Commit Graph

5514 Commits

Author SHA1 Message Date
3830423b3a refactor: correct typo
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-05-21 13:01:05 +02:00
c39b750790 performance: Cache host resources when mitigate dc flush
Related-To: NEO-10556

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2024-05-21 10:05:29 +02:00
cf8f0b9cd8 feature: 2-Tile device memory chunking independent of KMD migration
Related-To: NEO-10916

Signed-off-by: John Falkowski <john.falkowski@intel.com>
2024-05-21 08:07:10 +02:00
94a4bbac57 Revert "fix: change gmm resource for externalHostPtr"
This reverts commit 63843862df.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-05-21 07:43:53 +02:00
b755a30919 fix: patch correct payload offset value in relocation
Related-To:: GSD-8116
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2024-05-21 06:12:03 +02:00
63843862df fix: change gmm resource for externalHostPtr
Resolves: NEO-10157

Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2024-05-21 00:43:29 +02:00
4116dd5c9e feature: Improving elf rewriter - preserving strings
Original string section needs to be preserved so that
symbol table does not get broken.

Related-To: NEO-10190

Signed-off-by: Chodor, Jaroslaw <jaroslaw.chodor@intel.com>
2024-05-20 22:29:51 +02:00
49cc1a0ba0 fix: use llx for fprintf and IoctlFunctions
Changed format for address printing from %lx to %llx for
fprintf introduced in drm_neo.cpp, and then use
IoctlFunctions::fprintf instead of std::printf to avoid
errors on gcc.

Changed formate for address printing from %lx to %llx for
snprintf introduced in drm_test.cpp, and then type casted
to long long unsigned int explictly to avoid errors.

Related-To: GSD-7611
Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
2024-05-20 16:39:49 +02:00
ce7f578548 refactor: rename pipelined thread thread arbitration policy variable
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-05-20 15:05:52 +02:00
59829aeb11 fix(ocl): Force rebuild if binary contains bindless kernel
Related-To: NEO-7063
Signed-off-by: Fabian Zwoliński <fabian.zwolinski@intel.com>
2024-05-20 14:15:17 +02:00
90927135f9 feature: Adding elf rewriter utility
It will aid in various elf patching tasks both in ocloc
and UMD

Related-To: NEO-10190

Signed-off-by: Chodor, Jaroslaw <jaroslaw.chodor@intel.com>
2024-05-20 13:44:11 +02:00
bfc3e8fcf0 fix: assign unique ContextId for secondary engines
- add secondary engines to allRegisteredEngines

Related-To: NEO-11219

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2024-05-20 13:19:20 +02:00
d79c952d2b fix: set single localMemBanksCount when no MemoryInfo
Support legacy case.

Related-To: NEO-9754
Signed-off-by: Maciej Bielski <maciej.bielski@intel.com>
2024-05-20 12:22:16 +02:00
90df4b298b fix: cache flush dependency for queue blocked
Related-to: NEO-9872, HSD-18038461954
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2024-05-20 11:55:24 +02:00
986ec4f7cc refactor: add functions to store extra engines
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-05-20 10:46:23 +02:00
b0a3f16fc3 feature: brand string update
Related-To: NEO-7929

Updating device brand strings
Based on public support in Windows driver 31.0.101.5522
https://www.intel.com/content/www/us/en/download/785597/

Signed-off-by: ocldev <ocldev@intel.com>
2024-05-20 09:59:55 +02:00
a236171f0d performance(ocl): enable device usm alloc reuse
Enabling on MTL+
Limited to use max 2% of global device memory.

Related-To: NEO-6893, NEO-11463

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2024-05-17 13:32:45 +02:00
cb9977b8f4 feature: create copy offload queue under debug flag
Related-To: NEO-11376

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-05-17 11:04:35 +02:00
34f53d5d94 Revert "performance: Reuse GPU timestamps by default"
This reverts commit 7aceed58ca.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-05-16 21:52:10 +02:00
e204d27190 fix: print to stdout for disable scratch page
Modified to print out error messages to stdout when disable scratch page
is used.

Related-To: GSD-7611
Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
2024-05-16 15:05:07 +02:00
44ef4c1ef2 refactor: add defaultThreadArbitrationPolicy in cmdlist
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-05-16 13:28:20 +02:00
12ab4e5446 feature: Add 3 channel 8/16/32 bit formats
Related-To: NEO-11292

Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>
2024-05-16 13:28:04 +02:00
dc32da2507 feature: prepare StorageInfo for Drm-specific customization
Get the local memory regions count from the primary source (MemoryInfo)
and store for further use when using DrmMemoryManager.
Add a point of dispatch (virtual `createStorageInfoFromProperties`) for
further Drm-specific customizations related to StorageInfo. As the
function became virtual, move one of its callers
(`isLocalMemoryUsedForIsa()`) from the constructor of `MemoryManager` to
respective constructors of all derivative classes. This prevents
bypassing the virtual call dispatch.

