Lines Matching full:ccs

489  * DOC: Flat-CCS - Memory compression for Local memory
491 * On Xe-HP and later devices, we use dedicated compression control state (CCS)
495 * The memory required for the CCS of the entire local memory is 1/256 of the
497 * for the CCS data and a secure register will be programmed with the CCS base
500 * Flat CCS data needs to be cleared when a lmem object is allocated.
501 * And CCS data can be copied in and out of CCS region through
502 * XY_CTRL_SURF_COPY_BLT. CPU can't access the CCS data directly.
504 * I915 supports Flat-CCS on lmem only objects. When an objects has smem in
506 * content into smem. If the lmem object is Flat-CCS compressed by userspace,
508 * for such decompression. Hence I915 supports Flat-CCS only on lmem only objects.
510 * When we exhaust the lmem, Flat-CCS capable objects' lmem backing memory can
511 * be temporarily evicted to smem, along with the auxiliary CCS state, where
514 * the backing memory back to lmem, which includes restoring the saved CCS state,
518 * {lmem, smem}, objects are treated as non Flat-CCS capable objects.
549 * The XY_CTRL_SURF_COPY_BLT instruction is used to copy the CCS in emit_copy_ccs()
550 * data in and out of the CCS region. in emit_copy_ccs()
558 * 1024 blocks of 256 bytes of CCS represent a total 256KB of CCS. in emit_copy_ccs()
560 * 256 KB of CCS represents 256 * 256 KB = 64 MB of LMEM. in emit_copy_ccs()
647 * will be taken for the blt. in Flat-ccs supported in calculate_chunk_sz()
725 * When there is a eviction of ccs needed smem will have the in intel_context_migrate_copy()
726 * extra pages for the ccs data in intel_context_migrate_copy()
851 * should be safe to directly copy the CCS in intel_context_migrate_copy()
853 * initialised the CCS aux state when first in intel_context_migrate_copy()
857 * need to copy the CCS state as-is. in intel_context_migrate_copy()
865 * While we can't always restore/manage the CCS in intel_context_migrate_copy()
867 * the CCS state from the previous user, so make in intel_context_migrate_copy()
1052 * ccs surface in intel_context_migrate_clear()