Thread overview
Compiler debug enhancement request
Nov 20, 2001
Damian
Nov 20, 2001
Walter
Nov 21, 2001
Damian
Nov 21, 2001
Walter
Nov 21, 2001
Damian
Nov 21, 2001
Walter
Nov 22, 2001
Christof Meerwald
Nov 22, 2001
Walter
Nov 30, 2001
Damian
November 20, 2001
Walter,

Any chance of getting the compiler to output where it saw the previous definition?

This would make my life a lot easier.

D:\CVS\STL\STLport-1115\stlport\stl/_construct.h(56) : Error: '_Destroy' is already defined

Thanks
Damian

November 20, 2001
Best I have at the moment is grep. :-(


"Damian" <damiandixon@netscape.net> wrote in message news:3BFA891B.3090405@netscape.net...
> Walter,
>
> Any chance of getting the compiler to output where it saw the previous definition?
>
> This would make my life a lot easier.
>
> D:\CVS\STL\STLport-1115\stlport\stl/_construct.h(56) : Error: '_Destroy'
> is already defined
>
> Thanks
> Damian
>


November 21, 2001
Walter,

I think I may have asked for the wrong enhancement.

I believe that '_Destroy' may be a compiler generated construct?

If it is is there away of getting of compiler to say so? (enhancement :> please, or possible list the definitions the compiler assumes/creates)

I already use grep... but I am spoilt with the HP-UX ANSI C++ compiler, which tells me where a previous definition was defined and also gives suggestions of possible matches if it can't find a definition (inconsistent spelling).

Regards
Damian

Walter wrote:

> Best I have at the moment is grep. :-(
> 
> 
> "Damian" <damiandixon@netscape.net> wrote in message
> news:3BFA891B.3090405@netscape.net...
> 
>>Walter,
>>
>>Any chance of getting the compiler to output where it saw the previous
>>definition?
>>
>>This would make my life a lot easier.
>>
>>D:\CVS\STL\STLport-1115\stlport\stl/_construct.h(56) : Error: '_Destroy'
>>is already defined
>>
>>Thanks
>>Damian
>>
>>
> 
> 

November 21, 2001
No, _Destroy() is defined in stl_construct.h

I agree that the compiler would be improved if it did say where the other one was.

"Damian" <damiandixon@netscape.net> wrote in message news:3BFB5E17.6020905@netscape.net...
> Walter,
>
> I think I may have asked for the wrong enhancement.
>
> I believe that '_Destroy' may be a compiler generated construct?
>
> If it is is there away of getting of compiler to say so? (enhancement :> please, or possible list the definitions the compiler assumes/creates)
>
> I already use grep... but I am spoilt with the HP-UX ANSI C++ compiler, which tells me where a previous definition was defined and also gives suggestions of possible matches if it can't find a definition (inconsistent spelling).
>
> Regards
> Damian
>
> Walter wrote:
>
> > Best I have at the moment is grep. :-(
> >
> >
> > "Damian" <damiandixon@netscape.net> wrote in message news:3BFA891B.3090405@netscape.net...
> >
> >>Walter,
> >>
> >>Any chance of getting the compiler to output where it saw the previous definition?
> >>
> >>This would make my life a lot easier.
> >>
> >>D:\CVS\STL\STLport-1115\stlport\stl/_construct.h(56) : Error: '_Destroy'
> >>is already defined
> >>
> >>Thanks
> >>Damian
> >>
> >>
> >
> >
>


November 21, 2001
Walter,

The problem is more complex then I originally thought.

It appears to be related to template instantiation. I will try and
look deeper and create a small example.

Is there anyway of getting the compiler to print out information about which templates it has found?

Regards
Damian


Damian wrote:

> Walter,
> 
> I think I may have asked for the wrong enhancement.
> 
> I believe that '_Destroy' may be a compiler generated construct?
> 
> If it is is there away of getting of compiler to say so? (enhancement :> please, or possible list the definitions the compiler assumes/creates)
> 
> I already use grep... but I am spoilt with the HP-UX ANSI C++ compiler, which tells me where a previous definition was defined and also gives suggestions of possible matches if it can't find a definition (inconsistent spelling).
> 
> Regards
> Damian
> 
> Walter wrote:
> 
>> Best I have at the moment is grep. :-(
>>
>>
>> "Damian" <damiandixon@netscape.net> wrote in message
>> news:3BFA891B.3090405@netscape.net...
>>
>>> Walter,
>>>
>>> Any chance of getting the compiler to output where it saw the previous
>>> definition?
>>>
>>> This would make my life a lot easier.
>>>
>>> D:\CVS\STL\STLport-1115\stlport\stl/_construct.h(56) : Error: '_Destroy'
>>> is already defined
>>>
>>> Thanks
>>> Damian
>>>
>>>
>>
>>
> 

November 21, 2001
How I find template instantiation bugs is by whittling away the source till I find the smallest that shows it. The stl is just too big and complicated to isolate things down otherwise.

