Commit Graph

2643 Commits

Author SHA1 Message Date
Zbigniew Zdanowicz
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
Lukasz Jobczyk
b050a83242 performance: Use lock pointer copy for dc flush mitigation
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2024-10-31 10:53:39 +01:00
Compute-Runtime-Validation
3fcb9b18ee Revert "performance: Use lock pointer copy for dc flush mitigation"
This reverts commit b8be102455.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-10-31 05:15:07 +01:00
Szymon Morek
cf58be4142 performance: use staging buffer when writing to an image
Related-To: NEO-12968

Also, don't import usm/mapped allocations for image
operations

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-10-30 17:02:13 +01:00
Lukasz Jobczyk
b8be102455 performance: Use lock pointer copy for dc flush mitigation
Resolves: NEO-12898

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2024-10-29 21:37:53 +01:00
Zbigniew Zdanowicz
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
Oskar Hubert Weber
fd45b2883d fix: neo ULT build with optimizations disabled 1/n
Related-To: NEO-8116

Signed-off-by: Oskar Hubert Weber <oskar.hubert.weber@intel.com>
2024-10-28 15:38:28 +01:00
Compute-Runtime-Validation
f5c433c8f8 Revert "performance: Use lock pointer copy with sfence for dc flush mitigation"
This reverts commit 8c3c703ec0.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-10-26 14:28:26 +02:00
Kamil Kopryk
3891e887c1 feature: use heapless builtins for images
Related-To: NEO-12744
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-10-25 16:42:27 +02:00
Zbigniew Zdanowicz
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
Lukasz Jobczyk
8c3c703ec0 performance: Use lock pointer copy with sfence for dc flush mitigation
Resolves: NEO-12898

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2024-10-25 16:24:40 +02:00
Dominik Dabek
22a3ec948b fix: ulls controller sleep, windows
Request higher resolution for windows periodic timers for ulls
controller sleep.

Allows for controller thread to sleep with granularity of 1ms.

Related-To: NEO-10800

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2024-10-24 18:07:04 +02:00
Bartosz Dunajski
4d6265a46c refactor: pass additional enqueue params
Related-To: NEO-12591

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-10-21 17:55:41 +02:00
Damian Tomczak
e495b2238a refactor: isStateless on init
Related-to: NEO-12938

Signed-off-by: Damian Tomczak <damian.tomczak@intel.com>
2024-10-18 10:53:17 +02:00
Mateusz Jablonski
4154e6666b fix: correct loading L0 loader functions
- don't load ze_loader.dll from file system
- to perform self-open on Windows use getModuleHandleA with proper module name
- don't free library loaded with getModuleHandleA
- as loader may be not available during runtime teardown:
- load translate handle function during global setup
- load setDriverTeardown function during global teardown
- when loader is not available during teardown, unset translate handle
function

Related-To: GSD-10147

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-10-17 15:39:20 +02:00
Compute-Runtime-Validation
f117b5fb68 Revert "fix: correct loading L0 loader functions"
This reverts commit 9c7b3c5e19.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-10-17 08:20:19 +02:00
Mateusz Jablonski
9c7b3c5e19 fix: correct loading L0 loader functions
on Windows use getModuleHandleA with proper module name
don't load ze_loader.dll from file system

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-10-16 19:53:05 +02:00
Compute-Runtime-Validation
f195e977ff Revert "feature: stateless enqueueCopyBufferToImage"
This reverts commit 9dbace61eb.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-10-15 08:54:24 +02:00
Damian Tomczak
bd500b5b85 feature: stateless enqueueCopyImageToBuffer
Related-to: NEO-11859

Signed-off-by: Damian Tomczak <damian.tomczak@intel.com>
2024-10-14 10:33:46 +02:00
Damian Tomczak
9dbace61eb feature: stateless enqueueCopyBufferToImage
Related-to: NEO-11859

Signed-off-by: Damian Tomczak <damian.tomczak@intel.com>
2024-10-14 10:21:02 +02:00
Damian Tomczak
d8f0b1c17d refactor: stateless enqueueCopyBufferRect
Related-to: NEO-11859

Signed-off-by: Damian Tomczak <damian.tomczak@intel.com>
2024-10-14 10:14:15 +02:00
Damian Tomczak
0c8af82d1b fix: stateless suffix for stateless kernels
Resolves: NEO-12847

Signed-off-by: Damian Tomczak <damian.tomczak@intel.com>
2024-10-11 17:40:10 +02:00
Wojciech Konior
6b40f9bc5a refactor: engineInstancedType removed
Related-To: NEO-12594

Signed-off-by: Wojciech Konior <wojciech.konior@intel.com>
2024-10-09 16:30:48 +02:00
Zbigniew Zdanowicz
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
Compute-Runtime-Validation
680e62d333 Revert "performance: Set dispatch all for small TG"
This reverts commit 0dc2870513.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-10-02 05:41:13 +02:00
Szymon Morek
b2fd1972a4 fix: add cpu alloc to eviction list only once
Related-To: NEO-12572

Also, before migration to GPU domain, remove it from this list

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-10-01 11:47:32 +02:00
Lukasz Jobczyk
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
Maciej Plewka
33a346838e fix: return error when flush in waitForEvents return error
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2024-09-27 09:17:07 +02:00
Maciej Plewka
1dafb66463 fix: return success from release calls after platform teardown
Related-To: NEO-11282
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2024-09-26 19:30:26 +02:00
Radoslaw Jablonski
c53393152a fix: remove Systolic field from Xe2 Compute Walker
The field has been removed from Xe2.

