Re: [PATCH RFC 1/2] m68k: Add tracirqs

[Jean-Michel Hautbois <jeanmichel.hautbois@yoseli.org>]

Hi Geert, all,

On 23/10/2024 11:31, Jean-Michel Hautbois wrote:
> Hi Geert,
> 
> On 23/10/2024 11:13, Geert Uytterhoeven wrote:
>> Hi Jean-Michel,
>>
>> On Wed, Oct 23, 2024 at 11:07 AM Jean-Michel Hautbois
>> <jeanmichel.hautbois@yoseli.org> wrote:
>>> On 23/10/2024 10:47, Steven Rostedt wrote:
>>>> On Tue, 22 Oct 2024 11:21:34 +0200
>>>> Jean-Michel Hautbois <jeanmichel.hautbois@yoseli.org> wrote:
>>>>> I was not really expecting you to review the m68k one no :-).
>>>>> I think I have other issues which might have impact on ftrace too.
>>>>> For instance, when I launch cyclictest I have a warning about 
>>>>> HRTIMERS:
>>>>> # cyclictest -p 99
>>>>> WARN: stat /dev/cpu_dma_latency failed: No such file or directory
>>>>> WARN: High resolution timers not available
>>>>> policy: fifo: loadavg: 1.21 0.40 0.14 1/122 245
>>>>>
>>>>> T: 0 (  245) P:99 I:1000 C:  11203 Min:     92 Act:  623 Avg:  775 
>>>>> Max:
>>>>>      3516
>>>>>
>>>>> The latencies are quite high...
>>>>
>>>> Yes, if you don't have high resolution timers, the latency will be 
>>>> high.
>>>
>>> According to my config, I should have those:
>>> CONFIG_HIGH_RES_TIMERS=y
>>
>> Does your hardware have a high resolution timer, and do you have
>> a driver for it?
>>
>> $ git grep hrtimer -- arch/m68k/
>> $
>>
> 
> No, there is nothing with hrtimer. But, the architecture has four dma 
> timers, with a 8ns resolution at 125MHz says the documentation. I will 
> try to find a way to implement the missing part.
> 

I gave it a hacky try. And it seems to be *a lot* better:
# cyclictest -p 99 -l 10000
WARN: stat /dev/cpu_dma_latency failed: No such file or directory
policy: fifo: loadavg: 1.18 0.71 0.33 1/122 258

T: 0 (  258) P:99 I:1000 C:  10000 Min:    118 Act: 1104 Avg:  418 Max: 
   1858

I will try to clean it a bit for another patch set :-).

Thanks,
JM