Commit Graph

4375 Commits

Author SHA1 Message Date
Fabian Zwoliński
c832a37bac feature: use global bindless allocator in L0 on DG2
Related-To: NEO-7063, NEO-11533
Signed-off-by: Fabian Zwoliński <fabian.zwolinski@intel.com>
2024-08-09 15:59:52 +02:00
Bartosz Dunajski
d76ac1d1de fix: scratch controller residency
Related-To: HSD-18039519400

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-08-09 14:47:56 +02:00
Kamil Kopryk
8fe9751449 refactor: remove duplicated code for genCmdCast functions
Related-To: NEO-10641
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-08-09 14:23:47 +02:00
Compute-Runtime-Validation
a860adc457 Revert "fix: unblock xekmd recoverable pagefaults vmbind"
This reverts commit cc3e6d5055.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-08-09 14:17:41 +02:00
Szymon Morek
7ebf2e1994 performance: unlock shared mutex during evict
Related-To: NEO-11755

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-08-09 10:35:12 +02:00
Naklicki, Mateusz
cc3e6d5055 fix: unblock xekmd recoverable pagefaults vmbind
Related-To: HSD-13011898606
Signed-off-by: Naklicki, Mateusz <mateusz.naklicki@intel.com>
2024-08-09 09:10:27 +02:00
Kamil Kopryk
4048cbc0c2 test: refactor hw parse cmds
Related-To: NEO-10641
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-08-09 08:56:16 +02:00
Bartosz Dunajski
c42d0fbe49 refactor: improve xe mock engine testing
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-08-08 12:37:31 +02:00
Maciej Bielski
e434c5595f feature: get HW IP version from drm_xe_gt
Make use of `drm_xe_gt` members to determine HW IP version when running
with XeKMD. Fall back to current values when the respective ioctl call
does not provide relevant information.

Related-To: NEO-10773

Signed-off-by: Maciej Bielski <maciej.bielski@intel.com>
2024-08-08 12:06:54 +02:00
Mateusz Hoppe
4c172a406d fix: correct maxOsContextCount with multiple subdevices and context group
Related-To: NEO-12257

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2024-08-08 11:04:36 +02:00
Bartosz Dunajski
bca8478bd4 feature: use iotclhelper to get num media engines
Related-To: NEO-9777

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-08-08 09:36:28 +02:00
Kamil Kopryk
8210997f17 refactor: move csrBaseAddress cmdCast to separate file
Related-To: NEO-10641
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-08-08 08:32:53 +02:00
Filip Hazubski
22a382b7a2 test: Minor test fixes related to alignment
Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2024-08-08 08:26:12 +02:00
Szymon Morek
556a116987 fix: make paging fence address volatile
Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-08-07 15:09:34 +02:00
Dominik Dabek
ad229377b9 fix: disable indirect detection if any stack calls
Don't know if kernels will be initialized in the order needed to check
for indirect accesses in stack calls.

Remove now unused functionPointerWithIndirectAccessExists and reading
this value from zebin.

Related-To: NEO-12235

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2024-08-07 14:48:58 +02:00
Szymon Morek
d7d6996464 performance: initialize timeout params once
Currently this is done per each enqueue
which is not really needed

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-08-07 14:35:12 +02:00
Bartosz Dunajski
7fc7fbfc47 refactor: improve ulls prints
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-08-07 14:18:04 +02:00
Szymon Morek
0d6c506c0b performance: enable wait on paging fence on semaphore
Related-To: NEO-12197

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-08-07 10:20:03 +02:00
Szymon Morek
d4c1631ac7 performance: don't wait for paging fence on user thread
Related-To: NEO-12197

Currently for new resources user thread must wait before submitting
actual workload. With this commit, instead of waiting on user thread,
request is sent to background ULLS controller thread and additional
semaphore is programmed. ULLS controller will perform actual wait
and signal semaphore when paging fence reaches required value.

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-08-07 08:30:51 +02:00
Bartosz Dunajski
24cfd203ab fix: dont download tbx allocations on heapless first device submission
Related-To: HSD-18039476929

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-08-06 14:03:42 +02:00
Michal Mrozek
d52ca080bd Revert "performance: improve pool handling"
This reverts commit a3c3b6533a.

Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
2024-08-06 13:04:02 +02:00
shubham kumar
0002eb3fcc feature: adding eu stall support on windows
Related-To: NEO-12174


Signed-off-by: shubham kumar <shubham.kumar@intel.com>
2024-08-06 06:47:11 +02:00
Kamil Kopryk
2a9bcdeb83 refactor: pass outImplicitArgs to patchImplicitArgs function
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-08-05 17:31:47 +02:00
Fabian Zwoliński
674c4a15ad fix: use correct gpu address when bindless heaps helper is enabled
Related-To: NEO-7063
Signed-off-by: Fabian Zwoliński <fabian.zwolinski@intel.com>
2024-08-05 15:09:57 +02:00
Dominik Dabek
619b47e3d5 fix: disable usm host alloc recycle pre xe-hpg
Related-To: NEO-12182, NEO-12207

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2024-08-05 14:50:02 +02:00
Oskar Hubert Weber
3ccab79ed8 test: dir create
Related-To: NEO-11500

Signed-off-by: Oskar Hubert Weber <oskar.hubert.weber@intel.com>
2024-08-05 13:58:25 +02:00
Filip Hazubski
db50903051 refactor: correct typo
Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2024-08-04 21:12:09 +02:00
Dominik Dabek
26428d5af3 feature: track used memory by allocations
Track memory used by memory allocations. System and local per device.
Will be used for heuristics in memory pooling.

