Re: [V2 PATCH 0/6] x86, NMI: give NMI handler a face-lift

[Jason Wessel <jason.wessel@windriver.com>]

On 11/19/2010 05:09 PM, Peter Zijlstra wrote:
> On Fri, 2010-11-19 at 17:59 -0500, Don Zickus wrote:
>> 
>> @@ -1371,6 +1385,12 @@ void __init init_hw_perf_events(void)
>> 
>>         pmu_check_apic();
>> 
>> +       /* sanity check that the hardware exists or is emulated */
>> +       if (!check_hw_exists()) {
>> +               pr_cont("no PMU driver, software events only.\n");
>> +               return;
>> +       }
>
> Maybe report something like this:
>   "Broken PMU hardware detected, software events only."
>
> Because this is really not something that's supposed to happen.

The kgdb test suite is passing with Don's perf detect logic, so we are
back to good.  I am in agreement with Peter about the message
indicating that it is broken hardware.  We don't in any way shape or
form want leave the illusion this works in the VM.

# dmesg |grep Per
<6>Performance Events: no PMU driver, software events only.


Tested on qemu and kvm, several revisions worth, because it is all
automated.

Tested-by: Jason Wessel <jason.wessel@windriver.com>

Thanks,
Jason.