There was a follow on replacement patch against the prior
"kgdb: Timeout if secondary CPUs ignore the roundup".
See: https://lkml.org/lkml/2015/1/7/442
This patch is the delta vs the patch that was committed upstream:
  * Fix an off-by-one error in kdb_cpu().
  * Replace NR_CPUS with CONFIG_NR_CPUS to tell checkpatch that we
    really want a static limit.
  * Removed the "KGDB: " prefix from the pr_crit() in debug_core.c
    (kgdb-next contains a patch which introduced pr_fmt() to this file
    to the tag will now be applied automatically).
Cc: Daniel Thompson <daniel.thompson@linaro.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
                   online_cpus)
                cpu_relax();
        if (!time_left)
-               pr_crit("KGDB: Timed out waiting for secondary CPUs.\n");
+               pr_crit("Timed out waiting for secondary CPUs.\n");
 
        /*
         * At this point the primary processor is completely
 
        /*
         * Validate cpunum
         */
-       if ((cpunum > NR_CPUS) || !kgdb_info[cpunum].enter_kgdb)
+       if ((cpunum >= CONFIG_NR_CPUS) || !kgdb_info[cpunum].enter_kgdb)
                return KDB_BADCPUNUM;
 
        dbg_switch_cpu = cpunum;