Commit Graph

226 Commits

Author SHA1 Message Date
a06fbd2077 Remove device enqueue part 10
- remove DeviceQueue

Related-To: NEO-6559
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2022-01-19 17:41:06 +01:00
97765cd39f Remove device enqueue part 7
- mainly remove BlockKernelManager and ReflectionSurfaceHelper

Related-To: NEO-6559
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2022-01-19 10:41:57 +01:00
59683ec491 Remove device enqueue part 6
- isParentKernel, peekParentKernel, parentKernel
- structs: AUBParentKernelFixture, MockParentKernel,
ParentKernelCommandQueueFixture

Related-To: NEO-6559
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2022-01-17 13:56:29 +01:00
182042b04d Revert "Update default thread arbitration policy"
This reverts commit 8c3e9ace69.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2022-01-13 21:32:29 +01:00
8c3e9ace69 Update default thread arbitration policy
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2022-01-13 16:24:11 +01:00
45ae4fe881 Remove device enqueue part 3
- isSchedulerKernel

Related-To: NEO-6559
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2022-01-12 13:50:18 +01:00
967d8a56fe Don't reduce LWS if SIMD32 is being used
Don't divide max kernel work group size by 2
if it's using SIMD 32

Resolves: NEO-6429

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2022-01-11 14:03:49 +01:00
df7723411f Patch image descriptor fields with dword width
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2021-12-27 17:55:42 +01:00
2643eb691a New thread arbitration mode token support
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2021-12-09 16:38:22 +01:00
68aea5bf62 Rename compression flags and helpers
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2021-12-03 18:09:02 +01:00
5c266f9ab0 Compression refactor [3/n]
Dont use allocation type for compression preference

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2021-12-03 17:20:07 +01:00
f90932cca7 Use references instead copy ctors
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2021-11-15 14:42:33 +01:00
f20cdac7d3 Zebin: add support for has_dpas
Signed-off-by: Krystian Chmielewski <krystian.chmielewski@intel.com>
2021-10-28 09:59:21 +02:00
1f2aa82c5b Remove UNUSED_VARIABLE macro
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2021-10-22 18:56:33 +02:00
9614d435c5 Simplify code
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2021-10-20 11:29:49 +02:00
279181a384 Remove not needed include header
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2021-10-06 21:32:40 +02:00
0363749948 Disjoint local work size code from hw command generation
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2021-10-06 01:09:56 +02:00
85a52b7702 Fix typo
Signed-off-by: Rafal Maziejuk <rafal.maziejuk@intel.com>
2021-10-05 14:45:24 +02:00
1028295a02 Move per thread data to shared directory
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2021-09-30 17:57:41 +02:00
25c71a6c13 Move kernel info to shared directory
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2021-09-29 23:49:49 +02:00
2abcef2c35 Revert "Disable stateless compression in sba on indirect host access flag"
Related-To: NEO-5107

This reverts commit 658fa6d85419f4a1f4876b40ef74f77095c29125.

Signed-off-by: Milczarek, Slawomir <slawomir.milczarek@intel.com>
2021-09-29 13:07:07 +02:00
5c4c02d2e7 Disable stateless compression in sba on indirect host access flag
Signed-off-by: Milczarek, Slawomir <slawomir.milczarek@intel.com>
2021-09-21 16:57:34 +02:00
a4b3dc29e4 Remove unnecessary method
Signed-off-by: Maciej Dziuban <maciej.dziuban@intel.com>
2021-09-21 11:34:27 +02:00
0a8b473f3c Add resolves for stateless compression with shared buffer
Related-To: NEO-5107

Signed-off-by: Milczarek, Slawomir <slawomir.milczarek@intel.com>
2021-09-17 09:18:45 +02:00
e193589582 Add implicit args support for OCL
Related-To: NEO-5081

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2021-09-13 10:54:19 +02:00
9d99ee1636 Print driver diagnostics if aux translation of allocation in SVM allocs manager
Related-To: NEO-5107

Signed-off-by: Milczarek, Slawomir <slawomir.milczarek@intel.com>
2021-09-13 10:23:32 +02:00
a204b11c8f Do not require aux translation on built ins
Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2021-09-10 14:38:04 +02:00
bbfbf19a02 Move allowStatelessCompression from HwHelper to HwInfoConfig
Signed-off-by: Rafal Maziejuk <rafal.maziejuk@intel.com>
Related-To: NEO-4541
2021-09-09 11:37:47 +02:00
7f920139b4 Correct patching private surface in cloned kernel
Related-To: NEO-5081
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2021-09-07 13:50:22 +02:00
3d6d4acda2 Update isCooperativeDispatchSupported
Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2021-09-02 17:36:58 +02:00
3e6b3cd780 Revert "Create single point for selecting engine for builtin ops"
This reverts commit 6513bd371a69cb821ea45c7c1964ea1f51dd75b2.

