Commit Graph

4691 Commits

Author SHA1 Message Date
Compute-Runtime-Validation bcc5a8502b Revert "feature: add functions to get/set VA space"
This reverts commit 0550c80d4f.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2023-11-03 02:06:36 +01:00
Young Jin Yoon 0550c80d4f feature: add functions to get/set VA space
Added an extension version of zeContextGetVirtualAddressSpaceIpcHandle
and zeContextPutVirtualAddressSpaceIpcHandle, which return
ZE_RESULT_ERROR_UNSUPPORTED_FEATURE for now.

Related-To: NEO-8259

Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
2023-11-02 18:42:27 +01:00
Dominik Dabek 43841fd2ef fix: clear AllocationsList tail on free all
Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2023-11-02 18:06:13 +01:00
Zbigniew Zdanowicz 93cf33fda0 performance: remove redundant tlb flush on monitor fence
Related-To: NEO-8395

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2023-11-02 12:39:05 +01:00
Mateusz Hoppe 4a5d9eeae7 refactor: do not trigger debug break when mkdir fails with EEXIST
Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2023-11-02 12:26:32 +01:00
Aravind Gopalakrishnan 99abb40a45 feature: Add extensions to expose dot product capabilities
- Add new extension property for user to query platform DP support.
- Add new blackbox and ULTs for extension.

Related-To: NEO-8786

Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>
2023-10-31 21:54:39 +01:00
Zbigniew Zdanowicz db10e85526 refactor: add new diagnostic mode for device state
Related-To: NEO-8395

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2023-10-31 16:02:53 +01:00
Compute-Runtime-Validation c9d6615231 Revert "feature: Add extensions to expose dot product capabilities"
This reverts commit e988fe66e3.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2023-10-31 05:06:06 +01:00
Michal Mrozek 060fc40938 performance: move constants out of the loop.
Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
2023-10-30 18:16:35 +01:00
Mateusz Jablonski d56abe6f60 refactor: correct variable naming
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-10-30 17:26:20 +01:00
Aravind Gopalakrishnan e988fe66e3 feature: Add extensions to expose dot product capabilities
- Add new extension property for user to query platform DP support.
- Add new blackbox and ULTs for extension.

Related-To: NEO-8786

Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>
2023-10-30 15:15:51 +01:00
Mateusz Jablonski 036d1e4814 refactor: correct variable naming
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-10-30 12:48:51 +01:00
tkordek f8ad191b16 feature: Caller ID introduced to aubstream
... to distinguish environment of capturing.

Related-To: NEO-9235

Signed-off-by: Tomasz Kordek <tomasz.kordek@intel.com>
2023-10-30 12:37:49 +01:00
Mateusz Jablonski 80e59ff344 fix: don't call virtual method in ctor
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-10-30 12:26:57 +01:00
Michal Mrozek 80e3632c41 performance: do not free deferred allocations when not needed
Level Zero doesn't use deferred allocations so no point in paying the
price to check for them in cleanAllocationList.

Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
2023-10-27 18:02:43 +02:00
Dominik Dabek 39cf653959 performance(ocl): cmd buffer prealloc per cmdqueue
Add mechanism to preallocate cmd buffer allocations in command stream
receiver reusable allocations list per command queue initialized.

This should limit additional allocations during hot loop.

Needs to be enabled in subsequent commits by setting product helper
method.

Related-To: NEO-8152

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2023-10-27 16:56:29 +02:00
Mateusz Jablonski 1e18bf0392 fix: define timestamp mask based on timer resolution
Related-To: HSD-18034577551
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-10-27 11:55:20 +02:00
Mateusz Jablonski 2eebbcd41e fix: don't transfer FtrTile64Optimization to gmmlib
Related-To: HSD-18034338825
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-10-27 08:07:58 +02:00
Mateusz Hoppe 5d572b9c8f feature: allow freeing memory in aubstream
Related-To: NEO-2707

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2023-10-26 17:16:23 +02:00
Compute-Runtime-Validation fca2159430 Revert "fix: if device hierarchy is flat then getSubDevicesCount return 1u"
This reverts commit cb0bb57f49.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2023-10-26 15:40:29 +02:00
Baj, Tomasz cb0bb57f49 fix: if device hierarchy is flat then getSubDevicesCount return 1u
Related-To: NEO-9167

