Dear Sir,

Please refer to the following Linux PCB code and let me know why State has been declared Long.

/* memory management info */
struct mm_struct *mm;
/* open file information */
struct files_struct *files;
/* tss for this task */
struct thread_struct tss;
int pid;
volatile long state; /* -1
unrunnable, 0 runnable, >0 stopped
*/
long priority;
unsigned short uid,euid,suid,fsuid;
#ifdef __SMP__
int processor;
#endif
struct task_struct *p_opptr, *p_pptr,
*p_cptr, *p_ysptr, *p_osptr;
/* limits */
struct rlimit rlim[RLIM_NLIMITS];
long utime, stime, cutime, cstime,
start_time;

You'll have to look at how 'state' is used to find out. 'long' is typically used for a variable which may be 32-bit on a 32-bit CPU and 64-bit on a 64-bit CPU (at least GCC allows this behavior).

Thanks for the reply.

But please see the below code:

struct task_struct{
long state; /*0 for dispatchable*/ pid_t pid;
long priority;
unsigned long signal;
unsigned long blocked;
unsigned long flags;

Here we could also have taken int as we might be using 0 value or we could also have taken Double but it seems like there is a reason we used Long everywhere. Please if anyone could answer me this, it will be a great help to me.

Thanks.

Well since it is used as a flag then an integer type (rather than a float) makes sense. Since the comment indicates >0 maybe the use of long is to avoid an overflow problem. To confirm this you will need to look into the code.

long is used to maintain architecture alignment; as mentioned, this will be 32-bit on 32-bit platforms, and 64-bit on 64-bit platforms. If int were used, 64-bit platforms would be accessing unaligned data (relative to register size).