On Tue, Nov 3, 2020 at 9:03 AM Marc Kleine-Budde <mkl@xxxxxxxxxxxxxx> wrote: > > On 11/2/20 11:59 PM, Drew Fustini wrote: > > I modified devm_gpiod_get_optional() in mcp251xfd_probe() to look for > > "rx-int" instead of "microchip,rx-int". Is there a reason it was > > looking for "microchip,rx-int"? > > The DT property was changed why upstreaming the driver. So the overlay has to be > patched, not the code. > > > I am using branch v5.4-rpi/mcp251xfd-20201022-54 at commit 9e02abd4fe59. > > I'll update the overlay. > > > While it does now see the property, it does run into this issue: > > [ 6.830900] mcp251xfd spi0.0 (unnamed net_device) (uninitialized): > > RX_INT active after softreset, disabling RX_INT support. > > > > Any suggestions as to how to troubleshoot this? > > The waveshare board doesn't have the INT1 connected. So chose a free GPIO on you > rpi, add this as the parameter to the overlay and connect the GPIO to 3v3. > > Or, if you want to solder, connect the INT1 pin of one of the mcp2517fd with a > GPIO on the waveshare module. Thanks, I used rx-int of 27 for spi1 and connected to 3v3. The driver successfully registered the rx-int setting: [ 6.983684] mcp251xfd spi0.0 can1: MCP2517FD rev0.0 (+RX_INT +MAB_NO_WARN +CRC_REG +CRC_RX +CRC_TX +ECC -HD c:40.00MHz m:20.00MHz r:17.00MHz e:0.00MHz) successfully initialized. -Drew
