search

CHIP-SPV / chipStar

chipStar is a tool for compiling and running HIP/CUDA on SPIR-V via OpenCL or Level Zero APIs.
Other
182 stars 29 forks source link

Aborts and Asserts failing #768

Open pvelesko opened 7 months ago

pvelesko commented 7 months ago

Previously these tests were disabled due to a bug in the i915 driver so it seems like there's been some bitrot

╭─pvelesko@cupcake /space/pvelesko/chipStar/main/build ‹check-resovled-d8› 
╰─$ CHIP_BE=opencl ctest -R "abort|TestAssert"                                                                                  130 ↵
Test project /space/pvelesko/chipStar/main/build
    Start 1300: TestAssert
1/4 Test #1300: TestAssert .......................***Failed    0.51 sec
    Start 1301: TestAssertFail
2/4 Test #1301: TestAssertFail ...................***Failed    0.52 sec
    Start 1312: abort
3/4 Test #1312: abort ............................***Failed    0.01 sec
    Start 1313: abort2
4/4 Test #1313: abort2 ...........................***Failed    0.01 sec

0% tests passed, 4 tests failed out of 4

Total Test time (real) =   1.08 sec

The following tests FAILED:
        1300 - TestAssert (Failed)
        1301 - TestAssertFail (Failed)
        1312 - abort (Failed)
        1313 - abort2 (Failed)
Errors while running CTest
Output from these tests are in: /space/pvelesko/chipStar/main/build/Testing/Temporary/LastTest.log
Use "--rerun-failed --output-on-failure" to re-run the failed cases verbosely.
╭─pvelesko@cupcake /space/pvelesko/chipStar/main/build ‹check-resovled-d8› 
╰─$                                                                                                                               8 ↵
╭─pvelesko@cupcake /space/pvelesko/chipStar/main/build ‹check-resovled-d8› 
╰─$ CHIP_BE=level0 ctest -R "abort|TestAssert"                                                                                  130 ↵
Test project /space/pvelesko/chipStar/main/build
    Start 1300: TestAssert
1/4 Test #1300: TestAssert .......................***Failed    1.34 sec
    Start 1301: TestAssertFail
2/4 Test #1301: TestAssertFail ...................***Failed    1.19 sec
    Start 1312: abort
3/4 Test #1312: abort ............................***Failed    0.62 sec
    Start 1313: abort2
4/4 Test #1313: abort2 ...........................   Passed    0.61 sec

25% tests passed, 3 tests failed out of 4

Total Test time (real) =   3.79 sec

The following tests FAILED:
        1300 - TestAssert (Failed)
        1301 - TestAssertFail (Failed)
        1312 - abort (Failed)
Errors while running CTest
Output from these tests are in: /space/pvelesko/chipStar/main/build/Testing/Temporary/LastTest.log
Use "--rerun-failed --output-on-failure" to re-run the failed cases verbosely.
pjaaskel commented 2 weeks ago

What's in the log?

pvelesko commented 2 weeks ago

In cases where it fails it's either nothing or things being out of order.

pvelesko commented 1 week ago

Since these work properly on dGPUs, moving this to 1.3