[Mono Android] Runtime tests failed due to Android app crashing

[fanyang-mono]

Description

There were a bunch of tests failed due to Android app crashing. The root cause might be different. Need further investigation.

Some of them are GC tests. Some of them are under JIT/Methodical, JIT/Performance. Refer to https://github.com/dotnet/runtime/blob/master/src/tests/issues.targets for a detailed list of tests which are marked with this issue.

Configuration

Runtime tests building and running instructions could be found here

Regression?

No. Android is a new target for runtime tests.

[Dotnet-GitSync-Bot]

I couldn't figure out the best area label to add to this issue. If you have write-permissions please help me learn by adding exactly one area label.

[ghost]

Tagging subscribers to this area: @ViktorHofer
See info in area-owners.md if you want to be subscribed.

Issue Details

Description:	Description There were a bunch of tests failed due to Android app crashing. The root cause might be different. Need further investigation. Some of them are GC tests. Some of them are under JIT/Methodical, JIT/Performance. Refer to https://github.com/dotnet/runtime/blob/master/src/tests/issues.targets for a detailed list of tests which are marked with this issue. Configuration Runtime tests building and running instructions could be found here Regression? No. Android is a new target for runtime tests.
Author:	fanyang-mono
Assignees:	fanyang-mono
Labels:	area-Infrastructure, os-android, untriaged
Milestone:	-

[ghost]

Tagging subscribers to this area: @directhex
See info in area-owners.md if you want to be subscribed.

Issue Details

Description:	Description There were a bunch of tests failed due to Android app crashing. The root cause might be different. Need further investigation. Some of them are GC tests. Some of them are under JIT/Methodical, JIT/Performance. Refer to https://github.com/dotnet/runtime/blob/master/src/tests/issues.targets for a detailed list of tests which are marked with this issue. Configuration Runtime tests building and running instructions could be found here Regression? No. Android is a new target for runtime tests.
Author:	fanyang-mono
Assignees:	fanyang-mono
Labels:	area-Infrastructure-mono, os-android
Milestone:	-

[imhameed]

As part of getting the iOS simulator running the runtime tests, I think I stumbled on something related. With my PR in its current state, throwinfinallyintryfilter2.dll fails because it cannot load common.dll. So I started looking around for common.dlls.

Here's all the common.dlls available inside JIT/jit64:

imhameed@ms-ih-mbp ~/ms/dnr-9.x-ios/runtime/artifacts/tests/coreclr/iOSSimulator.x64.Release/JIT/jit64$ find . -iname common.dll -exec shasum -a 256 {} \;
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/common/common/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/call/call05_large/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/call/call05_small/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/call/call05_dynamic/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/ehverify/eh05_dynamic/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/ehverify/eh06_dynamic/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/ehverify/eh07_dynamic/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/ehverify/eh07_large/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/ehverify/eh08_dynamic/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/ehverify/eh07_small/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/eh/eh03_large/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/eh/eh03_dynamic/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/eh/eh04_small/common.dll
1e59f9481a841e9e2acdff6cac4bad0f079b6d8fa2c194be81e33334884b74cb  ./localloc/eh/eh03_small/common.dll
20d90a783ea937a100086c71d317ff3f2b35b1d1dd3ee1c48dfc037ddad1c392  ./hfa/main/dll/common/common.dll
20d90a783ea937a100086c71d317ff3f2b35b1d1dd3ee1c48dfc037ddad1c392  ./hfa/main/testC/hfa_nd1C_r/common.dll
bb44ad02dacb84b60eb77e56eff3e6a86578731e7edd5ddbdfd6de6371e45787  ./eh/basics/throwinfinallyintryfilter2/common.dll
bb44ad02dacb84b60eb77e56eff3e6a86578731e7edd5ddbdfd6de6371e45787  ./eh/Common/common/common.dll

Presumably an arbitrary one (I have no idea what, if any order MSBuild imposes on globs) survives the grand overwriting tournament caused by

runtime/src/tests/run.proj

Line 455 in d156b9e

<AssembliesInTestDirs Include="%(AllCMDsPresent.RelativeDir)*.dll" Exclude="@(TestAssemblies)"/>

So assuming that MSBuild ends up issuing copies in a random order, it is highly likely that throwinfinallyintryfilter2.dll and hfa_nd1C_r.dll will fail, and these two are exactly the tests from JIT.jit64 that are disabled for Android in issues.targets.

[fanyang-mono]

More detailed issues have been created. Closing this.