Commit Graph

768 Commits

Author SHA1 Message Date
Bellekallu Rajkiran
24f73f4686 fix(Sysman): Support for fabric port health change event.
Related-To: LOCI-4053

Signed-off-by: Bellekallu Rajkiran <bellekallu.rajkiran@intel.com>
2023-04-12 06:46:19 +02:00
Mateusz Hoppe
e26ebfc51b fix: check exception reason for stopped threads
- before marking interrupt request check exception reason. If there is
exception other than forced exception or forced external halt treat
thread as stopped and generate distinct event for it.

Related-To: NEO-7869

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2023-04-05 12:01:20 +02:00
Singh, Prasoon
71fe65b327 [Sysman] Replace normal pointers with smart pointers (10/n)
Replacing normal pointers by smart pointers in fan module of L0 sysman.

Related-To: LOCI-2810

Signed-off-by: Singh, Prasoon <prasoon.singh@intel.com>
2023-04-05 11:05:44 +02:00
Bellekallu Rajkiran
310380119a feat(sysman): Support for registering events during listen
Related-To: LOCI-3682

Signed-off-by: Bellekallu Rajkiran <bellekallu.rajkiran@intel.com>
2023-04-04 07:07:19 +02:00
Singh, Prasoon
42a422d9f4 [Sysman] Replace normal pointers with smart pointers (8/n)
Replacing normal pointers by smart pointers in fabric_port module of L0 sysman.

Related-To: LOCI-2810

Signed-off-by: Singh, Prasoon <prasoon.singh@intel.com>
2023-04-03 09:34:52 +02:00
Mayank Raghuwanshi
d0fd7c1667 Improve Sysman windows ult coverage
Related-To: LOCI-4141

Signed-off-by: Mayank Raghuwanshi <mayank.raghuwanshi@intel.com>
2023-03-31 15:00:15 +02:00
Mayank Raghuwanshi
3816b85fa0 Add check for memory type before calculating ras hbm errors
Related-To: LOCI-3500

Signed-off-by: Mayank Raghuwanshi <mayank.raghuwanshi@intel.com>
2023-03-31 13:47:41 +02:00
Mateusz Hoppe
bfb59cc573 fix: do not resume threads that have pending THREAD_STOPPED events
- threads with pending events cannot be resumed by resume(all) as they
they were not yet discovered by debugger as stopped.
- set reportAsStopped when THREAD_STOPPED event is read
- when resuming, only resume a thread when it was reported

Related-To: NEO-7817

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2023-03-31 12:42:25 +02:00
Singh, Prasoon
1a6f03c36e [Sysman] Replace normal pointers with smart pointers (4/n).
Replacing normal pointers by smart pointers in temperature module of L0 sysman.

Related-To: LOCI-2810

Signed-off-by: Singh, Prasoon <prasoon.singh@intel.com>
2023-03-30 13:07:02 +02:00
Mayank Raghuwanshi
065232eac7 Add support for ras l3 fabric errors
Related-To: LOCI-3966

Signed-off-by: Mayank Raghuwanshi <mayank.raghuwanshi@intel.com>
2023-03-30 12:47:45 +02:00
Singh, Prasoon
ac929eaf61 (Sysman):Changes made in standby API.
Replacing normal pointers by smart pointers in standby module of L0 sysman.

Related-To: LOCI-2810

Signed-off-by: Singh, Prasoon <prasoon.singh@intel.com>
2023-03-30 06:45:00 +02:00
Singh, Prasoon
49827b7122 (Sysman):Changes made in engine API.
Replacing normal pointers by smart pointers in engine module of L0 sysman.

Related-To: LOCI-2810

Signed-off-by: Singh, Prasoon <prasoon.singh@intel.com>
2023-03-30 06:41:26 +02:00
Bari, Pratik
b161af5886 Improve coverage in Fs Access
- Added ULTS to improve the code coverage in fs_access.
- Used the shared functions in the ULTs.
- Added the function for pwrite in the shared code.
Related-To: LOCI-2117

Signed-off-by: Bari, Pratik <pratik.bari@intel.com>
2023-03-30 04:21:18 +02:00
Singh, Prasoon
fa2bb2509b [Sysman] Replace normal pointers with smart pointers (6/n)
Replacing normal pointers by smart pointers in scheduler module of L0 sysman.

