162306a36Sopenharmony_ciTODO 262306a36Sopenharmony_ci. Where should we do eviction (detatch_pages())? We aren't necessarily 362306a36Sopenharmony_ci accessing the pages via a GART, so maybe we need some other threshold 462306a36Sopenharmony_ci to put a cap on the # of pages that can be pin'd. 562306a36Sopenharmony_ci . Use mm_shrinker to trigger unpinning pages. 662306a36Sopenharmony_ci . This is mainly theoretical since most of these devices don't actually 762306a36Sopenharmony_ci have swap or harddrive. 862306a36Sopenharmony_ci. GEM/shmem backed pages can have existing mappings (kernel linear map, 962306a36Sopenharmony_ci etc..), which isn't really ideal. 1062306a36Sopenharmony_ci. Revisit GEM sync object infrastructure.. TTM has some framework for this 1162306a36Sopenharmony_ci already. Possibly this could be refactored out and made more common? 1262306a36Sopenharmony_ci There should be some way to do this with less wheel-reinvention. 1362306a36Sopenharmony_ci . This can be handled by the dma-buf fence/reservation stuff when it 1462306a36Sopenharmony_ci lands 1562306a36Sopenharmony_ci 1662306a36Sopenharmony_ciUserspace: 1762306a36Sopenharmony_ci. git://anongit.freedesktop.org/xorg/driver/xf86-video-omap 1862306a36Sopenharmony_ci 1962306a36Sopenharmony_ciCurrently tested on 2062306a36Sopenharmony_ci. OMAP3530 beagleboard 2162306a36Sopenharmony_ci. OMAP4430 pandaboard 2262306a36Sopenharmony_ci. OMAP4460 pandaboard 2362306a36Sopenharmony_ci. OMAP5432 uEVM 24