Signed-off-by: Radoslaw Jablonski <radoslaw.jablonski@intel.com>
2024-09-24 13:40:03 +02:00
Maciej Plewka
80f75ceace fix: submit dummy exec to pin memory during zeContextMakeMemoryResident call
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2024-09-23 14:43:59 +02:00
Mateusz Hoppe
4a068c8eab fix: correclty program StateBaseAddress in global bindless mode
- prepare bindful ssh when kernel requires ssh heap and
SurfaceStateBaseAddress
- remove lastAppendedKernelBindlesMode - local ssh heap may be needed
for bindless kernels with scratch or misaligned buffer args
- use ssh heap gpu address to program SurfaceStateBaseAddress, global base is
used for BindlessSurfaceState and DynamicState

Related-To: NEO-7063

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2024-09-20 11:57:05 +02:00
Michal Mrozek
363aca782f refactor: remove not needed code.
Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
2024-09-19 18:50:16 +02:00
Filip Hazubski
ebc19b4a70 feature: Add logic to disable bindless addressing via AIL
Add mockable Device functions to get ReleaseHelper and AILConfiguration.

Resolves: NEO-12699

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2024-09-18 13:49:51 +02:00
Mateusz Jablonski
16458ca0ac refactor: remove not used image transforming logic
image transformer logic was used on gen9 and gen11

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-09-17 16:19:48 +02:00
Mateusz Jablonski
5912b43841 refactor: remove dead code
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-09-17 13:41:50 +02:00
Artur Harasimiuk
73382b9b6c refactor: remove legacy code
Related-To: NEO-6683
Signed-off-by: Artur Harasimiuk <artur.harasimiuk@intel.com>
2024-09-16 19:04:13 +02:00
Mateusz Jablonski
d5812f49d7 refactor: remove not needed vme builtin related code
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-09-16 15:27:49 +02:00
Mateusz Jablonski
8e7959b243 refactor: remove not needed code
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-09-16 14:55:55 +02:00
Mateusz Jablonski
ab090b515c refactor: remove not needed code
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-09-16 14:20:35 +02:00
Maciej Bielski
5bdc4a55cc fix: avoid nullptr dereference
Related-To: NEO-12651

Signed-off-by: Maciej Bielski <maciej.bielski@intel.com>
2024-09-13 19:59:04 +02:00
Jaroslaw Warchulski
05250e7460 fix: Remove patchtokens fallback AIL
Related-To: NEO-12358
Signed-off-by: Jaroslaw Warchulski <jaroslaw.warchulski@intel.com>
2024-09-12 18:26:19 +02:00
Zbigniew Zdanowicz
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
Lukasz Jobczyk
48dc0a823d fix: Check dst pitch in copy rect misaligned case
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2024-09-11 11:45:10 +02:00
Zbigniew Zdanowicz
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
Dominik Dabek
b2fc7345cf performance: redesign usm alloc reuse mechanism
Dedicated pools for different allocations size ranges.
Additional reused allocations will create their own pools.
Do not reuse allocations >256MB.

Related-To: NEO-6893, NEO-12299, NEO-12349

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2024-09-10 18:12:14 +02:00
Bartosz Dunajski
4f1262645b refactor: pass extra walker params
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-09-10 15:30:03 +02:00
Filip Hazubski
beaddb55e0 fix: Minor code changes
Remove needless checks for module and cmdList.
Add explicit check for surfaceState in setArgImageWithMipLevel.
Explicitly initialize WddmResidencyController::csr.

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2024-09-10 13:31:15 +02:00
Mateusz Jablonski
54bda0e986 fix: In Linux CL/GL sharing
- always issue flush request before export

Apparently it's expected to flush the object (which might convert them
from one format to another for export, or remove aux buffer uses or
anything not supported by export).

- use modifier to select tiling mode

Previously we just assumed that whatever tiling mode was picked by mesa
will match the one picked by GMMLIB but that's not always the case
and in particular on Arc and Xe it doesn't work ... Mesa picks Tile4
and GMMLIB picks Tile64 ...

Fixes: #761
Fixes: #736

Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-09-06 21:40:18 +02:00
Mateusz Hoppe
9d5f8eb587 refactor: save hpCopyEngine in Device
- remove loop iteration to get hpCopyEngine

Related-To: NEO-11983

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2024-09-06 17:56:53 +02:00