x86/mce: Reenable CMCI banks when swiching back to interrupt mode
authorXie XiuQi <xiexiuqi@huawei.com>
Wed, 12 Aug 2015 16:29:41 +0000 (18:29 +0200)
committerIngo Molnar <mingo@kernel.org>
Thu, 13 Aug 2015 08:12:52 +0000 (10:12 +0200)
commit1b48465500611a2dc5e75800c61ac352e22d41c3
tree95d6d3ab35478660cfbff295de82596bd88006f6
parent8838eb6c0bf3b6a6494a163947ab3d1700ab45d2
x86/mce: Reenable CMCI banks when swiching back to interrupt mode

Zhang Liguang reported the following issue:

1) System detects a CMCI storm on the current CPU.

2) Kernel disables the CMCI interrupt on banks owned by the
   current CPU and switches to poll mode

3) After the CMCI storm subsides, kernel switches back to
   interrupt mode

4) We expect the system to reenable the CMCI interrupt on banks
   owned by the current CPU

   mce_intel_adjust_timer
   |-> cmci_reenable
       |-> cmci_discover     # owned banks are ignored here

  static void cmci_discover(int banks)
...
for (i = 0; i < banks; i++) {
...
if (test_bit(i, owned)) # ownd banks is ignore here
continue;

So convert cmci_storm_disable_banks() to
cmci_toggle_interrupt_mode() which controls whether to enable or
disable CMCI interrupts with its argument.

NB: We cannot clear the owned bit because the banks won't be
polled, otherwise. See:

  27f6c573e0f7 ("x86, CMCI: Add proper detection of end of CMCI storms")

for more info.

Reported-by: Zhang Liguang <zhangliguang@huawei.com>
Signed-off-by: Xie XiuQi <xiexiuqi@huawei.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: <stable@vger.kernel.org> # v3.15+
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: huawei.libin@huawei.com
Cc: linux-edac <linux-edac@vger.kernel.org>
Cc: rui.xiang@huawei.com
Link: http://lkml.kernel.org/r/1439396985-12812-10-git-send-email-bp@alien8.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>
arch/x86/kernel/cpu/mcheck/mce_intel.c