Clarify documenation with respect to CPU binding

doc/man/man1/salloc.1

@@ -1257,6 +1257,12 @@ The reservation ID on Cray systems running ALPS/BASIL only.
 \fBSLURM_CPU_BIND\fR
 Set to value of the \-\-cpu_bind\fR option.
 .TP
+\fBSLURM_CPU_BIND_LIST\fR
+\-\-cpu_bind map or mask list (list of SLURM CPU IDs or masks for this node,
+CPU_ID = Board_ID x threads_per_board +
+Socket_ID x threads_per_socket +
+Core_ID x threads_per_core + Thread_ID).
+.TP
 \fBSLURM_JOB_ID\fR (and \fBSLURM_JOBID\fR for backwards compatibility)
 The ID of the job allocation.
 .TP

doc/man/man1/sbatch.1

@@ -1388,6 +1388,12 @@ The reservation ID on Cray systems running ALPS/BASIL only.
 \fBSLURM_CPU_BIND\fR
 Set to value of the \-\-cpu_bind\fR option.
 .TP
+\fBSLURM_CPU_BIND_LIST\fR
+\-\-cpu_bind map or mask list (list of SLURM CPU IDs or masks for this node,
+CPU_ID = Board_ID x threads_per_board +
+Socket_ID x threads_per_socket +
+Core_ID x threads_per_core + Thread_ID).
+.TP
 \fBSLURM_JOB_ID\fR (and \fBSLURM_JOBID\fR for backwards compatibility)
 The ID of the job allocation.
 .TP

doc/man/man1/srun.1

@@ -1878,7 +1878,10 @@ if specified on the execute line.
 \-\-cpu_bind type (none,rank,map_cpu:,mask_cpu:)
 .TP
 \fBSLURM_CPU_BIND_LIST\fR
-\-\-cpu_bind map or mask list (<list of IDs or masks for this node>)
+\-\-cpu_bind map or mask list (list of SLURM CPU IDs or masks for this node,
+CPU_ID = Board_ID x threads_per_board +
+Socket_ID x threads_per_socket +
+Core_ID x threads_per_core + Thread_ID).
 
 .TP
 \fBSLURM_CPU_FREQ_REQ\fR

doc/man/man5/gres.conf.5

@@ -30,7 +30,6 @@ otherwise the default value is one. A suffix of "K", "M" or "G" may be used
 to mulitply the number by 1024, 1048576 or 1073741824 respectively.
 Note that Count is a 32\-bit field and the maximum value is 4,294,967,295.
  
-
 .TP
 \fBCPUs\fR
 Specify the CPU index numbers for the specific CPUs which can
@@ -42,6 +41,12 @@ delimited list or a range may be specified using a "\-" separator
 If not specified, then any CPU can be used with the resources.
 If any CPU can be used with the resources, then do not specify the
 \fBCPUs\fR option for improved speed in the SLURM scheduling logic.
+Since SLURM must be able to perform resource management on heterogeneous
+clusters having various CPU ID numbering schemes, use the SLURM CPU index
+numbers here
+(CPU_ID = Board_ID x threads_per_board +
+Socket_ID x threads_per_socket +
+Core_ID x threads_per_core + Thread_ID).
 
 .TP
 \fBFile\fR