[00:00] headlessness was also discussed. I seem to recall they thought they had a way around that but don't remember the specifics. [00:27] RAOF: so yeah the SDP's have different pci ids than shipping hardware will and userspace doesnt support it incase you play around with the sdp, need to patch libdrm and mesa to work with 8086:0c16 [00:27] Sarvatt: Heh. I noticed your patch. [00:28] thats just the mobile one, libdrm and mesa dont support the desktop version already [00:28] I should turn on the IGFX again, see if i915.ko actually binds to the hardware. [00:34] it didnt before? didnt you have a desktop one? [00:35] 8086:0c16? === Amaranthus is now known as Amaranth [08:42] RAOF: ohh shiny new libdrm also has the prime calls, I want shiny :D [08:42] but it can wait a little [10:05] mlankhorst: What can we *do* with those prime calls? :) [10:26] oh for nouveau they just seem to be wrappers, but mostly provides an api around prime. === cnd` is now known as cnd === yofel_ is now known as yofel [18:48] RAOF: do you still want to have fun with r600? === pzanoni_ is now known as pzanoni [20:14] bryceh, hey [20:14] seb128, hey [20:14] bryceh, how are you? [20:14] seb128, fine, you? [20:14] bryceh, I'm good thanks [20:15] bryceh, could you put some lines in https://wiki.ubuntu.com/QuantalQuetzal/TechnicalOverview/Alpha3 about what happened in xorg world between a2 and a [20:15] 3 [20:15] seb128, certainly [20:15] bryceh, thanks! ;-) [20:16] seb128, under Graphics and Display, or Desktop? [20:16] bryceh, graphics and display is fine [20:16] bryceh, hum, wait [20:16] oh that's for known issues [20:17] bryceh, no, under "Ubuntu" [20:17] ok [20:17] right ;-) [20:26] and swapped out my wrt54gl for a wndr3800 [20:27] I created a -an (5ghz) network and a 2.4ghz -guest network that should only connect to internet, but is passwordless. You use ssl for everything right? [21:06] seb128, ok, think I've got the highlights there. Poking through changelogs... lots been updated but not much that users would notice. Think I got the main points. [21:08] bryceh: I think optimus support will land only properly in 3.7 kernel, so it's going to have to wait for the r-release to be useful. I could probably keep things in a ppa for quantal though [21:10] a bunch of important things will go in at 3.6 though. intel flushing list removal, deferred fput, some cleanups to the api. [21:11] seb128, I didn't mention anything on wayland; robert or raof should probably add a couple sentences on that if it's ready for users [21:12] bryceh, well, those are a3 notes, so those ppa packages don't really qualify [21:14] mlankhorst, *nod* yeah I didn't say anything about hybrid support although I know people will want to know status on that. Sounds like 12.10 won't be a solution for hybrid users [21:15] bryceh: actually all the parts except kernel will be in for 12.10 ;-) [21:16] X 1.13 will support it, libdrm needs a bump but will support it, drivers need some love but adding support is trivial [21:17] so likely those will be ready in time [21:21] mlankhorst, interesting; so what exactly is needed on the kernel side? [21:21] dma-fence and dmabufmgr [21:22] and some code to hook it up for i915 and nouveau [21:22] RAOF called dibs on r600, i might take him up on it soon :) [21:22] any of it potentially backportable? [21:22] probably but I think we would really want to wait until it's upstream first [21:24] so at least the api is reviewed [21:26] at first I had a total crap api for fences hardcoded directly into dmabufmgr, but now I dumbed down the dma-fence that was created in response to it and hooked it up to dmabufmgr. So it's actually usable and understandable by others. :-)