Avoid eternal requeue if the SlurmctldProlog fails

Reset the flag before requeuing, otherwise once SlurmctldProlog fails
once for the job the job will never be launched successfully.

Bug 11574.

NEWS

@@ -21,6 +21,7 @@ documents those changes that are of interest to users and administrators.
  -- srun - fix broken node step allocation in a heterogeneous allocation.
  -- Fail step creation if -n is not multiple of --ntasks-per-gpu.
  -- job_container/tmpfs - Fix slowdown on teardown.
+ -- Fix problem with SlurmctldProlog where requeued jobs would never launch.
 * Changes in Slurm 20.11.7
 ==========================

src/slurmctld/prep_slurmctld.c

@@ -74,6 +74,7 @@ extern void prep_prolog_slurmctld_callback(int rc, uint32_t job_id)
 
 	/* all async prologs have completed, continue on now */
 	if (job_ptr->prep_prolog_failed) {
+		job_ptr->prep_prolog_failed = false;
 		if ((rc = job_requeue(0, job_id, NULL, false, 0))) {
 			info("unable to requeue JobId=%u: %s", job_id,
 			     slurm_strerror(rc));