Re: RFC: i2c designware gpio recovery

[Tim Sander <tim@krieglstein.org>]

Good Day Phil

Am Mittwoch, 3. Mai 2017, 09:30:50 CEST schrieb Phil Reid:
> G'day Tim,
> 
> On 1/05/2017 21:31, Tim Sander wrote:
> > Good Day Phil
> > 
> > Am Montag, 1. Mai 2017, 09:57:35 CEST schrieb Phil Reid:
> >>> So i took a look into the device tree file socfpga.dtsi and found that
> >>> the
> >>> reset lines where not defined (although available in the corresponding
> >>> reset manager). Is there a reason for this? Other components are
> >>> connected.
> >> 
> >> There's a few thing like that where the bootloader has been expected to
> >> setup the resets etc.
> > 
> > Yes, but if the resets are not connected in the device tree, the linux
> > drivers are not going to use them?
> 
> Yes, so they should be added. I don't think we should assume the bootloader
> sets things up. But that doesn't seem to have been the assumption with the
> Alter SOC's.
I will prepare a patch for this.

> >>> However with the patch below my previously sent patch works!
> >>> 
> >>> If there is interest in would cleanup the patch and send it in for
> >>> mainlining. I think the most unacceptable part would be this line:
> >>> +       ret = gpio_request_one(bri->scl_gpio, //GPIOF_OPEN_DRAIN |
> >>> My gpio drivers refuse to work as output as they have no open drain
> >>> mode.
> >>> So i wonder how to get this solved in a clean manner.
> >> 
> >> I thought the gpio system would emulate open drain by switching the pin
> >> between an input and output driven low in this case. How are you
> >> configuring the GPIO's in the FPGA?
> > 
> > I don't switch to GPIO mode. As the I2C logic is only pulling active low,
> > i only do a wired and with the gpio (implemented in the fpga) port output
> > on the output enable line for the SCL output.  SDA is only an additional
> > input for the second in fpga gpio port.
> > 
> > A picture should make it a clearer:
> > 
> > gpio scl--\
> > i2c   scl --&---i2c mode output pin (configured as fpga loan)
> > 
> > In my case the original i2c pins where occupied by some other logic
> > conflicting so the i2c pins had to be shifted to some other pins using
> > fpga logic. So it was just a matter of adding a two port gpio port.
> 
> I think I understand. What soft core gpio controller are you using?
I am using the standard altera fpga gpios.

> >> Given a couple of days I can test this on some flack i2c hardware I have
> >> with a Cyclone-V SOC. I'm interested in the functionality as well.
> > 
> > Sounds good. If you need some further input how i have configured the fpga
> > drop me a line.
> > 
> >> For i2c that are connected to the dedicated HPS pins it should be
> >> possible
> >> to reconfigure the pin mux controller (see system manager) in the HPS to
> >> avoid the need to go thru the fpga to get direct control. The docs say
> >> this
> >> is "unsupport" but I've done some test and it does seem to work.
> > 
> > As far as i know the internal jtag chain is only used in the bootloader
> > and there is no linux driver? But it shouldn't be a too big problem to
> > port it to linux.
> > 
> > What i am unsure about is the fact that the internal jtag chain which
> > controls the pinmuxing might wreak havoc on other pin states if you
> > reconfigure it?
> Have a look at the Cyclone V handbook "pin mux control Group REgister
> Descriptions" From what I can see the chain is used to configure IO
> standards and drive strength. But not the actual muxes
Mh, there is not much to see in Volume 3. Just one paragraph and then a 
very encouraging closing line:
"Do not modify the pin multiplexing selection registers after I/O configuration."

I find the following lines in my favorite bootloader a little more enlightening:
The following function:
https://git.pengutronix.de/cgit/barebox/tree/arch/arm/mach-socfpga/system-manager.c
get feed with data from e.g.:
https://git.pengutronix.de/cgit/barebox/tree/arch/arm/boards/terasic-de0-nano-soc/pinmux_config.c
which doesn't look like beeing very memory mapped?

> >> I'm guess
> >> the no support is in a similar vain to the emac ptp FPGA interface
> >> couldn't
> >> be used when the HPS pin where used. But that got changed when the user's
> >> proved otherwise. There's just no pin ctrl driver yet to manage it.
> > 
> > I am interested in this ptp solution too. Is there anything on the way to
> > mainline?
> This was working the last time I tried it. I submitted a couple of minor
> patches for it a while ago. My hardware has a DSA switch attached to the
> ethernet port and so far I haven't figured out how to enable ptp when using
> the virtual lan ports on the DSA. But it worked fine when directly
> connected to a phy.
Thanks, will take a look.

Best regards
Tim