-Walter

"Damian" <damiandixon@netscape.net> wrote in message news:3BFB610E.2020303@netscape.net...
> Walter,
>
> The problem is more complex then I originally thought.
>
> It appears to be related to template instantiation. I will try and look deeper and create a small example.
>
> Is there anyway of getting the compiler to print out information about which templates it has found?
>
> Regards
> Damian
>
>
> Damian wrote:
>
> > Walter,
> >
> > I think I may have asked for the wrong enhancement.
> >
> > I believe that '_Destroy' may be a compiler generated construct?
> >
> > If it is is there away of getting of compiler to say so? (enhancement :> please, or possible list the definitions the compiler assumes/creates)
> >
> > I already use grep... but I am spoilt with the HP-UX ANSI C++ compiler, which tells me where a previous definition was defined and also gives suggestions of possible matches if it can't find a definition (inconsistent spelling).
> >
> > Regards
> > Damian
> >
> > Walter wrote:
> >
> >> Best I have at the moment is grep. :-(
> >>
> >>
> >> "Damian" <damiandixon@netscape.net> wrote in message news:3BFA891B.3090405@netscape.net...
> >>
> >>> Walter,
> >>>
> >>> Any chance of getting the compiler to output where it saw the previous definition?
> >>>
> >>> This would make my life a lot easier.
> >>>
> >>> D:\CVS\STL\STLport-1115\stlport\stl/_construct.h(56) : Error:
'_Destroy'
> >>> is already defined
> >>>
> >>> Thanks
> >>> Damian
> >>>
> >>>
> >>
> >>
> >
>


November 22, 2001
On Wed, 21 Nov 2001 08:08:46 +0000, Damian wrote:
> The problem is more complex then I originally thought.
> 
> It appears to be related to template instantiation. I will try and look deeper and create a small example.

Are you trying to compile STLport with namespaces enabled? Then I guess it's the same bug I reported on Nov 11 (subject: "template functions and namespaces") in this newsgroup.


bye, Christof

-- 
http://cmeerw.cjb.net                             JID: cmeerw@jabber.at mailto cmeerw at web.de

...and what have you contributed to the Net?
November 22, 2001
I'll get to the namespaces. I've deferred them for the moment until the template stuff works 100% to the standard. -Walter

"Christof Meerwald" <cmeerw@web.de> wrote in message news:9thoi0$q6f$1@digitaldaemon.com...
> On Wed, 21 Nov 2001 08:08:46 +0000, Damian wrote:
> > The problem is more complex then I originally thought.
> >
> > It appears to be related to template instantiation. I will try and look deeper and create a small example.
>
> Are you trying to compile STLport with namespaces enabled? Then I guess
it's
> the same bug I reported on Nov 11 (subject: "template functions and
> namespaces") in this newsgroup.
>
>
> bye, Christof
>
> --
> http://cmeerw.cjb.net                             JID: cmeerw@jabber.at mailto cmeerw at web.de
>
> ...and what have you contributed to the Net?


November 30, 2001
Actually I think the problem is with templates....

Still looking slowly at the problem.

I have run out of project work, so I've been rewriting our training material and giving training course around Europe.

The problem with doing training is you seem to have to know a lot more then you think and it takes up so much time, each customer seems to want training in different aspects of the products!!

Regards
Damian

Walter wrote:

> I'll get to the namespaces. I've deferred them for the moment until the
> template stuff works 100% to the standard. -Walter
> 
> "Christof Meerwald" <cmeerw@web.de> wrote in message
> news:9thoi0$q6f$1@digitaldaemon.com...
> 
>>On Wed, 21 Nov 2001 08:08:46 +0000, Damian wrote:
>>
>>>The problem is more complex then I originally thought.
>>>
>>>It appears to be related to template instantiation. I will try and
>>>look deeper and create a small example.
>>>
>>Are you trying to compile STLport with namespaces enabled? Then I guess
>>
> it's
> 
>>the same bug I reported on Nov 11 (subject: "template functions and
>>namespaces") in this newsgroup.
>>
>>
>>bye, Christof
>>
>>--
>>http://cmeerw.cjb.net                             JID: cmeerw@jabber.at
>>mailto cmeerw at web.de
>>
>>...and what have you contributed to the Net?
>>
> 
>