craycc: allocate(allocator(...), align(8): t) in a target region fails at the device link (__kmpc_aligned_alloc undefined)

// # Draft finding: omp52-c-clause-allocate-align on cce20
//
// - compiler: cce20
// - toolchain: cce20; arch: gfx942; machine: tuolumne; memory mode: unified; opt: -O2
// - program: catalog omp52-c-clause-allocate-align (c, version 52, sha256 f07c5cb195c9a71b)
// - expected: None (None; )
// - actual: rejected  (link_undefined_symbol: ld.lld: error: undefined symbol: __kmpc_aligned_alloc)
// - verdict: compiler_bug by rule R10b-own-link-or-hang: a link failure on symbols the compiler's own generated code needs, or a compile that does not finish, on a legal program: the compiler's defect (never a rejection)
// - reason: 
// - citations: RUL-COMPILE-TIME (rulings.md)
// - legality: legal (legal by construction: offload.engine.gen / c_gen / acc_gen check every program against the catalog rules and the offload scope as it is built (no compiler))
// - witnesses: none
// - withdrawn: none
//
// Generated by compliance.c_gen (OpenMP 5.2, C11): the C port (offload.engine.f2c) of omp52-clause-allocate-align. Do not edit.
// id: omp52-c-clause-allocate-align
// primary atom: omp.clause.allocate.align
// legal: yes
// observable: no (unverified: the check covers the device kernel, not the atom)
// ran_on_host: inside the target region of interest
// evidence: unverified
// needs: none
// std: c11
// unverified (oracle_not_written): the program does not check the private copy's address: a check of its address modulo an alignment larger than any default (e.g. align(4096)) would fail without the modifier (C mutant survived on AFAR)
#include <stdio.h>
#include <stdbool.h>
#include <omp.h>

enum { n = 64 };

int main(void) {
  int a[n];
  int b[n];
  int ref[n];
  int host_flag;
  bool ok;
  int t;
  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 map(to: a) map(from: b) map(tofrom: host_flag) \
      uses_allocators(omp_default_mem_alloc)
  {
    host_flag = (omp_is_initial_device() ? 1 : 0);
    #pragma omp parallel for default(none) shared(a, b) private(t) \
        allocate(allocator(omp_default_mem_alloc), align(8): t)
    for (int i = 1; i <= n; i++) {
      t = 2 * a[i - 1];
      b[i - 1] = t + 1;
    }
  }
  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;
}
// spec (requirement): The allocate clause accepts the allocator and align modifiers with the syntax and semantics of
//   the allocate directive's allocator and align clauses.
// "The allocator-complex-modifier and align-modifier have the same syntax and semantics for the allocate clause as
//   the allocator and align clauses have for the allocate directive." (5.2 §6.6 p.178 l.16-18)
// spec (requirement): The align modifier makes every allocation at least as aligned as its value.
// "each allocation is byte aligned to at least the maximum of the value to which alignment evaluates, the alignment
//   trait of the allocator being used for the allocation, and the alignment required by the base language for the
//   type of the variable that is allocated." (5.2 §6.3 p.175 l.3-6)
// spec (requirement): On a target construct and in its region an allocate clause must name its allocator, which the
//   program does, with uses_allocators for the predefined allocator.
// "allocate clauses that appear on a target construct or on constructs in a target region must specify an
//   allocator-simple-modifier or allocator-complex-modifier unless a requires directive with the dynamic_allocators
//   clause is present in the same compilation unit." (5.2 §6.6 p.179 l.8-10)
// spec (this program): allocate(allocator(omp_default_mem_alloc), align(8): t) on a parallel for in a target region
//   is a legal allocate clause with both modifiers; craycc fails at the device link because __kmpc_aligned_alloc is
//   undefined.