Commit Graph

11734 Commits

Author SHA1 Message Date
d653779098 Revert "L0 debug - Fix thread creation for windows DSS"
This reverts commit 3724807eed.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2022-10-22 14:38:24 +02:00
052164c559 Revert "Use total SLM instead of inline in error checking"
This reverts commit e1c49d0f23.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2022-10-22 09:41:36 +02:00
9ad3f6190f do not sleep in ULTs
Signed-off-by: Artur Harasimiuk <artur.harasimiuk@intel.com>
2022-10-21 19:37:52 +02:00
3724807eed L0 debug - Fix thread creation for windows DSS
Signed-off-by: Yates, Brandon <brandon.yates@intel.com>
2022-10-21 18:47:49 +02:00
518016f20f L0 Debug - Zero init debug surface in windows
Signed-off-by: Yates, Brandon <brandon.yates@intel.com>
2022-10-21 18:36:14 +02:00
ade323ce90 infra update
Signed-off-by: Artur Harasimiuk <artur.harasimiuk@intel.com>
2022-10-21 15:37:04 +02:00
001d3bbe11 igc revision update
Signed-off-by: ocldev <ocldev@intel.com>
2022-10-21 13:53:18 +02:00
c777f43909 Enable heap sharing for XE_HPG
Related-To: NEO-7142

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2022-10-21 13:43:44 +02:00
b103b0c43f Reduce the SLM time waiting on ready CMD to 100 uSec
Signed-off-by: Matias Cabral <matias.a.cabral@intel.com>
2022-10-21 12:50:58 +02:00
e1c49d0f23 Use total SLM instead of inline in error checking
Related-To: NEO-7280
Signed-off-by: Fabian Zwolinski <fabian.zwolinski@intel.com>
2022-10-21 12:39:48 +02:00
380e2dcc35 [WSL2] Avoid gdi calls while process exit
Related-To: NEO-7380
Signed-off-by: Kamil Diedrich <kamil.diedrich@intel.com>
2022-10-21 12:37:07 +02:00
9b26e96b11 Do not copy std::vector of OS handles when it is not needed
This change replaces unneeded copying of std::vectors
with usage of const references. Furthermore, it adds
reserve() call before filling the container via push_back().

Signed-off-by: Patryk Wrobel <patryk.wrobel@intel.com>
2022-10-21 10:28:58 +02:00
40ddeeb3ba Revert "Add support for USM shared in WSL for dGPU"
This reverts commit 1044a8aeee.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2022-10-21 05:59:34 +02:00
e65d2b6c0e Correct ZEX_NUMBER_OF_CCS documentation
Signed-off-by: Jaime Arteaga <jaime.a.arteaga.molina@intel.com>
2022-10-20 19:36:54 +02:00
1a08240474 Add ze_eu_count_t to get total number of EUs
Related-To: LOCI-2667

Signed-off-by: Jaime Arteaga <jaime.a.arteaga.molina@intel.com>
2022-10-20 19:12:16 +02:00
b1731626d0 Correct spelling mistakes
Signed-off-by: Kacper Nowak <kacper.nowak@intel.com>
2022-10-20 15:53:23 +02:00
3384b2fed0 Add debug flag to override zeDeviceCanAccessPeer return value.
Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
2022-10-20 14:39:12 +02:00
22dbae1771 feat(zebin): Add support for .kernel_arg_metadata section.
This commit adds support for parsing new .kernel_arg_metadata zeInfo's section,
which will be parsed only on demand (it won't get parsed on initial
zeInfo parsing).

Usage of populated structs will be added in the next commit.
Implemented section's parsing, decoding & populating corresponding fields in
kernelDescriptor.

Related-To: NEO-7372
Signed-off-by: Kacper Nowak <kacper.nowak@intel.com>
2022-10-20 14:20:32 +02:00
1044a8aeee Add support for USM shared in WSL for dGPU
This patch force KMD allocation path for USM host allocation
and also for host part of USM shared allocation

Related-To: NEO-6913
Signed-off-by: Kamil Diedrich <kamil.diedrich@intel.com>
2022-10-20 11:26:24 +02:00
90bc1a69d2 L0Debug - wait for the thread to start
Related-To: NEO-7322
Signed-off-by: Warchulski, Jaroslaw <jaroslaw.warchulski@intel.com>
2022-10-20 10:44:16 +02:00
a93edd8620 Support for FabricEdge APIs
Related-To: LOCI-3377

Signed-off-by: Joshua Santosh Ranjan <joshua.santosh.ranjan@intel.com>
2022-10-20 10:43:25 +02:00
553e194379 Ocloc: Enhance ocloc ULTs - ensure compatibility with zebin
This commit fixes ocloc ULT failure with zebin enabled by default.
Remove depencendy on precompiled binaries, instead use mock CTNI binary.

Related-To: NEO-7383
Signed-off-by: Kacper Nowak <kacper.nowak@intel.com>
2022-10-20 07:50:18 +02:00
4a3862b514 Revert "Add ze_eu_count_t to get total number of EUs"
This reverts commit 2b963c7359.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2022-10-20 06:53:14 +02:00
789a53e8f8 L0 Debug - Don't pollute debug log with event timeouts
Signed-off-by: Yates, Brandon <brandon.yates@intel.com>
2022-10-19 17:26:31 +02:00
510ba12cbc Use DSS when calculating max wgs on TGLLP and DG1
Use DualSubSlice count instead of SubSlice count when calculating wgs on
TigerLake LP and DashG1.