Related-To: NEO-9754
Signed-off-by: Maciej Bielski <maciej.bielski@intel.com>
2024-05-16 12:24:48 +02:00
84ba633098 refactor: use legacy versions of injectMMIOList and setTbxServerIp
Related-To: NEO-7539
Signed-off-by: Pawel Cieslak <pawel.cieslak@intel.com>
2024-05-16 10:47:35 +02:00
03d87d27ef fix: generate per process aub file name by default
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-05-16 09:03:21 +02:00
6b04cd4a1e fix: Release the Dummy Allocation before Root Device Environment
Related-To: GSD-6398, GSD-9027

Signed-off-by: Bari, Pratik <pratik.bari@intel.com>
2024-05-15 18:08:41 +02:00
7aceed58ca performance: Reuse GPU timestamps by default
Related-To: NEO-10615

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-05-15 17:51:42 +02:00
8a7444e798 fix: Undo accidental changes in wddm memory manager
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2024-05-15 16:11:30 +02:00
e9b81037a9 refactor: add heaplessStateInitialized flag
Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2024-05-15 15:39:52 +02:00
b4d839fe29 performance(usm): l0, add usm host memory pooling
Disabled by default.

Related-To: NEO-11356

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2024-05-15 15:20:51 +02:00
ae8c7589dc refactor: move implicit arg has indirect access
Move implicit arg has indirect access boolean to kernelAttributes

Related-To: NEO-11396

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2024-05-15 13:11:04 +02:00
a70aaa72ed refactor: add debug message about the zero engine info size
Signed-off-by: Krzysztof Gibala <krzysztof.gibala@intel.com>
2024-05-15 09:07:40 +02:00
06faaab5bb refactor: read scratch page options during init
Change scratch page logic to initialize during Drm::create.

Related-To: GSD-7742
Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
2024-05-15 08:56:14 +02:00
fd47030ad6 fix: use igc indirect detection v3
Update to use igc indirect detection v3. Fix for not detecting indirects
passed as implicit arguments.

Related-To: NEO-11396

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2024-05-15 06:35:42 +02:00
0b0f7bd4cd feature: Enable support for cl_intel_subgroup_2d_block_io
This change enables support for cl_intel_subgroup_2d_block_io extension for
PVC and later.

Related-To: GSD-7825
Signed-off-by: Ratajewski, Andrzej <andrzej.ratajewski@intel.com>
2024-05-14 15:16:49 +02:00
014720fc29 refactor: use DeviceBitfield where appropriate
Replace less type-safe uint32_t and reduce conversions between both
types.

Related-To: NEO-9754
Signed-off-by: Maciej Bielski <maciej.bielski@intel.com>
2024-05-14 09:13:11 +02:00
2c488d9e84 fix: check reset status after completion
Added a logic to check the reset status after the completion to make
sure we go through the logic at least once

Related-To: GSD-8902
Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
2024-05-14 06:38:07 +02:00
aa0441bc63 fix: Iterate from oldest allocation to latest
Related-To: NEO-11409

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-05-13 19:05:11 +02:00
ebf09afc42 Revert "performance: Stop direct submission before removing host ptrs"
This reverts commit 4fd219cdbb.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-05-13 10:10:51 +02:00
de789ac7e5 Revert "performance: remove page size limit for sharing ISAs"
This reverts commit e7c036a91b.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-05-12 09:00:32 +02:00
b37c2970ce test: Rename regkey BcsNumberOverride to BlitterEnableMaskOverride
BlitterEnableMaskOverride is a bitmask with BCS engines available on device

Related-To: NEO-11152

Signed-off-by: Slawomir Milczarek <slawomir.milczarek@intel.com>
2024-05-10 21:18:44 +02:00
e35b951a00 performance: Allow indirect allocs as pack on OpenCL
Related-To: NEO-11228

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-05-10 17:57:42 +02:00
6a24610ed1 Revert "fix: Make makeResident explicitly blocking on linux"
This reverts commit 9095c3ed8a.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-05-10 11:03:13 +02:00
e7c036a91b performance: remove page size limit for sharing ISAs
Related-To: NEO-9403

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-05-10 08:56:49 +02:00
77b817a15c fix: fail when invalid zebin is decoded
- bindless + bindful accesses in single kernel are not supported

Related-To: NEO-7063

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2024-05-09 15:53:00 +02:00
1707434591 refactor: remove unused variable
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-05-09 15:41:50 +02:00
9095c3ed8a fix: Make makeResident explicitly blocking on linux
Currently this call is blocking on Windows and
xe kmd, this commit also makes it blocking on i915.

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-05-09 14:13:42 +02:00
e39893485c fix: add cache flush as dependency for bcs ccs synchronization
Related-to: NEO-9872
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2024-05-09 13:43:39 +02:00
10ed479b16 performance: share inter-module ISA allocations
Related-To: NEO-10258

Currently each module has it's own GA
for kernel ISA's. This change allows new modules to
reuse existing allocation.

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-05-09 08:43:55 +02:00
76af952fcb feature: Bindless sampled image support
Related-To: NEO-10352

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2024-05-08 17:50:37 +02:00