Related-To: LOCI-2810

Signed-off-by: Singh, Prasoon <prasoon.singh@intel.com>
2023-03-29 16:01:08 +02:00
Singh, Prasoon
4475e2228d [Sysman] Replace normal pointers with smart pointers (7/n)
Replacing normal pointers by smart pointers in memory  module of L0 sysman.

Related-To: LOCI-2810

Signed-off-by: Singh, Prasoon <prasoon.singh@intel.com>
2023-03-29 10:59:40 +02:00
Yates, Brandon
0f6140deb0 fix(l0debug): windows topology map should always be created
Topology map was only being created when ZET_ENABLE_PROGAM_DEBUGGING was
set.  This was not correct.  Now it is unconditionally created at init,
and debug attach will fail if it is not valid.

Related-to: LOCI-3937

Signed-off-by: Yates, Brandon <brandon.yates@intel.com>
2023-03-27 16:24:57 +02:00
Bari, Pratik
fb82af429b Improve ULT function coverage in Fan Component
As the fan component is not supported on Linux, hence for the ULTs a
fan handle is created and used which improves the coverage for the Fan
component.
Related-To: LOCI-2113

Signed-off-by: Bari, Pratik <pratik.bari@intel.com>
2023-03-27 10:06:32 +02:00
Bari, Pratik
e813ae0f4a Added changes for maxBandwidth computation in Memory Component.
The memoryBusWidth used for the computation of maxBandwidth has been
changed to 128 bytes from 128 bits. Hence the final maxBandwidth need
not be further divided by 8 to convert it into bytes/sec.

Related-To: LOCI-3833

Signed-off-by: Bari, Pratik <pratik.bari@intel.com>
2023-03-23 20:20:59 +01:00
Kulkarni, Ashwin Kumar
9d866dac9e Support maxBandwidth in zesDevicePciGetState for windows
Fix to update zes_pci_speed_t.maxBandwidth

Related-To: LOCI-4056

Signed-off-by: Kulkarni, Ashwin Kumar <ashwin.kumar.kulkarni@intel.com>
2023-03-20 06:12:44 +01:00
Philip Meyer
2606788250 Sysman - Modify Consumption of KMD Energy Counter Timestamp
Signed-off-by: Philip Meyer philip.meyer@intel.com

Resolves: HSD-14018885283
2023-03-20 06:05:24 +01:00
Mateusz Jablonski
659cacf2c9 refactor l0 cmake: reduce include directories
Related-To: NEO-7507
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-03-17 13:41:55 +01:00
Mateusz Jablonski
a7830eb478 refactor l0 cmake: add CMakeLists.txt files to solution
Related-To: NEO-7507
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-03-17 12:09:03 +01:00
Mateusz Jablonski
cb7437b6b7 refactor l0 tools: cleanup cmake files
append sources in current directory

Related-To: NEO-7507
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-03-16 14:05:14 +01:00
Filip Hazubski
0025cb25f4 test: Correct logic of disabling slices in gtSystemInfo
Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2023-03-16 09:13:47 +01:00
Mateusz Hoppe
f2275caf3a performance: L0 debug - improve single-step time
- check threads with ATT only when more than one thread is resumed
- if single thread is resumed - check only that thread
- delay thread state check - allowing thread to reenter SIP

Related-To: NEO-7750

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2023-03-15 09:48:35 +01:00
Bari, Pratik
adbda5906a Added support for Power APIs
- Added support for the Power APIs in the new sysman design.
- Added ULTs for the Power APIs in the new sysman design.
- Added change to measure the average power in the black box test for
Power.

Related-To: LOCI-3891

Signed-off-by: Bari, Pratik <pratik.bari@intel.com>
2023-03-15 08:23:45 +01:00
Joshua Santosh Ranjan
e2177f44f3 [Metrics]Fix OA buffer size request for streamer
This patch uses maximum OA buffer size for the
platform when opening an io stream.

