Commit Graph

990 Commits

Author SHA1 Message Date
Jack Myers
f2b5126598 feature: enable tbx fault manager by default
Related-To: NEO-13748
Signed-off-by: Jack Myers <jack.myers@intel.com>
2025-06-23 09:59:32 +02:00
Aleksandra Nizio
f0780df9be fix: Remove unused AubMemDump logic
Related-To: NEO-14718
Signed-off-by: Aleksandra Nizio <aleksandra.nizio@intel.com>
2025-06-20 19:54:48 +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
Kamil Kopryk
bb52c358b9 refactor: flush Task method
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
Related-To: NEO-15011

Also correct ults
2025-06-11 16:36:56 +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
Compute-Runtime-Validation
1a50e8e7c0 Revert "performance: use RESOURCE_BARRIER as stalling barrier"
This reverts commit 556c0b64c6.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2025-05-30 18:00:34 +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
Szymon Morek
556c0b64c6 performance: use RESOURCE_BARRIER as stalling barrier
Related-To: NEO-14943

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2025-05-28 12:30:01 +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
Aleksandra Nizio
1dfc9227c4 fix: Removind address_mapper.h
Related-To: NEO-14718
Signed-off-by: Aleksandra Nizio <aleksandra.nizio@intel.com>
2025-05-27 20:37:56 +02:00
Aleksandra Nizio
e43ec2bbfd fix: Removing stream
Related-To: NEO-14718
Signed-off-by: Aleksandra Nizio <aleksandra.nizio@intel.com>
2025-05-27 18:53:02 +02:00
Aleksandra Nizio
31fe1978d4 fix: Removing streamProvider and addressMapper
Related-To: NEO-14718
Signed-off-by: Aleksandra Nizio <aleksandra.nizio@intel.com>
2025-05-26 16:01:18 +02:00
Lukasz Jobczyk
86e7d5b276 performance: Skip fence after PC when no post sync write
Related-To: NEO-14642

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-05-23 10:33:28 +02:00
Compute-Runtime-Validation
8839d62c79 Revert "performance: use RESOURCE_BARRIER as stalling barrier"
This reverts commit 2a63853349.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2025-05-16 20:19:06 +02:00
Szymon Morek
2a63853349 performance: use RESOURCE_BARRIER as stalling barrier
Related-To: NEO-14943

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2025-05-16 14:12:18 +02:00
Szymon Morek
bff8f74de7 refactor: remove unused parameter
Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2025-05-15 12:23:27 +02:00
Aleksandra Nizio
fa1e3fd6a2 fix: Remove old aub_mem_dump logic and related tests
Related-To: NEO-14718

Signed-off-by: Aleksandra Nizio <aleksandra.nizio@intel.com>
2025-05-13 08:38:59 +02:00
Compute-Runtime-Validation
4ab91eab1e Revert "refactor: move empty sba programming in heapless prolog to base class"
This reverts commit 809feca620.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2025-05-12 14:00:19 +02:00
Kamil Kopryk
809feca620 refactor: move empty sba programming in heapless prolog to base class
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2025-05-09 16:46:02 +02:00
Lukasz Jobczyk
5cd5bbafc5 refactor: Split waitpkg params for ulls light and default
Related-To: NEO-14866

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-05-09 12:40:36 +02:00
Michal Mrozek
ad99f00ba4 refactor: remove unused method
Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
2025-05-09 10:24:37 +02:00
Lukasz Jobczyk
1d1414febc refactor: remove unused dc flush mitigation
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-05-09 08:26:26 +02:00
Lukasz Jobczyk
6f4a56d440 refactor: pass product helper to isFenceAllocationRequired
Related-To: NEO-14642

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-04-28 14:09:02 +02:00
Szymon Morek
3596522637 refactor: remove unused logic in ULLS controller
Related-To: NEO-13843

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2025-04-17 18:35:20 +02:00
Jaroslaw Warchulski
3e1aa33924 refactor: cleanup headers
Related-To: NEO-5548
Signed-off-by: Jaroslaw Warchulski <jaroslaw.warchulski@intel.com>
2025-04-14 14:59:40 +02:00
Mateusz Hoppe
3204411aca refactor: use deviceBitfield from CSR when creating engine
Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2025-04-11 17:30:29 +02:00
Filip Hazubski
504440fc4d feature: Add ftrHeaplessMode flag
Pass hwInfo to isHeaplessModeEnabled and isForceBindlessRequired functions.

Related-To: NEO-14526

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2025-04-02 21:06:05 +02:00
Brandon Yates
a48d66ad75 feature: Add programExceptions stub to CSR
Related-to: NEO-12967

Signed-off-by: Brandon Yates <brandon.yates@intel.com>
2025-04-01 18:33:40 +02:00
Szymon Morek
62964a0b08 fix: invalidate caches when heap is placed into reuse list
Related-To: NEO-9004

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2025-03-31 12:30:29 +02:00
Jack Myers
0aa2c4f0cb feature: allow removal of heapful code paths
Related-To: NEO-13007

Signed-off-by: Jack Myers <jack.myers@intel.com>
2025-03-27 01:34:35 +01:00
Lukasz Jobczyk
60b551758c performance: Adjust waitpkg threshold for discrete devices
Related-To: NEO-14336

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-03-26 14:59:19 +01:00
Kamil Kopryk
73795ced64 refactor: add setupTimestampPacketFlushL3 function
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2025-03-24 13:10:03 +01:00
Lukasz Jobczyk
54cb0e24f8 performance: Switch waitpkg use to tpause for ULLS light
Related-To: NEO-13922, NEO-14336

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-03-21 18:09:37 +01:00
Lukasz Jobczyk
8a85a96ed2 feature: Add 3-level wait scheme with tpause intrinsic
Related-To: NEO-14336

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-03-21 12:12:57 +01:00
Jack Myers
0e25970853 fix: re-add switch case for once writable query
A change related to the tbx fault manager
incorrectly removed a switch case from
`AubHelper::isOneTimeAubWritableAllocationType`.

