This is the mail archive of the gcc-help@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Fw: I'd like to get warnings on loss of data during implicit conversions


Jeff:

I believe that the standard just says that sizeof(short) <= sizeof(int) <= 
sizeof(long).  It does not guarantee that "int" on machine A is the same 
as "int" on machine B. Same fro "short" and same for "long". As a result, 
it is possible that a narrowing cast (that's what this is) loses precision 
on machine A but not on machine B. This depends on the machine, the 
compiler, and the moon phase when the compiler was built.

Bill


----- Forwarded by William Mahoney/FACSTAFF/UNO/UNEBR on 09/20/2010 10:18 
AM -----

From:
Jeff Saremi <jeffsaremi@yahoo.com>
To:
gcc-help@gcc.gnu.org
Date:
09/20/2010 10:15 AM
Subject:
I'd like to get warnings on loss of data during implicit conversions
Sent by:
gcc-help-owner@gcc.gnu.org



When i compile my c or C++ code under Microsoft compiler I get warnings 
like the following:
warning C4244: 'argument' : conversion from 'unsigned long' to 'uint16
', possible loss of data

However, compiling the same code with Gcc/G++ does not produce this 
warning at all?

My question is:
- is this a standard (i mean ISO) warning? or is it a courtesy warning?
- Does gcc have something similar to this? If so which option is that?

I have tried the following individiaully and in combination to no avail:

-Wall, -Wextra, -Wconversion
thanks
jeff




Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]