Skip to content

Assert failure: brick_table[b] != 0 #90452

@cshung

Description

@cshung

This happens when running this program:

// Licensed to the .NET Foundation under one or more agreements.
// The .NET Foundation licenses this file to you under the MIT license.

using System;
using System.Reflection;
using System.Diagnostics;

namespace CoreLab
{
    internal sealed class Finalizer
    {
        private char[] data;
        private static Finalizer resururector;
        private static int MAX = 10000;
        private int index;

        public Finalizer(int i)
        {
            index = i;
            data = new char[i];
        }

        ~Finalizer()
        {
            // For every 10 objects, resurect 1 object
            // this is cover the resurecting scenario of finalizer objects.
            if(index % 10 == 0)
            {
                resururector = this;
            }
        }

        public static void Main()
        {
            for(int i = 0; i < MAX; ++i)
            {
                new Finalizer(i * 16);
                GC.Collect();
                GC.WaitForPendingFinalizers();
            }
        }
    }
}

with set DOTNET_gcServer=1, set DOTNET_GCDynamicAdaptationMode=1 and // #define STRESS_DYNAMIC_HEAP_COUNT uncommented in gcpriv.h.

Why this happen is unclear yet, but STRESS_DYNAMIC_HEAP_COUNT doesn't do anything except with messing around with some measurement variables, it is perceivable that this could manifest as a real bug in the product under certain situations.

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

    Milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions