Fix problems reported by valgrind, several memory leaks and one
  memory reference after free.

This is needed to support requeue and other operations on the
  pack job as a whole

Make sure the pack_job_id values match, just in case something
  bad happended with the job records. If they don't match, log
  an error and skip the records.

Always start the script in the correct component's allocation
  and get the script from the corrrect location (could work before
  but dependent upon race condition).

If num_tasks == NO_VAL but min_cpus != NO_VAL, then num_cpus gets
reset to NO_VAL, which results in select/cons_res getting stuck in
a very slow loop (counting up to NO_VAL).

There are a couple more left to handle

Copy the strings into job request descriptor rather than just copying the
variable pointers to avoid over-writing contents

Test for unset job node_bitmap as the job may not yet be scheduled

Work still needed on scheduling, launch, env vars, etc.
This just creates the job records

-v for every pack job component combined in first pass of parsing

Still need to document the API, flesh out slurmctld response and
  more work on sbatch logic

Make _opt_default() function in salloc and sbatch reetrant.
Re-initializing variables used to hold changable fields for
each component of the heterogeneous job.

Document the new pack job APIs slurm_allocate_pack_job_blocking()
  and slurm_pack_job_lookup().

Used the wrong variable name in one place

No change in logic

Also minor logic changes for env vars

This sets most env vars, but not all

Operations performed on pack leader are applied to all components
  of the heterogeneous job.
Requests to operate on any other heterogeneous job component returns
  an error.

Remove logic that was intended for temporary use while debugging
  commit d763675c

Moved logic so it could be used for both pack job and normal job

This is the equivalent to slurm_allocation_lookup() for pack jobs.