AMD Starts Rolling Out New Linux Driver Model, But Many Issues Remain 61
An anonymous reader writes: With the upcoming Linux 4.2 kernel will be the premiere of the new "AMDGPU" kernel driver to succeed the "Radeon" DRM kernel driver, which is part of AMD's long talked about new Linux driver architecture for supporting the very latest GPUs and all future GPUs. Unfortunately for AMD customers, there's still much waiting. The new open-source AMDGPU Linux code works for Tonga/Carrizo GPUs but it doesn't yet support the latest R9 Fury "Fiji" GPUs, lacks re-clocking/DPM for Tonga GPUs leading to low performance, and there are stability issues under high-load OpenGL apps/games. There's also the matter that current Linux users need to jump through hoops for now in getting the code into a working state with the latest kernel and forked versions of Mesa, libdrm, new proprietary microcode files, and the new xf86-video-amdgpu user-space driver.
Intel's linux support is impeccable (Score:4, Funny)
Re:Intel's linux support is impeccable (Score:5, Interesting)
Graphics is a patent minefield - it's one of the most legally aggressive areas in computing. Both parties have a strong commercial incentive to keep their technology secret - both to stop their rival stealing ideas, and to avoid inadvertently revealing any code that infringes upon a patent they were not aware of at the time. This slows down open-source development as every change needs to be examined by legal experts, and most legal experts are very cautious in their advice.
Re: (Score:1)
No. AMD is afraid that someone at Nvidia, IBM, or elsewhere, will look through their open source code and see that they are violating one of Nvidia/IBM/whomever's patent, and sues them for not just the linux driver, that opens the door to discovery on all their other platforms as well.
Re: (Score:2)
Graphics is a patent minefield - it's one of the most legally aggressive areas in computing. Both parties have a strong commercial incentive to keep their technology secret - both to stop their rival stealing ideas, and to avoid inadvertently revealing any code that infringes upon a patent they were not aware of at the time...
Sorry, no, it's not that at all. If there was any serious suspicion about that then the weenies would be all over the binary blobs with decompilers. Cost is no object for that sort of thing. The real reason is... a mystery. Fear of giving the competition a leg up on optimization algorithms? Maybe, but it seems a long shot. The real reason is probably just manager idiocy. I mean it's not like manager idiocy is in short supply.
Re: (Score:2)
Sorry, no, it's not that at all. If there was any serious suspicion about that then the weenies would be all over the binary blobs with decompilers. Cost is no object for that sort of thing.
I have no proof, but i also have no doubt this already happens. That, and reverse engineering the actual silicon.
Re: (Score:2)
Is reverse engineering not illegal. If it is illegal then any discovered result is unusable in court
It is if you happen to implement reverse-engineered tech covered by patents. As SuricouRaven succinctly put it, on graphics this is a minefield area.
Re: (Score:2)
Sorry, no, it's not that at all. If there was any serious suspicion about that then the weenies would be all over the binary blobs with decompilers. Cost is no object for that sort of thing.
I have no proof, but i also have no doubt this already happens. That, and reverse engineering the actual silicon.
And in-circuit emulators. Cost of one is a drop in the bucket compared to other efforts, often non-engineering, that goes into these sorts of things.
Re: (Score:1)
Because, beyond being able to display a negotiated resolution, Intels graphics chips don't do anything graphically impressive compared to what AMD & Nvidia are offering. To say Intels Graphics offering are on the same playing field as Nvidia or AMD, is apples to tomatoes.
Now, on the scope of driver model in Linux, yes. Intel does fare better. However, Intel isn't vying for improved gaming performance and support that AMD & Nvidia are. That is, unless Intel has decided to do a 180 with their stance o
Re: (Score:3, Informative)
Was impeccable. Every Intel chip post-Broadwell will require binary blobs, and Intel has yet to release the inteldrm module for Broadwell chips. Intel /used/ to be great, but now they're no better than nVidia.
