mx27 running too slow

[Fabio Estevam]

On Tue, Apr 8, 2014 at 11:31 AM, Alexander Shiyan <shc_work at mail.ru> wrote:
> Tue, 8 Apr 2014 10:49:38 -0300 от Fabio Estevam <festevam at gmail.com>:
>> On Wed, Apr 2, 2014 at 7:38 PM, Fabio Estevam <festevam at gmail.com> wrote:
>> > On Wed, Apr 2, 2014 at 7:32 PM, Fabio Estevam <festevam at gmail.com> wrote:
>> >> Hi Alexander,
>> >>
>> >> What is the BogoMIPs value you get on mx27 running 3.14?
>> >>
>> >> I am running Shawn's for-next branch and it is terribly slow on a mx27pdk:
>> >>
>> >> sched_clock: 32 bits at 13MHz, resolution 75ns, wraps every 322929872820ns
>> >> CPU identified as i.MX27, silicon rev 2.1
>> >> Console: colour dummy device 80x30
>> >> Calibrating delay loop (skipped), value calculated using timer
>> >> frequency.. 26.60 BogoMIPS (lpj=133000)
>> >>
>> >> ARM9 running at 400MHz should give around 200 BogoMIPs.
>> >
>> > Nevermind. On a clean 3.14 kernel I get the expected value:
>> >
>> > Calibrating delay loop... 199.06 BogoMIPS (lpj=995328)
>> >
>> > So this behaviour shows up only on Shawn's branch.
>>
>> Now I am running linux-next 20140408 and I am also getting the very
>> slow BogoMIPS:
>>
>> Calibrating delay loop (skipped), value calculated using timer
>> frequency.. 26.60 BogoMIPS (lpj=133000)
>>
>> Alexander,
>>
>> Do you see the same on Phytec mx27 board?
>
> Calibrating delay loop (skipped), value calculated using timer frequency.. 29.55 BogoMIPS (lpj=147777)
>
> Can you inspect code with "git-bisect"?

This is the result:

$ git bisect bad
1119c84aa3053bd415ba731ada1ecef24c8f82a2 is the first bad commit
commit 1119c84aa3053bd415ba731ada1ecef24c8f82a2
Author: Sebastian Andrzej Siewior <bigeasy at linutronix.de>
Date:   Wed Jan 22 12:35:44 2014 +0100

    ARM: imx: enable delaytimer on the imx timer

    The imx can support timer-based delays, so implement this.
    Skips past jiffy calibration.

    Signed-off-by: Sebastian Andrzej Siewior <bigeasy at linutronix.de>
    Signed-off-by: Shawn Guo <shawn.guo at linaro.org>