This fixes that and refactors some APIs to prevent
similar mistakes from happening again by cleaning
up logic.

Addresses show stopper for pre-si pytorch workflows.

Resolves: NEO-14399
Signed-off-by: Jack Myers <jack.myers@intel.com>
2025-03-19 09:54:54 +01:00
Compute-Runtime-Validation
5f7f0dd785 Revert "performance: Enable waitpkg"
This reverts commit 8ec5434460.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2025-03-18 13:28:29 +01:00
Maciej Plewka
36fa6d66ae fix: lock csr in stopDirectSubmission if needed
Related-To: NEO-13875, NEO-14143, HSD-16026538384, HSD-16026780358
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2025-03-18 09:30:38 +01:00
Jack Myers
5f78147e16 fix: hotfix for svmcpu tbx uploads
Test program in the linked, related issue
is crashing in tbx mode. Tbx server indicated
upload of invalid memory was made before exit.

Running with debug messages showed that the
problematic upload was an svmcpu buffer when
running neo with separate cpu and gpu
buffers for shared memory management.

Using this info, the problem was narrowed down
to a missing unprotect call in page fault manager
related code, resulting in a protected(invalid)
memory region getting uploaded to tbx.

It is unclear yet why this unprotect call was not made,
since other svmcpu buffers were uploaded without issue.

This hotfix forces the unprotect call in the fault handler,
which allows the test program to run to completion. However,
there is now a failing test case.

Considering the critical nature of the associated
NEO issue and that this patch should unblock
the work depending on the fix, this hotfix should
get merged regardless of the failing test case.

In the meantime, I will continue triaging the
failing test and will implement a proper fix
once the root cause is isolated.

Related-To: NEO-13404
Signed-off-by: Jack Myers <jack.myers@intel.com>
2025-03-14 04:47:21 +01:00
Zbigniew Zdanowicz
ddc0b0d03b feature: disable flat ring buffer for command list append operation
Related-To: NEO-10356

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2025-03-13 06:06:55 +01:00
Lukasz Jobczyk
8ec5434460 performance: Enable waitpkg
Resolves: NEO-14336

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-03-11 14:43:01 +01:00
Zbigniew Zdanowicz
cd904269ed fix: request for task count should enable monitor fence dispatch
Related-To: NEO-10356

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2025-03-10 18:25:35 +01:00
Lukasz Jobczyk
53062056ec performance: Enable wait pkg for ULLS light
Related-To: NEO-13922

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2025-03-07 12:18:38 +01:00
Damian Tomczak
81b0cac65f fix: raytracing heapless missing allocation
Related-to: NEO-12737

Signed-off-by: Damian Tomczak <damian.tomczak@intel.com>
2025-03-06 17:26:09 +01:00
Brandon Yates
64b027f71c feature: Add gfxCoreHelper for StateSip required
Related-to: NEO-12967

Signed-off-by: Brandon Yates <brandon.yates@intel.com>
2025-03-05 20:24:17 +01:00
Zbigniew Zdanowicz
ae1eb076b7 feature: add optional epilogue to flush task method
Related-To: NEO-10356

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2025-03-05 15:15:59 +01:00
Zbigniew Zdanowicz
27d7d72033 feature: add pipeline state management for append command list operation
- command list append state is managed from internal queue and can be skipped
- initial state configuration should be processed by both kernel and non-kernel
- only kernel operation can process required state, as non-kernel cannot change

Related-To: NEO-10356

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2025-03-04 20:43:22 +01:00
Szymon Morek
ff4da67979 fix: signal notify field before KMD wait
Related-To: NEO-13870

Currently all monitor fences are triggering
interrupt due to Notify Enable field.
With this change, such field is programmed
right before KMD wait.

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2025-03-04 15:19:16 +01:00
Zbigniew Zdanowicz
db99c25c79 feature: add support to dispatch epilogue commands into dedicated stream
Related-To: NEO-10356

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2025-03-03 14:39:34 +01:00
Zbigniew Zdanowicz
08b13750a1 fix: set stall cmd flag for bcs flush task count flag
Related-To: NEO-10356

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2025-02-28 12:24:51 +01:00
Zbigniew Zdanowicz
cae3bb1d0a feature: add internal interfaces to manage all dispatch models of command lists
- add new enum type for command list flush from immediate
- add new argument for flushing immediate command list - regular command list
- add capability to provide additional stream for epilogue commands
- add pointer to provide external csr mutex to lock both execution and flush

Related-To: NEO-10356

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2025-02-28 12:02:22 +01:00