OpenMP 6.0 requires device_safesync and defaultmap(private) are rejected (not in GCC's implementation status table)

// # Finding (GCC 16.2): requires device_safesync and defaultmap(private) are rejected, and GCC's table does not list them (C); the
//   reproducer is the catalog program omp60-c-clause-device_safesync
//
// - toolchains: gcc_g90 (GCC 16.2.0, AMD Instinct MI250X gfx90a, Tioga) and gcc_sm80 (GCC 16.2.0,
//   NVIDIA A100 sm_80, wingtip); separate memory only; -O2; OMP_TARGET_OFFLOAD=mandatory
// - build: gcc -fopenmp -foffload=amdgcn-amdhsa -foffload-options=amdgcn-amdhsa=-march=gfx90a -O2 (gcc_g90); -foffload=nvptx-none
//   -foffload-options=nvptx-none=-march=sm_80 -foffload-options=nvptx-none=-latomic -O2 (gcc_sm80)
// - failure: rejected
// - observed: `error: expected 'unified_address', 'unified_shared_memory', 'self_maps', 'dynamic_allocators', ...` at
//   device_safesync; `error: expected 'alloc', 'to', 'from', 'tofrom', 'firstprivate', 'none' or 'default' before 'private'` at
//   defaultmap(private); both GPUs.
// - expected: compiles; `RESULT ran_on_host=0 check=pass`
// - GCC 16.2 documentation (libgomp.texi of the GCC 16.2.0 release tarball, "OpenMP Implementation Status", OpenMP 6.0 table): no
//   row names the device_safesync requirement clause or private as a defaultmap implicit-behavior (the table's defaultmap row is
//   "self modifier to map and self as defaultmap argument"). Undocumented.
// - spec: the device_safesync clause is a 6.0 requirement clause (6.0 §5.4 p.173 lists it among the clauses that take a
//   directive-name modifier); defaultmap's implicit-behavior private is new in 6.0.
// - other families (catalog matrix): requires device_safesync: CCE 19/20/21 and NVHPC reject it; AFAR amdflang compiles and runs
//   the Fortran program, AFAR amdclang's C compile does not finish (afar-clang-requires-device-safesync-hang.c).
//   defaultmap(private): AFAR amdclang and CCE 21 run the C program correctly, CCE 20 the Fortran one; CCE 19 and NVHPC reject it.
// - gate and confirmer: 8 cells (4 gcc_g90, 4 gcc_sm80), each confirmed in the gate's four environments (a-d;
//   results/gcc/gate-tioga/ and gate-wingtip/ verdicts.jsonl); confirmer verdicts (results/gcc/confirm/batch.gcc_*.jsonl):
//   unsupported (R11a-newer-form-unclaimed) 8
// - catalog programs (4): omp60-c-clause-defaultmap-private, omp60-c-clause-device_safesync,
//   omp60-c-clause-device_safesync-directive_name, omp60-c-clause-device_safesync-required
// - legality: the catalog programs are legal by construction (offload.engine.gen, c_gen and acc_gen
//   check every program against the catalog rules and the offload scope as it is built; no
//   compiler). GCC is a compiler under test only: never a legality witness, here or for another
//   family's case (catalog/machines.yaml under_test_only).
// Generated by compliance.c_gen (OpenMP 6.0, C11): the C port (offload.engine.f2c) of omp60-clause-device_safesync. Do not edit.
// id: omp60-c-clause-device_safesync
// primary atom: omp.clause.device_safesync
// legal: yes
// observable: no (requires only changes semantics: no checked device effect)
// ran_on_host: inside the target region of interest
// evidence: construct
// needs: none
// std: c11
#include <stdio.h>
#include <stdbool.h>
#include <omp.h>

enum { n = 64 };
#pragma omp requires device_safesync

int main(void) {
  int a[n];
  int b[n];
  int ref[n];
  int host_flag;
  bool ok;
  for (int i = 1; i <= n; i++) {
    a[i - 1] = i;
    b[i - 1] = 0;
    ref[i - 1] = 2 * i + 1;
  }
  host_flag = 1;
  #pragma omp target teams distribute parallel for map(to: a) map(from: b) \
      map(tofrom: host_flag) default(none) shared(a, b, host_flag)
  for (int i = 1; i <= n; i++) {
    b[i - 1] = 2 * a[i - 1] + 1;
    if (i == 1) host_flag = (omp_is_initial_device() ? 1 : 0);
  }
  bool t1_ = true;
  for (int k2_ = 0; k2_ < n; k2_++)
    t1_ = t1_ && (b[k2_] == ref[k2_]);
  ok = t1_;
  printf("RESULT ran_on_host=%d check=%s\n", host_flag, ok ? "pass" : "fail");
  return 0;
}