Commit Graph

3315 Commits

Author SHA1 Message Date
Mateusz Hoppe
e6ed42d056 fix: apply relocations in isa segments for builtin kernels
- use correct patched isa segments when transferring isa for builtins to
preserve applied relocations
- do not set requiresImplicitArgs to false for builtins, zebin defines
is implicit arg buffer is required

Related-To: NEO-14667, NEO-15276

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2025-06-16 17:51:55 +02:00
Maciej Plewka
cd27bb32c5 fix: In order signaling with pipe control for fill operation
Related-To: NEO-15192
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2025-06-16 15:58:44 +02:00
Mateusz Jablonski
0a8248fc9f fix: expose P2P property access flag based on canAccessPeer result
simplify case with same root device indices

Related-To: NEO-14938
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2025-06-16 15:47:01 +02:00
Aravind Gopalakrishnan
253764f440 refactor: Use handle member in IpcHandleTracking
Related-To: NEO-15185

Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>
2025-06-16 13:17:38 +02:00
Dominik Dabek
e5477e614e fix: prepare for l0 host usm pool
enable only if all devices support hostUsmPooling

split product helper method by api type for pre PTL products

add ult helper function to cleanup and initialize pools -
used when replacing memory managers or normal cleanup during driver
destruction would access already freed resources

adjust ults

Related-To: NEO-6893

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2025-06-16 12:13:45 +02:00
Mateusz Jablonski
f165e713a5 fix: cache canAccessPeer result of each pair of devices
query peer access only once per pair of devices

Related-To: NEO-14938

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2025-06-16 12:04:08 +02:00
Kamil Kopryk
cdf087f3da fix: correct bindless offsets for L0 bindless images extension with heapless
Related-To: NEO-14710
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2025-06-16 11:50:18 +02:00
Oskar Hubert Weber
0b5c9125ac fix: skip cleanup in DllMain when terminating process
Per https://learn.microsoft.com/en-us/windows/win32/dlls/dllmain,
it's not safe to cleanup resources in DllMain when
fdwReason == DLL_PROCESS_DETACH and lpvReserved != NULL (the process is
terminating), e.g. can result in a hang in some cases.

Related-To: NEO-14121

Signed-off-by: Oskar Hubert Weber <oskar.hubert.weber@intel.com>
2025-06-16 11:08:43 +02:00
Bartosz Dunajski
9fc244c83a refactor: remove not needed imm cmd list flush task code
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2025-06-16 08:59:51 +02:00
Lukasz Jobczyk
747a708213 performance: Allow optimized CB events on MCL
Resolves: NEO-14195

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-06-13 18:44:29 +02:00
Mateusz Jablonski
d842363c8b refactor: remove not needed function
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2025-06-13 17:21:48 +02:00
Bartosz Dunajski
7d5679cbfe fix: enable flush task submission for internal cmd lists
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2025-06-12 19:44:14 +02:00
Damian Tomczak
5cd1423bab feature: packed surface format
Resolves: NEO-13669

Signed-off-by: Damian Tomczak <damian.tomczak@intel.com>
2025-06-12 17:32:16 +02:00
Zbigniew Zdanowicz
762dc5ef8e refactor: pass heap allocation as pointer and not as reference to encoder
Related-To: NEO-10492

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2025-06-12 11:04:09 +02:00
Raiyan Latif
b6c2c59d8f feature: Add product helper for PidFdOrSocketForIPC
Related-To: NEO-15183

Signed-off-by: Raiyan Latif <raiyan.latif@intel.com>
2025-06-12 02:02:14 +02:00
Mateusz Hoppe
534f7837fe refactor: add validateWorkgroupSize() method
Related-To: NEO-14209

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2025-06-11 18:17:57 +02:00
Lukasz Jobczyk
f0bf8ade4d performance: Enable async builtin init xe2+, linux and windows
Related-To: NEO-14821

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-06-11 11:25:26 +02:00
Mateusz Jablonski
c01ef3d9ad refactor: remove redundant exec environment pointer from L0::DeviceImp
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2025-06-10 22:10:02 +02:00
Bartosz Dunajski
9505230c50 refactor: move prefetch methods to different files
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2025-06-10 16:54:53 +02:00
Compute-Runtime-Validation
523bd92db4 Revert "refactor: remove redundant exec environment pointer from L0::DeviceImp"
This reverts commit c7893fb3ed.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2025-06-10 10:04:06 +02:00
Bartosz Dunajski
ff7d0a76cb feature: prework to enable memory prefetch
Related-To: NEO-14703

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2025-06-10 01:05:41 +02:00
John Falkowski
805a716fe3 feature: Implement appendMemoryCopy/Fill for Shared System USM
Related-To: NEO-13697

Signed-off-by: John Falkowski <john.falkowski@intel.com>
2025-06-09 16:18:56 +02:00
Bartosz Dunajski
17a15a3d9d refactor: dedicated prefetch cmd list method
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2025-06-09 13:33:02 +02:00
Mateusz Jablonski
c7893fb3ed refactor: remove redundant exec environment pointer from L0::DeviceImp
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2025-06-09 12:03:58 +02:00
Compute-Runtime-Validation
9fd521f471 Revert "refactor: add validateWorkgroupSize() method"
This reverts commit 708b81bab5.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2025-06-07 17:25:49 +02:00
Bartosz Dunajski
6d7148ee5e refactor: move prefetch to dedicated method
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2025-06-06 16:49:19 +02:00
Zbigniew Zdanowicz
b6018010c4 refactor: add option to skip to add to residency event external allocation
Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2025-06-06 15:55:03 +02:00
Mateusz Hoppe
708b81bab5 refactor: add validateWorkgroupSize() method
Related-To: NEO-14209

