intel / intel/pti-gpu

[pti-1.0.1] Unittests fail on Ubuntu 26.10 with dpclang-7.0.1

Open
#118 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
C++
Stars
273
Forks
70
PR merge metrics
No merged PRs in 30d

Description

On https://github.com/intel/pti-gpu/tree/pti-1.0.1, Ubuntu 26.10, Intel PVC

```
ARG IMAGE=ubuntu:26.10
FROM $IMAGE

ENV DEBIAN_FRONTEND=noninteractive

RUN apt-get update && \
apt-get install -y --no-install-recommends \
ca-certificates \
cmake \
curl \
git \
gcc \
g++ \
gnupg \
make \
ninja-build \
pkg-config \
python3 \
python3-dev \
python3-venv \
python3-pip \
software-properties-common \
tree

RUN apt-get update && \
apt-get install --no-install-recommends -y \
dpclang-7 \
libxptifw-dev \
libze-dev \
ocl-icd-opencl-dev

RUN curl -LsSf https://astral.sh/uv/install.sh | sh

ENV PATH=/root/.local/bin:$PATH
```

The following tests are failing:
```
405 - VecsqaddMetricsFixtureTest.ConfigureGoodMetricGroup (Failed) functional hw-metrics metrics
406 - VecsqaddMetricsFixtureTest.ConfigureAlreadyConfiguredMetricGroup (Failed) functional hw-metrics metrics
407 - VecsqaddMetricsFixtureTest.StartCollectionAndStop (Failed) functional hw-metrics metrics
410 - VecsqaddMetricsFixtureTest.StartAlreadyStartedCollection (Failed) functional hw-metrics metrics
411 - VecsqaddMetricsFixtureTest.StartPausedAlreadyStartedCollection (Failed) functional hw-metrics metrics
412 - VecsqaddMetricsFixtureTest.StartCollectionPausedAndStop (Failed) functional hw-metrics metrics
415 - VecsqaddMetricsFixtureTest.StartAlreadySartedInPauseModeCollection (Failed) functional hw-metrics metrics
416 - VecsqaddMetricsFixtureTest.StartInPausedModeAlreadySartedInPausedModeCollection (Failed) functional hw-metrics metrics
417 - VecsqaddMetricsFixtureTest.ResumeAlreadyRunningCollection (Failed) functional hw-metrics metrics
418 - VecsqaddMetricsFixtureTest.ResumeNotStartedCollection (Failed) functional hw-metrics metrics
421 - VecsqaddMetricsFixtureTest.ResumePausedCollection (Failed) functional hw-metrics metrics
422 - VecsqaddMetricsFixtureTest.StopNotStartedCollection (Failed) functional hw-metrics metrics
425 - VecsqaddMetricsFixtureTest.StopAlreadyStoppedCollection (Failed) functional hw-metrics metrics
426 - VecsqaddMetricsFixtureTest.StopPausedCollection (Failed) functional hw-metrics metrics
427 - VecsqaddMetricsFixtureTest.CalculateNotStartedCollection (Failed) functional hw-metrics metrics
433 - VecsqaddMetricsFixtureTest.CalculateEnabledCollection (Failed) functional hw-metrics metrics
434 - VecsqaddMetricsFixtureTest.CalculatePausedCollection (Failed) functional hw-metrics metrics
435 - VecsqaddMetricsFixtureTest.CalculateEmptyCollection (Failed) functional hw-metrics metrics
436 - VecsqaddMetricsFixtureTest.StartCollectionPausedAndResumeAndCalculateBeforeStop (Failed) functional hw-metrics metrics
437 - VecsqaddMetricsFixtureTest.StartCollectionPausedAndResumeAndStopAndCalculate (Failed) functional hw-metrics metrics
440 - MetricsMultiThreadingTest.ConcurrentCollectionDifferentDevices (Failed) functional hw-metrics-multi-thread metrics
441 - MetricsMultiThreadingTest.ConcurrentCollectionManagement (Failed) functional hw-metrics-multi-thread metrics
471 - GemmMetricsScopeFixtureTest.ScopeReconfigureMetricCount (Failed) functional hw-metrics-scope metrics
