Tweet
Hi
Im testing remotely without a graphics card device, specifically an Amazon AWS EC2 instance. My theory is this - I should be able to remote desktop into a standard Ubuntu 12.04 64-bit EC2 instance and utilise its cores (I understand llvmpipe takes advantage of up to 8) for 3D rendering with LLVMpipe. I posted on Oibafs forum, also Mesa-3D list and ec2ubuntu groups, no once seems to know.
So on physical systems there are standard drivers such as flgrx, nouveau, intel etc which are loaded in X then we can force llvmpipe with $ LIBGL_ALWAYS_SOFTWARE=1 appname
But what about a system without a graphics card? Can we access llvmpipe through the vesa fallback? As that is what the EC2 instance seems to be using. The is no pci hardware under $ lspci nor loaded kernal modules related to video. I have the gnome desktop manager going gdm, and gnome-shell installed, and a remote session using XVNC on display :1. And there is no display :0. How can I get llvmpipe going?
Im testing remotely without a graphics card device, specifically an Amazon AWS EC2 instance. My theory is this - I should be able to remote desktop into a standard Ubuntu 12.04 64-bit EC2 instance and utilise its cores (I understand llvmpipe takes advantage of up to 8) for 3D rendering with LLVMpipe. I posted on Oibafs forum, also Mesa-3D list and ec2ubuntu groups, no once seems to know.
So on physical systems there are standard drivers such as flgrx, nouveau, intel etc which are loaded in X then we can force llvmpipe with $ LIBGL_ALWAYS_SOFTWARE=1 appname
But what about a system without a graphics card? Can we access llvmpipe through the vesa fallback? As that is what the EC2 instance seems to be using. The is no pci hardware under $ lspci nor loaded kernal modules related to video. I have the gnome desktop manager going gdm, and gnome-shell installed, and a remote session using XVNC on display :1. And there is no display :0. How can I get llvmpipe going?