89c3aab321
refactor: Extract PostSyncType to outer abstract layer
...
This modification serves to simplify the integration of generated
compute walker structures.
Related-To: NEO-13147
Signed-off-by: Vysochyn, Illia <illia.vysochyn@intel.com>
2024-11-28 18:09:21 +01:00
fa58073095
refactor: remove not used usings/typedefs/variables
...
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-11-28 16:19:39 +01:00
6b7235cd6c
refactor: change parameter names into more meaningful
...
- change additional size into local region size
- change walk order into dispatch walk order to distinguish for local id walk
Related-To: NEO-13350
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-27 16:15:02 +01:00
1ddcb92f1d
refactor: pass device to encoder functions
...
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-11-25 16:25:04 +01:00
906aa485a8
refactor: unify encoder methods
...
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-11-22 16:41:46 +01:00
1f6e39af0e
refactor: cleanup for defaultWalker
...
Making use of defaultWalker typedef more consistent
Related-To: NEO-12747
Signed-off-by: Chodor, Jaroslaw <jaroslaw.chodor@intel.com>
2024-11-17 00:24:11 +01:00
f07ee9a8cf
refactor: reorganize command encode classes 7/n
...
- add dedicated inl files
- unify setGrfInfo
- move methods for scratch data programming
- move methods for encodeEuSchedulingPolicy and appendSemaphoreCommand
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-14 15:48:53 +01:00
71c6bfc439
refactor: reorganize command encode classes 6/n
...
- unify programBarrierEnable
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-13 15:49:08 +01:00
d7f310dd5a
refactor: reorganize command encode classes 5/n
...
- move various methods from specialization in cpp into correct inl files
- EncodeAtomic, EncodeMiFlushDW, EncodeMemoryPrefetch, EncodeDispatchKernel
- move methods from generic inl files that are used on single platform to cpp
- EncodeMiFlushDW
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-13 09:18:30 +01:00
051648e215
refactor: reorganize command encode classes 4/n
...
- share xe hpc methods in generic way
- move gen12lp and xe hpg methods into dedicated inl
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-12 12:15:54 +01:00
ead001360b
refactor: reorganize command encode classes 3/n
...
- shift xe hpg inl file methods into specialization and generic inl files
- remove gen12lp specific implementation from main inl file
- remove not needed xe hpg inl file
- remove not needed compression inl files
- remove not used raytracing inl file
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-08 14:38:31 +01:00
eb435acb22
refactor: reorganize command encode classes 2/n
...
- add new inl files for selected platform families
- split image compression methods into correct inl files
- remove dedicated compression inl files
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-08 13:07:03 +01:00
9536510c5b
refactor: reorganize command encode classes 1/n
...
- remove obsolete file
- move methods from redundant compute mode file into dedicated platform files
- group same implementation into platform specific inl files
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-07 13:06:28 +01:00
89b40b7ec4
refactor: change additional walker fields encoder 6/n
...
- change order of members of additional walker struct
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-05 08:28:57 +01:00
2e2b7a473a
refactor: change additional walker fields encoder 5/n
...
- move compute dispatch all walker into dedicated encoder
- group same implementations into single file
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-04 13:11:59 +01:00
32fd00e150
refactor: change additional walker fields encoder 4/n
...
- move post sync system fence into dedicated encoder
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-10-31 14:41:55 +01:00
6f4994c269
refactor: change additional walker fields encoder 1/n
...
- move encoding l3 prefetch field into dedicated function
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-10-29 21:37:38 +01:00
2dccdd886a
refactor: change encoder for thread group over dispatch 2/n
...
- bind algorithms to input arguments
- use thread group count array for regular kernels
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-10-29 21:20:59 +01:00
022f9e642d
Revert "refactor: change encoder for thread group over dispatch 2/n"
...
This reverts commit 046631767b
2024-10-29 01:51:52 +01:00
046631767b
refactor: change encoder for thread group over dispatch 2/n
...
- bind algorithms to input arguments
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-10-28 10:11:13 +01:00
bbdf1ac7b6
refactor: change encoder for thread group over dispatch 1/n
...
- change method name to more meaningful
- add all inputs of the algorithm as explicit function arguments
- position all implementations accordingly
- rename unit test names to fit new method name
- fix unit test to have correct initial command values
- fix unit test to have consistent input values with command values
- fix unit test to change input values together with command values
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-10-25 16:31:39 +02:00
227ce579fe
fix: align total slm size for preferred size
...
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-10-10 20:44:19 +02:00
f4746a5b07
fix: remove not needed limitations for preferred slm
...
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-10-10 19:40:54 +02:00
ac6472b105
fix: limit slm size based on hw informations
...
- add override default values method
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-10-09 17:34:56 +02:00
e690217227
refactor: remove not needed slm alignment
...
