Re: [Boost-users] crc.hpp and visual studio run time checking

From: Peter Dimov (pdimov_at_[hidden])
Date: 2005-12-18 14:10:52

Daryle Walker wrote:
> On 12/16/05 2:03 PM, "Clarke, Trevor" <tclarke_at_[hidden]> wrote:
>
>> It's not a mask to stop data loss but to tell VS you want the data
>> loss. I.E.
>>
>> unsigned short a = 0x1234;
>> return (a << 8) & 0xffff;
> [TRUNCATE past message context]
>
> That mask retains 100% of the bits (if a "short" is 16 bits), so it
> effectively does nothing.  That's a waste of cycles if the masking
> isn't optimized away.  I don't get how this shuts up the RTC unless
> this construct was specifically created as a workaround.

There is no unsigned short arithmetic in C/C++; in a << 8 a is promoted to 
int or unsigned int first, then the result is implicitly converted to 
unsigned short by the return, and this causes high-order bits to be lost.