Dunajski, Bartosz b4f79e036f Map/unmap enqueue fixes [5/n]: Unify offset calculation ...

Change-Id: I53eafe89532d43c5cf5139ed3fac0a87619dc7a3

2018-02-21 20:12:52 +01:00

..

CMakeLists.txt

Map/unmap enqueue fixes [1/n]: Unify Buffer and Image paths

2018-02-09 17:35:03 +01:00

command_queue_hw.h

For HostPtr surfaces of enqueue calls use GPU address

2018-02-14 15:44:27 +01:00

command_queue_hw.inl

Map/unmap enqueue fixes [1/n]: Unify Buffer and Image paths

2018-02-09 17:35:03 +01:00

command_queue.cpp

Map/unmap enqueue fixes [5/n]: Unify offset calculation

2018-02-21 20:12:52 +01:00

command_queue.h

Hint for map/unmap when queue is blocked

2018-02-16 15:57:44 +01:00

cpu_data_transfer_handler.cpp

Map/unmap enqueue fixes [5/n]: Unify offset calculation

2018-02-21 20:12:52 +01:00

dispatch_walker_helper.h

Add Preemption WA waModifyVFEStateAfterGPGPUPreemption for WALKER replay issue

2018-01-02 17:45:40 +01:00

dispatch_walker_helper.inl

Add Preemption WA waModifyVFEStateAfterGPGPUPreemption for WALKER replay issue

2018-01-02 17:45:40 +01:00

dispatch_walker.h

Instrumentation refactoring

2018-02-15 23:58:24 +01:00

enqueue_barrier.h

Initial commit

2017-12-21 00:45:38 +01:00

enqueue_common.h

Improve thread arbitration policy programming

2018-02-20 11:05:54 +01:00

enqueue_copy_buffer_rect.h

For HostPtr surfaces of enqueue calls use GPU address

2018-02-14 15:44:27 +01:00

enqueue_copy_buffer_to_image.h

For HostPtr surfaces of enqueue calls use GPU address

2018-02-14 15:44:27 +01:00

enqueue_copy_buffer.h

Initial commit

2017-12-21 00:45:38 +01:00

enqueue_copy_image_to_buffer.h

For HostPtr surfaces of enqueue calls use GPU address

2018-02-14 15:44:27 +01:00

enqueue_copy_image.h

For HostPtr surfaces of enqueue calls use GPU address

2018-02-14 15:44:27 +01:00

enqueue_fill_buffer.h

Initial commit

2017-12-21 00:45:38 +01:00

enqueue_fill_image.h

For HostPtr surfaces of enqueue calls use GPU address

2018-02-14 15:44:27 +01:00

enqueue_kernel.h

Add Preemption WA waModifyVFEStateAfterGPGPUPreemption for WALKER replay issue

2018-01-02 17:45:40 +01:00

enqueue_marker.h

Initial commit

2017-12-21 00:45:38 +01:00

enqueue_migrate_mem_objects.h

Initial commit

2017-12-21 00:45:38 +01:00

enqueue_read_buffer_rect.h

For HostPtr surfaces of enqueue calls use GPU address

2018-02-14 15:44:27 +01:00

enqueue_read_buffer.h

Map/unmap enqueue fixes [4/n]: Return slice/row pitch

2018-02-20 14:30:35 +01:00

enqueue_read_image.h

For HostPtr surfaces of enqueue calls use GPU address

2018-02-14 15:44:27 +01:00

enqueue_svm.h

Initial commit

2017-12-21 00:45:38 +01:00

enqueue_write_buffer_rect.h

For HostPtr surfaces of enqueue calls use GPU address

2018-02-14 15:44:27 +01:00

enqueue_write_buffer.h

Map/unmap enqueue fixes [4/n]: Return slice/row pitch

2018-02-20 14:30:35 +01:00

enqueue_write_image.h

For HostPtr surfaces of enqueue calls use GPU address

2018-02-14 15:44:27 +01:00

finish.h

Fix resource destruction scheme on device closure.

2018-01-15 18:45:48 +01:00

flush.h

Initial commit

2017-12-21 00:45:38 +01:00

local_id_gen_avx2.cpp

Initial commit

2017-12-21 00:45:38 +01:00

local_id_gen_sse4.cpp

Initial commit

2017-12-21 00:45:38 +01:00

local_id_gen.cpp

Initial commit

2017-12-21 00:45:38 +01:00

local_id_gen.h

Initial commit

2017-12-21 00:45:38 +01:00

local_id_gen.inl

Initial commit

2017-12-21 00:45:38 +01:00

local_work_size.cpp

For devices with small HW thread count, limit the available pool of LWS.

2018-01-24 14:30:39 +01:00