with UXA 2D acceleration more than a year ago. There were initial problems for other Intel users as well when switching to this overhauled driver stack -- to the point that it killed the netbook experience -- but those problems were quickly worked away. But for those using Intel's oldest supported hardware under Linux, the problems to this day remain. To circumvent this issue there's been the approach to add back user-space mode-setting to the Intel driver with EXA 2D acceleration to simply avoid these problems rather than correct the actual issues with KMS/GEM/UXA, but now another alternative has emerged.
Back in July we reported on driver work done by Intel's Chris Wilson to add back user-space mode-setting support to the Intel X.Org DDX driver (xf86-video-intel) to allow those with older Intel (i8xx) chipsets where kernel mode-setting can be buggy to at least have a decent experience with UMS. Intel was quick to strip out user mode-setting support from their X.Org driver once their KMS support was stabilized, but it turns out that old Intel hardware with UXA (the GEM-ified EXA) and kernel mode-setting was buggy and could lead to artifacts and stability issues. These problems had led Ubuntu and other distribution vendors to use old Intel drivers so that they wouldn't be shafting a small percent of their users with vintage Intel hardware.
Towards the end of last month we reported on GEM-free UMS support for the Intel driver that was worked on by Intel's Chris Wilson to hopefully address the stability issues and other problems that have challenged owners of old Intel i8xx hardware running the newer Intel driver stack, which is presently limited to kernel mode-setting support with GEM (the Graphics Execution Manager) memory management. However, it seems the work invested into adding back user-space mode-setting support to the Intel driver without the kernel memory management still doesn't resolve the i8xx issues at hand.
MeeGo 1.0 for netbooks was released back in May (and an update already released to that) while a month ago there was an early release of MeeGo 1.1 For Handsets released with the official release coming later in the year with MeeGo 1.1 for netbooks. Today "MeeGo 1.0 IVI" has been released and this is designed for in-vehicle infotainment systems.
While it may have seemed like PowerTop was idling by for a while without a new release or any major advancements to this open-source utility for analyzing power consumption to find programs causing more wake-ups than necessary and to provide other power savings tips on Intel-based Linux systems, a new release has emerged. Intel released PowerTop v1.13 recently and it adds a few new features to the power table along with a number of bug-fixes.
It seems like it was just yesterday, but Intel introduced the Graphics Execution Manager more than two years after they had a falling out with TTM. In switching over to using GEM for their in-kernel video memory management, and subsequently supporting kernel mode-setting and then introducing UXA to GEM-ify the EXA 2D acceleration architecture, there was a lot of problems. Fortunately, most of these problems were worked out as this more advanced Intel Linux driver stack matured and for the better part of a year now the experience has been pleasant for users of most Intel GMA chipsets.
Earlier this month an Intel employee began asking about making optimizations to Mesa's shader compiler (on the Mesa-dev list). This Intel employee was not one of their usual Open-Source Technology Center developers commonly working on their Linux graphics stack as part of Keith Packard's team, but instead it was an uncommon name: Benjamin Segovia. Ben is from Intel's Advanced Graphics Lab team where previously he worked on ray-tracing techniques, but as of late seems to be at least dedicating some of his Intel effort towards optimizing Mesa.
Yesterday we found it interesting that Intel is not even able to ship their own Linux driver for their own hardware with their MeeGo operating system. The driver in question is their new EMGD driver for the Menlow and Tunnel Creek platforms that have a graphics core that's designed by Imagination Technologies rather than their own in-house intellectual property. The EMGD driver from Intel currently requires signing a Non-Disclosure Agreement with them to gain access to this driver, but it's not the only driver available that targets the Intel GMA 500 / GMA 600 graphics core that's derived from the Imagination Technologies PowerVR SGX 535.
With the introduction of Intel's Poulsbo (GMA 500) chipset it marked a point at which Intel's Linux graphics support was no longer stellar, but as they had outsourced the graphics IP from Imagination Technologies, they could not provide an open-source driver stack like they do with their in-house IGPs. Not only was this Intel Poulsbo Linux driver closed-source, but the level of support was appalling and it was a bloody mess of a situation. The overall situation since has only become worse and even MeeGo (their own Linux OS) will be shipping without Intel's EMGD driver.
MeeGo, the joint project between Nokia and Intel that married the Maemo and Moblin operating systems together as one superior Linux-based operating system for Intel Atom netbooks and select Nokia devices, has a new stable 1.0 series update. This is actually the first update since MeeGo 1.0 was released for these mobile devices since the inaugural release in late May.
While Oracle may not be capable of releasing an OpenSolaris update (or even communicating their intentions), the Intel-Nokia team working on the MeeGo operating system that marries Moblin and Maemo is capable of getting work done. Today they have released their first spin of MeeGo Handset Project, which is the MeeGo variant designed for handsets like the Nokia N900.
While Intel had only released its xf86-video-intel 2.12 release candidate ten days ago and there was only one RC, yesterday afternoon they decided to go forward and make the final release. The xf86-video-intel 2.12.0 DDX driver is now available and they have also tagged their 2010Q2 driver package.
While there is the "Clover" branch of Mesa started by Zack Rusin for providing an OpenCL state tracker that can be used by Gallium3D hardware drivers, it hasn't yet amounted to much. The OpenCL state tracker is not yet working, hasn't been touched in months, and has yet to be integrated in the mainline Mesa code-base. However, as another GPGPU alternative, it looks like a CUDA back-end that's specific to Intel's open-source driver may end up being worked on.
Last October we reported on new capabilities within the Phoronix Test Suite that allowed performance regressions to be located within any code-base by leveraging our open-source testing framework with the git-bisect command to automatically traverse a tree in a binary manner until the Phoronix Test Suite spots the regression-causing commit. As an example we tracked down a regression in the mainline Linux kernel with the EXT4 file-system without any manual intervention. Now not only can we automatically track down any performance regressions, but the Phoronix Test Suite can also help you spot any functional regressions.
Intel has rolled out their first release candidate for their forthcoming xf86-video-intel 2.12 DDX driver update. The Intel 2.12 X.Org driver, as we have already mentioned twice brings faster 2D performance, among other changes.
MeeGo 1.0 was released last week and we found its netbook performance to be competitive and beat Ubuntu Netbook Remix, Moblin, and Fedora. While the performance of this joint project between Intel and Nokia may be nice, it doesn't improve the situation for the Poulsbo graphics situation under Linux.
A week ago we reported that Intel's next X.Org driver (the xf86-video-intel 2.12 DDX) would render text/glyphs faster thanks to optimizations done by Chris Wilson, but this was not all that was in store for this Intel Linux driver that's updated quarterly. With the most recent Git, there are more performance optimizations.
1022 Intel news articles published on Phoronix.