And if so why? What does X do for me beyond piping my rendering commands to the graphics card driver?
I'm not clear on the relationship X - OpenGL. I've searched the internet but couldn't find a concise answer.
If it matters, assuming a minimal modern distribution, like a headless Ubuntu 13 machine.
With the current drivers: Yes.
And if so why?
Because the X server is the host for the actual graphics driver talking to the GPU. At the moment Linux GPU drivers require a X server that gives them an environment to live in and a channel to the kernel interfaces to talk through with the GPU.
On the DRI/DRM/Gallium front a new driver model has been created that allows to use the GPU without an X server, for example using the EGL-API. However only a small range of GPUs is supported by this right now; most Intel and AMD; none NVidia.
I'm not clear on the relationship X - OpenGL
I covered that in detail in the SO answers found at https://stackoverflow.com/a/7967211/524368 and https://stackoverflow.com/a/8777891/524368
In short the X server acts like a "proxy" to the GPU. You send the X server commands like "open a window" or "draw a line there". And there's an extension to the X protocol called "GLX", where each OpenGL command gets translated into a stream of GLX/X opcodes and the X server executes those commands on the GPU on behalf of the calling client. Also most OpenGL/GLX implementations provide a mechanism to bypass the X server if the client process could actually talk directly to the GPU (because it runs on the same machine as the X server and has permissions to access the kernel API); that is called Direct Rendering. It however still requires the X server for opening the window, creating the context and to general housekeeping.
Update due to comment
Also if you can live without GPU acceleration, you can use Mesa3D using the osmesa (off-screen mesa) mode and the LLVMpipe software rasterizer.
With Linux 3.12: Not any more.
Offscreen rendering is what DRM render nodes are for, according to the commit. See the developer's blog for a better explanation.
TLDR:
A render node (/dev/dri/renderD<num>) appears as a GPU with no screens attached.
As for how exactly one is supposed to make use of this, the (kernel) developer only has very general advice for userspace infrastructure. Nevertheless, it is fair to assume the feature to be nothing short of a show-enabler for Wayland and Mir, as clients won't be able to render on-screen any more.
The wikipedia entry has some more pointers.
Related
Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.
This question does not appear to be about programming within the scope defined in the help center.
Closed 2 years ago.
The community reviewed whether to reopen this question 5 months ago and left it closed:
Original close reason(s) were not resolved
Improve this question
OpenGL and Windows Remote don't play along nicely.
Solutions for this are dependent on the use case and answers are fragmented across the vast depths of the net.
This is a write-up I wish existed when I started researching this, both for coders and non-coders.
Problem:
A RDP session of Windows does not expose the graphics card, at least not directly. For instance you cannot change the desktop resolution and GraphicsCard drivers usually just disable their setting menus. Starting a OpenGL context higher than v1.1 fails because of this. The, especially in support IRCs, often suggested "Don't use WindowsRemote" is unfortunately not an option for many. In many corporate environments Windows Remote is a constantly used tool and an app has to work there as well.
Non-Coder workarounds
You can start the OpenGL program, allowing it to see the graphics card, create an opengl context and then connect via WindowsRemote. This always works, as Windows remote just transfers the window content. This can be accomplished by:
A batch script, that closes the session and starts the program, allowing you to connect to the program already running. (Source)
Using VNC or other to remote into the machine, start the program and then switch to Windows Remote. (Simple VNC programm, also with a portable client)
Coder workarounds
(Only for OpenGL ES)Translate OpenGL to DirectX. DirectX works under Windows Remote flawselly and even has a Software rendering fallback built into DX11 if something fails.
Use the ANGLE Project to do this at run-time. This is what QT officially suggests you do and how Chrome and Firefox implement WebGL. (Source)
Switch to software rendering as a fall back. Some CAD software like 3dsMax does this for instance:
Under SDL2 you can use SDL_CreateSoftwareRenderer (Source)
Under GLFW version 3.3 will release OSMesa (Mesa's off screen rendering), in the mean time you can build the Github version with -DGLFW_USE_OSMESA=TRUE, but I personally still struggle to get that running (Source)
Directly use Mesa's LLVM pipe for a fast OpenGL implementation. (Source)
Misc:
Use OpenGL 1.1: Windows has a built in implementation of OpenGL 1.1 and
earlier. Some game engines have a built in fall back to this and thus
work under Windows Remote.
