| Rob Clark | 8bb0daf | 2013-02-11 12:43:09 -0500 | [diff] [blame] | 1 | TODO |
| 2 | . Where should we do eviction (detatch_pages())? We aren't necessarily |
| 3 | accessing the pages via a GART, so maybe we need some other threshold |
| 4 | to put a cap on the # of pages that can be pin'd. |
| 5 | . Use mm_shrinker to trigger unpinning pages. |
| 6 | . This is mainly theoretical since most of these devices don't actually |
| 7 | have swap or harddrive. |
| 8 | . GEM/shmem backed pages can have existing mappings (kernel linear map, |
| 9 | etc..), which isn't really ideal. |
| 10 | . Revisit GEM sync object infrastructure.. TTM has some framework for this |
| 11 | already. Possibly this could be refactored out and made more common? |
| 12 | There should be some way to do this with less wheel-reinvention. |
| 13 | . This can be handled by the dma-buf fence/reservation stuff when it |
| 14 | lands |
| 15 | |
| 16 | Userspace: |
| 17 | . git://anongit.freedesktop.org/xorg/driver/xf86-video-omap |
| 18 | |
| 19 | Currently tested on |
| 20 | . OMAP3530 beagleboard |
| 21 | . OMAP4430 pandaboard |
| 22 | . OMAP4460 pandaboard |
| 23 | . OMAP5432 uEVM |