intel/compute-runtime
1
0
Fork 0
mirror of https://github.com/intel/compute-runtime.git synced 2025-12-21 01:04:57 +08:00
Code Issues Packages Projects Releases Wiki Activity
9,062 Commits 42 Branches 655 Tags
06be26fd1e4fa2c07f94fe197bfeebe8c8111bd2
Commit Graph

508 Commits

Author SHA1 Message Date
Bartosz Dunajski
94f0a535a1 Fix passing resource48Bit flag 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-12-09 17:01:51 +01:00
Bartosz Dunajski
4dde9393ad Extend engine checkers to support more engine types 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-12-09 13:41:25 +01:00
Bartosz Dunajski
d07c76c237 unTypedDataPortCacheFlush pipe_control helper support 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-12-09 13:00:10 +01:00
Jim Snow
2dfb7df63b Add RTDispatchGlobals allocation for ray tracing 
If a kernel has ray tracing calls, we allocate and initialize
per-device RTDispatchGlobals if needed, and hand off pointer to
the same into a running kernel via an implicit parameter.

Related-To: NEO-5384
Signed-off-by: Jim Snow <jim.m.snow@intel.com>
 2021-12-08 09:44:13 +01:00
Milczarek, Slawomir
b16438de56 Add regkey to override MOCS index in surface state for scratch space 
Introduce the debug regkey OverrideMocsIndexForScratchSpace
to control MOCS index in surface state for scratch space

Related-To: NEO-6509

Signed-off-by: Milczarek, Slawomir <slawomir.milczarek@intel.com>
 2021-12-07 16:27:33 +01:00
Bartosz Dunajski
7c610ed5bb Add new Engine Group types 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-12-07 15:28:15 +01:00
Bartosz Dunajski
dfdd3c597a Remove redundant BUFFER_COMPRESSED allocation type 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-12-07 13:35:49 +01:00
Bartosz Dunajski
2b1aa8b331 Compilation fix: Add missing LrcaHelper types 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-12-07 13:19:36 +01:00
Maciej Dziuban
e3bb526067 Optimize timestamp packet dependencies 
- Clear dependencies even if last engine changed
- Do no program semaphore waiting for blit when blit is submitted with gpgpu
- Track barrier timestamps to correctly synchronize blits in OOQ

Related-To: NEO-6444
Signed-off-by: Maciej Dziuban <maciej.dziuban@intel.com>
 2021-12-07 11:47:50 +01:00
Bartosz Dunajski
f20236c7f2 Initial PVC support 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>

Related-To: NEO-5542
 2021-12-07 10:22:41 +01:00
Katarzyna Cencelewska
3979af1dcc Move slice enable to setHwInfoValuesFromConfig 
Fix problem with not enabled slices in mode SetCommandStreamReceiver=1

Related-To: NEO-6507
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
 2021-12-07 09:59:17 +01:00
Compute-Runtime-Validation
b7cee61bde Revert "Fail build program in shared system USM + stateful access case" 
This reverts commit 0ad8afc0b3.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
 2021-12-07 01:00:59 +01:00
Kamil Kopryk
0ad8afc0b3 Fail build program in shared system USM + stateful access case 
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
Related-To: NEO-6075

After this change driver will fail clBuildProgram/zeModuleCreate api calls
whenever stateful access is discovered and device has shared system usm caps
enabled.This is required since in this case allocation greater than 4GB
will not work.
If user still wants to use stateful addressing mode,
-cl-opt-smaller-than-4GB-buffers-only / -ze-opt-smaller-than-4GB-buffers-only
build option should be passed as build option, but then user can not use
buffers greater than 4GB.
 2021-12-06 15:10:16 +01:00
Bartosz Dunajski
68aea5bf62 Rename compression flags and helpers 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-12-03 18:09:02 +01:00
Zbigniew Zdanowicz
47dbe359bf Add command encoder for store data command 
Related-To: NEO-6262

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
 2021-12-02 20:56:07 +01:00
Zbigniew Zdanowicz
3e1023fa1a Unify memory layout for all multi tile post sync operations 
Related-To: NEO-6262

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
 2021-12-02 18:00:40 +01:00
Lukasz Jobczyk
63116e4100 Use wait utils while waiting for timestamps 
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
 2021-12-02 17:49:05 +01:00
Mateusz Hoppe
8b233f7f45 Support for bindless mode in L0 - improvements 
Related-To: NEO-6448

- add new IGC compilation flag when bindless mode used
- fix SBA programming of BindlessSurfaceStateSize -
always set maximum surface state count
- fix residency of global DSH heap on gen9 - gen12lp
in bindless mode
- add L0 aub test with bindless kernel - disabled
- partial fixes in OCL aub tests


Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
 2021-12-02 16:30:58 +01:00
Bartosz Dunajski
55959d4d1d Helper method to check if allocation is compressed 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-12-02 16:13:53 +01:00
Bartosz Dunajski
7b55d38e34 Refactor [1/n]: Make compression preference allocation property 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-12-02 15:24:35 +01:00
Katarzyna Cencelewska
fa16cbceb9 Change size calculating for scratch on dg2 
Also add proper init for SliceInfo on Linux

