Microsoft Specific

Generate an instruction which examines bit b of the address a and returns its current value before setting it to 1.

unsigned char _interlockedbittestandset(  
   long *a,  
   long b  
);  
unsigned char _interlockedbittestandset_acq(  
   long *a,  
   long b  
);  
unsigned char _interlockedbittestandset_HLEAcquire(  
   long *a,  
   long b  
);  
unsigned char _interlockedbittestandset_HLERelease(  
   long *a,  
   long b  
);  
unsigned char _interlockedbittestandset_nf(  
   long *a,  
   long b  
);  
unsigned char _interlockedbittestandset_rel(  
   long *a,  
   long b  
);  
unsigned char _interlockedbittestandset64(  
   __int64 *a,  
   __int64 b  
);  
unsigned char _interlockedbittestandset64_HLEAcquire(  
   __int64 *a,  
   __int64 b  
);  
unsigned char _interlockedbittestandset64_HLERelease(  
   __int64 *a,  
   __int64 b  
);  

[in] a
A pointer to the memory to examine.

[in] b
The bit position to test.

The value of the bit at position b before it is set.

On x86 and [!INCLUDEvcprx64] processors, these intrinsics use the lock bts instruction to read and set the specified bit to 1. The operation is atomic.

On ARM processors, use the intrinsics with _acq and _rel suffixes for acquire and release semantics, such as at the beginning and end of a critical section. The ARM intrinsics with an _nf ("no fence") suffix do not act as a memory barrier.

On Intel processors that support Hardware Lock Elision (HLE) instructions, the intrinsics with _HLEAcquire and _HLERelease suffixes include a hint to the processor that can accelerate performance by eliminating a lock write step in hardware. If these intrinsics are called on processors that do not support HLE, the hint is ignored.

These routines are only available as intrinsics.

END Microsoft Specific

Compiler Intrinsics
Conflicts with the x86 Compiler