QMessageBox memory problem

[Jeffb]

Hi guys

Have been using this code with no problem up until now.
Now its suddenly started crashing when it gets to aMB.exec();
with memory bad access.
Signal name :
EXC_BAD_ACCESS

Has anyone got any ideas what might be happening.
I have cleaned and rebuilt and even turned the machine off and back on again.
Am using Mac OS X Lion.

Thanks.
Jeff

void JB_Utilities::debugMB(const QString &text,
                           QWidget* parent,
                           Qt::WindowModality aWindowModality)
{
    cout << "ERROR: " << qPrintable(text) << endl;
    QMessageBox aMB(QMessageBox::Warning,
                    "ERROR",
                    text,
                    QMessageBox::Ok,
                    parent);
    aMB.setWindowModality(aWindowModality);
    aMB.setDefaultButton(QMessageBox::Ok);
    aMB.exec();
}

void JB_Utilities::debugMB(const QString &text,
                           QWidget* parent,
                           Qt::WindowModality aWindowModality)
{
    cout << "ERROR: " << qPrintable(text) << endl;
    QMessageBox aMB(QMessageBox::Warning,
                    "ERROR",
                    text,
                    QMessageBox::Ok,
                    parent);
    aMB.setWindowModality(aWindowModality);
    aMB.setDefaultButton(QMessageBox::Ok);
    aMB.exec();
}

To copy to clipboard, switch view to plain text mode 

[ChrisW67]

Check that parent is a valid pointer or null/zero.

[Jeffb]

@Chris

I'm definitely passing zero into the method for the parent.

[amleto]

presumably this is with debug build? what is the call stack? do you have multiple qt installs?

[ChrisW67]

Does the problem persist after a clean and full rebuild?

[Jeffb]

Yes it continued after a complete clean and rebuild.
Its being called from another method of JB_Utilities.

When called from an entirely different object, it has no problems.

The call stack shows it getting to .exec and then crashing.
There is only one qt install.

The code is simply there to tell me when there is a problem.
In this case I can get around it by making sure the code that calls it never gets executed (which is the way it should be).
But I still hate not knowing what is going on.

Cheers
Jeff

[Lykurg]

May be this article helps: http://blogs.kde.org/node/3919.