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2025-06-06 09:45:45 +02:00
Bartosz Dunajski
d9ec4b1a18 refactor: improve thread data size getter
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2025-06-04 11:27:56 +02:00
Aravind Gopalakrishnan
953639411a fix: Return additional nodes correctly to tag allocator
Related-To: NEO-13003

Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>
2025-06-04 10:57:06 +02:00
Mateusz Jablonski
f487cbec7a refactor: update L0 base handle type definiton to ensure std layout
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2025-06-03 20:35:47 +02:00
Mateusz Hoppe
f1f3ee13db refactor: set argsExtended in EncodeWalkerArgs
Related-To: NEO-14209

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2025-06-03 11:45:39 +02:00
Dominik Dabek
cf87684644 performance: enable l0 device usm pool
Disable if debugging is enabled.

Related-To: NEO-6893

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2025-06-03 10:32:08 +02:00
Zbigniew Zdanowicz
2efa5cb907 refactor: add command list state base address hook method
Related-To: NEO-10492

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2025-06-02 11:59:23 +02:00
Chodor, Jaroslaw
46c60290b9 refactor: introducing igfxfmid_wrapper
Signed-off-by: Chodor, Jaroslaw <jaroslaw.chodor@intel.com>
2025-05-30 19:26:04 +02:00
Mateusz Jablonski
13a5e90124 refactor: split definitions from cmdlist_launch_params.h
Moving defintion of:
- CmdListKernelLaunchParamsExt to cmdlist_launch_params_ext.h
- CopyOffloadMode to copy_offload_mode.h
- MemoryCopyParams to cmdlist_memory_copy_params.h
- CommandToCopy to command_to_copy.h

Store LaunchParamsExt as non-pointer member of CmdListKernelLaunchParams

Related-To: NEO-14560
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2025-05-30 17:39:35 +02:00
Katarzyna Cencelewska
828d6bafa7 fix: return proper value for zeKernelSuggestGroupSize
Resolves: HSD-18042274687
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2025-05-30 16:35:21 +02:00
Lukasz Jobczyk
332340b02b refactor: Prework for release fence removal leaving acquire fence
Related-To: NEO-14642

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-05-30 12:01:28 +02:00
Krzysztof Sprzaczkowski
b0db130df8 fix: Add coverage for all memory types in L0 API
Related-To: NEO-12807
Signed-off-by: Krzysztof Sprzaczkowski <krzysztof.sprzaczkowski@intel.com>
2025-05-29 17:35:50 +02:00
Katarzyna Cencelewska
e357e7e404 fix: correct limitation for num threads per thread group
taking into account the max work group limit
when simd1 don't use the same limitation as for simd32

Resolves: NEO-14922
Related-To: NEO-11881
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2025-05-29 16:34:40 +02:00
Dominik Dabek
ad4d7a2ce8 fix: l0 usm pooling, ipc handling
Move getting usm pool for ptr to its own method.
Move trying to free via pooling to its own method.

Use base ptr of pool for tracking in IPCHandleMap.
Track ipc handle refcount in freeMem.
Return error when trying to use a not allocated ptr within pool.

Related-To: NEO-6893

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2025-05-29 16:07:19 +02:00
Bartosz Dunajski
cdc725e0fe fix: flush split task count
Related-To: NEO-14641

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2025-05-29 11:30:38 +02:00
Szymon Morek
0c605f2da1 performance: remove cache flush after transfers
Related-To: NEO-15024

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2025-05-29 11:06:37 +02:00
Young Jin Yoon
2773bd8ec0 fix: estimate blitter commands for memory fill
Added getNumberOfBlitsForColorFill, getNumberofBlitsForByteFill,
and getNumberOfBlitsForFill in BlitCommandsHelper to estimate
total number of blitter command generated for
dispatchBlitMemoryColorFill, dispatchBlitMemoryByteFill, and
dispatchBlitMemoryFill, respectively.

Use getNumberOfBlitsForColorFill to determine whether
additionalTimestampNode is required for appendMemoryBlitFill.

Related-To: NEO-13003
Signed-off-by: Young Jin Yoon <young.jin.yoon@intel.com>
2025-05-29 00:21:35 +02:00
Dominik Dabek
cdb088b477 fix: bindless image from usm pooled ptr
Add handling for creating bindless image from usm pooled ptr.
Make relevant ULTs test USM device pool enabled/disabled.

Related-To: NEO-6893

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2025-05-28 15:38:07 +02:00
Szymon Morek
ce1bcb536c performance: don't reprogram SBA when not needed
Related-To: NEO-15024

When defer to pat is enabled, all mocs are
set to 0, so no check for uncached mocs is needed

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2025-05-28 15:13:34 +02:00
Maciej Plewka
e6f3ebce5d fix: add pc with stall before barrier with post sync on bmg
Related-To: NEO-14491
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2025-05-28 09:24:21 +02:00
Kamil Kopryk
b7681a2e7b refactor: redirect flush L3 host usm to external by default
Related-To: NEO-13163

Rename ForceL3FlushAfterPostSync to EnableL3FlushAfterPostSync
Rename DisableFlushL3ForHostUsm to RedirectFlushL3HostUsmToExternal

Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2025-05-28 08:50:01 +02:00
Bartosz Dunajski
9c19e9eb96 fix: enable single temporary allocations list mode 2
Related-To: NEO-14641

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2025-05-27 17:35:17 +02:00
Lukasz Jobczyk
0306b77297 performance: Always allocate l0 events in LMEM on Xe2
Resolves: NEO-13422

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-05-27 10:39:26 +02:00