¿ªÔÆÌåÓý

Cross Fix to ALSA-LIB and GPIO PTT issues


 

? With the simple correction to the problem having been identified (thank you Dan Cross) the 04-May-2023 file on the site is not the updated version. The Git source now has the fix however.
? I'm running Direwolf on a Wandboard Dual (iMX6DL) ARM machine with ARCH linux. This with YAAC on an Intel Atom machine with Fedora. Direwolf on the Intel was processor intensive (50%) however on the ARM machine it's taking only 10%. Both machines are ancient but were handy and adapted for a mobile station. My current problem is finding the GPIO connection for PTT control of a Yaesu radio (PTT GPIO_19). The linux boot record shows a number of GPIO chips (0-6) and a dynamic allocation (MMIO?) "gpio gpiochip0: Static allocation of GPIO base is deprecated, use dynamic allocation." which may negate the DIrewolf interface. Is there anyone running a similar ARM system (PI 32 bit)?

Hasta,
Dave



 

¿ªÔÆÌåÓý


Hello Dave,

Thank you for the report on the ALSA fixes and now we wait for the downstreams to get this important fix.? I imagine that will take a long time to happen so a lot of people will get impacted.

To the second part of your email, it's phrased in such a way that's rather confusing to me.? It seems you have a an ARM-based Wandboard machine, an Intel Atom based machine, yet then ask about "Pi 32bit".?? Direwolf natively supports the Raspberry PI via the direct "GPIO" directive.? For example:

?? --
?? PTT GPIO 25
?? --

I suppose this is considered "static" but this is actively supported from the Raspberry Pi foundation.? That said, I take it you're looking for Direwolf support of non-Raspberry Pi ARM based GPIO access methods?

--David
KI6ZHD


On 08/24/2023 07:40 AM, Dave Cowell via groups.io wrote:

? With the simple correction to the problem having been identified (thank you Dan Cross) the 04-May-2023 file on the site is not the updated version. The Git source now has the fix however.
? I'm running Direwolf on a Wandboard Dual (iMX6DL) ARM machine with ARCH linux. This with YAAC on an Intel Atom machine with Fedora. Direwolf on the Intel was processor intensive (50%) however on the ARM machine it's taking only 10%. Both machines are ancient but were handy and adapted for a mobile station. My current problem is finding the GPIO connection for PTT control of a Yaesu radio (PTT GPIO_19). The linux boot record shows a number of GPIO chips (0-6) and a dynamic allocation (MMIO?) "gpio gpiochip0: Static allocation of GPIO base is deprecated, use dynamic allocation." which may negate the DIrewolf interface. Is there anyone running a similar ARM system (PI 32 bit)?

Hasta,
Dave




 

Hi Dave,
? I went over the ptt.c file and compared the debug output I had for the Direwolf startup. There didn't seem to be any errors as it follows the libgpiod syntax completely. My guess right now is that the GPIO pins aren't identified in the dtb table for the kernel startup. I've translated three versions and see no reference in the dtc's for a simple GPIO I/O interface on the external pins. Most of the defined pins are used for chip-chip interaction. So now I have to "re-invent the wheel" and do a dtc blob for my GPIO-19 pin on gpio@209c000 (the first bank of GPIO modules) and a gpiopin@209c000 definition with the needed pinctrl and pin-gpios data. The iMX6 reference data from Technexion is forthcoming as I've hung on my access and login, TBD. Unfortunately the other iMX6 ARM users don't leave their dtc files lying around for me to plagiarize.

Hasta,
Dave
KI6KPP


 

? Finally found the issue with using the GPIO pins on the Wandboard ARM machine. The GPIO pin I expected to use was muxed for use with a second SD drive. That they identified it in the diagrams for external use was the problem. Using a different external pin and re-designating it in the .conf file did the trick!? The 3.0v output drives an opto-isolator to run the PTT on a Yaesu FT-1800 with some current limiting resistors.
Curiously, I discovered I still have the KAM Plus and Yaesu FT270 I used in the 80's.