Apparently there is a middle-ware, that allows for even OpenGL 4 over Windows Remote, but it's part of a bigger package and is a commercial solution. (Source)
Any other solutions or corrections are greatly appreciated.
[10] Nvidia -> https://www.khronos.org/news/permalink/nvidia-provides-opengl-accelerated-remote-desktop-for-geforce-5e88fc2035e342.98417181
According to this article it seems that now RDP handles newer versions of Direct3D and OpenGL on Windows 10 and Windows Server 2016, but by default it is disabled by Group Policy.
I suppose that for performance reasons, using a hardware graphics card is disabled, and RDP uses a software-emulated graphics card driver that provides only some baseline features.
I stumbled upon this problem when trying to run Ultimaker CURA over standard Remote Desktop from a Windows 10 client to a Windows 10 host. Cura shouted "cannot initialize OpenGL 2.0 context". I also noticed that Repetier Host's "preview" window runs terribly slow, and Repetier detects only an OpenGL 1.1 card. Pretty much fits the "only baseline features" description.
By running gpedit.msc then navigating to
Local Computer Policy\Computer Configuration\Administrative Templates\Windows Components\Remote Desktop Services\Remote Desktop Session Host\Remote Session Environment
and changing the value of
Use hardware graphics adapters for all Remote Desktop Services sessions
I was able to successfully run Ultimaker CURA via with no issues, and Repetier-Host now displays OpenGL 4.6, and everything finally runs fast as it should.
Note from genpfault:
As usual, this Policy is kept in the HKLM registry group in
HKEY_LOCAL_MACHINE\SOFTWARE\Policies\Microsoft\Windows NT\Terminal Services
Set REG_DWORD:bEnumerateHWBeforeSW to 1 to turn ON using GPUs in RDP.
OpenGL works great by RDP with professional Nvidia cards without anything like virtual machines and RemoteFX. For Quadro (Quadro 4000 tested) you need driver 377.xx. For M60 you can use the same driver. If you want to use last driver with M60, you have to change the driver mode to WDDM mode (see c:\Program Files\NVIDIA Corporation\NVSMI\nvidia-smi.1.pdf). It is possible that there are some problems with licensing in this last case.
Some people recommend using "tscon.exe" if you can: https://stackoverflow.com/a/45723167/32453 or using a scheduler to do it on native hardware: https://stackoverflow.com/a/41839102/32453 or creating a group policy:
https://community.esri.com/thread/225251-enabling-gpu-rendering-on-windows-server-2016-windows-10-rdp
maybe copy opengl32.dll (or opengl64.dll) to your executable's dir: https://blender.stackexchange.com/a/73014 and newer version of the dll: https://fdossena.com/?p=mesa/index.frag
Remote Desktop and OpenGL does not play very well. When you connect to a Windows box the OpenGL Driver is unloaded and you end up with software emulation of OpenGL.
When you disconnect from the Windows box the OpenGL driver is not reloaded. This causes issues when you are running tests on the machine as you have to physically login to the machine to reset the drivers.
The solution I ended up using was to:
Disable Remote Desktop.
Delete all other software for remote desktop access. Because if it's used for logging in remotely the current set of drivers loaded may be messed up.
Install NoMachine
NoMachine is my personal favourite (when it does not play up) for a number of reasons:
Hardware acceleration of compression (video of desktop).
Works on Windows and Linux.
Works well on low-bandwidth connections especially if the client and server have the necessary hardware for compression of the data stream.
On Linux you get your desktop as you last left it when you were sitting in front of the machine.
On Windows it does not affect OpenGL.
currently free for personal and commercial use. Do check the licence in case it's changed.
When NoMachine plays up it hogs the CPU but this happens rarely. It is however in active development
Others to consider:
TurboVNC
TightVNC
TeamViewer - only free for personal use.
I'm trying too use DirectX desktop duplication API.
I tried running exmaples from
http://www.codeproject.com/Tips/1116253/Desktop-Screen-Capture-on-Windows-via-Windows-Desk
And from
https://code.msdn.microsoft.com/windowsdesktop/Desktop-Duplication-Sample-da4c696a
Both of these are examples of screen capture using DXGI.
