Test failure: baseservices/callconvs/TestCallingConventions/TestCallingConventions.cmd (AV in NativeLibrary.LoadByName under gcstress0x3)

[JulieLeeMSFT]

Summary:
baseservices/callconvs/TestCallingConventions crashes with a native access violation (0xC0000005) in NativeLibrary.g____PInvoke during Program.BlittableFunctionPointers(), under gcstress0x3.

Build Information:

• runtime-coreclr gcstress0x3-gcstress0xc 20260613.1

Error Message:

Fatal error.

0xC0000005

   at System.Runtime.InteropServices.NativeLibrary.<LoadByName>g____PInvoke|1_0(UInt16*, System.Runtime.CompilerServices.QCallAssembly, Int32, UInt32, Int32)

   at Program.BlittableFunctionPointers()

Affected test legs:

runtime-coreclr gcstress0x3-gcstress0xc
- coreclr windows x64 Checked gcstress0x3 @ Windows.10.Amd64.Open
    - baseservices/callconvs/TestCallingConventions/TestCallingConventions.cmd
- coreclr windows x64 Checked gcstress0xc @ Windows.10.Amd64.Open
    - baseservices/callconvs/TestCallingConventions/TestCallingConventions.cmd

Console log: Console Log

Source: runtime-coreclr gcstress0x3-gcstress0xc / coreclr windows x64 Checked gcstress0x3 @ Windows.10.Amd64.Open / baseservices/callconvs/TestCallingConventions/TestCallingConventions.cmd

Analysis:
A native access violation (0xC0000005) occurs in the NativeLibrary. P/Invoke stub while running the blittable function-pointer calling-convention test under GC stress (gcstress0x3). The process faults in NativeLibrary.g____PInvoke|1_0 during Program.BlittableFunctionPointers().

Milestone: 11.0.0

Note

This issue was generated with assistance from GitHub Copilot (AI).

[dotnet-policy-service]

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

[JulieLeeMSFT]

Similar issue: #112328

[dhartglassMSFT]

This seems to no longer fail on top-of-tree, but does fail on the 6/13 build. Trying to find the change.

[dhartglassMSFT]

Closing this.

Failure hit only on 613.1, passed on 614.1, and passed on the previous weekends runs.

Confirmed offline that the failure is gone with #129362 (reverting #128728) but is present with the previous commit

So this failure was introduced by #128728