On 07/04/15 08:52, Lee Jones wrote: > On Mon, 06 Apr 2015, Bert Vermeulen wrote: > >> The SPI-connected CPLD chip controls access to the main NAND flash >> chip and five LEDs. >> >> Signed-off-by: Bert Vermeulen <bert@xxxxxxxx> >> --- >> arch/mips/include/asm/mach-ath79/rb4xx_cpld.h | 49 +++++ >> drivers/mfd/Kconfig | 7 + >> drivers/mfd/Makefile | 1 + >> drivers/mfd/rb4xx-cpld.c | 279 ++++++++++++++++++++++++++ >> 4 files changed, 336 insertions(+) >> create mode 100644 arch/mips/include/asm/mach-ath79/rb4xx_cpld.h >> create mode 100644 drivers/mfd/rb4xx-cpld.c > > This device doesn't look like an MFD, it rather looks like a CPLD > driver. We had a recent submission like this [1], perhaps this will > provide another argument for drivers/programmables or something. > > [1] https://lkml.org/lkml/2015/2/17/42 Yup, got bounced into drivers/mfd after initially submitting it as an SPI protocol driver (where it lives in openwrt). Indeed it's not a great fit anywhere -- not even programmables: this thing has its firmware on board, nothing ever feeds it on startup. Drivers for CPLDs don't necessarily have anything in common -- these are customized chips basically. In this case it's a NAND controller and GPIO/LED expander rolled into one. -- Bert Vermeulen bert@xxxxxxxx