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-10-09 16:45:19 +02:00
46a63d3e0e
refactor: unify programming of preferred slm size 5/n
...
- remove xe2 hpg encode preferred slm size
- add xe2 release helper preferred slm array
- add dedicated method to calculate thread count per sub slice
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-10-09 09:39:35 +02:00
552930a75f
fix: don't setup preemption surface when debugger is active
...
Related-To: NEO-12878
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-10-08 13:58:11 +02:00
d6016e1b91
refactor: unify programming of preferred slm size 3/n
...
- add shared implementation to encode preferred slm size
- add pvc release helper preferred slm array
- drop pvc preproduction steppings values for preferred slm size
- remove obsolete product helper method
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-10-07 12:10:27 +02:00
8687a88d06
refactor: unify programming of preferred slm size 2/n
...
- add interface that will provide all preferred slm sizes for all releases
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-10-04 11:50:44 +02:00
49371cb13e
refactor: unify programming of preferred slm size 1/n
...
- rename function to reflect actual task function does
Related-To: NEO-12639
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-10-03 18:52:01 +02:00
680e62d333
Revert "performance: Set dispatch all for small TG"
...
This reverts commit 0dc2870513
2024-10-02 05:41:13 +02:00
0dc2870513
performance: Set dispatch all for small TG
...
Resolves: NEO-11814
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2024-10-01 09:28:55 +02:00
8e7959b243
refactor: remove not needed code
...
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-09-16 14:55:55 +02:00
78604bd475
refactor: remove not needed code
...
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-09-16 12:12:43 +02:00
6263b9599f
feature: Enabling Blitter offset for Remap
...
Related-To: NEO-11934
Signed-off-by: Andrzej Koska <andrzej.koska@intel.com>
2024-09-16 09:04:28 +02:00
f58b5b6ac3
Revert "feature: Enabling Blitter Remapping for REGISTER_MEM"
...
This reverts commit d4f59c1b09
2024-09-12 23:50:59 +02:00
c0d9cb4d84
Revert "feature: Enabling Blitter Remapping for REGISTER_REG"
...
This reverts commit 4f2e61841f
2024-09-12 16:43:46 +02:00
4f2e61841f
feature: Enabling Blitter Remapping for REGISTER_REG
...
Related-To: NEO-11934
Signed-off-by: Andrzej Koska <andrzej.koska@intel.com>
2024-09-11 19:02:16 +02:00
8bb92ff445
performance: avoid reading from gfx memory when modifying walker command
...
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-09-11 18:58:33 +02:00
7ce4a8adc2
performance: replace virtual calls with native class methods
...
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-09-11 11:10:40 +02:00
d4f59c1b09
feature: Enabling Blitter Remapping for REGISTER_MEM
...
Related-To: NEO-11934
Signed-off-by: Andrzej Koska <andrzej.koska@intel.com>
2024-09-11 09:14:44 +02:00
4f1262645b
refactor: pass extra walker params
...
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-09-10 15:30:03 +02:00
a54a3bf624
performance: Optimize heap handling when mitigate dc flush
...
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2024-09-06 04:33:41 +02:00
b0e7a11e9a
refactor: Improving information transfer about the copy engine
...
Related-To: NEO-11934
Signed-off-by: Andrzej Koska <andrzej.koska@intel.com>
2024-09-05 16:11:52 +02:00
dc84b163b5
Revert "performance: Optimize heap handling when mitigate dc flush"
...
This reverts commit 9249c5c65c
2024-09-03 08:33:20 +02:00
9249c5c65c
performance: Optimize heap handling when mitigate dc flush
...
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2024-09-02 18:12:19 +02:00
4b3b1be191
fix: add heap to residency container instead of calling make resident
...
Related-To: NEO-11212
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
There was make resident call outside lock on csr, so in multi thread
scenarios assigned residency task count may be not correct.
Instead of calling make resident allocation is added to
residency container
2024-08-23 14:52:24 +02:00
360a5bbc96
fix: Use consistently proper size of implicit args buffer
...
Fixes discrepancy of implicit args buffer allocation and programming
Resolves: NEO-12316
Signed-off-by: Chodor, Jaroslaw <jaroslaw.chodor@intel.com>
2024-08-23 13:32:17 +02:00
4c3a0d8344
fix: add missing sba capture when debug enabled
...
- In global bindless mode SBA may be programmed for scratch allocation,
missing sba capturing caused issues for kernels with scratch.
- this fix adds SBA capturing after SBA command
Related-To: NEO-7063
Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2024-08-20 09:32:17 +02:00
9b652f4a34
Revert "feature: Improving information transfer about the copy engine"
...
This reverts commit 17ffdff4f1
2024-08-15 22:06:31 +02:00
Vysochyn, Illia