Ok, now I understand what is happening. -(-32768) results in -32768 which when
converted to the long unsigned type results in 4294934528. This is the same
behavior on my PC. (Now knowing the unary operator promotes to an int first.)
#include<stdio.h>
int main(void)
{
unsigned long y = ((short) -32768);
printf("%lu\n", y);
}
Output: 4294934528
And, yes casting to a long first gets my code working the way I want. It is
good to know and I will have to be careful of this in the future. Thanks for
your comments.
Carl
On Wed, 24 Feb 2010 11:02:11 +1000
"Wayne Uroda" <[email protected]> wrote:
>According to the C standard, when you apply the unary operator - to x, x
>is first promoted to an integer.
>I think the difference between the PC and the MSP430 is that on the
>MSP430, the variable is promoted to 16bits which is the same size as the
>short. On the PC, x is first promoted to 32bits.
>
>This is a common annoyance of mine on the msp430 when dealing with 32bit
>types. On any line giving a 32bit result, try casting all the operands
>to long (or unsigned long) before you work on them:
>
>short x = -32768;
>unsigned long y = -((long)x);
>
>does this fix your problem?
>
>- Wayne
>
>-----Original Message-----
>From: Carl [mailto:[email protected]]
>Sent: Wednesday, 24 February 2010 10:01 AM
>To: GCC for MSP430 - http://mspgcc.sf.net
>Subject: Re: [Mspgcc-users] problem converting signed shortto abs
>unsigned long
>
>On Tue, 2010-02-23 at 23:37 +0000, Grant Edwards wrote:
>> On 2010-02-23, Carl <[email protected]> wrote:
>> > On Tue, 2010-02-23 at 23:08 +0000, Grant Edwards wrote:
>> >> The values being passed to printf and then displayed are
>> >> "unsigned". On the '430, that's 16 bits. On the PC that's 32
>> >> bits.
>> >
>> > A short on both my system and 430 is 16 bits. A long on both my
>> system
>> > and the 430 are both 32 bits. I can inspect the memory with gdb. The
>> > problem is the upper 16 bits of y on the 430 are set. Why does it do
>> > this?
>> >
>> > On PC:
>> >
>> > (gdb) p sizeof(y)
>> > $1 = 4
>> > (gdb) p /x y
>> > $2 = 0x8000
>> >
>> > On 430:
>> >
>> > (gdb) p sizeof(y)
>> > $1 = 4
>> > (gdb) p /x y
>> > $2 = 0xffff8000
>>
>> Unless you provide the code that was run up to that point,
>> there's no way to answer your question.
>>
>> But, I suspect that on the msp430, the value in question was an
>> "int" at some point and then got converted to a long.
>>
>> On the MSP430, that resulted in the value being sign-extended
>> when it got converted from 16 to 32 bits. On a PC, an "int"
>> and "long" are both 32 bits, so there is no sign extension when
>> a value is converted from an int to a long.
>
>
>Sorry, here is the code for the msp430:
>
>#include <io.h>
>int main(void)
>{
> short x = -32768;
> unsigned long y = -x;
> LPM0;
>}
>
>And same code as my previous posting for the PC:
>
>#include <stdio.h>
>int main()
>{
> short x = -32768;
> unsigned long y = -x;
> printf("sizeof short =%d, x=%d, y=%lu\n", sizeof(short), x, y);
>}
>
>As you can see there is no int type used.
>
>Carl
>
>
>
>
>------------------------------------------------------------------------
>------
>Download Intel® Parallel Studio Eval
>Try the new software tools for yourself. Speed compiling, find bugs
>proactively, and fine-tune applications for parallel performance.
>See why Intel Parallel Studio got high marks during beta.
>http://p.sf.net/sfu/intel-sw-dev
>_______________________________________________
>Mspgcc-users mailing list
>[email protected]
>https://lists.sourceforge.net/lists/listinfo/mspgcc-users
>
>------------------------------------------------------------------------------
>Download Intel® Parallel Studio Eval
>Try the new software tools for yourself. Speed compiling, find bugs
>proactively, and fine-tune applications for parallel performance.
>See why Intel Parallel Studio got high marks during beta.
>http://p.sf.net/sfu/intel-sw-dev
>_______________________________________________
>Mspgcc-users mailing list
>[email protected]
>https://lists.sourceforge.net/lists/listinfo/mspgcc-users
