Re: [LTP] [PATCH] Add ftrace-stress-test to LTP

[Garrett C. <yan...@gm...>]

On Mon, Sep 6, 2010 at 9:16 AM, Subrata Modak
<su...@li...> wrote:
> Garret,
>
> Did you get a chance to review these tests further?

(see below for bottom post)

> On Tue, 2010-08-24 at 19:41 -0700, Garrett Cooper wrote:
>> On Tue, Aug 24, 2010 at 6:25 PM, Li Zefan <li...@cn...> wrote:
>> > Subrata Modak wrote:
>> >> Thanks. I need to test this on a suitable kernel. But doubtful if i can
>> >> add in the default run.
>> >>
>> >
>> > Why? Because it costs 90 secs? This is configurable, and some other
>> > test suites like cpuctl cost more time.
>> >
>> > If a test suite is not run by default, I guess people won't know its
>> > existence, probably.
>>
>> I think that more of the concern stems from the fact that it's a
>> performance test, and a lot of people complain if and when their test
>> machines slow down to a halt (take hackbench, proc01, for example).
>>
>> There are issues with the code, but it's dealing with the fact that
>> it's not really written in a portable way. Give me a bit of time and
>> I'll comment on what needs to be changed.

Li needs to fix some of my concerns which are valid (shell portability
by removing bashisms). Once that's done, then we can move on to the
next stage of review and potentially commit.

Thanks,
-Garrett

PS I still think that tst_kvercmp should be used because it's a lot
more versatile than what's added here, and will result in less
bitrotted code in the future.