Media
/
compute-runtime
mirror of https://github.com/intel/compute-runtime.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
9,935 Commits 35 Branches 603 Tags 163 MiB
Commit Graph

298 Commits

Author SHA1 Message Date
Mateusz Hoppe beff0019d1 SBA tracking for single address space 
Related-To: NEO-6539


Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
 2022-04-01 15:24:11 +02:00
Bartosz Dunajski 08e3853982 Debug flag to add extra MI_MEM_FENCE for DirectSubmission 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2022-03-30 16:07:25 +02:00
Rafal Maziejuk c7d8915dd4 Add debug variable to print XY_BLOCK_COPY_BLT command details 
Signed-off-by: Rafal Maziejuk <rafal.maziejuk@intel.com>
 2022-03-29 18:15:13 +02:00
Bartosz Dunajski 68351249d1 Add debug flag to read back command buffer pointer 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2022-03-29 13:08:18 +02:00
Mateusz Jablonski 3ac0581f2e Add debug flags to read back cmd/ring buffer before unblocking semaphore 
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
 2022-03-25 20:36:29 +01:00
Bartosz Dunajski be50afb930 Add and enable Direct Submission ring switch tag update WA 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2022-03-25 12:26:29 +01:00
Mateusz Jablonski 8a8b4866cb XeHPC: force local memory for command/ring/semaphore buffer 
require 48bit resource for ring/semaphore buffer
for multi tile allocations select first tile
for single tile allocation select preferred tile

Related-To: NEO-6698
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
 2022-03-24 10:54:26 +01:00
Daniel Chabrowski 28b867f883 Move and extend local memory DRM allocations 
Signed-off-by: Daniel Chabrowski daniel.chabrowski@intel.com
Related-To: NEO-6591
 2022-03-22 13:47:05 +01:00
Mateusz Jablonski 3792481d33 XeHPC Implicit scaling: put command/ring/semaphore buffer to first memory bank 
In direct submission scenario command/ring/semaphore buffer allocations
are placed in the same memory bank to ensure that their memory is updated in
correct order

Related-To: NEO-6698
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
 2022-03-21 08:10:50 +01:00
Lukasz Jobczyk f8449fb216 Add round robin engine assign controls 
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
 2022-03-14 10:57:20 +01:00
Bartosz Dunajski 79fedd59e4 Add BB_START prepatching WA - disabled by default 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2022-03-10 17:05:53 +01:00
Bartosz Dunajski e24322f266 Debug flag to control MI_ARB_CHECK prefetcher 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
 2022-03-10 12:50:05 +01:00
Patryk Wrobel 835b344968 Add debug flag to disable GPU hang detection 
This change introduces the new flag called DisableGpuHangDetection.
By default it is disabled. When someone wants to disable hang checking,
then this flag can be set to true.

Related-To: NEO-6681
Signed-off-by: Patryk Wrobel <patryk.wrobel@intel.com>
 2022-03-09 18:07:42 +01:00
Milczarek, Slawomir c0b7f05897 Add memory prefetch for kmd migrated shared allocations 
This feature is disabled by default, controlled with the knob
AppendMemoryPrefetchForKmdMigratedSharedAllocations

Related-To: NEO-6740

Signed-off-by: Milczarek, Slawomir <slawomir.milczarek@intel.com>
 2022-03-09 16:02:18 +01:00
Filip Hazubski dd01cff879 Unify logic determining thread arbitration policy value 
Related-To: NEO-6728

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
 2022-03-08 13:14:56 +01:00
Lukasz Jobczyk 3c30e1b02b Add AssignBCSAtEnqueue debug flag 
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
 2022-03-01 12:43:36 +01:00
Lukasz Jobczyk b096422e78 BCS round robin adjustments 
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
 2022-03-01 09:45:38 +01:00
Lukasz Jobczyk 090bfb9642 Reuse kernel allocation 
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
 2022-02-28 19:26:19 +01:00
Jaime Arteaga 8b56a76b25 Fix typo 
Signed-off-by: Jaime Arteaga <jaime.a.arteaga.molina@intel.com>
 2022-02-23 23:46:30 +01:00
Aravind Gopalakrishnan c1970942aa Update debug key to override timestamp event settings 
Related-To: NEO-5968

Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>
 2022-02-23 23:38:05 +01:00
Igor Venevtsev 086a70ca52 Add ForceAllResourcesUncached debug variable 
Resolves: NEO-6563

Signed-off-by: Igor Venevtsev <igor.venevtsev@intel.com>
 2022-02-23 09:43:48 +01:00
Compute-Runtime-Validation e984760423 Revert "Move and extend local memory DRM allocations" 
This reverts commit b84b0015d5.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
 2022-02-22 08:48:46 +01:00
Daniel Chabrowski b84b0015d5 Move and extend local memory DRM allocations 
Signed-off-by: Daniel Chabrowski daniel.chabrowski@intel.com
Related-To: NEO-6591
 2022-02-21 16:04:53 +01:00
Mateusz Jablonski b9d613995d Move createDrmVirtualMemory function to drm_neo.cpp 
Related-To: NEO-6591
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
 2022-02-16 11:34:52 +01:00
Mateusz Jablonski 4a24195e99 Move bind/unbind buffer objects methods to drm_neo.cpp 
use ioctl helper to handle binding properly

Related-To: NEO-6591
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
 2022-02-16 08:37:16 +01:00
Mateusz Jablonski cf27583264 Move createDrmContextExt to drm_neo.cpp 
add ioctl helper functions to create dedicated drm contexts

createContextWithAccessCounters
createCooperativeContext

Related-To: NEO-6591

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
 2022-02-15 09:16:36 +01:00
Compute-Runtime-Validation c5c3e865f0 Revert "Fail build program on PVC with stateful accesses" 
This reverts commit 9466113cef.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
 2022-02-14 18:55:14 +01:00
Kamil Kopryk 9466113cef Fail build program on PVC with stateful accesses 
Related-To: NEO-6075

After this change driver will fail clBuildProgram/zeModuleCreate api calls
whenever stateful access is discovered on PVC.
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
bufers greater than 4GB.


Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
 2022-02-14 13:44:22 +01:00
Compute-Runtime-Validation 6fe91498d8 Revert "Extend DRM memory manager" 
This reverts commit dec634c2cd.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
 2022-02-12 23:41:31 +01:00
Krzysztof Gibala 02c87fd8b9 Refactor naming around additional PC before NP state command 
Rename:
- debug flag ProgramPipeControlPriorToNonPipelinedStateCommand
to ProgramExtendedPipeControlPriorToNonPipelinedStateCommand
- local variables

Related-To: NEO-6615
Signed-off-by: Krzysztof Gibala <krzysztof.gibala@intel.com>
 2022-02-11 19:24:14 +01:00
Daniel Chabrowski dec634c2cd Extend DRM memory manager 
Signed-off-by: Daniel Chabrowski daniel.chabrowski@intel.com
Related-To: NEO-6591
 2022-02-11 17:28:32 +01:00
Mateusz Jablonski 2e32bcb2e1 Extend bindDrmContext function 
move implementation to drm_neo.cpp
respect multiple CCS and BCS engines

Related-To: NEO-6591
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
 2022-02-11 14:50:48 +01:00
Daniel Chabrowski 23999110ed Add prelim DRM page fault allocation check support 
Signed-off-by: Daniel Chabrowski daniel.chabrowski@intel.com
Related-To: NEO-6591
 2022-02-10 18:57:52 +01:00
Daniel Chabrowski 85eb6d7403 [CP-1] Extend CLOS cache mechanism for DRM allocation 
Signed-off-by: Daniel Chabrowski daniel.chabrowski@intel.com
Related-To: NEO-6591
 2022-02-09 16:51:48 +01:00
Lukasz Jobczyk b60d963ff5 Add debug variable to force default heap allocation size 
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
 2022-02-09 13:27:05 +01:00
Lukasz Jobczyk 538e0aea87 Add debug flag to bind at creation time 
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
 2022-02-09 09:26:02 +01:00
Krzysztof Gibala f7a5d29085 Introduce debug flag ForceExtendedUSMBufferSize usage for L0 
Forces extended buffer size by adding pageSize specify by number when
debug flag is >=1 in L0 USM calls

