On 2018-06-15 02:57 AM, Michel Dänzer wrote: > On 2018-06-14 09:49 PM, Leo Li wrote: >> On 2018-06-14 12:57 PM, Michel Dänzer wrote: >>> On 2018-06-01 06:03 PM, sunpeng.li at amd.com wrote: >>>> From: "Leo (Sunpeng) Li" <sunpeng.li at amd.com> >>>> >>>> This ended up being different enough from v2 to warrant a new >>>> patchset. Per >>>> Michel's suggestions, there have been various optimizations and >>>> cleanups. >>>> Here's what's changed: >>>> >>>> * Cache DRM color management property IDs at pre-init, >>>>     * instead of querying DRM each time we need to modify color >>>> properties. >>>> >>>> * Remove drmmode_update_cm_props(). >>>>     * Update color properties in drmmode_output_get_property() instead. >>>>     * This also makes old calls to update_cm_props redundant. >>>> >>>> * Get rid of fake CRTCs. >>>>     * Previously, we were allocating a fake CRTC to configure color >>>> props on >>>>       outputs that don't have a CRTC. >>>>     * Instead, rr_configure_and_change_cm_property() can be easily >>>> modified to >>>>       accept NULL CRTCs. >>> >>> Is it currently ever the case in the hardware / kernel, or expected to >>> ever be, that colour management is supported for some but not all CRTCs >>> of a GPU? >>> >> >> This was more of an effort to align with what DRM allows, which is >> per-CRTC cm support and LUT sizes. I'm not aware of any current or >> future hardware where this is the case though. It was a relatively >> simple thing to implement, so I thought, might as well. >> >>> If not, the LUT sizes could be tracked once instead of per CRTC, and at >>> least the (DE)GAMMA_LUT_SIZE properties could always return the proper >>> values, even if the output currently isn't associated with a CRTC. >>> >> >> My original take was that it's best to support what the framework >> allows. But it does sound like this would have more utility to the >> client, and simplify the code as well. > > Right. In contrast to the generic modesetting driver, we can take > advantage of properties of our GPUs in this driver. > Ping! FYI, all the new patches are v2, with the exception of 6/7, which is a v3. (On second thought, should have started a new thread :) ) Leo >
