Commit Graph

2946 Commits

Author SHA1 Message Date
Compute-Runtime-Validation
37eeef5ba4 Revert "fix: Remove the check for ze-take-global-address and update errors"
This reverts commit 156e219e7f.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-12-14 04:50:41 +01:00
Wenbin Lu
17e537b694 feature: enable event synchronization GPU status check by default
Related-To: GSD-10187

Signed-off-by: Wenbin Lu <wenbin.lu@intel.com>
2024-12-13 20:36:15 +01:00
Mateusz Jablonski
8f7bacdd95 feature: add eudebug interface class
eudebug interface is now hidden under EuDebugInterface class
shared code uses generic object and param values

layout of structs is guarded by static asserts

eudebug support is guarded by cmake flags:
- NEO_ENABLE_XE_EU_DEBUG_SUPPORT - enables eudebug in general
- NEO_USE_XE_EU_DEBUG_EXP_UPSTREAM - registers exp upstream uAPI support
- NEO_ENABLE_XE_PRELIM_DETECTION - registers prelim uAPI support

This way we can support two different xe-eudebug interfaces within
single binary.

In unit tests there is mock eudebug interface enabled (even if no
eudebug support is enabled by cmake flag).

Related-To: NEO-13472
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-12-13 17:07:12 +01:00
Neil R. Spruit
156e219e7f fix: Remove the check for ze-take-global-address and update errors
Related-to: NEO-13458

- With new IGC, ze-take-global-address is not needed to ensure global
pointers are allowed. Updates the error message to provide the
correct information on why the query failed and remove the requirement
check for the flag.

Signed-off-by: Neil R. Spruit <neil.r.spruit@intel.com>
2024-12-13 15:52:36 +01:00
Maciej Plewka
8151224501 fix: add microsecond resolution for timeout
Related-To: NEO-13445
Signed-off-by: Maciej Plewka <maciej.plewka@intel.com>
2024-12-13 09:32:18 +01:00
Wenbin Lu
c19df80bd8 feature: add key to force GPU status check in event synchronization
Related-To: GSD-10187

Signed-off-by: Wenbin Lu <wenbin.lu@intel.com>
2024-12-12 22:24:36 +01:00
Bartosz Dunajski
084ea8b687 fix: unset InOrderExecInfo after appendEventReset
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-12-12 16:26:53 +01:00
Bartosz Dunajski
071712b835 feature: enable new relaxed ordering heuristic
Related-To: NEO-13431

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-12-12 14:49:04 +01:00
Compute-Runtime-Validation
6c5d9a6ed7 Revert "feature: extend TBX page fault manager from CPU implementation"
This reverts commit 51c0e80299.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-12-12 12:30:22 +01:00
Bartosz Dunajski
5f6b2de3dc feature: clear standalone timestamp before submission
Related-To: HSD-18040896547

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-12-12 10:55:10 +01:00
Chandio, Bibrak Qamar
8cf4804fcd fix: Overhead in zeDeviceGetGlobalTimestamps
Related-To: NEO-11908

There is overhead when submission method is used for
zeDeviceGetGlobalTimestamps. This fixes it.

Signed-off-by: Chandio, Bibrak Qamar <bibrak.qamar.chandio@intel.com>
2024-12-12 08:54:19 +01:00
Compute-Runtime-Validation
b060cb6c8c Revert "fix: ensure regular event completion in in-order mode"
This reverts commit 27c36c4998.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-12-12 03:18:13 +01:00
Bartosz Dunajski
27c36c4998 fix: ensure regular event completion in in-order mode
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-12-11 17:46:34 +01:00
Lukasz Jobczyk
f2725f217e refactor: Introduce debug flags to manipulate event's signal visibility
-Add AbortHostSyncOnNonHostVisibleEvent which abort when waiting for non
host visible event from host
-Add ForceHostSignalScope which forces add or clear of host scope to
event's signal scope

Related-To: NEO-13441

Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2024-12-11 14:07:43 +01:00
Bartosz Dunajski
eca3d5a677 feature: debug flag to clear timestamp before submission
Related-To: HSD-18040896547

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-12-11 13:56:49 +01:00
Jack Myers
51c0e80299 feature: extend TBX page fault manager from CPU implementation
In TBX mode, the host could not write to host buffers after access from device
code due to the lack of a migration mechanism post-initial TBX upload.
Migration is unnecessary with real hardware, but required for TBX.

