I've added the ML back on Cc, due to the SPI discussion below... On 12/5/19 4:47 PM, Dan Murphy wrote: >>> As far as a device what country do you reside in? >> Germany, I'll drop you the address in a private Mail. I've send it in a separate mail. > Sounds good! I emailed our HW guys to see if we can get you a complete > test setup. Nice, I just need the SPI-Chip. > I had a MSP430 with firmware on it that transmit and received CAN frames > so I could run the cangen commands. I've several CAN devices at home test against, I've no need for microcontrollers yet. :D > But I make no promises on that level of hardware support. At the very > least maybe an EVM. > > When I get time I am also going to try to get rid of the regmap calls in > the driver to speed up the through put. Regmap is good, If you want to speedup things you should read stuff in bulk. See the mcp25xxfd driver I'm currently re-writing. But the mcp25xxfd is a SPI-only device, which makes things easier. Here I read all pending RX mailboxes in one or two goes (depending if we have a wrap around in the RAM) and then handle all in one go. https://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next.git/tree/drivers/net/can/spi/mcp25xxfd/mcp25xxfd-core.c?h=mcp25xxfd#n1148 Doing individual SPI register accesses kills performance, not regmap. regards, Marc -- Pengutronix e.K. | Marc Kleine-Budde | Embedded Linux | https://www.pengutronix.de | Vertretung West/Dortmund | Phone: +49-231-2826-924 | Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
Attachment:
signature.asc
Description: OpenPGP digital signature
