[Coin-lpsolver] Re: Coin-lpsolver Digest, Vol 24, Issue 7

[Darío Alejandro Guzik]

Lou Hafer wrote:
> Anureet,
>
> 	I'd better grab this one, OsiDylp is my responsibility. I'm aware of the
> problem but don't have access to a 64-bit system to do testing.
>
> 	The warnings should not be significant.
> 	
> 	The error messages (there are several occurrences) are due to a set of
> informational messages. You should be able to get them to go away by redoing
> the configure with --without-osidylp-info, or by editing the file 
> Osi/inc/config_osi.h. Simply comment out the line
>
> #define ODSI_INFOMSGS 1
>
> or replace it with
>
> #undef ODSI_INFOMSGS
>
> You'll see the pattern in config_osi.h. Autoconf goes one step further, putting
> comments around the undef to avoid possible symbol conflicts.
>
> If you're willing, could you please try the following?
> Each occurence of the problem will look something like this:
>
> #   ifdef ODSI_INFOMSGS
>     hdl->message(ODSI_ATTACH,messages_)
>       << "resolve" << reinterpret_cast<int>(this)
>       << CoinMessageEol ;
> #   endif
>
> The problem is the reinterpret_cast --- it needs to be something large enough
> to hold an address. I'm guessing that int is 32 bits and addresses are now 64 
> bits. Replacing `int' with `long int' might fix the problem.
>
>   
I tried this but it didn't work. I  will look up the error since I
didn't wrote it up when I tried, but I recall it was something about
ambiguity (I mean the change of int with long int, the undef method
worked fine)
(Gentoo amd64 + gcc4.1.1)
> 	Let me know which works. I can suggest some more radical solutions if
> necessary.
>
> 							Lou
>
> _______________________________________________
> Coin-lpsolver mailing list
> Coin-lpsolver at list.coin-or.org
> http://list.coin-or.org/mailman/listinfo/coin-lpsolver
>
>   

-- 
Darío Alejandro Guzik
email: dguzik at papyro.com

Braier & Asociados Consultores
www.papyro.com
4795-9666