Oops, I sent v4 again, so it had no changes at all. Can't I just resend v5 instead of calling it v6? September 24, 2018 10:48 AM, "Johan Hovold" <johan@xxxxxxxxxx> wrote: > On Sun, Sep 23, 2018 at 06:03:30PM +0200, Karoly Pados wrote: > >> This patch allows using the CBUS pins of FT-X devices as GPIO in CBUS >> bitbanging mode. There is no conflict between the GPIO and VCP >> functionality in this mode. Tested on FT230X and FT231X. >> >> As there is no way to request the current CBUS register configuration >> from the device, all CBUS pins are set to a known state when the first >> GPIO is requested. This allows using libftdi to set the GPIO pins >> before loading this module for UART functionality, a behavior that >> existing applications might be relying upon (though no specific case >> is known to the authors of this patch). >> >> Signed-off-by: Karoly Pados <pados@xxxxxxxx> >> --- >> Changelog: >> - v2: Fix compile error when CONFIG_GPIOLIB is not defined. >> - v3: Incorporate review feedback. >> - v4: Include linux/gpio/driver.h unconditionally. >> Replace and invert gpio_input with gpio_output. >> Make ftdi_gpio_direction_get return 0/1. >> Change dev_err msg in ftdi_set_bitmode_req. >> Change formatting of error checking in ftdi_gpio_get. >> Drop dev_err in ftdi_gpio_set. >> Remove some line breaks and empty lines. >> Change error handling in ftdi_read_eeprom (and adjust caller). >> Replace SIO->FTX in FTDI_SIO_CBUS_MUX_GPIO macro name. >> - v5: Read only 4 bytes from eeprom in ftx_gpioconf_init. >> Compare ftdi_read_eeprom result with 0 instead of eq. cehck. >> Reserve 4 GPIOs even for FT234X. >> Release CBUS after gpiochip deregister to avoid possible race. >> Adjust comment on FTDI_SIO_SET_BITMODE macro. >> Protect GPIO value/dir setting with mutex. > > This patch doesn't add any locking so I'm assuming you posted the wrong > version of the patch. > >> Add support for gpiochip.get_multiple and set_multiple. >> Add names to GPIO lines. > > I'll wait for v6. > > Thanks, > Johan
