Hi Phillip,
I've had a KEAZ128AMLK board sitting on my desk for like 6 months and just
haven't gotten around to starting to hack on it yet, so I'm glad somebody
is moving forward :). Let me dig through the TRMs and see if the EA8
family shares a flash controller (it's an automotive part), I'd love to
just provide support in your driver if possible.
Jeff
On Fri, Feb 19, 2016 at 4:39 AM Tomas Vanek <tom_...@users.sourceforge.net>
wrote:
> On 19.02.2016 8:50, Phillip Pearson wrote:
>
> Thanks Tom!
>
> I have a few questions about coding style / acceptable changes in OpenOCD.
>
> 1. It would be super super useful if struct target had a void *driver_priv
> member, like struct flash_bank. The target detection code in kinetis.c is
> a bit scattered currently, and it stores target-specific stuff like the
> SIM_SDID register value using flash_bank.driver_priv (struct
> kinetis_flash_bank). This is manageable when only targetting K and L
> devices, but E devices have a very different register layout, so there's a
> lot of redundant probing without anywhere to cache it. Would it be OK if I
> added a void *driver_priv to struct target?
>
>
> Please do not regard kinetis.c as coding style quality example. IMHO parts
> of this code would need a substantial
> rewrite (and yes, SDID detection should be processed once per MCU and not
> per flash bank as you pointed out). Unfortunately without support from
> Freescale it is very hard to test changes with sufficient coverage
> of Kinetis portfolio. So my personal policy in this particular driver is
> not to touch what is working.
>
> There is not a lot of common between KE and other Kinetis series and also
> automatic detection KE and others
> could be problematic. I would recommend to write a new, simple and clean
> flash driver for KE only.
>
>
> 2. Is it OK to create helper functions for things that currently have a
> lot of boilerplate code? For example, allocating a work area, copying in
> code, initializing r0/r1/r2, and executing the code.
>
> No problem. Submit the code and reviews show.
>
> 3. I've written a Python script that will run a C file declaring a naked
> function through arm-none-eabi-gcc to generate an assembly listing, then
> extract the opcodes and output them as a C file that defines a static const
> uint8_t[] array, like kinetis_flash_write_code in kinetis.c. Where would
> be the best place to put this? Should I cut and paste the code into
> kinetis.c (following existing convention) or #include it from elsewhere?
>
> There is a simpler way: src/helper/bin2char.sh - see
> http://openocd.zylin.com/#/c/3134/
> and http://openocd.zylin.com/#/c/3235/
>
> Thanks!
> Phil
>
> Not at all
>
> Tom
>
>
>
> On Thu, Feb 18, 2016 at 11:30 PM, Tomas Vanek <
> tom_...@users.sourceforge.net> wrote:
>
>> Welcome on board, Phil.
>>
>> Jeff wrote a mail with subject "Kinetis-KEx flash support" 4 month ago.
>> There is no KE-related work submitted in gerrit so I suppose nobody is
>> working on it.
>>
>> I have a custom testing board with MKE06Z128. I will help with driver
>> testing/review.
>>
>> Please follow HACKING file in OpenOCD root directory for instructions how
>> to submit a change.
>>
>> Tom
>>
>>
>> On 19.02.2016 5:09, Phillip Pearson wrote:
>>
>> Hi,
>>
>> I've been hacking up a flash driver for OpenOCD for the Kinetis E series
>> of chips. I have it currently working for the MKE02Z64VLD2 (FTMRH flash
>> controller), programming and debugging via Kinetis Design Studio.
>>
>> Is anyone else working on Kinetis support right now? I have some
>> MKE04Z8VTG4 boards I can use to program support for the FTMRE controller,
>> and have ordered a bunch of samples of different Kinetis E chips (different
>> sizes and speeds of MKE02/04/06) so I can test a bit better.
>>
>> I gather the procedure for getting code into OpenOCD is to upload it to
>> Gerrit for review?
>>
>> Cheers,
>> Phil
>>
>>
>> ------------------------------------------------------------------------------
>> Site24x7 APM Insight: Get Deep Visibility into Application Performance
>> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
>> Monitor end-to-end web transactions and take corrective actions now
>> Troubleshoot faster and improve end-user experience. Signup
>> Now!http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140
>>
>>
>>
>> _______________________________________________
>> OpenOCD-devel mailing
>> listOpenOCD-devel@lists.sourceforge.nethttps://lists.sourceforge.net/lists/listinfo/openocd-devel
>>
>>
>>
>
>
> ------------------------------------------------------------------------------
> Site24x7 APM Insight: Get Deep Visibility into Application Performance
> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
> Monitor end-to-end web transactions and take corrective actions now
> Troubleshoot faster and improve end-user experience. Signup Now!
> http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140
> _______________________________________________
> OpenOCD-devel mailing list
> OpenOCD-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/openocd-devel
>
------------------------------------------------------------------------------
Site24x7 APM Insight: Get Deep Visibility into Application Performance
APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
Monitor end-to-end web transactions and take corrective actions now
Troubleshoot faster and improve end-user experience. Signup Now!
http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140
_______________________________________________
OpenOCD-devel mailing list
OpenOCD-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/openocd-devel
