RE: [reSIProcate] Really Strange compiler behaviour

["Scott Godin" <slgodin@xxxxxxxxxxxx>]

I don’t really have time to read and absorb this in detail right now – but repro overrides the ServerAuthManager without and problems.

 

---

From: Matthias Moetje - TERASENS GmbH [mailto:moetje@xxxxxxxxxxxx]
Sent: Friday, April 21, 2006 11:19 AM
To: Scott Godin; Alexander Altshuler; resiprocate-devel@xxxxxxxxxxxxxxxxxxx
Subject: RE: [reSIProcate] Really Strange compiler behaviour

 

Scott,

 

are you using a ServerAuthManager? For me everything

worked fine until I implemented server authentication.

Even this worked as long as there is no OK message

sent (for auth success). Only in this case the problem

occurs because the target of the dum command is an

invalid object.

 

I now have an idea why the target command is not

traversed in the ServerAuthManager's constructor:

 

I use the following to construct my ServerAuthManager:

 

MyServerAuthManager::MyServerAuthManager(CTSPILineConnection* pLine, CTSPIAddressInfo* pAddr) :

ServerAuthManager(*pLine->GetDum(),pLine->GetDum()->dumIncomingTarget())

{

m_pLine = pLine;

m_pAddr = pAddr;

}

The problem might be that dum->dumIncomingTarget() returns an
object that is actually an abstract class i.e. it can't exist as an
object but only as a pointer.

When I try the following:

resip::DialogUsageManager::IncomingTarget mTarget(mDum->dumIncomingTarget());

I get an error as well (cannot instantiate abstract class). So if this

fails, I think it's no wonder that it fails in the constructor, too. In fact

if I debug the dumIncomingTarget() function I can see that it

returns an invalid object, so I think this is where the problem really

lies.

 

I am not sure if the following is valid anyway:

IncomingTarget* mIncomingTarget;

TargetCommand::Target& dumIncomingTarget()

{

        return *mIncomingTarget;

}

So this function actually returns IncomingTarget& but not Target&.
Can we expect that this is automatically casted correctly?
It seems not, because this function does not return anything valid,
at least not with VS 2005.

I would tend to change this stuff to use pointers, but this would surely

break some applications (e.g. repro). Or rather try to make

TargetCommand::Target non-abstract, could this help. What do you think?

 

Matthias

 

---

From: Scott Godin [mailto:slgodin@xxxxxxxxxxxx]
Sent: Friday, April 21, 2006 4:40 PM
To: Matthias Moetje - TERASENS GmbH; Alexander Altshuler; resiprocate-devel@xxxxxxxxxxxxxxxxxxx
Subject: RE: [reSIProcate] Really Strange compiler behaviour

I’m using VS.NET 2005 and don’t have these problems – I have no idea why you are seeing this.  Have you tried to eliminate your application / settings from the picture and see if BasicCall behaves the same way?

 

Scott

 

---

From: resiprocate-devel-bounces@xxxxxxxxxxxxxxxxxxx [mailto:resiprocate-devel-bounces@xxxxxxxxxxxxxxxxxxx] On Behalf Of Matthias Moetje - TERASENS GmbH
Sent: Friday, April 21, 2006 10:11 AM
To: Alexander Altshuler; resiprocate-devel@xxxxxxxxxxxxxxxxxxx
Subject: RE: [reSIProcate] Really Strange compiler behaviour

 

Alex,

 

thanks for the tip, but this didn't solve the problem.

 

I have now converted the TargetCommand class to

use a pointer to dum instead of passing it by value

and now it works.

 

Seems that VS 2005 is a bit more restrictive with

passing objects by value (or it just has a bug..?).

At least it is said to be more standards conformant

than all previous versions.

 

 

Unfortunately this is not all. A similar problem occurs

in the constructor of ServerAuthManager when it

should be initialized with dumIncomingTarget().

 

I suspect all this might be due to the fact that the

IncomingTarget and Outgoing target classes are

declared privately within DialogUsageManager?

 

I will now try to move these outside of the dum

class...

Best regards,

Matthias Moetje

TERASENS GmbH
Augustenstraße 24
80333 Munich
GERMANY

 

Phone: Fax: e-mail: Web:

+49.89.143370-0 +49.89.143370-22 info@xxxxxxxxxxxx www.terasens.com

 

 

---

From: resiprocate-devel-bounces@xxxxxxxxxxxxxxxxxxx [mailto:resiprocate-devel-bounces@xxxxxxxxxxxxxxxxxxx] On Behalf Of Alexander Altshuler
Sent: Friday, April 21, 2006 8:45 AM
To: resiprocate-devel@xxxxxxxxxxxxxxxxxxx
Subject: RE: [reSIProcate] Really Strange compiler behaviour

Delete all files (.obj, .pdb etc) in Debug directory by hand.

Recompile the project.

Sometimes it works.

 

Alex

 

-----Original Message-----
From: resiprocate-devel-bounces@xxxxxxxxxxxxxxxxxxx [mailto:resiprocate-devel-bounces@xxxxxxxxxxxxxxxxxxx] On Behalf Of Matthias Moetje - TERASENS GmbH
Sent: Friday, April 21, 2006 6:31 AM
To: resiprocate-devel@xxxxxxxxxxxxxxxxxxx
Subject: [reSIProcate] Really Strange compiler behaviour

 

Hi,

 

I am experiencing some really strange behaviour on the

following lines in the constructor of DialogUsageManager:

 

mIncomingTarget = new IncomingTarget(*this);

mOutgoingTarget = new OutgoingTarget(*this);

 

Actually the objects are created through _nh_malloc_dbg
when I debug through the generic runtime implementation
of the new operator; afterwards the constructors of
the object and the inherited objects are called. Though,
in the end the result from the new operator is not assigned
to the pointer variable i.e. in the end the pointer variable
is NULL.

But if I note the pointer from the operator new implementation
and assign it to the variable(s) manually in the debugger, everything
is fine!

Seems very strange to me! I'm using VS.NET 2005. All I could
think of here is probably the way the dum object itself is
passed into the constructor (*this)..?

Does anyone have an idea why this happens? I thought of
passing dum as a pointer instead, but that would require a
change to dum itself...

I would be very thankful for any hints on this, I have no other
idea about that...

Best regards,

Matthias Moetje