I have NVIDIA GeForce GTX 1060 with Windows 10 Pro on the machine. It has Intel™ Core i7-6700HQ processor.
These examples work perfectly fine when NVIDIA Control Panel > 3D Settings is selected to Auto select processor.
However if I set the setting manually to NVIDIA Graphics Card the samples stop working.
Error occurs at the following line.
//IDXGIOutput1* DxgiOutput1
hr = DxgiOutput1->DuplicateOutput(m_Device, &m_DeskDupl);
Error in hr(HRESULT) is DXGI_ERROR_UNSUPPORTED 0x887A0004
I'm new to DirectX and I don't know the issue here, is DirectX desktop duplication not supported on NVIDIA ?
If that's the case then is there a way to select a particular processor at the start of program so that program can run with any settings ?
#Edit
After looking around I asked the developer (Evgeny Pereguda) of the second sample project on codeproject.com
Here's a link to the discussion
https://www.codeproject.com/Tips/1116253/Desktop-Screen-Capture-on-Windows-via-Windows-Desk?msg=5319978#xx5319978xx
Posting the screenshot of the discussion on codeproject.com in case original link goes down
I also found an answer on stackoverflow which unequivocally suggested that it could not be done with the desktop duplication API referring to support ticket at microsoft's support site https://support.microsoft.com/en-us/help/3019314/error-generated-when-desktop-duplication-api-capable-application-is-ru
Quote from the ticket
This issue occurs because the DDA does not support being run against
the discrete GPU on a Microsoft Hybrid system. By design, the call
fails together with error code DXGI_ERROR_UNSUPPORTED in such a
scenario.
However there are some applications which are efficiently duplicating desktop on windows in both modes (integrated graphics and discrete) on my machine. (https://www.youtube.com/watch?v=bjE6qXd6Itw)
I have looked into the installation folder of the Virtual Desktop on my machine and can see following DLLs of interest
SharpDX.D3DCompiler.dll
SharpDX.Direct2D1.dll
SharpDX.Direct3D10.dll
SharpDX.Direct3D11.dll
SharpDX.Direct3D9.dll
SharpDX.dll
SharpDX.DXGI.dll
SharpDX.Mathematics.dll
Its probably an indication that this application is using DXGI to duplicate desktop, or may be the application is capable of selecting a specific processor before it starts.
Anyway the question remains, is there any other efficient method of duplicating desktop in both modes?
The likely cause is certain internal limitation for Desktop Duplication API, described in Error generated when Desktop Duplication API-capable application is run against discrete GPU:
... when the application tries to duplicate the desktop image against the discrete GPU on a Microsoft Hybrid system, the application may not run correctly, or it may generate one of the following errors:
Failed to create windows swapchain with 0x80070005
CDesktopCaptureDWM: IDXGIOutput1::DuplicateOutput failed: 0x887a0004
The article does not suggest any other workaround except use of a different GPU (without more specific detail as for whether it is at all achievable programmatically):
To work around this issue, run the application on the integrated GPU instead of on the discrete GPU on a Microsoft Hybrid system.
Microsoft introduced a registry value that can be set programmatically to control which GPU an application runs on. Full answer here.
I'm trying to understand the userland part of the Raspberry Pi graphics driver code from https://github.com/raspberrypi/userland
My understanding so far is:
- a firmware blob runs in the GPU and offers an OpenGL-like interface which, on lower levels, is based on message (byte-array) passing on top of one of multiple 28-bit-word FIFOs called VCHIQ (the other VCHIQ queues are irrelevant for graphics)
- on the CPU part, OpenGL calls are turned into messages to the GPU. Access to the low-level facility (either the message queue or VCHIQ -- I haven't found that part yet in the code) requires a Linux kernel module, but no high-level logic happens in there.
- the GPU part is closed, but that's okay for my purposes. The (ARM) CPU part is, AFAIK, open
My ultimate goal is to get communication with the GPU working on bare metal (without Linux), but with the closed firmware blob intact. As a first goal, I want to understand how an OpenGL call is actually passed to the GPU. Anything beyond that is not part of this question.
