Curious, why the second cast?

[Pommie]

I was looking at some Arduino software and came across two rather interesting defines,

#define LOBYTE(w)           ((unsigned char)(w))
//! High nibble of 16bit variable
#define HIBYTE(w)           ((unsigned char)(((unsigned short)(w) >> 8) & 0xFF))

Both look very useful but I wonder about the second cast in the HIBYTE definition. Why the cast to unsigned short? In my experience, short equals 24 bits. Can someone explain this?

Thanks,

Mike.

 

[Nigel Goodwin]

An unsigned short is normally 16 bits - presumably used here because you're wanting the high byte of the word in w.

Try removing the second cast, and see if you get a warning message.

 

[Pommie]

I'd assumed that w is already 16 bit or it's kinda pointless. I suppose if you feed the macro (define) an 8 bit value it'll stop it falling over. I now always use the uintn_t type bit safe defines so I know how big something is. Never thought unsigned short was 16 bit, learn't something today. Thanks.

Mike.

 

[Pommie]

I guess coming from an 8 bit background (games in the 80s) and now being mainly XC8 based makes for a very limited knowledge of systems. Am I right in thinking that a short in XC8 is 24 bits?

Mike.

 

[Ian Rogers]

LOL I looked up XC's version of Generic types

/* INT is processor specific in length may vary in size */
typedef signed int          INT;
typedef signed char         INT8;
typedef signed short int    INT16;
typedef signed long int     INT32;

/* MPLAB C Compiler for PIC18 does not support 64-bit integers */
#if !defined(__18CXX)
__EXTENSION typedef signed long long    INT64;
#endif

/* UINT is processor specific in length may vary in size */
typedef unsigned int        UINT;
typedef unsigned char       UINT8;
typedef unsigned short int  UINT16;
/* 24-bit type only available on C18 */
#if defined(__18CXX)
typedef unsigned short long UINT24;
#endif
typedef unsigned long int   UINT32;     /* other name for 32-bit integer */
/* MPLAB C Compiler for PIC18 does not support 64-bit integers */

unsigned short long UINT24... I was thinking Short round in Indiana Jones...

 

[Pommie]

So, we now have two definitions and neither is short. We have short int and short long. Bed time for me.

Thanks all,

Mike.

 

[Nigel Goodwin]

I guess coming from an 8 bit background (games in the 80s) and now being mainly XC8 based makes for a very limited knowledge of systems. Am I right in thinking that a short in XC8 is 24 bits?

Mike.

Nope, 16 bits in XC8.

 

[Pommie]

Damb, more stupid than I thought. However madam, in the morning .....

 

[Pommie]

I wasn't calling you a madam, it was a quote.

 

[Nigel Goodwin]

I wasn't calling you a madam, it was a quote.

I've been called worse

Doesn't the quote continue "I'll be sober"?.

 

[Pommie]

Doesn't the quote continue "I'll be sober"?.

Indeed, allegedly by Churchill.

Mike.

 

[Nigel Goodwin]

Indeed, allegedly by Churchill.

Mike.

I though pretty well all quotes were allegedly by Churchill (or Lincoln)

 

[granddad]

Built in __uint24 __int24 can be used in XC8 , I use it as SPI , FRAM addresses

/*                FRAM SPI DRIVERS 2Mb FRAM
  * uint8_t Fram_READ(__uint24 FramAddr);
  * READ_address 0000 to FFFF, page 00 to 03     
  *
  * void Fram_WRITE(__uint24 FramAddr,uint8_t FramData); 
  *  WRITE  address 0000 to FFFF, page 00 to 03 ,WRITE DATA .   
  *
  * void Fram_WREN();
  *  SET WRITE ENABLE LATCH
  *
  * uint8_t Fram_RDSR();
  *
  * void Fram_WRSR();
  *
  *  * Name   description            Op-code
 * WREN Set Write Enable Latch   0000 0110 B
 * WRDI Reset Write Enable Latch 0000 0100 B
 * RDSR Read Status Register     0000 0101 B
 * WRSR Write Status Register    0000 0001 B
 * READ Read Memory Code         0000 0011 B
 * WRITE Write Memory Code       0000 0010 B
 * RFID Read Device ID           1001 1111 B
 * FSTRD Fast Read Memory Code   0000 1011 B
 * SLEEP Sleep Mode              1011 1001 B
 *
 */