[PyQt] Regression with guiqwt somewhere after PyQt4 4.10

[Phil Thompson]

On 2013-12-08 11:44, Dmitry Shachnev wrote:
> On Wed, Nov 13, 2013 at 3:54 PM, Dmitry Shachnev <mitya57 at ubuntu.com> 
> wrote:
>> The crash happens here (src/corelib/kernel/qmetaobject.cpp:525):
>>
>>     for (const QMetaObject *m = *baseObject; m; m = m->d.superdata) 
>> {
>>         int i = (MethodType == MethodSignal && 
>> priv(m->d.data)->revision >= 4)
>>                 ? (priv(m->d.data)->signalCount - 1) :
>> (priv(m->d.data)->methodCount - 1);
>>
>> The problem is that m->d.data is null on the fourth iteration:
>>
>> (gdb) p **baseObject
>> $12 = {d = {superdata = 0x8527d80, stringdata = 0x8712588
>> "LevelsHistogram", data = 0x8fddb48, extradata = 0x0}}
>> (gdb) p *(baseObject->d.superdata)
>> $13 = {d = {superdata = 0x8ec6f00, stringdata = 0x88123b0 
>> "CurvePlot",
>> data = 0x8e156f8, extradata = 0x0}}
>> (gdb) p *(baseObject->d.superdata->d.superdata)
>> $14 = {d = {superdata = 0xb13244e0, stringdata = 0x8ecd1a0 
>> "BasePlot",
>> data = 0x8ef04e0, extradata = 0x0}}
>> (gdb) p *(baseObject->d.superdata->d.superdata->d.superdata)
>> $15 = {d = {superdata = 0xb12bc4ec, stringdata = 0x0, data = 0x0,
>> extradata = 0xb12bc4cd}}
>>
>> A backtrace is attached.
>>
>> On Sat, Nov 9, 2013 at 9:47 PM, Scott Kitterman 
>> <debian at kitterman.com> wrote:
>>> Below is a part of an Ubuntu bug report.  You can find a 
>>> reproduction test case
>>> at 
>>> https://bugs.launchpad.net/ubuntu/+source/python-qt4/+bug/1249571/+attachment/3903922/+files/test_guiqwt.py
>
> Hi Phil,
>
> I continue getting pings about that. Do you have any thoughts on the
> issue? Sounds like a problem in PyQt.

Then I need a PyQt test case that demonstrates the problem. I have no 
idea what guiqwt is.

Phil