This patch introduces a new page fault manager type that extends the original
CPU fault manager, enabling automatic migration of host buffers in TBX mode.

Refactoring was necessary to avoid diamond inheritance, achieved by using a
template parameter as the base class for OS-specific fault managers.

Related-To: NEO-12268
Signed-off-by: Jack Myers <jack.myers@intel.com>
2024-12-11 09:09:50 +01:00
Maciej Bielski
1fafd44af5 refactor: use level-specific name for CacheInfo instances
Related-To: NEO-12837

Signed-off-by: Maciej Bielski <maciej.bielski@intel.com>
2024-12-10 11:45:11 +01:00
Dunajski, Bartosz
37e81d2a11 feature: new heuristic to enable relaxed ordering 2
Related-To: NEO-13431

Signed-off-by: Dunajski, Bartosz <bartosz.dunajski@intel.com>
2024-12-09 11:58:42 +01:00
Michal Mrozek
080488e243 performance: add prefetch capabilities to L0.
Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
2024-12-09 11:11:29 +01:00
Compute-Runtime-Validation
af8ad3aa7a Revert "feature: new heuristic to enable relaxed ordering"
This reverts commit 526f9c5e81.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-12-08 16:01:32 +01:00
Bartosz Dunajski
526f9c5e81 feature: new heuristic to enable relaxed ordering
Related-To: GSD-10308

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-12-06 17:04:39 +01:00
Zbigniew Zdanowicz
75139d2322 fix: zero sync and region barrier buffers before use on regular command lists
Related-To: NEO-13350

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-12-06 15:54:50 +01:00
Bartosz Dunajski
b9beb9becc fix: clear TS data before dispatching relaxed ordering split operation
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-12-05 17:32:14 +01:00
Bartosz Dunajski
e575bc52c2 fix: disable non-walker signaling skip for implicit non-timestamp CB Events
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-12-05 12:45:04 +01:00
Chandio, Bibrak Qamar
ab2e831a4a fix: zeDeviceGetGlobalTimestamp to use submisison
Related-To: GSD-10253, GSD-9467, GSD-9381, NEO-11908

When EnableGlobalTimestampViaSubmission is set then
zeDeviceGetGlobalTimestamp uses immediate cmd submission
method to get GPU time.

Signed-off-by: Chandio, Bibrak Qamar <bibrak.qamar.chandio@intel.com>
2024-12-04 19:10:07 +01:00
shubham kumar
69f81d5d57 fix: Check for root device flag in multi-device usecase
Related-To: NEO-12381

Signed-off-by: shubham kumar <shubham.kumar@intel.com>
2024-12-03 07:03:39 +01:00
Szymon Morek
e6d11eb04b performance: stop ULLS for BCS during migration
Related-To: NEO-13340

When regular copy CSR has enabled direct submission,
stop it before migration on internal CSR.

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-12-02 17:57:12 +01:00
Bartosz Dunajski
9629ab3cc3 fix: disable fence wait if not supported on given CSR type
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-12-02 15:32:23 +01:00
Zbigniew Zdanowicz
901f7c2cb7 feature: block sync buffer and region barrier allocations in command view mode
Related-To: NEO-13350

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-12-02 09:11:34 +01:00
Zbigniew Zdanowicz
56b15f17f7 refactor: unify further calculation to get max work group count
- move available device calculcation into common helper
- change interface to have code available where no descriptor is available
- expand unit test for implementation of new inteface

Related-To: NEO-13350

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-29 17:51:05 +01:00
Bartosz Dunajski
c6accb8d53 fix: select correct CSR for submission if copy offload is not allowed
Related-To: GSD-10194

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-11-29 11:41:53 +01:00
Zbigniew Zdanowicz
c5ed6bf73c refactor: split sync buffer and region allocation creation code
- split the allocation code from command list or kernel
- allow to call allocation code in all parts of the driver

Related-To: NEO-13350

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-29 11:41:39 +01:00
Bartosz Dunajski
5e1fa75676 refactor: adjust code to compile with c++20
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-11-29 10:27:29 +01:00
Vysochyn, Illia
afd22999cc refactor: Adjust RENDER_SURFACE_STATE structures naming
Performs minor renaming (mostly capitalization) in order to align with
specification.

