Re: Regression in kernel 4.13-rcX for 32-bit x86 system - bisected to commit fc8dffd379ca

[Larry Finger]

On 08/14/2017 02:21 AM, Thomas Gleixner wrote:

On Mon, 14 Aug 2017, Larry Finger wrote:

My Del Latitude D600 (single 32-bit CPU) shows the following locking message
when booted:

============================================
WARNING: possible recursive locking detected
4.12.0-rc2-00029-gfc8dffd #159 Not tainted
--------------------------------------------
systemd-udevd/153 is trying to acquire lock:
  (cpu_hotplug_lock.rw_sem){.+.+.+}, at: [<c030fc26>] stop_machine+0x16/0x30

but task is already holding lock:
  (cpu_hotplug_lock.rw_sem){.+.+.+}, at: [<c0234353>] mtrr_add_page+0x83/0x470

....

  cpus_read_lock+0x48/0x90
  ? stop_machine+0x16/0x30
  stop_machine+0x16/0x30
  mtrr_add_page+0x18b/0x470
  mtrr_add+0x3e/0x70

This problem was bisected to commit fc8dffd379ca ("cpu/hotplug: Convert
hotplug locking to percpu rwsem"). I am very confident in the correctness of
the bisection.

Fix below.

Thanks,

        tglx

8<--------------------

diff --git a/arch/x86/kernel/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c
index c5bb63be4ba1..40d5a8a75212 100644
--- a/arch/x86/kernel/cpu/mtrr/main.c
+++ b/arch/x86/kernel/cpu/mtrr/main.c
@@ -237,6 +237,18 @@ set_mtrr(unsigned int reg, unsigned long base, unsigned 
long size, mtrr_type typ
        stop_machine(mtrr_rendezvous_handler, &data, cpu_online_mask);
  }

+static void set_mtrr_cpuslocked(unsigned int reg, unsigned long base,

+                               unsigned long size, mtrr_type type)
+{
+       struct set_mtrr_data data = { .smp_reg = reg,
+                                     .smp_base = base,
+                                     .smp_size = size,
+                                     .smp_type = type
+                                   };
+
+       stop_machine_cpuslocked(mtrr_rendezvous_handler, &data, 
cpu_online_mask);
+}
+
  static void set_mtrr_from_inactive_cpu(unsigned int reg, unsigned long base,
                                      unsigned long size, mtrr_type type)
  {
@@ -370,7 +382,7 @@ int mtrr_add_page(unsigned long base, unsigned long size,
        /* Search for an empty MTRR */
        i = mtrr_if->get_free_region(base, size, replace);
        if (i >= 0) {
-               set_mtrr(i, base, size, type);
+               set_mtrr_cpuslocked(i, base, size, type);
                if (likely(replace < 0)) {
                        mtrr_usage_table[i] = 1;
                } else {
@@ -378,7 +390,7 @@ int mtrr_add_page(unsigned long base, unsigned long size,
                        if (increment)
                                mtrr_usage_table[i]++;
                        if (unlikely(replace != i)) {
-                               set_mtrr(replace, 0, 0, 0);
+                               set_mtrr_cpuslocked(replace, 0, 0, 0);
                                mtrr_usage_table[replace] = 0;
                        }
                }
@@ -506,7 +518,7 @@ int mtrr_del_page(int reg, unsigned long base, unsigned 
long size)
                goto out;
        }
        if (--mtrr_usage_table[reg] < 1)
-               set_mtrr(reg, 0, 0, 0);
+               set_mtrr_cpuslocked(reg, 0, 0, 0);
        error = reg;
   out:
        mutex_unlock(&mtrr_mutex);

That fix works for me.

Thanks,

Larry