On Mon, Nov 11, 2019 at 12:00 PM Drew Fustini <drew@xxxxxxxx> wrote: > > On Mon, Nov 11, 2019 at 11:28:14AM +0100, Marc Kleine-Budde wrote: > > I successfully got a mcp25625 (which is basically a mcp2515 with > > internal phy) running on a rapsi using the frequency printed on the > > oscillator in the DT-overlay. > > I came across this patch by Andy on the linux-can archives: > > [PATCH v2] can: mcp251x: Get rid of legacy platform data > https://www.spinics.net/lists/linux-can/msg02407.html > > The description: > Instead of using legacy platform data, switch to use device properties. > For clock frequency we are using well established clock-frequency property. > > I reads the clock-frequency property from device tree: > > freq = clk_get_rate(clk); > - if (freq == 0 && pdata) > - freq = pdata->oscillator_frequency; > + if (freq == 0) > + device_property_read_u32(&spi->dev, "clock-frequency", &freq); > > > I feel like this is currently missing from mcp251x.c. I don't see how it is > reading the clock frequency from Device Tree as I don't see any code in the > current mcp251x.c to read device tree properties. > > I would appreciate any insights to improve my understanding. Also, I meant to ask as to why the clock for MCP2515 Oscillator comes from the spi device: clk = devm_clk_get(&spi->dev, NULL); line 1034 from: https://elixir.bootlin.com/linux/latest/source/drivers/net/can/spi/mcp251x.c#L1034 It seem like there should be separate clocks for the SPI frequency and the MCP2515 oscillator frequency. Am I missing something? Thanks, Drew
