Commit Graph

3204 Commits

Author SHA1 Message Date
Mateusz Jablonski 6af9856e33 fix: reorder members in BufferObject class
Related-To: HSD-13011781488
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-25 11:44:38 +01:00
Young Jin Yoon 27a3307bb0 feature: support explicit memory locking
Added lockMemory in context to explicitly locking memory,
Added a boolean flag in graphics_allocation to indicate the allocation
is locked, and modified memory_operations_handler to add lock().
Change the logic to work correctly with makeResident() when lock() is
called previously for the same memory region

Related-To: NEO-8277
Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
2024-03-25 09:49:18 +01:00
Compute-Runtime-Validation df164174b4 Revert "fix: pass FtrTile64Optimization as-is"
This reverts commit 22d08dabc4.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-03-22 23:56:28 +01:00
Mateusz Jablonski 8d38a0c726 fix: remove not needed hardcoded caps reported by ioctl helper xe
Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-22 19:00:28 +01:00
Krzysztof Gibala 4d797890af fix: Use FtrHwScheduling for setting ftrWddmHwQueues
Related-To: NEO-10212

Signed-off-by: Krzysztof Gibala <krzysztof.gibala@intel.com>
2024-03-22 18:36:24 +01:00
Mateusz Jablonski d94be09020 refactor: remove not needed check for exec softpin
Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-22 17:30:49 +01:00
Mateusz Jablonski 8e6240ac70 fix: don't use fake userptr flag in ioctl helper xe
Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-22 15:59:56 +01:00
Compute-Runtime-Validation a75db375a1 Revert "fix: remove hardcoded caps reported by ioctl helper xe"
This reverts commit 0b2e8e2848.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-03-22 15:15:19 +01:00
Dominik Dabek 2b964254d6 performance: debug key for adjust ULLS on battery
ULLS controller timeout settings will be adjusted based on ac line
status and lowest queue throttle from submissions.

Lowest queue throttle is reset when controller stops ULLS.

Related-To: NEO-10800

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2024-03-22 14:24:00 +01:00
Mateusz Jablonski ec19ce536a refactor: store userptr value in buffer object
Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-22 12:49:13 +01:00
Katarzyna Cencelewska da7b03dd15 fix: to always use grfs count in calculateNumThreadsPerThreadGroup
grf size != grf count

Related-To: GSD-8437
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2024-03-22 11:03:18 +01:00
Mateusz Jablonski 0b2e8e2848 fix: remove hardcoded caps reported by ioctl helper xe
Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-22 10:41:54 +01:00
Katarzyna Cencelewska dd1d52259e refactor: add param rootDeviceEnvironment to calculateNumThreadsPerThreadGroup
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2024-03-21 22:25:14 +01:00
Young Jin Yoon ec009cf9e3 fix: abort only when disabling scratch page
Modifed getResetStatus to abort only when scratch page is disabled
Removed an incorrect UNRECOVERABLE_IF statement based on the status:
validPageFault can be true when banned flag is not set, if CAT error
does not occur as a result of page fault.

Related-To: GSD-5673
Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
2024-03-21 21:55:25 +01:00
Mateusz Jablonski 4df0dd7894 fix: remove hardcoded caps reported by ioctl helper xe
Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-21 21:09:20 +01:00
Mateusz Jablonski 92d37b20a6 fix: setup gpu address space based on config info from xe kmd
Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-21 18:49:19 +01:00
Mateusz Jablonski 1e343053ba refactor: remove redundant recreating vector of engines in xe kmd path
make ContextParamEngine structure more generic and populate engines
by drm specific methods

Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-21 17:55:39 +01:00
Mateusz Jablonski 22d08dabc4 fix: pass FtrTile64Optimization as-is
Related-To: NEO-10623
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-21 16:52:27 +01:00
Aravind Gopalakrishnan 04b99de4d6 refactor: Force tlb flush during TC after copy
Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>
2024-03-21 07:25:46 +01:00
Mateusz Hoppe 0cec5ccc32 fix: check if any of devices has BindlessHeapHelper enabled
- report Bindless_images extension in driver only when at least one
device has global bindless heaps enabled