Related-To: NEO-5719

Signed-off-by: Krystian Chmielewski <krystian.chmielewski@intel.com>
2022-10-19 12:23:51 +02:00
5bd4b9eb48 Do not call DebuggerOpen ioctl again on EBUSY
Resolves: NEO-7429

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2022-10-19 10:26:35 +02:00
abe4e0cb91 Fix case style
Test fixture classes will use correct camel case

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2022-10-19 09:41:45 +02:00
d1a6054af9 enable create subBuffer from pooled buffer
Allow creating subBuffer from buffer from buffer pool allocator
by redirecting the call to the pool buffer and adjusting offset

Related-To: NEO-7332

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2022-10-19 09:33:10 +02:00
f06df021b5 feat(zebin): Add support for indirect stateless count
Add support for newly added indirect statelss count check;
populate related field in kernelInfo.
- Move hasIndirectStatelessAccess check from KernelInfo to
KernelDescriptor.

Related-To: NEO-7428
Signed-off-by: Kacper Nowak <kacper.nowak@intel.com>
2022-10-19 07:58:02 +02:00
2b963c7359 Add ze_eu_count_t to get total number of EUs
Related-To: LOCI-2667

Signed-off-by: Jaime Arteaga <jaime.a.arteaga.molina@intel.com>
2022-10-19 07:17:25 +02:00
e744116bbb Revert "Add support for USM shared in WSL for dGPU"
This reverts commit 68d0523ccf.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2022-10-19 05:05:55 +02:00
fb453f5190 Revert "Enable flush task for immediate command lists"
This reverts commit a4eb78d524.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2022-10-19 03:32:53 +02:00
0b2491f9e6 createSampler returns error for PVC
Signed-off-by: John Falkowski <john.falkowski@intel.com>
2022-10-18 18:13:34 +02:00
a4eb78d524 Enable flush task for immediate command lists
Enabling for all platforms

Related-To: LOCI-3379

Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>
2022-10-18 17:17:05 +02:00
565d820933 Replace virtual method call for DC flush with stored bool value 3/n
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2022-10-18 15:04:02 +02:00
68d0523ccf Add support for USM shared in WSL for dGPU
This patch force KMD allocation path for USM host allocation
and also for host part of USM shared allocation


Related-To: NEO-6913
Signed-off-by: Kamil Diedrich <kamil.diedrich@intel.com>
2022-10-18 14:00:14 +02:00
6a73001a3f Implement zesSchedulerComputeUnitDebugMode sysman Api
Related-To: LOCI-866

Signed-off-by: Jitendra Sharma <jitendra.sharma@intel.com>
2022-10-18 13:43:29 +02:00
c224fd0666 Require LWS to be in ascending format X >= Y >= Z
This change fixes problem with memory locality.
When calculating work group size do not take into account
work group sizes where there's bigger number of elements in
higher dimensions namely: Y>X or Z>Y.

Related-To: NEO-5719

Signed-off-by: Krystian Chmielewski <krystian.chmielewski@intel.com>
2022-10-18 13:23:57 +02:00
0fdb308207 Use correct context for validateHostPtr
- with per context VMs, context with subdeviceBitfield must
be selected

Resolves: NEO-7426

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2022-10-18 13:10:28 +02:00
3fced92766 L0 debugger - decanonize SDI address before adding to buffer
Signed-off-by: Yates, Brandon <brandon.yates@intel.com>
22.43.24558
2022-10-17 20:22:17 +02:00
6bcb2ea063 infra update
Signed-off-by: Artur Harasimiuk <artur.harasimiuk@intel.com>
2022-10-17 18:28:41 +02:00
4affe9907c Update SLM access offset behavior
Related-To: NEO-5998

Signed-off-by: Matias Cabral <matias.a.cabral@intel.com>
2022-10-17 18:07:27 +02:00
3d1dba3d10 Update test cmake file
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2022-10-17 18:00:18 +02:00
fc1a1b975a Remove check for symmetrical SKU on XE_HPC_CORE
Signed-off-by: Rafal Maziejuk <rafal.maziejuk@intel.com>
Related-To: NEO-6989
2022-10-17 17:47:35 +02:00
7865975aca Correct check for symmetrical SKU on XE_HPC_CORE
Signed-off-by: Rafal Maziejuk <rafal.maziejuk@intel.com>
Related-To: NEO-6989
2022-10-17 16:03:58 +02:00
2dd3c11248 Do not copy ze_ipc_mem_handle_t in internal functions of NEO
The size of this type is at least 64 bytes due to
ZE_MAX_IPC_HANDLE_SIZE. This change introduces usage
of const reference instead of copying.

Signed-off-by: Patryk Wrobel <patryk.wrobel@intel.com>
2022-10-17 13:51:03 +02:00
90597995d2 Do not copy ze_ipc_event_pool_handle_t in internal functions of NEO
The size of this type is at least 64 bytes due to
ZE_MAX_IPC_HANDLE_SIZE. This change introduces usage
of const reference instead of copying.

Signed-off-by: Patryk Wrobel <patryk.wrobel@intel.com>
2022-10-17 13:17:24 +02:00
5e36b1fcbf Revert "Calculate CS timestamp based on OA timestamp and frequencies ratio"
This reverts commit 03c528382f.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2022-10-16 11:14:46 +02:00
b2c80be6f9 Fix typo and error check when retrieving pci information
Fix
1. typo PhysialDevicePciSpeedInfo to PhysicalDevicePciSpeedInfo
2. error check to verify open() call

Signed-off-by: Joshua Santosh Ranjan <joshua.santosh.ranjan@intel.com>
22.42.24548
2022-10-14 14:57:07 +02:00
95505d87a5 L0Debug - fix interrupt
- pass deviceIndex based on deviceBitfield
- do not call ioctl again on EBUSY error

Resolves: NEO-7414

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2022-10-14 13:56:57 +02:00