Keith Packard Making Progress With DRM Leases, Adding Vulkan Support
Keith Packard's latest hacking in the open-source world has been around DRM leases support as part of his work under contract with Valve for better supporting VR HMDs on Linux.
He's been making good progress on this Valve work and today issued a new status update. His proposed Linux kernel and RandR changes are being reviewed and looking good, so next he's been looking at figuring out how to tie in his DRM leasing capabilities into Vulkan, for being able to hand off the VR HMD device to be used directly by Vulkan-powered VR Linux games.
Keith has been experimenting with a VK_KEITHP_kms_display Vulkan extension for getting information about KMS-driven displays, including DRM node / lease information. There's also the existing VK_KHR_display extension for being able to access a raw hardware device.
He has a hacked-up Vulkan cube demo working, but commented of the Mesa Vulkan code, "I'm not satisfied with the mesa code at this point; there's a bunch of code in the radeon driver which should be in the vulkan WSI bits, and the vulkan WSI bits should probably not have the KMS interfaces wired in. I'll ask around and see what other Mesa developers think I should do before restructuring it further; I'll probably have to rewrite it all at least one more time before it's ready to upstream."
More details via his post at KeithP.com.
He's been making good progress on this Valve work and today issued a new status update. His proposed Linux kernel and RandR changes are being reviewed and looking good, so next he's been looking at figuring out how to tie in his DRM leasing capabilities into Vulkan, for being able to hand off the VR HMD device to be used directly by Vulkan-powered VR Linux games.
Keith has been experimenting with a VK_KEITHP_kms_display Vulkan extension for getting information about KMS-driven displays, including DRM node / lease information. There's also the existing VK_KHR_display extension for being able to access a raw hardware device.
He has a hacked-up Vulkan cube demo working, but commented of the Mesa Vulkan code, "I'm not satisfied with the mesa code at this point; there's a bunch of code in the radeon driver which should be in the vulkan WSI bits, and the vulkan WSI bits should probably not have the KMS interfaces wired in. I'll ask around and see what other Mesa developers think I should do before restructuring it further; I'll probably have to rewrite it all at least one more time before it's ready to upstream."
More details via his post at KeithP.com.
1 Comment