Re: (Score:2)
Intel has yet to release the inteldrm module for Broadwell chips.
Let me hazard a guess why:
"the inteldrm module"
"inteldrm"
"drm"
Digital restrictions management strokes again.
Re: (Score:2, Informative)
Direct Rendering Manager?
Re: (Score:2, Flamebait)
Let me spell it out: Intel probably is having a hard time figuring out how to release a Direct Rendering Manager module that upholds its Digital Restrictions Management obligations.
Re: (Score:2)
You're not the first to be confused by the coincidence in terms.
ROI (Score:1)
Will investing hundreds of thousands of dollars supporting an OS with 1.6% [netmarketshare.com] of market share ever return a profit?
Re: (Score:1)
If AMD can land more semi-custom design wins for products that run the Linux kernel this investment will likely pay off in spades.
Re: (Score:2)
Re: (Score:3, Interesting)
Re: (Score:2)
Re: (Score:2)
ABI (Score:4, Insightful)
Stick to what you know, hairyfeet. Linux has no ABI because it does not want to encourage having random binary blobs on the users' systems. It's an ideological principle that will never change.
You clearly only understand the desktop arena, which is fine, but most Linux users are pretty happy with its current niche. The Linux ecosystem is probably better described as a set of tools for building an OS, and so you see things made from it like Android, Maemo/Meego/Sailfish/Whatever, SteamOS, various container-style projects, and the normal run of distributions. It's a development platform, mostly for server applications. Yeah, it would be nice in some ways if everyone was programmer enough to use Linux, but it's essential to no one, and the minority of people who have aftermarket video cards is not sufficient justification to bend over backwards for closed development models. Gamers may keep you in business, but they are a tiny part of the computing market. Now if you could point to something like ABI compatibility being an issue with GPU supercomputing, that might be more compelling.
There are arguments for a stable ABI. They are never going to get traction in a very successful open development paradigm, and desktop market share is not one of them. I'm sorry you've been riding this hobby horse for however many years, and I hate to tell you how silly you look doing it, but if that's your thing I guess you can keep it up until doomsday.
Re: (Score:2)
Re: (Score:2)
Failed driver model? By what definition of failure? Linux exists on millions upon millions of devices. Many Linux advocates may recommend NVIDIA hardware for gaming, but who cares? It's a tiny market. NVIDIA and AMD both have great Linux support for workstations and GPU computing, which is where the money is.
I hope that Linux never gets above 2% of the desktop, personally. Non-coders like you would ruin the platform, with exactly these kinds of moronic suggestions. I'd lay odds the majority of linux instanc
Re: ABI (Score:3)
As Tenebrousedge already stated, you're talking about websites related to gaming that are recommending NVIDIA.
I personally have used Linux exclusively (both on desktop and laptop computers) since 2003. I have also used AMD GPU & CPUs exclusively since 2008. The most graphics intensive thing I've ever run is Youtube in Firefox. My current video card in my primary workstation is from 2011.
My wife and young children have no problem navigating my linux based systems. We as a family use it for regular every
Re: (Score:3)
Is it worth investing hundreds of thousands of dollars on software to support tens of billions of dollars of research and development in a hardware products evelotion. Why yes, yes it is.
If you dont understand the benefits of open source why are you even posting here ?
Re: (Score:2)
Sorry but we are not talking about open source. We are talking about a company with shareholders who are interested in revenue spending lots of money for no direct return. They most probably have a budget and are limited in what they can do.
How does making drivers so Linux games play better support research?
Re: (Score:2)
They have spent money or research developing the HARDWARE. Hardware is useless without software.
Spending money on hardware and not software is like designing a car without an engine.
Open source drivers allows them to provide a better "user experience" on Linux platforms, but more importantly it enables "peer review" of the driver which helps to indentify bugs, promotes trust, and provides a better "user experience" on some platforms (ie Linux).
