llvm icon indicating copy to clipboard operation
llvm copied to clipboard

[UR][CUDA] enqueue//enqueue-test/urEnqueueKernelLaunchIncrementMultiDeviceTest fails

Open yingcong-wu opened this issue 7 months ago • 1 comments

Describe the bug

CI job Unified Runtime Pre Commit / Adapters (CUDA, UR_CUDA, -u 1001 --privileged --cap-add SYS_ADM failed and still fails for multiple rerun.

  • https://github.com/intel/llvm/actions/runs/15703193260/job/44305438571?pr=19009

Failures:

FAIL: Unified Runtime Conformance :: enqueue//enqueue-test/24/31 (101 of 492)
******************** TEST 'Unified Runtime Conformance :: enqueue//enqueue-test/24/31' FAILED ********************
Script(shard):
--
GTEST_OUTPUT=json:/__w/llvm/llvm/build/test/conformance/enqueue/enqueue-test-Unified Runtime Conformance-6017-24-31.json GTEST_SHUFFLE=0 GTEST_TOTAL_SHARDS=31 GTEST_SHARD_INDEX=24 /__w/llvm/llvm/build/test/conformance/enqueue/enqueue-test
--

Script:
--
/__w/llvm/llvm/build/test/conformance/enqueue/enqueue-test --gtest_filter=urEnqueueKernelLaunchIncrementMultiDeviceTest.Success/UR_BACKEND_CUDA__NVIDIA_CUDA_BACKEND_ID0__NoUseEventWaitRunBackgroundCheck
--
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 3
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 4
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 5
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 6
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 7
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 8
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 9
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:330: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 3

/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 3
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 4
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 5
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 6
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 7
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 8
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 9
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:330
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 3


********************
Testing:  0.. 10.. 
FAIL: Unified Runtime Conformance :: enqueue//enqueue-test/23/31 (102 of 492)
******************** TEST 'Unified Runtime Conformance :: enqueue//enqueue-test/23/31' FAILED ********************
Script(shard):
--
GTEST_OUTPUT=json:/__w/llvm/llvm/build/test/conformance/enqueue/enqueue-test-Unified Runtime Conformance-6017-23-31.json GTEST_SHUFFLE=0 GTEST_TOTAL_SHARDS=31 GTEST_SHARD_INDEX=23 /__w/llvm/llvm/build/test/conformance/enqueue/enqueue-test
--

Script:
--
/__w/llvm/llvm/build/test/conformance/enqueue/enqueue-test --gtest_filter=urEnqueueKernelLaunchIncrementMultiDeviceTest.Success/UR_BACKEND_CUDA__NVIDIA_CUDA_BACKEND_ID0__UseEventWaitNoRunBackgroundCheck
--
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:330: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 3

/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:330
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 3


********************
Testing:  0.. 10.. 
FAIL: Unified Runtime Conformance :: enqueue//enqueue-test/22/31 (107 of 492)
******************** TEST 'Unified Runtime Conformance :: enqueue//enqueue-test/22/31' FAILED ********************
Script(shard):
--
GTEST_OUTPUT=json:/__w/llvm/llvm/build/test/conformance/enqueue/enqueue-test-Unified Runtime Conformance-6017-22-31.json GTEST_SHUFFLE=0 GTEST_TOTAL_SHARDS=31 GTEST_SHARD_INDEX=22 /__w/llvm/llvm/build/test/conformance/enqueue/enqueue-test
--

Script:
--
/__w/llvm/llvm/build/test/conformance/enqueue/enqueue-test --gtest_filter=urEnqueueKernelLaunchIncrementMultiDeviceTest.Success/UR_BACKEND_CUDA__NVIDIA_CUDA_BACKEND_ID0__UseEventWaitRunBackgroundCheck
--
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 3
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 4
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 5
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 6
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 7
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 8
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 9
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:330: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 3

/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 3
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 4
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 5
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 6
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 7
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 8
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:116
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 9
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:330
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 3


********************
Testing:  0.. 10.. 
FAIL: Unified Runtime Conformance :: enqueue//enqueue-test/25/31 (108 of 492)
******************** TEST 'Unified Runtime Conformance :: enqueue//enqueue-test/25/31' FAILED ********************
Script(shard):
--
GTEST_OUTPUT=json:/__w/llvm/llvm/build/test/conformance/enqueue/enqueue-test-Unified Runtime Conformance-6017-25-31.json GTEST_SHUFFLE=0 GTEST_TOTAL_SHARDS=31 GTEST_SHARD_INDEX=25 /__w/llvm/llvm/build/test/conformance/enqueue/enqueue-test
--

Script:
--
/__w/llvm/llvm/build/test/conformance/enqueue/enqueue-test --gtest_filter=urEnqueueKernelLaunchIncrementMultiDeviceTest.Success/UR_BACKEND_CUDA__NVIDIA_CUDA_BACKEND_ID0__NoUseEventWaitNoRunBackgroundCheck
--
/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:330: Failure
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 3

/__w/llvm/llvm/unified-runtime/test/conformance/enqueue/urEnqueueKernelLaunchAndMemcpyInOrder.cpp:330
Expected equality of these values:
  reinterpret_cast<uint32_t *>(SharedMem[i])[j]
    Which is: 2
  ExpectedValue
    Which is: 3


********************
Testing:  0.. 10.. 20.. 30.. 40.. 50.. 60.. 70.. 80.. [90](https://github.com/intel/llvm/actions/runs/15703193260/job/44305438571?pr=19009#step:12:91).. 
********************
Failed Tests (4):
  Unified Runtime Conformance :: enqueue//enqueue-test/urEnqueueKernelLaunchIncrementMultiDeviceTest/Success/UR_BACKEND_CUDA__NVIDIA_CUDA_BACKEND_ID0__NoUseEventWaitNoRunBackgroundCheck
  Unified Runtime Conformance :: enqueue//enqueue-test/urEnqueueKernelLaunchIncrementMultiDeviceTest/Success/UR_BACKEND_CUDA__NVIDIA_CUDA_BACKEND_ID0__NoUseEventWaitRunBackgroundCheck
  Unified Runtime Conformance :: enqueue//enqueue-test/urEnqueueKernelLaunchIncrementMultiDeviceTest/Success/UR_BACKEND_CUDA__NVIDIA_CUDA_BACKEND_ID0__UseEventWaitNoRunBackgroundCheck
  Unified Runtime Conformance :: enqueue//enqueue-test/urEnqueueKernelLaunchIncrementMultiDeviceTest/Success/UR_BACKEND_CUDA__NVIDIA_CUDA_BACKEND_ID0__UseEventWaitRunBackgroundCheck

To reproduce

  1. Include code snippet as short as possible
  2. Specify the command which should be used to compile the program
  3. Specify the command which should be used to launch the program
  4. Indicate what is wrong and what was expected

Environment

  • OS: [e.g Windows/Linux]
  • Target device and vendor: [e.g. Nvidia GPU]
  • DPC++ version: [e.g. commit hash or output of clang++ --version]
  • Dependencies version: [e.g. the output of nvidia-smi and sycl-ls --verbose]

Additional context

No response

yingcong-wu avatar Jun 18 '25 02:06 yingcong-wu

Tagging @pbalcer and @kbenzie for awareness. Please help dispatch to the right owner. Thanks.

yingcong-wu avatar Jun 18 '25 02:06 yingcong-wu

I've created https://github.com/intel/llvm/pull/19035 to disable the tests until its fixed.

kbenzie avatar Jun 18 '25 08:06 kbenzie