Messages in this thread | | | Date | Wed, 9 Jul 2014 16:01:12 -0700 | Subject | Re: [PATCH 6/6] x86-mce: ensure the MCP timer is not already set in the mce_timer_fn. | From | Havard Skinnemoen <> |
| |
On Wed, Jul 9, 2014 at 2:04 PM, Luck, Tony <tony.luck@intel.com> wrote: > + /* Ensure a CMCI interrupt can't preempt this. */ > + local_irq_save(flags); > if (mce_available(__this_cpu_ptr(&cpu_info))) { > machine_check_poll(MCP_TIMESTAMP, > &__get_cpu_var(mce_poll_banks)); > > Does this remove the problem that you fixed in part4? If a CMCI can > no longer interrupt a poll ... do we really need the locks you added?
It's possible this fixes part of the problem. But I still think there's a race between multiple CPUs polling shared banks.
Havard
| |