Open matouskozak opened 3 months ago
Tagging subscribers to 'arch-android': @vitek-karas, @simonrozsival, @steveisok, @akoeplinger See info in area-owners.md if you want to be subscribed.
Tagging subscribers to this area: @dotnet/area-system-security, @bartonjs, @vcsjones See info in area-owners.md if you want to be subscribed.
Could the out-of-memory issue be introduced by https://github.com/dotnet/runtime/pull/102167, causing that the imports are allocating more memory than the previous version @bartonjs ? Or it is possible that Mono is leaking memory somewhere and this change made it more prelevant...
The replacement loader changes the allocation pattern, which might affect what the peak is... but I don't have hard data here, so I can't authoritatively say. It shouldn't be anything with lasting effect.
I'd just disable this test on 32bit, we already do this for other tests with [ConditionalFact(typeof(PlatformDetection), nameof(PlatformDetection.Is64BitProcess))]
.
Build Information
Build: https://dev.azure.com/dnceng-public/public/_build/results?buildId=751577 Build error leg or test failing:
System.Security.Cryptography.Tests
Affected CI:android_arm_release_allsubsets_mono
(runtime-extra-platforms
) Range of commits: https://github.com/dotnet/runtime/compare/c5e8f83d...6effb8f6 (not 100% sure this is the correct range as the issue doesn't reproduce 100% of time)Stack trace example:
Error Message
Fill the error message using step by step known issues guidance.
Known issue validation
Build: :mag_right: https://dev.azure.com/dnceng-public/public/_build/results?buildId=751577 Error message validated:
[Scudo ERROR: out of memory trying to allocate 64 bytes
] Result validation: :x: Known issue did not match with the provided build. Validation performed at: 7/25/2024 6:27:11 AM UTCReport
Summary