However, I'm stuck at finding the actual code for this. The OpenGL calls use RPC_CALL* and in turn RPC_DO, which calls khronos_server_lock_func_table(). However, that function seems to be missing from the code, and to my surprise, I couldn't find anything useful about it on Google.
My questions:
- am I still on the ARM CPU side, or did I move to GPU land without noticing? If the latter is the case, where did I cross that line?
- Assuming I'm still on the CPU side -- where is the code for that function? Is it open at all, or do we actually have closed parts left around on the CPU side here? All sources on the web seem to indicate that the code for the CPU is 100% open.
- at which point does the implementation of the C OpenGL functions actually send a message to the GPU? I'm somewhat expecting a call to the kernel functionality that represents VCHIQ to be happening at some point, probably implemented as a device file.
I don't fully understand how do you intend to access the GPU without using Linux, and I am not that familiar with the technicalities, but some time ago I've been digging into the GPU for my private project so I'll tell you what I know.
The GPU is VideoCore IV and its documentation is available on Broadcom's website.
Also, on the Raspberry Pi Wiki you can see on the picture on the left that VCHIQ is in the kernel driver, so you might look for the implementation details in the kernel's source code.
Maybe this might be of some help too: VideoCore IV Programmer's Manual. About the document:
This is a independent documentation project based on a combination of static analysis and trial and error on real hardware. This work is 100% independent from and not sanctioned by or connected with Broadcom or its agents. No Broadcom documents or materials were used beyond those publicly available.
As for the software itself, The Khronos Group provides OpenGL ES and OpenVG implementation, but it's not open source. You can get the documentation from their website, but I doubt you'll find anything on such low level.
Hope it helps.
Is there a way to start an application with OpenGL >= 3 on a remote machine?
Local and remote machine run on Linux.
More precisely, I have the following problem:
I have an application that uses Qt for GUI stuff and OpenGL for 3D rendering.
I want to start this application on several remote machines because the program does some very time consuming computation.
Thus, I created a version of my program that does not raise a window. I use QGuiApplication, QOffscreenSurface, and a framebuffer object as rendertarget.
BUT: When I start the application on a remote machine (ssh -Y remotemachine01 myapp) I only have OpenGL version 2.1.2. When I start the application locally (on the same machine, I have opengl 4.4). I suppose the X forwarding is the problem.
So I need a way to avoid X forwarding.
Right now there's no clean solution, sorry.
GLX (the OpenGL extension to X11 which does the forwarding stuff) is only specified up to OpenGL-2.1, hence your inability to forward a OpenGL-3 context. This is actually a ridiculous situation, because the "OpenGL-3 way" is much better suited for indirected rendering, than old fashioned OpenGL-2.1 and earlier. Khronos really needs to get their act together and specify GLX-3.
Your best bet would be either to fall back to a software renderer on the remote side and some form of X compression. Or use Xpra backed by on GPU X11 server; however that only works for only a single user at a time.
In the not too far future the upcomming Linux graphics driver models will allow for remote GPU rendering execution by multiple users sharing graphics resources. But we're not there yet.
We are trying to setup a server with Multiple Tesla M2050 to run with OpenGL.
The current setup is as follows : Ubuntu 12.04 with NVidia Drivers. We have setup the xorg.conf with separate devices identified by BUS ID.
Now we have tied an X server each with display which in turn is tied to each device and our code is attached to each of these X servers. But somehow only one X session seems to work out alright. The other one produces garbled output and while watching it from nvidia-smi, we notice that when the garbled output is being produced the GPU's are not at all used.
Could someone verify that our setup seems reasonable? The other thing we noticed was that, it was only the first X server that was started is the one that has the issue.
EDIT : This is in headless mode.
A problem with multiple X servers is, that each server may grab the active VT and hence disable the other X server's rendering output. This can be avoided. But I think in your situation good ole' "Zaphod Mode" would suit your needs far better:
Zaphod mode is a single X server, controlling multiple Devices, each with its own Monitor forming a Screen, joined in a single screen layout. This is not TwinView or Xinerama! In Zaphod mode you can not move windows between Screens, i.e. each Screen acts on its own.