So yes we are talking about open source, and we are talkign abou
Re: (Score:2, Insightful)
Ok timeline time:
Linux AMD: catalyst for workstations, lets add a open source version for long term support, the open source version is pretty mature why don't we use the open source stuff and incorporate the workstation stuff in one driver and replace the closed source stuff with open source as it matures(this is today's news).
Linux nvidia: Closed driver for workstations.... Independent developers working on the side with zero nvidia support making open source driver.
At least give AMD credit for already be
Re: (Score:1)
Intel's Linux support is not, and has never been, "impeccable". Their lawyers strip out masses of useful information from documentation and vendor-supplied driver code just like AMD's. I've heard the impeccable line for years, even during the years when I couldn't even switch the resolution of X on my supposedly perfectly supported Intel hardware (fixed in the meantime, I imagine, but I insist on AMD/ATI again now). Which brings another point: why do you think AMD is "dragging their foot"? They provide
Re: (Score:2)
Not sure why AMD and nVidia keep dragging their foot. It makes no businesses sense
Whaaaaaaaaaaat? AMD and nVidia has a quite well optimized proprietary OpenGL engine following the latest standards, Intel has not. Intel is doing it open source because who could possibly benefit, Matrox? They have about 0.00% market share these days. AMD and nVidia already has engines doing it better. They got nothing to lose and a fair bit to gain from doing it open source. You have to be pretty deaf, blind and dumb to not see that AMD and nVidia has a competitive advantage they don't want to give away. M
Re:Intel's linux support is impeccable (Score:4, Interesting)
Not sure why AMD and nVidia keep dragging their foot. It makes no businesses sense
I have an older AMD card and the current AMD 'radeon' opensource driver in the kernel impeccably supports it.
In fact it's been about 5 years since I've had to ever even think about drivers for my video card. It just works, with great 3d performance, with every distro out of the box because it's supported by every default kernel and mesa libs. Even usb-flash distros.
Heck I've even run many steam games just to try them out. Had to install i386 packages for steam, but the AMD kernel driver, never had to worry about it.
If AMD can make the next driver arch as good and as open as the 'radeon' driver, I'll be buying AMD again soon.
By the way Nvidia, I wouldn't touch your junk with a 10ft pole. Come back when it's become mostly Free Software and is shipped default in every kernel and distro and I never have to worry about it.
Re: (Score:1)
Moo [deviantart.net]
What about (Score:1)
the systemd requirements?
Re: (Score:2)
systemd requires 0.0000000001553 Chuck Norris
SOP (Score:4, Insightful)
There's also the matter that current Linux users need to jump through hoops for now in getting the code into a working state with the latest kernel and forked versions ...
Seems the usual way Linux works.
Re: (Score:3)
Yeah, running "apt get update" which breaks the install and requires a full rebuild of a web server is such a good idea for a OS. It happened to me. I have yet to do an install or upgrade on a Linux server that does not refer to packages that I do not have or which may be incomparable with packages I do have.
I use Linux only because I have to (IIS is even worse).
Re: (Score:1)
So you've upgraded once and had an extremely unusual experience. Cool story bro.
Re: (Score:2)
no it was update as I was trying to update for heartbleed.
Re: (Score:2)
There's also the matter that current Linux users need to jump through hoops for now in getting the code into a working state with the latest kernel and forked versions ...
Seems the usual way Linux works.
AMD still beats Nvidia for stability. The Nvidia community driver that ships with the Linux kernel is unstable and the proprietary driver is a pain in the a**. I have never had stability issues with AMD drivers on Linux as long as I kept clear of the very bleeding edge.
Re: (Score:2)
the proprietary driver is a pain in the a**
Not sure how, at least for arch linux, pacman -S nvidia nvidia-libgl is a pain in the ass.
I suspect it's about as simple as this for most other distributions.
This is actually the main reason for me buying nvidia cards, they "just work" in linux and provide stellar performance.
Re: (Score:1)