Related-To: NEO-10352

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2024-03-20 20:50:58 +01:00
Mateusz Jablonski 614ed7ff5f fix: remove not needed appendBaseAddressOffset method
Related-To: NEO-10623
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-20 17:46:14 +01:00
Wenju He 03078541d7 feature: make global bindless heaps resident when created
Make bindless heaps resident right after heap allocation.
Motivation is that SYCL bindless image can be passed as a value argument
or through memory. Therefore, we're not able to make its bindless heap
resident during kernel initialization or setting kernel arguments.

This fixes SYCL bindless image read_write_*D.cpp tests on DG2.

Related-To: NEO-7063
Signed-off-by: Wenju He <wenju.he@intel.com>
2024-03-20 14:40:11 +01:00
Mateusz Hoppe 27b930cabc refactor: allow default setting for UseExternalAllocatorForSshAndDsh
- value of -1 selects driver default setting for external allocator

Related-To: NEO-7063

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2024-03-20 12:29:56 +01:00
Raiyan Latif 40471da591 fix: properly retrieve sub-device count in allocateRTDispatchGlobals
Previous method used for retrieiving sub-device count in this path
did not take into account device affinity mask nor device hiearchy
mode, resulting in segmentation faults when attempting to allocate
rtDispatchGlobals structure using improper deviceBitFields.

Related-To: NEO-8422

Signed-off-by: Raiyan Latif <raiyan.latif@intel.com>
2024-03-20 09:23:30 +01:00
Jaroslaw Warchulski 7c5640669c fix: incorrect check in parseAffinityMask function
Resolves: GSD-8333
Signed-off-by: Jaroslaw Warchulski <jaroslaw.warchulski@intel.com>
2024-03-19 09:20:00 +01:00
Mateusz Jablonski a2742492ab feature: enable xe drm detection by default
driver is built with xe drm support by default

added cmake flag to control xe eu debug API support
NEO_ENABLE_XE_EU_DEBUG_SUPPORT

This flag is disabled by default and uapi-eu-debug headers are not
needed for driver compilation as these headers are not a part of
upstream kernel yet.

Related-To: NEO-10780

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-19 08:47:13 +01:00
Zbigniew Zdanowicz d1041e2335 feature: find cpu base address from all command buffers of container
Related-To: NEO-10381

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-03-18 13:43:08 +01:00
Mateusz Jablonski 19dcc80e44 Revert "build: enable xe drm detection by default"
This reverts commit 973757a58d.

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-18 09:04:59 +01:00
Compute-Runtime-Validation 016c234893 Revert "feature: disable scratch page by default"
This reverts commit dab5469f81.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-03-16 01:52:00 +01:00
Mateusz Jablonski 1e1d675606 fix: disable passing FtrTile64Optimization to gmmlib
add debug key to control if the value should be passed

Related-To: NEO-10785
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-15 17:42:53 +01:00
Mateusz Jablonski 0270cd6a5b fix: respect gt id when getting engines for drm context under xe kmd
Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-15 16:02:47 +01:00
Bellekallu Rajkiran 9736313d10 feature: Support for ccs mode configuration via SysFs
Add support for configuring ccs mode for all applicable devices
before KMD is loaded.

Use ZEX_NUMBER_OF_CCS to configure ccs mode.

Format is as follows:

ZEX_NUMBER_OF_CCS=NumberOfCcs i,e Setting ZEX_NUMBER_OF_CCS
to 4 sets ccs mode to 4 for all devices for which configuration
is supported.