Usage:
ForceExtendedUSMBufferSize=2
size += (2 * pageSize)

Signed-off-by: Krzysztof Gibala <krzysztof.gibala@intel.com>
 2022-02-08 16:23:40 +01:00
Krzysztof Gibala e518a8f3f9 Add debug flag ForceExtendedUSMBufferSize 
Forces extended buffer size by adding pageSize specify by number when
debug flag is >=1 in:
- clHostMemAllocINTEL
- clDeviceMemAllocINTEL
- clSharedMemAllocINTEL

Usage:
ForceExtendedUSMBufferSize=2
size += (2 * pageSize)

Signed-off-by: Krzysztof Gibala <krzysztof.gibala@intel.com>
 2022-02-07 11:44:31 +01:00
Lukasz Jobczyk 0768bf7517 When select copy engine round robin over all copy engines 
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
 2022-02-04 14:55:46 +01:00
Daniel Chabrowski f2c4231a27 Extend DRM mock and queryTopology tests - part 1 
Signed-off-by: Daniel Chabrowski daniel.chabrowski@intel.com
Related-To: NEO-6591
 2022-02-04 12:02:49 +01:00
Katarzyna Cencelewska 3d0c065183 Remove device enqueue part 16 
-delete old unused flags

Related-To: NEO-6559
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
 2022-02-03 19:38:41 +01:00
Mateusz Jablonski 1671e0a1a7 Add new debug flags for filtering devices to open 
FilterBdfPath is used only on Linux as a filter for BDF
when opening from /dev/dri/by-path

FilterDeviceId is used on both OSes as a filter for device id

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
 2022-02-03 14:34:01 +01:00
Michal Mrozek 8f85d4b8f8 Add debug variable to override message simd. 
Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
 2022-02-03 11:07:42 +01:00
Aravind Gopalakrishnan 493a96b308 [3/n] L0 Immediate commandlist improvements 
Enable flushTask by default
Related-To: LOCI-1988

Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>
 2022-02-02 07:46:00 +01:00
Joshua Santosh Ranjan c317cff408 Make BDF method as default for UUID calculation 
This patch makes PCIE BDF as the default method for UUID
calculation.

Related-To: LOCI-2909

Signed-off-by: Joshua Santosh Ranjan <joshua.santosh.ranjan@intel.com>
 2022-01-31 15:47:21 +01:00
Katarzyna Cencelewska 58055aecdf Remove device enqueue part 12 
remove:
- debug flag ForceDeviceEnqueueSupport
- functions isDeviceEnqueueSupported, supportsDeviceEnqueue

Related-To: NEO-6559
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
 2022-01-28 13:38:34 +01:00
Lukasz Jobczyk 44596f6e4d Revert "Improve mmap logic in createAllocWithAlignment" 
This reverts commit adb5c58292165f1d5a144548c5d5148021d2152e.

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
 2022-01-26 08:52:12 +01:00
Compute-Runtime-Validation 15b9ec57d1 Revert "Enable Flush task by default for immediate commandlist" 
This reverts commit 22bbce42dd.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
 2022-01-26 04:53:03 +01:00
Michal Mrozek 6df17f5a30 [3/n] Optimize indirect allocations handling. 
Add new debug variable to trigger new mode.

Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
 2022-01-25 16:40:56 +01:00
Filip Hazubski 0424f30782 Improve mmap logic in createAllocWithAlignment 
Group small allocations and reuse mapped memory in order to keep map
count small.

Related-To: NEO-6417

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
 2022-01-25 15:30:29 +01:00