Resolves: NEO-6449
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
 2021-12-01 12:36:41 +01:00
Mateusz Jablonski
263becc3f8 refactor CFE state programming 
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
 2021-11-30 12:51:23 +01:00
Compute-Runtime-Validation
383daf9dbe Revert "Change size calculating for scratch on dg2" 
This reverts commit adc3ab5f24.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
 2021-11-30 04:22:55 +01:00
Fabian Zwolinski
5ee5e792e6 Fix hash counting 
Related-To: NEO-6209
Signed-off-by: Fabian Zwolinski <fabian.zwolinski@intel.com>
 2021-11-29 16:59:06 +01:00
Katarzyna Cencelewska
adc3ab5f24 Change size calculating for scratch on dg2 
Resolves: NEO-6449
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
 2021-11-29 14:07:38 +01:00
Filip Hazubski
3b820d41c6 Reorder engines to match API engines order 
Related-To: NEO-6219

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
 2021-11-29 09:59:01 +01:00
Bartosz Dunajski
995cb88bfa Improve ftr/wa flags packing 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-11-25 16:05:57 +01:00
Lukasz Jobczyk
fb376639ee Refactor isDirectSubmissionSupported 
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
 2021-11-25 13:40:41 +01:00
Mateusz Hoppe
de7195d174 Fix fusedEuDispatch programming and minimum wg size 
Related-To: NEO-6455

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
 2021-11-24 12:16:30 +01:00
Raiyan Latif
33de789531 Fix Blitter memory fill destination Pitch value 
Signed-off-by: Raiyan Latif <raiyan.latif@intel.com>
 2021-11-23 16:58:52 +01:00
Grzegorz Choinski
2e68f0abbd lint revision update 
Related-To: NEO-6441
Signed-off-by: Grzegorz Choinski <grzegorz.choinski@intel.com>
 2021-11-23 12:50:25 +01:00
Filip Hazubski
cd3930251e Align sync buffer address to max atomic type size 
Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
 2021-11-19 17:18:03 +01:00
Bartosz Dunajski
97af428866 Update PipeControlArgs struct 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-11-19 15:11:26 +01:00
Bartosz Dunajski
91dfa5c2ac Initial DG2 support 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-11-19 10:01:29 +01:00
Kamil Kopryk
f0cb1cc388 Remove isForceToStatelessRequired from compilerHwInfoConfig helper 
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
Related-To: NEO-6075
 2021-11-17 17:53:09 +01:00
Kamil Kopryk
7bbb43a563 Move sharedSystemMemCapabilities to hwInfo 
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
Related-To: NEO-6075
 2021-11-17 14:42:05 +01:00
Bartosz Dunajski
5be5cc33ce Update RayTracing helper values 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2021-11-16 10:43:43 +01:00
Lukasz Jobczyk
e8cc34d7b0 Add infrastructure to assign engine to command queue with round robin 
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
 2021-11-15 13:13:40 +01:00
Zbigniew Zdanowicz
78609cd9f5 Optimize number of calls for pipe control post syncs 
Related-To: NEO-6262

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
 2021-11-10 18:36:21 +01:00
Zbigniew Zdanowicz
36fd163837 Refactor pipe control post sync operations 
Related-To: NEO-6262

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
 2021-11-10 08:53:03 +01:00
Zbigniew Zdanowicz
61641bb70a Add implicit barriers capability to opencl 
Related-To: NEO-6262

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
 2021-11-10 08:19:50 +01:00
Compute-Runtime-Validation
908a7721b2 Revert "Assign engine to command queue using round robin algorithm" 
This reverts commit 2fff0f9059.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
 2021-11-10 07:25:44 +01:00
Lukasz Jobczyk
2fff0f9059 Assign engine to command queue using round robin algorithm 
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
 2021-11-09 11:05:16 +01:00
Mateusz Jablonski
c16eb0ff84 Correct getGpgpuEngineInstances function 
dont expose bcs engine if blitter operations not supported

Related-To: NEO-6325
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
 2021-11-09 08:36:05 +01:00
Compute-Runtime-Validation
9058de77cc Revert "Assign engine to command queue using round robin algorithm" 
This reverts commit 1c68ac1cbc.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
 2021-11-08 16:36:18 +01:00
Compute-Runtime-Validation
faea7915c2 Revert "Fail build program in shared system USM + statefull access case OCL" 
This reverts commit 9dabc2db0c.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
 2021-11-05 20:31:47 +01:00
Kamil Kopryk
9dabc2db0c Fail build program in shared system USM + statefull access case OCL 
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
Related-To: NEO-6075
 2021-11-05 14:55:05 +01:00
Mateusz Hoppe
ee418efadf Per-thread scratch offset calculation 
Related-To: NEO-6404

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
 2021-11-05 11:03:17 +01:00
Zbigniew Zdanowicz
23a7ab7593 Refactor implicit scaling barriers to add more cache flush options 
Related-To: NEO-6262

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
 2021-11-04 17:56:12 +01:00
Mateusz Jablonski
21ef249961 Move ocl extensions logic to shared 
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
 2021-11-04 17:38:15 +01:00