Related-To: NEO-10378

Signed-off-by: Bellekallu Rajkiran <bellekallu.rajkiran@intel.com>
2024-03-15 15:51:45 +01:00
Zbigniew Zdanowicz f496233462 refactor: use single vector to store used heap allocations
Related-To: NEO-10483

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-03-15 15:01:36 +01:00
Young Jin Yoon dab5469f81 feature: disable scratch page by default
Modified default values for disableScratch and gpuPageFault
to true and 10 respectively in drm_nep.cpp, in order to
disable scratch pages by default.

Related-To: GSD-5673
Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
2024-03-15 11:44:10 +01:00
Mateusz Jablonski 470def9be6 test: fix variable naming
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-15 10:06:46 +01:00
Young Jin Yoon 82728ff394 feature: add logic to iterate for all contexts to check GPU pagefault
Implemented to go through entire contexts in the process and then query
reset status to check the unexpected GPU segfault.

Added a new debug variable GpuFaultCheckThreshold to change the checking
frequency for each hang check for performance analysis.

Related-To: GSD-5673
Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
2024-03-15 07:48:39 +01:00
Compute-Runtime-Validation 94cc48f81b Revert "fix: don't use fake userptr flag in ioctl helper xe"
This reverts commit d3ab256f55.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-03-15 03:08:01 +01:00
Mateusz Jablonski d3ab256f55 fix: don't use fake userptr flag in ioctl helper xe
Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-14 18:41:17 +01:00
Neil R. Spruit b5f8a38f19 feature: Enable Per IP euStall Functionality
Related-To: NEO-10220

Signed-off-by: Neil R. Spruit <neil.r.spruit@intel.com>
2024-03-14 16:49:52 +01:00
Compute-Runtime-Validation ef7dbc99f1 Revert "fix: don't use fake userptr flag in ioctl helper xe"
This reverts commit 98824fdaf6.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-03-14 14:35:14 +01:00
Mateusz Jablonski 833fa6bce1 fix: correct querying engines from xe kmd
we get drm_xe_query_engines, not array of drm_xe_engine_class_instance

Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-14 12:06:25 +01:00
Mateusz Jablonski 98824fdaf6 fix: don't use fake userptr flag in ioctl helper xe
Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-14 10:07:38 +01:00
Young Jin Yoon 7b81c4e08f feature: abort when unexpected GPU page fault detected
If ResetStats from i915 is from the GPU page fault, abort
the entire process instead of disabling engines.
Added a fallback mechanism when prelim_drm_i915_reset_stats
fails.

Related-To: GSD-5673
Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
2024-03-14 08:14:59 +01:00
Mateusz Jablonski 0210e37f03 fix: respect gt id when finding xe engine info
Related-To: NEO-10496
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-03-13 20:52:36 +01:00
Bartosz Dunajski e400a5c624 feature: in-order host counter allocation pooling 2
Related-To: NEO-10507

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-03-13 18:59:02 +01:00
Compute-Runtime-Validation 9cce1183cd Revert "feature: use prelim reset_stats for detailed statisics"
This reverts commit 835dc8b594.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-03-13 14:31:57 +01:00
Aravind Gopalakrishnan 3f20dd3b49 refactor: Add optional user fence during unbind
Add optional fence and wait operations after unbind operation.

Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>
2024-03-13 12:47:44 +01:00
Young Jin Yoon 835dc8b594 feature: use prelim reset_stats for detailed statisics
Added getResetStats() in ioctl_helper.h to support extended header for
prelim_drm_i915_reset_stats.
Added new data structure to capture the fault data structure for prelim.

Related-To: GSD-5673
Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
2024-03-13 11:37:04 +01:00
Francois Dugast 5483e466e8 fix: Align on strings returned for unknown values
Related-to: NEO-10321

Signed-off-by: Francois Dugast <francois.dugast@intel.com>
2024-03-13 11:21:51 +01:00
Dominik Dabek d664995221 performance: enable localPreferred on dg2
Use localPreferred for buffer, svmGpu and image allocations on dg2.

Related-To: NEO-9695

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2024-03-12 22:47:47 +01:00