Re: [Paparazzi-devel] Starting point for SPI on Lisa/M2
From:
Gerard Toonstra
Subject:
Re: [Paparazzi-devel] Starting point for SPI on Lisa/M2
Date:
Tue, 9 Oct 2012 13:41:27 -0300
It's probably going to appear automatically for you, but I just created a branch derived from spi_driver in my repo on github (gtoonstra), called finish_spi_driver, which contains the changes up to so far.
It doesn't work yet, but I think it's getting somewhere. Please see the pull request comments for more details.
Rgds,
Gerard
On Tue, Oct 9, 2012 at 6:09 AM, Felix Ruess <address@hidden> wrote:
Hi Gerard,
I merged master into spi_driver, so it should be up-to-date regarding libopencm3, etc....
Please note that stm32 spi driver on this branch don't compile yet as they are not adapted to the improved spi structure with phase and polarity, etc...
Most of the imu drivers, etc. will need to be updated as well, but should then be architecture agnostic (work for both lpc21 and stm32).
So if you want to have a go at this, it would be appreciated!
Cheers, Felix
On Tue, Oct 9, 2012 at 4:13 AM, Gerard Toonstra <address@hidden> wrote:
I've tried to give this a go this morning to see if my device would run on what's available, but couldn't figure out a suitable starting point. I've tried the "spi_driver" branch, but on my machine this didn't compile, I think due to missing libopencm3 headers
(probably relied on a machine installation of those somewhere).
If it's easy enough to establish a starting point, please let me know how. I could have a go at spi_1 (external), but spi3 looks a lot more difficult,
since it has a number of different possible configurations.
Rgds,
Gerard
On Mon, Oct 8, 2012 at 10:01 AM, Gautier Hattenberger <address@hidden> wrote:
I did the lpc part of the new spi driver a long time ago, and it
would be very nice if someone could finish the stm part (I'm not
using this mcu, so I can't help here). It would make the integration
of new spi device easier and cleaner. So any help is welcome here
Looking at the arch/stm32/mcu_periph/spi_arch.c, not only the phase
and polarity are missing, but also only spi2 seems to be implemented
here.
Gautier
Le 08/10/2012 12:29, Felix Ruess a écrit :
Just real quick: the STM part is still not finished,
it is still missing the phase and polarity settings:
On Mon, Oct 8, 2012 at 11:24 AM,
Gerard Toonstra <address@hidden>
wrote:
Hi all,
I'm looking at the complexity of attaching another SPI
device for communications purposes on Lisa/M2.
The expectation was that spi_arch.c would be more of an
SPI driver that you could use to initialize a structure
with
and then pass along to different functions for
reading/writing and so on, things that would typically
be done inside a driver.
Reading the code, I get the impression that instead it
is a file specifically written for communication with
the IMU,
as there are specific references to SS pins, addresses
and DMA's and the likes.
The examples and tests had some specific SPI code in
there not using any of the "mcu arch" code, indicating
that others
have seen the same thing and just tried to get their
equipment working on the hardware.
What's the best starting point in either tests, examples
or existing code for doing anything related to SPI on
the current
code base, which ideally uses the standard existing SPI
connector on the main board?