x86/topology: Use total_cpus not nr_cpu_ids for logical packages

nr_cpu_ids can be limited on the command line via nr_cpus=. That can break the
logical package management because it results in a smaller number of packages,
but the cpus to online are occupying the full package space as the hyper
threads are enumerated after the physical cores typically.

total_cpus is the real possible cpu space not limited by nr_cpus command line
and gives us the proper number of packages.

Reported-by: Mike Galbraith <umgwanakikbuti@gmail.com>
Fixes: 1f12e32f4cd5 ("x86/topology: Create logical package id")
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Xiong Zhou <jencce.kernel@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Andreas Herrmann <aherrmann@suse.com>
Link: http://lkml.kernel.org/r/alpine.DEB.2.11.1603181254330.3978@nanos

diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
index 81e6a432f23cb64286e837ae0f3227d2c302ca29..b2c99f811c3ff437ddfb45c7c4a56ed73101eccd 100644 (file)
--- a/arch/x86/kernel/smpboot.c
+++ b/arch/x86/kernel/smpboot.c
@@ -325,9 +325,14 @@ static void __init smp_init_package_map(void)
         * By not including this we'll sometimes over-estimate the number of
         * logical packages by the amount of !present siblings, but this is
         * still better than MAX_LOCAL_APIC.
+        *
+        * We use total_cpus not nr_cpu_ids because nr_cpu_ids can be limited
+        * on the command line leading to a similar issue as the HT disable
+        * problem because the hyperthreads are usually enumerated after the
+        * primary cores.
         */
        ncpus = boot_cpu_data.x86_max_cores;
-       __max_logical_packages = DIV_ROUND_UP(nr_cpu_ids, ncpus);
+       __max_logical_packages = DIV_ROUND_UP(total_cpus, ncpus);
 
        /*
         * Possibly larger than what we need as the number of apic ids per