Signed-off-by: Maciej Dziuban <maciej.dziuban@intel.com>
2021-09-02 11:58:46 +02:00
a03b5f8a95 Create single point for selecting engine for builtin ops
- selectCsrForBuiltinOperation selects proper CSR
- selected CSR is passed to dispatchBcsOrGpgpuEnqueue
- Notifications such as notifyEnqueueReadBuffer are now made on correct engine
- Temporary allocs for hostptrs are now created on gpgpuCsr

Related-To: NEO-6057
Signed-off-by: Maciej Dziuban <maciej.dziuban@intel.com>
2021-08-31 13:29:17 +02:00
de1e4e0074 Add adjustMaxWorkGroupCount helper
Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2021-08-27 15:39:48 +02:00
63f8c9d98b Add new EngineUsage
Simplify verifying EngineUsage in tryGetEngine function.
Remove unused getGpgpuEnginesCount function.

Related-To: NEO-4940

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2021-08-23 17:41:13 +02:00
29c64c3dd0 Disable implicit scaling for cooperative kernels
When implicit scaling is disabled use useSingleSubdeviceValue = true.

Resolves: NEO-5757

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2021-08-18 14:56:37 +02:00
66ee5bd3c4 Correct typo
Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2021-08-03 20:08:18 +02:00
eb14d8458b Add helper function to enable stateless compression
Related-To: NEO-5107

Signed-off-by: Milczarek, Slawomir <slawomir.milczarek@intel.com>
2021-08-02 14:34:33 +02:00
f9f5c96ffe Add resolves for USM allocations when aux translation is required
Related-To: NEO-5107

Signed-off-by: Milczarek, Slawomir <slawomir.milczarek@intel.com>
2021-07-27 15:10:41 +02:00
a18fb402a0 Add resolves for SVM GPU allocations if aux translation is required
Related-To: NEO-5107

Signed-off-by: Milczarek, Slawomir <slawomir.milczarek@intel.com>
2021-07-27 10:21:11 +02:00
330856c9e9 Add resolves for unified memory when aux translation is required
Related-To: NEO-5107

Signed-off-by: Milczarek, Slawomir <slawomir.milczarek@intel.com>
2021-07-22 14:02:52 +02:00
6f3c89decb Add mechanism to migrate multi root device memory
invalidate TLB cache if kernel requires migration


Related-To: NEO-3691

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2021-07-19 19:49:22 +02:00
bbe599aa95 Selectively enable getMaxThreadsForWorkgroup WA
Related-To: NEO-6022

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2021-07-14 17:29:43 +02:00
bc121c09fa Use DSS count to callculate max work group size
This reverts commit 89b7a4c9cc0a69cfd2eb5a0bd138356f34c52f0d.

Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2021-07-05 10:14:20 +02:00
64eb82efac Add Kernel restrictions
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2021-07-02 17:41:03 +02:00
aed3fada28 Revert "Use Eu per dss to callculate max work group size"
This reverts commit 86b12dd23979db12e1898013c9162cb7106e40f1.

Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2021-07-02 17:16:06 +02:00
bcf93a6aeb Use Eu per dss to callculate max work group size
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2021-07-02 12:30:18 +02:00
e1f42c2ae1 Remove TSP atomic dependency tracking
This reverts commit d17668f023bee409ee68e766bbc7d5f16ce8d52b.

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2021-06-23 16:59:41 +02:00
5e8e066541 Bindless support ocl
Signed-off-by: Krystian Chmielewski <krystian.chmielewski@intel.com>
2021-06-23 13:56:12 +02:00
2ab961be49 Ensure that builtin kernel is patched and can't transform images
Related-To: NEO-3691
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2021-06-22 16:20:07 +02:00
843b65bb07 Add usingImages field to the Kernel class
Signed-off-by: Maciej Dziuban <maciej.dziuban@intel.com>
2021-06-22 15:25:22 +02:00