Re: Re: [omniORB] Porting of omniORB over RTOS

binod pal binod_pal at rediffmail.com
Thu Sep 9 14:31:26 BST 2010


Hi,

I would like to clarify that UserException is not a macro, it is a class. I had tried two options

Option-1
I commented virtual and inline functions in UserException class
then i got the following error

OMNIORB_DECLARE_USER_EXCEPTION_IN_CORBA(WrongTransaction, _dyn_attr)

error #54-D: too few arguments in macro invocation
    OMNIORB_DECLARE_USER_EXCEPTION_IN_CORBA(Bounds, _dyn_attr)
appears in various files like CORBA_NamedValue.h, CORBA_Context.h etc.

I think the error may be in macro expansion i.e arguments of macro may themselves be macro. Any comments for this?

Option-2
with virtual and inline lines without comment
following error appears
error #54-D: too few arguments in macro invocation
      virtual ~UserException();
error #279: invalid destructor declaration
      virtual ~UserException();
error #79: expected a type specifier
      virtual ~UserException();
error #407: a destructor may not have parameters
      virtual ~UserException();
error #79: expected a type specifier
      virtual ~UserException();
but UserException is class not a macro. I don't know why this is giving error.

Does anyone has idea what is wrong here.

Regards
Binod Pal 

On Thu, 09 Sep 2010 06:22:08 +0530  wrote
>2010/9/8 Thomas Lockhart :

>

>> 2. When i am compiling the code for INTEGRITY, i am getting the

>> following errors...

>> "D:toDelincludeomniORB4/CORBA_Exception.h", line 245: warning #76-D:

>> argument to macro is empty virtual ~UserException();

>

> It looks like "virtual" is declared as some kind of goofy macro in

> INTEGRITY. Too bad for you :/



No, I find that rather hard to believe, otherwise you could not code

in C++ under INTEGRITY, which is something we've been doing for many

years now. However, I can believe that UserException is a macro (hint:

so is DataBuffer). I have no idea why anyone thinks it's a good idea

to write macros in CamelCase...



We contemplated (briefly) porting omniORB to INTEGRITY, and I wish you

luck, Binod Pal. INTEGRITY does provide a POSIX wrapper around it's

API, but I would be afraid that the semantics of all the various

functions may not match the true POSIX semantics close enough for a

trouble-free port. But please keep us posted of your progress, I might

be interested in helping you.



Regards,

BN



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.omniorb-support.com/pipermail/omniorb-list/attachments/20100909/6152aa4a/attachment.htm


More information about the omniORB-list mailing list