[nfc][libomptarget] Refactor nvptx/target_impl.cu

[JonChesterfield]

Summary

[nfc][libomptarget] Refactor nxptx/target_impl.cu

Use __kmpc_impl_atomic_add instead of atomicAdd to match the rest of the file.
Alternatively, target_impl.cu could use the cuda functions directly. Using a mixture in this
file was an oversight, happy to resolve in either direction.

Removed some comments that look outdated.

Call __kmpc_impl_unset_lock directly to avoid a redundant diagnostic and remove an implict
dependency on interface.h.

[JonChesterfield]

• Avoid trip through interface for target_impl

[jdoerfert]

LGTM.

openmp/libomptarget/deviceRTLs/nvptx/src/target_impl.cu

 #include "common/debug.h"
 #include "common/target_atomic.h"
 
 #define __OMP_SPIN 1000
 #define UNSET 0u
 #define SET 1u
 
 EXTERN void __kmpc_impl_init_lock(omp_lock_t *lock) {
-  omp_unset_lock(lock);
+  __kmpc_impl_unset_lock(lock);
 }
 
 EXTERN void __kmpc_impl_destroy_lock(omp_lock_t *lock) {
-  omp_unset_lock(lock);
+  __kmpc_impl_unset_lock(lock);
 }
 
 EXTERN void __kmpc_impl_set_lock(omp_lock_t *lock) {
-  // int atomicCAS(int* address, int compare, int val);
-  // (old == compare ? val : old)
-
   // TODO: not sure spinning is a good idea here..
   while (__kmpc_atomic_cas(lock, UNSET, SET) != UNSET) {
     clock_t start = clock();
     clock_t now;
     for (;;) {
       now = clock();
       clock_t cycles = now > start ? now - start : now + (0xffffffff - start);
       if (cycles >= __OMP_SPIN * GetBlockIdInKernel()) {
         break;
       }
     }
   } // wait for 0 to be the read value
 }
 
 EXTERN void __kmpc_impl_unset_lock(omp_lock_t *lock) {
   (void)__kmpc_atomic_exchange(lock, UNSET);
 }
 
 EXTERN int __kmpc_impl_test_lock(omp_lock_t *lock) {
-  // int atomicCAS(int* address, int compare, int val);
-  // (old == compare ? val : old)
-  return atomicAdd(lock, 0);
+  return __kmpc_atomic_add(lock, 0u);
 }