This is the mail archive of the
ecos-bugs@sourceware.org
mailing list for the eCos project.
[Bug 1001397] I2C driver for Kinetis microcontrollers
- From: bugzilla-daemon at bugs dot ecos dot sourceware dot org
- To: ecos-bugs at ecos dot sourceware dot org
- Date: Wed, 24 Apr 2013 09:07:59 +0000
- Subject: [Bug 1001397] I2C driver for Kinetis microcontrollers
- Auto-submitted: auto-generated
- References: <bug-1001397-13 at http dot bugs dot ecos dot sourceware dot org/>
Please do not reply to this email, use the link below.
http://bugs.ecos.sourceware.org/show_bug.cgi?id=1001397
--- Comment #60 from Ilija Kocho <ilijak@siva.com.mk> ---
Created attachment 2206
--> http://bugs.ecos.sourceware.org/attachment.cgi?id=2206&action=edit
I2C driver print out actual frequency when trace enabled. Increment to
attachment 2188.
Mike
(In reply to comment #59)
> After setting CPU clock to 100Khz:
>
> Freq Mode Meas
> 100Khz Std 98Khz
> 100Khz Agr 111Khz
> 400Khz Std 363Khz
> 400Khz Agr 416Khz
Thank you for the measurements.
>
> From an application point of view, this is a little problematic, in that:
>
> - When you set the goal and value, you don't get feedback in the
> configuration that tells you what the result will be. You have to get a
> scope and measure it.
>
> - If you want to make it more accurate, you have to either guess the correct
> CPU clock, or reverse engineer the algorithm to calculate the required clock
> manually.
>
> One solution might be to have the configuration calculate the proper CPU
> clock so you can then manually change the clock to get a better result.
>
It would be good to have this convenience in CDL but it requires some Tcl
programming, which requires some time...
The second best is to enable some printout at run-time. The attached
incremental patch enables printout of actual I2C clock frequency and period
when trace is enabled.
Note: The printouts that I got are pretty close to what you have measured for
100kHz, but there is a large discrepancy for 400kHz. Can you please and repeat
measurements and check the printouts?
> I am ok with committing the solution as is because I can figure it out for
> 100 and 400 and will have the only answers I need. I think it is just a
> question of how much you want to help the application programmer.
If you are happy with proposed patch and there aren't objections, I would
proceed with commit.
Ilija
--
You are receiving this mail because:
You are on the CC list for the bug.