Bug 901500

Summary: GPU offload (dynamic hybrid graphics) support
Product: [Fedora] Fedora Documentation Reporter: Tomas Miljenović (TomasM) <Tomas>
Component: release-notesAssignee: Release Notes Tracker <relnotes>
Status: CLOSED WONTFIX QA Contact: Karsten Wade <kwade>
Severity: medium Docs Contact:
Priority: unspecified    
Version: develCC: me, wb8rcr, zach
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2013-02-10 18:32:54 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
sample glxinfo output (with DRI_PRIME set and unset) none

Description Tomas Miljenović (TomasM) 2013-01-18 10:35:23 UTC
Created attachment 682312 [details]
sample glxinfo output (with DRI_PRIME set and unset)

Fedora 18 supports GPU offloading for many systems with dynamic switchable graphics support (AMD PowerXPress, NVIDIA Optimus or similar). Setting DRI_PRIME=1 will render applications with the discrete graphics card on supported hardware.

Is this something that should be in the release notes?
- AFAIK Fedora 18 is the first distro released with dynamic gpu offload support - this is a big deal for most laptops with two graphics cards, as the more powerful card is just a power drain for most Linux users at the moment.
- As per IRC discussion with randomuser, I haven't found upstream docs on how to actually enable this functionality and think it would be useful to document somewhere.

I can pull snippets from kernel, Xorg, libXrandr and mesa changelogs to confirm all of this, but am hopeful a graphics dev (this is largely David Airlie's work) can verify this instead.  Dave has come across a laptop model which doesn't work with the release-day code; I'm not sure if his warrants some ambiguity or a technical disclaimer for 'supported hardware'.

Comment 1 Pete Travis 2013-02-10 18:32:54 UTC
This sounds like a valuable feature indeed, but perhaps an edge case. Documenting an implementation of this technology may be too in depth for the normal scope of the release notes, and we didn't come up with the upstream docs.  I'm looking forward to developers taking advantage of this, but it unfortunately did not make the Release Notes.