Related-To: NEO-12197
Currently for new resources user thread must wait before submitting
actual workload. With this commit, instead of waiting on user thread,
request is sent to background ULLS controller thread and additional
semaphore is programmed. ULLS controller will perform actual wait
and signal semaphore when paging fence reaches required value.
Signed-off-by: Szymon Morek <szymon.morek@intel.com>
Make sure local mem alloc size atomic array is initialized with 0.
Add debug breaks to catch possible overflow on unregistering
allocations.
Related-To: NEO-11356
Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
Track memory used by memory allocations. System and local per device.
Will be used for heuristics in memory pooling.
Related-To: NEO-11356
Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
computeMaxNeededSubSliceSpace is no longer needed as getHighestEnabledSubSlice
already determines maximum index from all enabled subslices
Related-To: NEO-12073
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
avoid redefinition issue in case of integrating multiple xe drm versions
define mock drm xe in inl file to provide xe definitions for mock members
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
The dependency towards `Drm` is unnecessary and only makes testing more
difficult. Instead, dependency towards `IoctlHelper` alone only is
sufficient.
Related-To: NEO-10158
Signed-off-by: Maciej Bielski <maciej.bielski@intel.com>
Implemented device property query API for determining
support capabilities regarding 2d-block-load-tranpose
features for which not all Intel devices support.
Related-To: NEO-11592
Signed-off-by: Jack Myers <jack.myers@intel.com>
add fallback to get max eu per ss from topology if not available in other way
Related-To: NEO-12073
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
firstly, setup hw info using product specific functions
secondly, query system info from GuC to setup max values
then, query memory info
then, query engine info as it depends on memory info
then, query topology as it depends on engine info
Related-To: NEO-12073
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
support for DRM_XE_VM_BIND_FLAG_IMMEDIATE and DRM_XE_VM_BIND_FLAG_READONLY
Fixes: #717
Related-To: NEO-10958
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
firstly, setup hw info using product specific functions
secondly, query system info from GuC to setup max values
thirdly, query topology to setup current topology data
Related-To: NEO-12073
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
- do not release resources derived from parent process
- zeInit() in child should initilize new driver
Related-To: NEO-11761
Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
Some OpenCL application need to render libva surface
with drm display and use the same libva surface as
OpenCL kernel output. Such applicatoins need to
use /dev/dri/card0 instead of /dev/dri/renderD128
for getting the display fd.
Related-To: NEO-11714
Signed-off-by: Elaine, Wang <elaine.wang@intel.com>
Removed if condition when calling vm_bind in order to
reflect changed in KMD.
Specifically, dii-4347 to dii-4349 include changes on
strict fence ordering on vm_bind, where user fence has
to be inserted regardless of using pagefault or explicit
residency.
Related-To: NEO-11966
Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
Bartosz Dunajski