Em Sat, Jan 30, 2016 at 09:06:51AM -0800, Andi Kleen escreveu:
> From: Andi Kleen <a...@linux.intel.com>
>
> Move the running/noise printing into printout to avoid
> duplicated code in the callers.
>
> v2: Merged with other patches. Remove unnecessary hunk.
> Readd hunk that ended in earlier patch.
> v3: Fix noise/running output in CSV mode
> v4: Merge with later patch that also moves not supported printing.
> Acked-by: Jiri Olsa <jo...@kernel.org>
So, the description makes it look like this is just elliminating code
duplication, but then, before we had:
[acme@jouet linux]$ perf.old stat -I 500 -e instructions,cycles sleep 1
# time counts unit events
0.500164711 548,376 instructions # 0.72
insn per cycle
0.500164711 756,872 cycles
1.000405470 <not counted> instructions
1.000405470 <not counted> cycles
1.000690913 156,623 instructions # 0.48
insn per cycle
1.000690913 226,630 cycles
[acme@jouet linux]$
And now we have instead:
[acme@jouet linux]$ perf stat -I 500 -e instructions,cycles sleep 1
# time counts unit events
0.500173751 555,962 instructions # 0.77
insn per cycle
0.500173751 722,964 cycles
1.000382613 0 instructions # 0.00
insn per cycle
1.000382613 0 cycles
1.001157068 154,765 instructions # 0.49
insn per cycle
1.001157068 223,609 cycles
[acme@jouet linux]$
So there is a change in behaviour.
- Arnaldo