482 - GemmMetricsScopeFixtureTest.StartStopCollection (Failed) functional hw-metrics-scope metrics
483 - GemmMetricsScopeFixtureTest.CompleteWorkflowWithGemm (Failed) functional hw-metrics-scope metrics
484 - GemmMetricsScopeFixtureTest.AutoDetectStartStop (Failed) functional hw-metrics-scope metrics
485 - GemmMetricsScopeFixtureTest.AutoDetectMultiThreadedAllDevices (Failed) functional hw-metrics-scope metrics
486 - GemmMetricsScopeMultiDeviceFixtureTest.MultiDeviceMultiThreadedConcurrentSubmission (Failed) functional hw-metrics-scope metrics
487 - GemmMetricsScopeMultiDeviceFixtureTest.MultiDeviceKernelOnlyOnOneDevice (Failed) functional hw-metrics-scope metrics
488 - GemmMetricsScopeMultiDeviceFixtureTest.MultiDeviceBufferRotationBothDevices (Failed) functional hw-metrics-scope metrics
489 - GemmMetricsScopeMultiDeviceFixtureTest.MultiDeviceReconfigureAfterStop (Failed) functional hw-metrics-scope metrics
492 - Configuration/GemmMetricsScopeConfigureTest.ScopeConfigureWithVariousMetrics/Metrics_from_different_groups_PVC (Failed) functional hw-metrics-scope metrics
494 - Configuration/GemmMetricsScopeConfigureTest.ScopeConfigureWithVariousMetrics/Valid_metrics_from_ComputeBasic_group (Failed) functional hw-metrics-scope metrics
495 - Configuration/GemmMetricsScopeConfigureTest.ScopeConfigureWithVariousMetrics/Single_valid_metric (Failed) functional hw-metrics-scope metrics
496 - Configuration/GemmMetricsScopeConfigureTest.ScopeConfigureWithVariousMetrics/Valid_metrics_from_TestOa_group_PVC (Failed) functional hw-metrics-scope metrics
733 - metrics_scope_multiprocess (Failed) functional hw-metrics-scope metrics multiprocess
734 - perf-profiling-overhead (Failed) performance
735 - perf-prof-gpu-overhead (Failed) performance
736 - perf-link-overhead (Failed) performance
742 - run-onemkl_gemm_exe (Not Run) functional samples
744 - run-iso (Not Run) functional samples
745 - run-metrics-iso (Not Run) functional hw-metrics metrics samples
747 - run-metrics-scope (Failed) functional hw-metrics-scope metrics samples
748 - run-omp_vec_add (Not Run) functional samples
751 - ZeGraphTestSuite.TestZeUsmGraphExecutionTracingGraphCreation (Subprocess aborted) functional graph level-zero
752 - ZeGraphTestSuite.TestZeUsmGraphExecutionWithoutTracingGraphCreation (Subprocess aborted) functional graph level-zero
```

Above is a sighting and requires a triage. If fixes will be available in later versions, that's all fine, but we need https://github.com/intel/pti-gpu/issues/117 address to check the later code.

CC: @mschilling0

Contributor guide

Open the contributing guide

Research direction

Reproduce the Ubuntu 26.10 container setup on the pti-1.0.1 branch, then inspect the failing VecsqaddMetricsFixtureTest, MetricsMultiThreadingTest, GemmMetricsScopeFixtureTest, performance, sample, and ZeGraphTestSuite cases. Compare with issue #117 and later code; done means identifying whether the failures are version-specific, recording the cause, and confirming any available later-version fix.

Written by the indexing model from the issue text.

Assessment

Tech stack
cmake, cpp, ubuntu
Domain
performance, testing-qa
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Active
Clarity
Mostly clear
Newbie friendliness
45/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.