Renames L1_CACHE_POLICY to L1_CACHE_CONTROL.

Related-To: NEO-13147

Signed-off-by: Vysochyn, Illia <illia.vysochyn@intel.com>
2024-11-29 09:43:11 +01:00
Compute-Runtime-Validation
bced7e4621 Revert "performance: stop ULLS for BCS during migration"
This reverts commit 81ba52aac4.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2024-11-29 04:42:26 +01:00
Mateusz Jablonski
fa58073095 refactor: remove not used usings/typedefs/variables
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2024-11-28 16:19:39 +01:00
Zbigniew Zdanowicz
92ebbe75c0 feature: add kernel getters for sync buffer and region barrier allocations
Related-To: NEO-13350

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-28 10:29:48 +01:00
Szymon Morek
81ba52aac4 performance: stop ULLS for BCS during migration
Related-To: NEO-13340

When regular copy CSR has enabled direct submission,
stop it before migration on internal CSR.

Signed-off-by: Szymon Morek <szymon.morek@intel.com>
2024-11-27 20:06:50 +01:00
Zbigniew Zdanowicz
6b7235cd6c refactor: change parameter names into more meaningful
- change additional size into local region size
- change walk order into dispatch walk order to distinguish for local id walk

Related-To: NEO-13350

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2024-11-27 16:15:02 +01:00
Katarzyna Cencelewska
194da4b7ef fix: do not duplicate calculations when printing ts enabled
Related-To: HSD-14023527252
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2024-11-21 15:36:52 +01:00
Katarzyna Cencelewska
4ad8c17db9 feature: add debug flags for timestamps
PrintCalculatedTimestamps - print ts in level zero paths
PrintTimestampPacketContents - add logging also to level zero paths
ForceUseOnlyGlobalTimestamps - force using a global ts

Related-To: HSD-14023527252
Signed-off-by: Katarzyna Cencelewska <katarzyna.cencelewska@intel.com>
2024-11-21 11:28:08 +01:00
Bartosz Dunajski
dab4166837 fix: add missing aub polls on sync points
Related-To: HSD-14023925176

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-11-21 09:17:54 +01:00
Aravind Gopalakrishnan
33f64edc31 feature: Add External Semaphore experimental stubs [1/N]
Related-To: NEO-11488

Signed-off-by: Raiyan Latif <raiyan.latif@intel.com>
Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>
2024-11-20 19:04:42 +01:00
Wenbin Lu
2ba80ce114 feature: support physical host memory
Related-To: NEO-11981

Signed-off-by: Wenbin Lu <wenbin.lu@intel.com>
2024-11-20 08:19:52 +01:00
Bartosz Dunajski
47665280f4 fix: check standalone CB event completion for profiling
Related-To: HSD-18040322890

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-11-19 16:03:38 +01:00
Joshua Santosh Ranjan
f589408848 feature: update metrics for v1.11
Related-To: NEO-12597

Signed-off-by: Joshua Santosh Ranjan <joshua.santosh.ranjan@intel.com>
2024-11-19 02:02:46 +01:00
Bartosz Dunajski
e363c871c7 fix: skip signaling remaining packets only for explicit CB Events
Related-To: HSD-18040605969

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-11-15 14:36:00 +01:00
Bartosz Dunajski
fea50b1b94 feature: adjust CB Event creation in BB tests
Related-To: NEO-11925

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-11-14 16:51:15 +01:00
Aravind Gopalakrishnan
6844dec1d1 feature: Append recorded commandlist into immediate (7/N)
fix relaxed ordering and stalling commands check

Related-To: NEO-10356

Signed-off-by: Aravind Gopalakrishnan <aravind.gopalakrishnan@intel.com>
2024-11-14 16:16:40 +01:00
Bartosz Dunajski
7bf22ed33e feature: counter based allocation peer sharing
Related-To: NEO-13079

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2024-11-13 15:01:32 +01:00