Related-To: NEO-11356

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2024-08-02 17:54:34 +02:00
Mateusz Jablonski
5a339aeb2c test: define mock drm xe in cpp file
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-08-02 14:36:16 +02:00
Compute-Runtime-Validation
e27efd701f Revert "fix: correct calculating max subslice space"
This reverts commit 67f2500c03.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-08-02 12:28:13 +02:00
Bartosz Dunajski
ec34656e0e fix: debug flag to defer first device submission
Related-To: HSD-18039343751

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-08-02 09:31:25 +02:00
Mateusz Jablonski
67f2500c03 fix: correct calculating max subslice space
computeMaxNeededSubSliceSpace is no longer needed as getHighestEnabledSubSlice
already determines maximum index from all enabled subslices

Related-To: NEO-12073
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-08-01 16:38:24 +02:00
Kamil Kopryk
27ba5f5089 refactor: move duplicated code to xeHpc and later
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-08-01 16:21:25 +02:00
Mateusz Jablonski
07a11e640f refactor: include xe header in dedicated namespace
avoid redefinition issue in case of integrating multiple xe drm versions

define mock drm xe in inl file to provide xe definitions for mock members

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-08-01 14:38:20 +02:00
Kamil Kopryk
ee2a688323 test: add flushTaskStateless mocks
Related-To: NEO-10641
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-08-01 14:37:34 +02:00
Zbigniew Zdanowicz
7ca0210ef2 fix: patch compute walker scratch address when scratch controller is changed
- scenario when command list is changed from normal to low priority queue
- save only scratch patch address, when changed, then enable patching
- kernels will re-use scratch patch address after update of compute walker

Related-To: NEO-11972

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-08-01 13:28:53 +02:00
Jemale Lockett
dd36ac8401 fix: Make debug area resident for subdevices
Resolves: NEO-9455

Signed-off-by: Jemale Lockett <jemale.lockett@intel.com>
2024-08-01 11:08:30 +02:00
Szymon Morek
375d65d8e5 refactor: remove unused code
Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-08-01 09:55:46 +02:00
Compute-Runtime-Validation
2d1b263e9a Revert "refactor: remove redundant function computeMaxNeededSubSliceSpace"
This reverts commit c0b96dcd6e.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-08-01 03:54:56 +02:00
Mateusz Jablonski
8edc40adbc fix: populate SliceInfo based on TopologyMap in drm path
pick minimal config in case of multi tile

Related-To: NEO-12073
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-07-31 15:55:21 +02:00
Mateusz Jablonski
c0b96dcd6e refactor: remove redundant function computeMaxNeededSubSliceSpace
use GfxCoreHelper::getHighestEnabledDualSubSlice instead

Related-To: NEO-12073
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-07-31 14:50:13 +02:00
Zbigniew Zdanowicz
b33fe6ccf1 feature: adding flag to block dispatch implicit scaling commands
- this feature is part of making compute walker command view
- compute walker is programed for implicit scaling but not dispatched
- together with new flag, comes the refactor to reduce number of arguments

Related-To: NEO-11972

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-07-31 14:24:27 +02:00
Kamil Kopryk
1971a87da2 test: disable heapless or heapless state init in inappropriate ults 2/n
Related-To: NEO-10641
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-07-31 12:45:39 +02:00
Mateusz Jablonski
aa32f7420b fix: correct getting max enabled dual subslice
when no dss info provided, assume max dss per slice are enabled

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-07-31 11:15:23 +02:00
Mateusz Jablonski
838894e074 fix: correct getting max enabled subslice
when no subslice info provided, assume max subslices per slice are enabled

Related-To: NEO-12073

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-07-31 10:42:57 +02:00
Maciej Bielski
790bb84841 refactor: decouple ClosCacheReservation from Drm
The dependency towards `Drm` is unnecessary and only makes testing more
difficult. Instead, dependency towards `IoctlHelper` alone only is
sufficient.

Related-To: NEO-10158
Signed-off-by: Maciej Bielski <maciej.bielski@intel.com>
2024-07-31 10:38:35 +02:00
Jack Myers
f5d00b2616 feature: 2d-block-load-transpose query
Implemented device property query API for determining
support capabilities regarding 2d-block-load-tranpose
features for which not all Intel devices support.

Related-To: NEO-11592
Signed-off-by: Jack Myers <jack.myers@intel.com>
2024-07-30 18:21:07 +02:00
Kamil Kopryk
c09d5b383f test: disable heapless or heapless state init in inappropriate ults
Related-To: NEO-10641
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-07-30 17:36:52 +02:00
Bartosz Dunajski
8e8b2582d8 fix: wait for first submission in TBX mode
Related-To: HSD-18039343751

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-07-30 16:28:17 +02:00
Kamil Kopryk
2735b06206 test: add getCommandWalkerCount method
Related-To: NEO-10641
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2024-07-30 16:21:12 +02:00