Signed-off-by: Baj, Tomasz <tomasz.baj@intel.com>
2023-10-25 15:51:52 +02:00
Andrzej Koska c4a24d8f31 fix: Avoiding races when modifying queue and Tag data
Related-To: NEO-8196
Signed-off-by: Andrzej Koska <andrzej.koska@intel.com>
2023-10-25 14:20:57 +02:00
Naklicki, Mateusz 1352bde5c8 fix: use pointer instead of std::function
MSVC Compiler found an issue that there could be an attempt to
reference a deleted function when using std::function.
As the functionality of ioctlHelper::getGpuTime is very straightforward,
simple function pointer should be sufficient to make the job and
it mitigates overhead of std::function.

Related-To: NEO-8324
Signed-off-by: Naklicki, Mateusz <mateusz.naklicki@intel.com>
2023-10-25 06:23:43 +02:00
Compute-Runtime-Validation 69f614a8c2 Revert "fix: allow legacy device binary validation logic for Blender on DG2 p...
This reverts commit d3d15542fb.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2023-10-24 21:00:19 +02:00
Jablonski, Mateusz ac5f64f5c6 fix: fix compilation error in clang on Windows (2/n)
Signed-off-by: Jablonski, Mateusz <mateusz.jablonski@intel.com>
2023-10-24 15:59:06 +02:00
Mateusz Jablonski 6d2d16d68e fix: avoid overflow of gpu time stamp in ns
Related-To: NEO-8394
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-10-24 14:56:15 +02:00
Francois Dugast 1fd9d29f54 feature: update Xe KMD header
xe_drm.h header is generated from the series "uAPI Alignment - take 1 v4"
from https://patchwork.freedesktop.org/series/124387/

Related-To: NEO-8324
Signed-off-by: Francois Dugast <francois.dugast@intel.com>
Signed-off-by: Kamil, Kopryk <kamil.kopryk@intel.com>
Signed-off-by: Naklicki, Mateusz <mateusz.naklicki@intel.com>
2023-10-24 12:03:45 +02:00
Kacper Nowak d3d15542fb fix: allow legacy device binary validation logic for Blender on DG2 platforms
Temporarily opt-out from additional compatibility checks
on DG2 for Blender AOT-compiled kernels.
This prevents a long kernel recompilation.

Additionally, same behavior can be enforced for other applications
manually via NEO debug key named DoNotUseProductConfigForValidationWa.

Signed-off-by: Kacper Nowak <kacper.nowak@intel.com>
Related-To: NEO-9240
2023-10-23 18:20:37 +02:00
Zbigniew Zdanowicz 0b42510f41 fix: add l1 cache flush to barrier operations when required
Related-To: NEO-8395

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2023-10-23 18:18:22 +02:00
Mateusz Hoppe 52b0f32688 fix: offset cpu address when writing chunk in simulated csr
- not only gpuAddress is offset but also cpu address with data needs
to be offset while writing memory.

Related-To: GSD-6604

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2023-10-23 17:01:25 +02:00
Mateusz Jablonski 8da4a9cbc7 fix: add debug flag to control non walker signalling in in-order cmdlist
set to false by default

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-10-23 14:43:40 +02:00
Kulkarni, Ashwin Kumar 875f653952 Revert "feature: support UUID feature for windows"
This reverts commit cb2b4214b4.

Related-To: NEO-8799

Signed-off-by: Kulkarni, Ashwin Kumar <ashwin.kumar.kulkarni@intel.com>
2023-10-23 14:08:03 +02:00
Zbigniew Zdanowicz 2e09b5ff66 refactor: reposition preamble helper implementation methods
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2023-10-20 13:18:29 +02:00
Fabian Zwolinski b8831e87b0 fix: iGPUs - increase size of reported global memory available on Linux to 94%
Related-To: NEO-8989, NEO-7229
Signed-off-by: Fabian Zwolinski <fabian.zwolinski@intel.com>
2023-10-20 10:51:47 +02:00
Fabian Zwolinski ede8663e41 fix: iGPUs - increase size of reported global memory available on Windows to 94%
Related-To: NEO-8989, NEO-7229
Signed-off-by: Fabian Zwolinski <fabian.zwolinski@intel.com>
2023-10-20 10:37:11 +02:00
Fabian Zwolinski 55abd8dc9a fix: reduce globalMemory by internalResourcesSize for iGPUs
internalResourcesSize is a constant 450 MB value.