Related-To: LOCI-4103

Signed-off-by: Joshua Santosh Ranjan <joshua.santosh.ranjan@intel.com>
2023-03-14 11:10:11 +01:00
Kamil Kopryk
fa8579602f refactor: rename product helper files n/n
Related-To: NEO-7703
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2023-03-10 13:24:38 +01:00
Mateusz Jablonski
fbe01e534b fix l0 debug: correct UNRECOVERABLE_IF condition
correct debug API tests to respect more than 8 threads per eu

Related-To: NEO-7442

Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-03-10 12:54:31 +01:00
Yates, Brandon
11bec57d10 fix(L0 debug): Fix singleThread conversion for some hw configs
Configs with low slices disabled produced wrong EuThread vector

Signed-off-by: Yates, Brandon <brandon.yates@intel.com>
2023-03-10 10:30:49 +01:00
Jitendra Sharma
266a495837 [Sysman]: Add support for Fabric APIs
Add support for Fabric APIs in new sysman design.
This design is independent of level zero core.

Related-To: LOCI-4091
Signed-off-by: Jitendra Sharma <jitendra.sharma@intel.com>
2023-03-10 07:31:21 +01:00
Bellekallu Rajkiran
f0c7eef667 Sysman: Fix and refactor event implementation
Remove unnecessary prelim implementation for events.
Modify uevent property name to check memory health
events.

Related-To: LOCI-4089

Signed-off-by: Bellekallu Rajkiran <bellekallu.rajkiran@intel.com>
2023-03-09 09:34:11 +01:00
Mateusz Hoppe
caecbb12f0 fix: Add ATT scan after interrupt in handling ATT event
- make sure to capture all threads after interrupt, also those that
could be lagging to enter SIP

Related-To: NEO-7776

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2023-03-07 11:53:25 +01:00
Zbigniew Zdanowicz
49def723b7 Unify layout of command list class
Related-To: NEO-5055

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2023-03-07 11:38:47 +01:00
Bellekallu Rajkiran
2282f26734 feature(sysman): Support events for multiple devices
Related-To: LOCI-3683

Signed-off-by: Bellekallu Rajkiran <bellekallu.rajkiran@intel.com>
2023-03-07 09:50:32 +01:00
Bari, Pratik
9acba4ea2a Added changes in the Temperature Component
- The 'paramInfo' for the sensor type ZES_TEMP_SENSORS_MEMORY has been
corrected from TemperatureDomainPackage to TemperatureDomainHBM.
- The ULTs for temperature have been enabled for windows.

Related-To: LOCI-3149

Signed-off-by: Bari, Pratik <pratik.bari@intel.com>
2023-03-07 09:47:34 +01:00
Mateusz Jablonski
5b2e1f619c fix hw config linux: set slice info based on slice count
Related-To: NEO-7622
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-03-06 15:47:40 +01:00
Mateusz Hoppe
802848a43f fix: L0Debug - allow access only for reported stopped threads
- read/write registers/memory only allowed for threads reported as
stopped by events
- threads newly stopped, accidentally, that are resumed immediately
are not allowed register/memory access

Related-To: NEO-7776

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2023-03-03 13:34:57 +01:00
Bari, Pratik
5d04c15297 Added alternate way to initialize sysman
A new environment variable ZELLO_SYSMAN_USE_ZESINIT is used to
distinguish the sysman initialization with zeInit and zesInit.
For sysman initialization using zeInit, ZES_ENABLE_SYSMAN is set
if it is not set initially.

Related-To: LOCI-3949

Signed-off-by: Bari, Pratik <pratik.bari@intel.com>
2023-03-02 09:08:29 +01:00
Mateusz Jablonski
5af9509877 Correct scope of WA for pipecontrol before nonpipelined state
- for DG2 platforms it is valid only for G10/G11/G12
- for MTL platforms it is valid only for 12.70.0 and 12.71.0

Additionally:
- setup default hw ip version for each platform
- merge dg2 specific product helper tests to single file

Related-To: HSD-14015808183, HSD-14015812625, HSD-14016015202
Related-To: HSD-14015812559, HSD-14015816823
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2023-02-24 09:27:12 +01:00
Mateusz Hoppe
6947220298 perf(debugger): Optimize registers access during debugging
Register read/write implementation was unnecessarily checking magic values
which were taking significant time. By removing those checks, register
access is speed-up by a factor of 3.

