[QUIC] [Linux] Segmentation fault in System.Net.Http.Functional.Tests

[matouskozak]

trafficstars

Build Information

Build: https://dev.azure.com/dnceng-public/public/_build/results?buildId=712564 Build error leg or test failing: System.Net.Http.Functional.Tests Affected CI: linux-arm Release Libraries_Release_CoreCLR (runtime-extra-platforms)

Error Message

Fill the error message using step by step known issues guidance.

{
  "ErrorPattern": ["arm", "Segmentation fault.*System.Net.Http.Functional.Tests"],
  "BuildRetry": false,
  "ExcludeConsoleLog": false
}

Known issue validation

Build: :mag_right: https://dev.azure.com/dnceng-public/public/_build/results?buildId=712564 Error message validated: [arm Segmentation fault.*System.Net.Http.Functional.Tests] Result validation: :white_check_mark: Known issue matched with the provided build. Validation performed at: 6/21/2024 6:18:04 AM UTC

Report

Build	Definition	Test	Pull Request
721722	dotnet/runtime	System.Net.Http.Functional.Tests.WorkItemExecution
720898	dotnet/runtime	System.Net.Http.Functional.Tests.WorkItemExecution
720816	dotnet/runtime	System.Net.Http.Functional.Tests.WorkItemExecution
719424	dotnet/runtime	System.Net.Http.Functional.Tests.WorkItemExecution
713848	dotnet/runtime	System.Net.Http.Functional.Tests.WorkItemExecution
712564	dotnet/runtime	System.Net.Http.Functional.Tests.WorkItemExecution

Summary

24-Hour Hit Count	7-Day Hit Count	1-Month Count
0	4	6

[dotnet-policy-service[bot]]

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

[jkotas]

Condition(s) not met: "IsChromium

This is not an error message. This message is in every test log. I have fixed the pattern.

[MihaZupan]

exit code 139 means SIGSEGV Illegal memory access. Deref invalid pointer, overrunning buffer, stack overflow etc. Core dumped.

The console output looks like a crash.

The System.Exception: Early EOF does appear a lot, but those tests are passing, XUnit is just reporting the ITestOutputHelper text for all the tests now, not just the failing ones -- see #103445.

[matouskozak]

exit code 139 means SIGSEGV Illegal memory access. Deref invalid pointer, overrunning buffer, stack overflow etc. Core dumped.

The console output looks like a crash.

The System.Exception: Early EOF does appear a lot, but those tests are passing, XUnit is just reporting the ITestOutputHelper text for all the tests now, not just the failing ones -- see #103445.

There is another console output without the System.Exception: Early EOF.

[rzikm]

cc @liveans

[jkotas]

Analysis for build 712564:

Crash in libmsquic.so at RecvDataReturn . Null pointer passed to InterlockedDecrement.

(gdb) bt
#0  0xe861799c in InterlockedDecrement (Addend=0x60) at /__w/1/s/src/inc/quic_platform_posix.h:115
#1  RecvDataReturn (RecvDataChain=0x0) at /__w/1/s/src/platform/datapath_epoll.c:2067
#2  0xe860f836 in CxPlatRecvDataReturn (RecvDataChain=<optimized out>) at /__w/1/s/src/platform/datapath_linux.c:338
#3  0xe85e7204 in QuicConnRecvDatagrams (Connection=Connection@entry=0xf686f278, Packets=0x0, Packets@entry=0xe43948c8, PacketChainCount=PacketChainCount@entry=2,
    PacketChainByteCount=PacketChainByteCount@entry=2333, IsDeferred=<optimized out>, IsDeferred@entry=0 '\000') at /__w/1/s/src/core/connection.c:5745
#4  0xe85e76aa in QuicConnFlushRecv (Connection=Connection@entry=0xf686f278) at /__w/1/s/src/core/connection.c:5826
#5  0xe85e9a2a in QuicConnDrainOperations (Connection=Connection@entry=0xf686f278) at /__w/1/s/src/core/connection.c:7575
#6  0xe85d578c in QuicWorkerProcessConnection (Worker=Worker@entry=0xf68b94c0, Connection=0xf686f278, ThreadID=<optimized out>, TimeNow=TimeNow@entry=0xe166fd88) at /__w/1/s/src/core/worker.c:506
#7  0xe85d5d20 in QuicWorkerLoop (Context=0xf68b94c0, State=0xe166fd88) at /__w/1/s/src/core/worker.c:658
#8  0xe860d7bc in CxPlatRunExecutionContexts (State=0xe166fd88, Worker=<optimized out>) at /__w/1/s/src/platform/platform_worker.c:395
#9  CxPlatRunExecutionContexts (Worker=<optimized out>, State=0xe166fd88) at /__w/1/s/src/platform/platform_worker.c:369
#10 0xe860d948 in CxPlatWorkerThread (Context=0xf6863370) at /__w/1/s/src/platform/platform_worker.c:492
#11 0xf7745dd6 in start_thread (arg=0x3d63daff) at pthread_create.c:442
#12 0xf779c8a0 in ?? () at ../sysdeps/unix/sysv/linux/arm/clone.S:74 from /lib/arm-linux-gnueabihf/libc.so.6

[ManickaP]

This is Arm32 again, isn't it. Similar issue: https://github.com/dotnet/runtime/issues/103404. cc @nibanks

[jkotas]

Hit in #104202 : https://dev.azure.com/dnceng-public/public/_build/results?buildId=726568&view=ms.vss-test-web.build-test-results-tab&runId=18247354&resultId=137390&paneView=dotnet-dnceng.dnceng-anon-build-release-tasks.helix-anon-test-information-tab

[jkotas]

Hit in https://github.com/dotnet/runtime/pull/104264

[liveans]

This issue, #103404, #91757 will be fixed with next release of msquic.