Related-To: NEO-8989
Signed-off-by: Fabian Zwolinski <fabian.zwolinski@intel.com>
2023-10-20 10:10:41 +02:00
Kozlowski, Marek 90987bd878 feature: add minimal interface for software tag injection
* Add new Software Tag `ArbitraryString`
* Add minimal interface for injecting software tags

Related-To: NEO-5550

Signed-off-by: Marek Kozlowski <marek.kozlowski@intel.com>
2023-10-20 01:49:33 +02:00
Mateusz Hoppe 97faeae16f feature: heapAssigner per root device
- create heapAssigner per root device in memory manager to allow per
device config

Related-To: NEO-7063

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2023-10-19 19:52:24 +02:00
Dunajski, Bartosz 25195ebc96 fix: capability to write memory chunk in aub/tbx mode
Related-To: GSD-6604

Signed-off-by: Dunajski, Bartosz <bartosz.dunajski@intel.com>
2023-10-19 19:13:11 +02:00
Mateusz Jablonski 4dfa12c8eb fix: add mechanism to detect gpu timestamp overflows
unify naming CpuGpu to GpuCpu

Related-To: NEO-8394
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-10-19 16:31:06 +02:00
Lukasz Jobczyk 750b5ba89a fix: flush necessary caches when dispatch pipe control
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2023-10-19 15:09:31 +02:00
Spruit, Neil R 58fa968273 fix: Calculate size of buffer surface state given mapped allocations
Related-To: NEO-8350

- given a virtual address part of a mapping to multiple physical
allocations, then the buffer surface state size is increased to
include the allocations which follow the current allocation, which
allows users access to the remainder of the mapped buffer.

Signed-off-by: Spruit, Neil R <neil.r.spruit@intel.com>
2023-10-19 13:38:51 +02:00
Katarzyna Cencelewska ff1a8770fe fix: when gt type media on xe kmd then do not inizialize
also use proper size when query config

Related-To: HSD-18034189281
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2023-10-19 12:37:36 +02:00
Spruit, Neil R 456f07212d fix: Return pageSize/Heap for reservation based on the size and devices
Related-To: NEO-9127,NEO-9142

- Change the heap/pageSize used based on the size of the reservation
required and the device heaps available.
- Return the page size required based on the size requested by the user
for virtual reservation and the devices on the system.
- Check the size passed in by the user in memory map and physical memory
allocation is valid for the heap and page size that is required.

Signed-off-by: Spruit, Neil R <neil.r.spruit@intel.com>
2023-10-19 10:44:13 +02:00
Mateusz Hoppe 5a4fa180d6 feature: control bindless compilation mode based on release
- check releaseHelper support when selecting bindless mode, if not
disabled, prefer bindless mode in L0 API
- bindless mode can be forced with DebugVariable: UseBindlessMode

Related-To: NEO-7063

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2023-10-19 10:11:11 +02:00
Compute-Runtime-Validation 096e424428 Revert "fix: initialize tile only when gt type main on xe kmd"
This reverts commit cbde47d638.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2023-10-19 09:47:58 +02:00
John Falkowski f156a74f54 fix: split chunking prefetch flags
Related-To: NEO-9120

Signed-off-by: John Falkowski <john.falkowski@intel.com>
2023-10-18 19:20:42 +02:00
Katarzyna Cencelewska cbde47d638 fix: initialize tile only when gt type main on xe kmd
also use proper size when query config

Related-To: HSD-18034189281
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2023-10-18 16:01:16 +02:00
Compute-Runtime-Validation 713aa05f27 Revert "performance: limit monitor fence dispatch only to task count dispatches"
This reverts commit f26f027ee7.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2023-10-18 11:15:40 +02:00
Mateusz Jablonski 15484c6778 fix: initialize member before using it
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-10-18 08:38:22 +02:00