Commit | Line | Data |
---|---|---|
cd5351f4 | 1 | TODO |
cd5351f4 | 2 | . add video decode/encode support (via syslink3 + codec-engine) |
50836770 RC |
3 | . NOTE: with dmabuf this probably could be split into different driver |
4 | so perhaps this TODO doesn't belong here | |
cd5351f4 RC |
5 | . where should we do eviction (detatch_pages())? We aren't necessarily |
6 | accessing the pages via a GART, so maybe we need some other threshold | |
7 | to put a cap on the # of pages that can be pin'd. (It is mostly only | |
8 | of interest in case you have a swap partition/file.. which a lot of | |
9 | these devices do not.. but it doesn't hurt for the driver to do the | |
10 | right thing anyways.) | |
11 | . Use mm_shrinker to trigger unpinning pages. Need to figure out how | |
12 | to handle next issue first (I think?) | |
13 | . Note TTM already has some mm_shrinker stuff.. maybe an argument to | |
14 | move to TTM? Or maybe something that could be factored out in common? | |
15 | . GEM/shmem backed pages can have existing mappings (kernel linear map, | |
16 | etc..), which isn't really ideal. | |
17 | . Revisit GEM sync object infrastructure.. TTM has some framework for this | |
18 | already. Possibly this could be refactored out and made more common? | |
19 | There should be some way to do this with less wheel-reinvention. | |
f7f9f453 RC |
20 | . Solve PM sequencing on resume. DMM/TILER must be reloaded before any |
21 | access is made from any component in the system. Which means on suspend | |
22 | CRTC's should be disabled, and on resume the LUT should be reprogrammed | |
23 | before CRTC's are re-enabled, to prevent DSS from trying to DMA from a | |
24 | buffer mapped in DMM/TILER before LUT is reloaded. | |
cd5351f4 RC |
25 | |
26 | Userspace: | |
27 | . git://github.com/robclark/xf86-video-omap.git | |
28 | ||
29 | Currently tested on | |
30 | . OMAP3530 beagleboard | |
31 | . OMAP4430 pandaboard | |
32 | . OMAP4460 pandaboard |