Resolves: NEO-7754

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2023-02-23 12:14:03 +01:00
Mateusz Hoppe
e8b4841577 perf(debugger): Optimize single-stepping of eu threads
Single-stepping threads is based on resume() and THREAD_STOPPED events
To speed up single-stepping, stopped threads must be discovered
immediately after resume() is called.
This change is adding ATT scan and event generation in resume()

Related-To: NEO-7750

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
2023-02-23 11:06:13 +01:00
Mayank Raghuwanshi
29ed6ea077 Add support l3_bank and subslice Ras errors
Related-To: LOCI-3662

Signed-off-by: Mayank Raghuwanshi <mayank.raghuwanshi@intel.com>
2023-02-22 09:14:15 +01:00
Bellekallu Rajkiran
ac5d719e25 Fix (sysman): Modify device wedged status retrieval mechanism
Related-To: LOCI-3551

Signed-off-by: Bellekallu Rajkiran <bellekallu.rajkiran@intel.com>
2023-02-20 23:21:24 +01:00
Mayank Raghuwanshi
eea0ff8e16 Fix clang tidy issues in fabric ults
Related-To: LOCI-4054

Signed-off-by: Mayank Raghuwanshi <mayank.raghuwanshi@intel.com>
2023-02-20 11:59:12 +01:00
Ranjan, Joshua Santhosh
8ae2e3dac9 Add per tile metric data support for stall sampling
Related-To: LOCI-3869

Signed-off-by: Ranjan, Joshua Santhosh <joshua.santosh.ranjan@intel.com>
2023-02-19 12:16:13 +01:00
Warchulski, Jaroslaw
b485c025d0 Cleanup includes 57
Related-To: NEO-5548
Signed-off-by: Warchulski, Jaroslaw <jaroslaw.warchulski@intel.com>
2023-02-17 11:19:59 +01:00
Warchulski, Jaroslaw
0556d543a3 Cleanup includes 56
Related-To: NEO-5548
Signed-off-by: Warchulski, Jaroslaw <jaroslaw.warchulski@intel.com>
2023-02-16 14:42:44 +01:00
Warchulski, Jaroslaw
89ddfbb83c Cleanup includes 54
Cleaned up files:
level_zero/tools/test/unit_tests/sources/sysman/windows/mock_sysman_fixture.h
shared/test/common/os_interface/windows/wddm_fixture.h
shared/test/unit_test/command_stream/compute_mode_tests.h
shared/test/unit_test/encoders/walker_partition_fixture_xehp_and_later.h
shared/test/unit_test/fixtures/preemption_fixture.h
shared/test/unit_test/helpers/blit_commands_helper_tests.inl
shared/test/unit_test/helpers/simd_helper_tests.inl

Related-To: NEO-5548
Signed-off-by: Warchulski, Jaroslaw <jaroslaw.warchulski@intel.com>
2023-02-15 13:59:56 +01:00
Warchulski, Jaroslaw
8c17313c8b Cleanup includes 53
Cleaned up files:
opencl/source/mem_obj/image.inl
shared/offline_compiler/source/decoder/zebin_manipulator.h
shared/source/aub_mem_dump/aub_alloc_dump.h
shared/source/compiler_interface/intermediate_representations.h
shared/source/helpers/blit_commands_helper_base.inl
shared/source/utilities/debug_file_reader.h
shared/source/utilities/software_tags.h
shared/source/xe_hpc_core/hw_cmds_pvc.h

Related-To: NEO-5548

Signed-off-by: Warchulski, Jaroslaw <jaroslaw.warchulski@intel.com>
2023-02-15 13:21:14 +01:00