Hi,

On 03/31/2014 11:08 PM, [email protected] wrote:
> Hi!
>> Make performance_counter01 run default. This performance_counter01 case
>> is to use perf_event_open() to test hardware CPU events.
>>
>> Do not let performance_counter02 run defalult. This test need to know
>> an upper bound on the number of hardware counters, which varies in
>> different platforms. So if knowing this value, user can execute:
>> ./performance_counter02 -C (upper bound) to run this test.
> The test seems to default to 8. Can we just make it return TCONF if
> there are less than 8 hw counters? Having tests that are not executed by
> default is pointless.
Here I actually do not know how to get the number of hardware counters.
I had googled this issue, but do not find any method to get the number.

Whether can we do this like below:
1, we assume the number of  hardware counters is 100, I think this is big 
enough.
   There is some statistic data( 
http://en.wikipedia.org/wiki/Hardware_performance_counter):
    processor   available HW counters
    UltraSparc ll        2
    Pentinum lll          2
    IA-64                   4
    Pentium              18
    ARM Cortex-A8    4

2, if the bug is not fixed, that is to say "If the task clock counters in the 
event groups
don't stop when their group gets taken off the PMU", then the ratio will be 
close to 100.
so if the real running result is greater than 99 or 98, we may think this test 
result is fail.

3, or the test result is successful and the ratio will almost be the real 
number of hardware counters.
 

> Also what about moving these tests to syscalls? I see no reason not to.

OK, it's reasonable, thanks.

Regards,
Xiaoguang Wang
>

------------------------------------------------------------------------------
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

Reply via email to